r «£* Volume II: Crosscutting Issues In Minority Health Report of Secretary's Task Force on Black & Minority Margaret M. Heckler Secretary U.S. Department of Health and Human Services PROPERTY OF THE NATIONAL LIBRARY OF MEDICINE Volume II: Crosscutting Issues In Minority Health Report of the Secretary's Task Force on Black & Minority Health Margaret M. Heckler Secretary U.S. Department of Health and Human Services August 1985 SECRETARY'S TASK FORCE ON BLACK AND MINORITY HEALTH MEMBERS Thomas E. Malone, Ph.D., Chairperson Katrina W. Johnson, Ph.D., Study Director Wendy Baldwin, Ph.D Betty Lou Dotson, J.D. Manning Feinleib, M.D., William T. Friedewald, Robert Graham, M.D. M. Gene Handelsman Jane E. Henney, M.D. Donald R. Hopkins, M.D. Stephanie Lee-Miller Dr.P.H. M.D. Jaime Manzano J. Michael McGinnis, M.D. Mark Novitch, M.D. Clarice D. Reid, M.D. Everett R. Rhoades, M.D. William A. Robinson, M.D., James L. Scott Robert L. Trachtenberg T. Franklin Williams, M.D. M.P.H ALTERNATES Shirley P. Bagley, M.S. Claudia Baquet, M.D., M.P.H. Howard M. Bennett Cheryl Damberg, M.P.H. Mary Ann Danello, Ph.D. Jacob Feldman, Ph.D. Marilyn Gaston, M.D. George Hardy, M.D. John H. Kelso James A. Kissko Robert C. Kreuzburg, M.D Barbara J. Lake Patricia L. Mackey, J.D. Delores Parron, Ph.D. Gerald H. Payne, M.D. Caroline I. Reuter Clay Simpson, Jr., Ph.D. Ronald J. Wylie ii VOLUME II: CROSSCUTTING ISSUES IN MINORITY HEALTH TABLE OF CONTENTS Introduction to the Task Force Report ............. Members of the Subcommittee on Data Development ......... PERSPECTIVES ON NATIONAL HEALTH DATA FOR MINORITIES: Report of the Subcommittee on Data Development ............ Supporting Papers 1. Benjamin S. Bradshaw, W. Parker Frisbie, Clayton W. Eifler: Excess and deficit mortality due to selected causes of death and their contribution to differences in life expectancy of Spanish-surnamed and other white males—1970 and 1980 ..... 43 2. Mary N. Haan, George A. Kaplan: The contribution of socioeconomic position to minority health .............. °9 3. M. Alfred Haynes, Girma Wolde-Tsadik, Paul Juarez: Associations of health problems with ethnic groups in ambulatory care visits ............... 4. Shiriki K. Kumanyika, Deborah L. Helitzer: Nutritional status and dietary patterns of racial minorities in the United States . . 118 5. Reiko Homma True: Health care service delivery in Asian . . 193 American communities ................ 6. Elena S. H. Yu, Ching-Fu Chang, William T. Liu, Stephen H. Kan: Asian-White mortality differences: Are there excess deaths? . . .209 7. Elena S. H. Yu, William T. Liu, Paul Kurzeja: Physical and mental health status indicators for Asian-American communities . . 255 MINORITY ACCESS TO HEALTH CARE IN THE MID-1980's.........287 HEALTH EDUCATION AMONG MINORITY POPULATIONS ........... 333 MINORITY AND OTHER HEALTH PROFESSIONALS SERVING MINORITY COMMUNITIES . . 277 iii INTRODUCTION TO THE TASK FORCE REPORT Background The Task Force on Black and Minority Health was established by Secretary of Health and Human Services Margaret M. Heckler in response to the striking differences in health status between many minority populations in the United States and the nonminority population. In January 1984, when Secretary Heckler released the annual report of the Nation's health, Health, United States, 1983, she noted that the health and longevity of all Americans have continued to improve, but the prospects for living full and healthy lives were not shared equally by many minority Americans. Mrs. Heckler called attention to the longstanding and persistent burden of death, disease, and disability experienced by those of Black, Hispanic, Native American, and Asian/Pacific Islander heritage in the United States. Among the most striking differentials are the gap of more than 5 years in life expectancy between Blacks and Whites and the infant mortality rate, which for Blacks has continued to be twice that of Whites. While the differences are particularly evident for Blacks, a group for whom information is most accurate, they are clear for Hispanics, Native Americans, and some groups of Asian/Pacific Islanders as well. By creating a special Secretarial Task Force to investigate this grave health discrepancy and by establishing an Office of Minority Health to implement the recommendations of the Task Force, Secretary Heckler has taken significant measures toward developing a coordinated strategy to improve the health status of all minority groups. Dr. Thomas E. Malone, Deputy Director of the National Institutes of Health, was appointed to head the Task Force and 18 senior DHHS executives whose programs affect minority health were selected to serve as primary members of the Task Force. While many DHHS programs significantly benefit minority groups, the formation of this Task Force was unique in that it was the first time that attention was given to an integrated, comprehensive study of minority health concerns. Charge Secretary Heckler charged the Task Force with the following duties: • Study the current health status of Blacks, Hispanics, Native Americans, and Asian/Pacific Islanders. • Review their ability to gain access to and utilize the health care system. • Assess factors contributing to the long-term disparities in health status between the minority and nonminority populations. 1 • Review existing DHHS research and service programs relative to minority health. • Recommend strategies to redirect Federal resources and programs to narrow the health differences between minorities and nonminorities. • Suggest strategies by which the public and private sectors can cooperate to bring about improvements in minority health. Approach After initial review of national data, the Task Force adopted a study approach based on the statistical technique of "excess deaths" to define the differences in minority health in relation to nonminority health. This method dramatically demonstrated the number of deaths among minorities that would not have occurred had mortality rates for minorities equalled those of nonminorities. The analysis of excess deaths revealed that six specific health areas accounted for more than 80 percent of the higher annual proportion of minority deaths. These areas are: • Cardiovascular and cerebrovascular diseases • Cancer • Chemical dependency • Diabetes • Homicide, suicide, and unintentional injuries • Infant mortality and low birthweight. Subcommittees were formed to explore why and to what extent these health differences occur and what DHHS can do to reduce the disparity. The subcommittees examined the most recent scientific data available in their specific areas and the physiological, cultural, and societal factors that might contribute to health problems in minority populations. The Task Force also investigated a number of issues that cut across specific health problem areas yet influence the overall health status of minority groups. Among those reviewed were demographic and social characteristics of Blacks, Hispanics, Native Americans, and Asian/Pacific Islanders; minority needs in health information and education; access to health care services by minorities; and an assessment of health professionals available to minority populations. Special analyses of mortality and morbidity data relevant to minority health also were developed for the use of Task Force. Reports on these issues appear in Volume II. Resources More than 40 scientific papers were commissioned to provide recent data and supplementary information to the Task Force and its subcommittees. Much material from the commissioned papers was incorporated into the subcommittee reports; others accompany the full text of the subcommittee reports. 2 An inventory of DHHS program efforts in minority health was compiled by the Task Force. It includes descriptions of health care, prevention, and research programs sponsored by DHHS that affect minority populations. This is the first such compilation demonstrating the extensive efforts oriented toward minority health within DHHS. An index listing agencies and program titles appears in Volume I. Volume VIII contains more detailed program descriptions as well as telephone numbers of the offices responsible for the administration of these programs. To supplement its knowledge of minority health issues, the Task Force communicated with individuals and organizations outside the Federal system. Experts in special problem areas such as data analysis, nutrition, or intervention activities presented up-to-date information to the Task Force or the subcommittees. An Hispanic consultant group provided inform- ation on health issues affecting Hispanics. A summary of Hispanic health concerns appears in Volume VIII along with an annotated bibliography of selected Hispanic health issues. Papers developed by an Asian/Pacific Islander consultant group accompany the data development report appearing in Volume II. A nationwide survey of organizations and individuals concerned with minority health issues was conducted. The survey requested opinions about factors influencing health status of minorities, examples of success- ful programs and suggestions for ways DHHS might better address minority health needs. A summary of responses and a complete listing of the organizations participating in the survey is included in Volume VIII. Task Force Report Volume I, the Executive Summary, includes recommendations for department-wide activities to improve minority health status. The recommendations emphasize activities through which DHHS might redirect its resources toward narrowing the disparity between minorities and nonminorities and suggest opportunities for cooperation with nonfederal structures to bring about improvements in minority health. Volume I also contains summaries of the information and data compiled by the Task Force to account for the health status disparity. Volumes II through VIII contain the complete text of the reports prepared by subcommittees and working groups. They provide extensive background information and data analyses that support the findings and intervention strategies proposed by the subcommittees. The reports are excellent reviews of research and should be regarded as state-of-the-art knowledge on problem areas in minority health. Many of the papers commissioned by the Task Force subcommittees accompany the subcommittee report. They should be extremely useful to those who wish to become familiar in greater depth with selected aspects of the issues that the Task Force analyzed. 3 The full Task Force report consists of the following volumes: Volume I: Executive Summary Volume II: Crosscutting Issues in Minority Health: Perspectives on National Health Data for Minorities Minority Access to Health Care Health Education and Information Minority and other Health Professionals Serving Minority Communities Volume III: Cancer Volume IV: Cardiovascular and Cerebrovascular Diseases Volume V: Homicide, Suicide, and Unintentional Injuries Volume VI: Infant Mortality and Low Birthweight Volume VII: Chemical Dependency Diabetes Volume VIII: Hispanic Health Issues Inventory of DHHS Program Efforts in Minority Health Survey of the Non-Federal Community 4 Perspectives On National Health Data For Minorities Report of the Subcommittee On Data Development SUBCOMMITTEE ON DATA DEVELOPMENT Robert Graham, M.D., Chairperson Assistant Surgeon General Administrator Health Resources and Services Administration Jane Delgado, Ph.D. Special Assistant on Minority Affairs Office of the Secretary Manning Feinleib, M.D., Dr.P.H. Director National Center for Health Statistics Donald Hopkins, M.D. Deputy Director Centers for Disease Control Stephanie Lee-Miller Assistant Secretary for Public Health Office of the Secretary J. Michael McGinnis, M.D. Assistant Surgeon General Deputy Assistant Secretary for Health Director, Office of Disease Prevention and Health Promotion Office of the Secretary William Robinson, M.D., M.P.H. Deputy Director Bureau of Health Professions Health Resources and Services Administration Staff Liaison: Clifford Patrick, Ph.D. 6 PERSPECTIVES ON NATIONAL HEALTH DATA FOR MINORITIES SUMMARY The Data Development Subcommittee served as an preliminary review committee whose purpose was to guide the Task Force, provide national data to other subcommittees and commission reports on subjects that cut across all health problem areas. The subcommittee performed initial analyses for the Task Force by comparing the mortality between minority groups and the nonminority population in the United States. The technique of "excess deaths", a methodology particularly suited to the Task Force needs, allowed the Task Force to set priorities among health issues. This technique highlighted health needs that affect large proportions of the minority population. As a result of this technique, six causes of death were selected for more intensive examination. The approach was limited by lack of complete data for Hispanic and Asian/Pacific Islander populations and by its dependence on the size and characteristics of the nonminority comparison group. Supplementary efforts to identify more specifically the health status and needs of minorities included regional data, commissioned papers, and a range of other morbidity datasets. Although these sources were not more complete than death certificate information, they confirmed the selection of priority health issues. The papers and other reports in this volume summarize health-related issues that cut across each subcommittee area and reflect additional information reviewed by all components of the Task Force. 7 PERSPECTIVES ON NATIONAL HEALTH DATA FOR MINORITIES FIGURES AND TABLES FIGURES Figure 1. Death Rates by Age at Death: White Males and White Females Average Annual Rates for All Causes of Death, U.S. 1979-81 Figure 2. Death Rates for Black Males Relative to White Males Average Annual Rates for All Causes of Death, U.S., 1979-81 Figure 3. Death Rates for Black Females Relative to White Females Average Annual Rates for All Causes of Death, U.S., 1979-81 Figure 4. Death Rates for Native American Males Relative to White Males Average Annual Rates for All Causes of Death, U.S., 1979-81 Figure 5. Death Rates for Native American Females Relative to White Females Average Annual Rates for All Causes of Death, U.S., 1979-81 Figure 6. Death Rates for Asian Males Relative to White Males Average Annual Rates for All Causes of Death, U.S., 1979-81 Figure 7. Death Rates for Asian Females Relative to White Females Average Annual Rates for All Causes of Death, U.S., 1979-81 9 TABLES Table 1. Average Annual Total and Excess Deaths in Blacks Selected Causes of Mortality, 1979-1981 Table 2. Life Expectancy for Males by Race at Selected Ages 1969-71 and 1979-81 Table 3. Life Expectancy for Females by Race at Selected Ages 1969-71 and 1979-81 Table 4. Relative Risk of Mortality for Blacks Compared with Whites All Causes of Death, 1979-81 Table 5. Relative Risk of Mortality for Blacks Compared with Whites Selected Causes of Death by Age and Gender, 1979-81 Table 6. Relative Risk of Morbidity for Blacks Compared with Whites Selected Conditions by Age and Gender, 1978-80 Table 7. Percent of U.S. Population Groups Visiting a Physician During the Past 12 Months (by Age and by Sex) Table 8. Bed Disability Days Per Person During One Year U.S. Population Groups (by Age and by Sex) Table 9. Percent with Activity Limitation During the Past Year U.S. Population Groups (by Age and by Sex) Table 10. Work Days Lost Per Currently Employed Person During the Past Year U.S. Population Groups (by Age and by Sex) Table 11. Prevalence of End Stage Renal Disease (ESRD) by Race and Diagnosis Dialysis and Transplant Patients, 1982 10 PERSPECTIVES ON NATIONAL HEALTH DATA FOR MINORITIES OVERVIEW The Data Development Subcommittee of the Secretary's Task Force on Black and Minority Health examined a variety of methods to measure the disparity between the health of minority and nonminority Americans. The analytical approach selected by the Subcommittee was critical in guiding the Task Force in its efforts to establish priorities in minority health, structure further Task Force activities, and recommend ways to improve the health of Black, Hispanic, Asian/Pacific Islander, and Native Americans. Initial examination of national data as reported in Health US, 1983 showed disparities between Black and White populations in life expectancy, infant mortality, several selected causes of death, and other indicators. While national data on the health status of Blacks are more complete, available data on Native Americans and Hispanics suggest that these populations suffer a greater disease and mortality burden than Whites. Charged with fuller exploration of health discrepancies for all minorities, the Task Force sought a methodology to guide their deliberations. After careful consideration of the data in Health US, 1983 as well as a more detailed presentation of national datasets, the Task Force concluded that additional analyses were required to explore why the discrepancy has continued for such a long period. At the June 1984 meeting of the Task Force, as a result of a presentation by Dr. M. Alfred Haynes, President and Dean of the Drew Postgraduate Medical School in Los Angeles, California, Task Force members decided that the statistical method of "excess deaths" was particularly suited for comparing minority and nonminority population groups. At that time, the Task Force established a Data Development Subcommittee, charged with applying the excess deaths technique to national mortality data and identifying other issues to be addressed. Dr. Robert Graham, Administrator, Health Resources and Services Administration, chaired the Subcommittee. At the July 1984 Task Force meeting, the results of the calculation of excess deaths by age, sex, race, and cause of death were presented by the Data Development Subcommittee to the Task Force membership. The excess deaths analysis clearly demonstrated that six health problem areas most significantly affect the health of minority populations. These areas are: cancer, cardiovascular disease (including stroke), cirrhosis and liver disease (indicators of chemical dependency), diabetes, homicide and unintentional injuries, and infant mortality (death under one year of age). Subcommittees addressing these areas were established. Where national data were missing or incomplete, papers were commissioned reviewing regional data or special topics. 11 Data and analyses reviewed by the Data Development Subcommi provided to the six subcommittees for further exploration. To augment and confirm the priorities elucidated by the excess deaths analysis, additional measures of mortality and morbidity were reviewed. Additional mortality indices included "excess" person-years of life lost, life expectancy, and relative risk of death by cause. Morbidity measures included prevalence rates of selected diseases, hospital admissions, physician visits, limitation of activity and self-assessed health status. These were presented and discussed at the December 1984 meeting of the Task Force. A number of scientific papers were commissioned by the Data Development Subcommittee on special topics not included in other subcommittee work; these include socioeconomic influences on health, nutrition, health of Asian subgroups and regional analyses of Hispanic populations. The Task Force recognized that other issues cut across health problem areas and influence the health status of minority Americans. Reports representing state-of-the-art expertise in these fields were developed. These issues include health education, access to health care, and availability of health professionals to minorities. The report of the Data Development Subcommittee describes the methodology and analytical processes that led to the findings and recommendations reported in Volume I. The principal health indices, limitations in the data and methodology, and selected findings drawn from the methodology used by the Data Development Subcommittee are included. SUBCOMMITTEE DATA SOURCES The measure of excess deaths is not a routinely used mortality indicator. The more traditional measures are usually presented as ratios (e.g. relative risk). They fail, however, to impart a sense of the magnitude of the disparity that is achieved when using the number of excess deaths. The Task Force used excess deaths, therefore, as a methodology that more dramatically characterizes the disparities in health status. Because it is less familiar than other measures of mortality, the technique must be explained in some detail. The excess deaths methodology used the nonminority (White) death rate as a baseline from which to calculate the actual number of minority deaths exceeding those which would have occurred if a minority group experienced the same mortality rates as the White population. A death rate, from all causes or for a specific cause of death, is commonly expressed as a rate of deaths per 100,000 population. Among all racial/ethnic groups, death rates vary by gender, age, and cause of death reported. 12 Figure 1 shows the overall average annual death rate for White males and females by age group. In general, women have lower death rates and longer life expectancy than men at any comparable age. Death rates for those under age one reflect infant mortality. Relatively few deaths are observed among children over one year of age. A slight increase is observed in death rates in early adult life, largely attributed to external causes including homicide and accidents. A close steep rise in the rate of mortality is seen for adults after age 65. Death rates are influenced by mortality patterns related to the different ages at which a specific cause of death occurs. Causes of death confined to infancy, for example, are a major proportion of the deaths occurring before adulthood, whereas deaths from cancer and heart disease are highest in middle and later life and eventually account for the greatest proportion of total deaths. Consequently, the contribution of a particular cause of death is related to the age at which it is most expected. Examination of major contributions to mortality during early, middle, or later life will reveal differences in the causes of death affecting young, middle-aged, or older individuals. Gender, age, and cause of death differences thus became important in excess deaths analysis to identify the health issues of greatest concern to a population group. The excess deaths analyses of the Data Development Subcommittee were derived from death rates and depend on the validity and reliability of those statistics. Data were computed from datasets supplied by the National Center for Health Statistics (NCHS) annual mortality microdata tapes. The Task Force employed mortality data for the years 1969-71 and 1979-81. Decennial census population data for 1970 and 1980 were used as the denominators to compute annualized mortality rates. National mortality data were available for Whites and Blacks, fewer and less complete data were available for Native Americans, Asian/Pacific Islanders, and Hispanics. EXCESS DEATHS METHODOLOGY The number of excess deaths was calculated by applying the age, sex, and cause-specific death rates for Whites (under age 70) to the comparable minority populations to derive an "expected" number of minority deaths. These were compared with the actual deaths that occurred in that minority group, computed separately by age, sex and cause of death. In the Task Force Report this calculation is shown as excess deaths = actual deaths - expected deaths. For those age or cause-of-death categories in which the minority death rate equalled the White rate, there were no "excess" deaths for the minority. For causes of death in which the minority rate was lower than the White rate, comparison with the White baseline resulted in a "negative" number of excess deaths. For example, Blacks under age 45 have a lower death rate from unintentional injuries than Whites. To 13 prevent distortion of the number of excess deaths that actually occur among minorities due to the arithmetical summation of negative and positive numbers, negative numbers were not included in calculation c excess deaths (see Volume 1, page 71, Table 6, final column). Because of variations in death rates across the age and cause-of-death categories, and because the mortality profile of each minority also varies across these categories, excess deaths do not occur uniformly each minority. Reasons for lack of excess deaths in some of these categories are discussed in the section on limitations of the data analysis. Although the actual death rate for any population varies according to age as shown in Figure 1, the excess deaths methodology uses the White rate for each sex, age, and cause of death as a baseline to evaluate minority mortality. Visualized as a graph, this comparative method portrays the minority death rates as they differ from the mortality rate of the White population for each age group. For example, Figure 2 shows the overall death rate for Whites as a baseline with he rate for Black males in each age category compared with the White rate Except for ages 15-19, the mortality rate for Black males is higher than for White males. The disparity between the two rates illustrates the deaths among Black males who would not have died if their death rate equalled the baseline rate, mortality of White males under age 70. Figure 3 shows excess deaths for Black females. In an analogous method, the Task Force calculated the disparity between each minority group for whom data are available and the White population for age- and sex-specific overall mortality. Figures 4 and 5 illustrate the disparity between Native American death rates and White death rates. As discussed in the Executive Summary, excess deaths for Native Americans occur principally among those under age 50. National data indicate that Asian/Pacific Islanders as an aggregate population group have an overall death rate lower than that of White Americans. Figures 6 and 7 show the Asian death rate as lower than the White baseline rate for both males and females and at all ages. This generalization however obscures the several small sub-groups of Asians with unique health needs. These issues are explored in the accompanying papers and subcommittee reports. As noted in other Task Force volumes, national data on mortality rates for Hispanic Americans are not uniformly or reliably available and excess deaths could not be calculated from this dataset. Following analysis of overall excess deaths the Task Force sought to determine the causes of death that account for the excess deaths. Subsequent analyses showed that more than 80% of the excess deaths occurring among minorities under age 70 fell into six areas. Table 1, reprinted from Volume I, Executive Summary, summarizes the contributions of the major causes of death to overall excess deaths 14 among the Black population. Although the actual contribution of each cause of death varies among the four minority groups, these six areas remain significant for most of the higher mortality reported among minorities. Based on examination of mortality patterns at different ages, the Task Force reported excess deaths only to age 70. A finding, not unique to this analysis, was the existence of lower mortality rates after age 70 for Blacks and other minorities compared with Whites. This "cross- over" effect was seen for overall death rates and most of the major causes of death, including cancer, cardiovascular disease and stroke, cirrhosis and liver disease, and unintentional injuries. Although the exact cause of the "crossover" effect is not known, this phenomenon may reflect inaccuracies in data collection and unreliability in reporting age or a hardiness among those minorities who survive to later life. Emphasis on mortality before age 70 in the Task Force analysis avoids confounding of data from mortality rates of the oldest population whose experiences are unlike those at earlier ages. LIMITATIONS OF DATA ANALYSES Limitations of the excess deaths analysis used by the Task Force fall into two categories: limitations of the data base and limitations inherent to the excess deaths methodology itself. Limitations of the Data Base The dataset analyzed by the Task Force was severely limited by the lack of accurate and complete national mortality data on the Hispanic population. National mortality data for Hispanics were not available because, although the current Standard Certificate of Death adopted in 1978 includes a race identification (White, Black, American Indian, etc.), it does not include an Hispanic identifier. The National Center for Health Statistics (NCHS) has recommended that states voluntarily add ethnic identifiers and has provided instructions for the Funeral Directors' Handbook on Death Registration and Fetal Death Reporting (1978). The reporting of Hispanic identifiers on death certificate data from 1979-1981, however, was not adequate for analysis by the Task Force. The revised Standard Certificate of Death and the Standard Report of Fetal Death, proposed for implementation effective with the 1988 data year, however, will include an Hispanic identifier. Until that time, NCHS will prepare a report on Hispanic mortality for those states with ethnic identifiers. The Subcommittee on Data Development attempted to overcome the absence of comparable national data for Hispanics by using the country of birth indicator (for Cuba or Mexico) on the death certificates along with Cuban- or Mexican-born population data from the 1980 U.S. Census to suggest possible areas of highest priority for the health of Hispanics. This approach, however, had severe limitations in its coverage, both in 15 the census populations and the death certificates. For example, significant undercount of Hispanic mortality under one year is apparent when using foreign-born death certificates since the infant mortality rate among those born in Cuba or Mexico but living long enough to reach the United States is very small. Foreign-born mortality should not be considered as representative of the Hispanic American health status and cannot be acknowledged as parallel to data on other groups. As another attempt to obtain comparable data on Hispanics, death certificates for the Spanish surname and White non-Spanish surname populations of Texas were analyzed for the Task Force to try to identify disparities in health for this Hispanic population. The analysis is included in this volume in the paper by Bradshaw. While not national, these data provided useful insights and directions for recommendations to the Task Force and its subcommittees. Compared to Hispanic Americans, data on Asians are more reliably reported on death certificates. Death certificate data for Asian/Pacific Islanders are limited, however, because they reflect mainly the group of Asians who are the healthiest and most numerous—Chinese and Japanese. Asian/Pacific Islanders whose health needs are reported as different compared with Whites, such as Filipinos (high levels of hypertension), native Hawaiians (cancer) and Southeast Asians (infant mortality), are few in number in national data. The Task Force attempted to obtain additional data on health problems of Asian Americans in the Pacific Basin through a review of data made available by the San Francisco Regional Health Administrator, Public Health Service (PHS/DHHS). This analysis, conducted by the University of Hawaii under contract to the PHS, was seriously hampered by an absence of information, especially reliable vital statistics, in all jurisdictions in the Basin except Hawaii. Again, the Task Force supplemented death certificate analyses by commissioning consultants to analyze and compare data on health status of Asian Americans with that of Whites. Their papers (included in this volume) aided in emphasizing the needs of Asian/Pacific Islanders by analysis of mortality and morbidity comparisons among Chinese, Japanese, and Filipino Americans. Data from the Indian Health Service (IHS) were used to augment death certificate analyses and confirm the findings of the excess deaths analysis. Native American mortality rates compared with other population groups show a high death rate at early ages (before age 45) but lower death rates after middle age. It has been suggested that Native American populations have an early "crossover" effect (Figures 4 and 5). IHS data emphasize health needs related to higher rates of suicide, unintentional injuries, and diabetes among Indian subgroups. 16 Limitations of the Methodology Limitations due to the technique of excess deaths are particularly important in interpreting the Task Force findings. Excess deaths are based on the nonminority death rate and are a comparative and a relative measure, dependent on both the death rates of the White population and the size of the minority group. The White death rate that is used as the base from which to estimate expected deaths, as shown in Figure 1, is not a constant rate, but varies according to characteristics of the White population. If a major cause of death among Blacks is also a major cause among Whites, such a cause can go unidentified with this comparative approach. For example, relatively higher rates of death from myocardial infarction among White males diminish the impact of excess deaths among Blacks from this cause irrespective of the actual number of individuals affected. Another limitation is that use of the number of excess deaths rather than death rates may underrepresent some concerns due to the dependence of the measure on the population size of the minority group and the small population base of minority subgroups. Consequently, the Task Force overall findings are shown most prominently for the Black population for which there are sufficient numbers within the population base to be relatively confident of priorities. Use of excess deaths is less appropriate to determine priorities for smaller minority groups. Similarly, the excess deaths technique is not useful for most smaller population units, for example, county or city-level analyses that have a small population base and fewer members of minority groups. Use of national data in excess deaths analysis is most appropriate to ensure the broadest population bases for calculation and to provide information for programmatic efforts that must apply to the entire country. Comparisons of minority and nonminority health status over time using the excess deaths techniques may be misleading unless changes in the White baseline rate are carefully evaluated. A change (increase or decrease) in death rates among the White population will modify the minority excess even with no change in the actual minority mortality rate. Analyses of contributors to excess deaths emphasize health needs that affect the greatest number of people within a population group and minimize needs of those with less common or rare conditions. That is, excess deaths combine measures of relative risk with public health impact as determined by the number of individuals at risk or affected by a condition. For example, conditions such as a sickle cell anemia may show great disparity in rates between the minority and nonminority population, but actually occur to relatively few individuals compared with those affected by heart disease. In planning national priorities, the Task Force selected areas which have a strong public health impact. 17 Less common health problems must still be addressed by programs at national or other levels. The technique of excess deaths as used by the Task Force was an appropriate tool for setting priorities relevant to policy formation, rather than as a methodology intended to replace other accepted epidemiological and investigatory techniques. Excess deaths analysis clearly highlighted the differences between minority and nonminority health status. While it allowed the Task Force initial identification of problem areas, it did not specify the causes of those disparities. Subcommittees of the Task Force were formed to do the next analytical step: the detailed examination of each area to determine what is known about the reasons for the persistence of inequities in each cause of death. The subcommittee reports contain findings and recommendations related to the causes of the disparity. OTHER MORTALITY MEASURES Other measures of mortality used by the Task Force to reinforce and confirm their basic findings are briefly reported here. Person-years of life lost has become a widely used indicator of the impact of mortality on public health. This measures the degree to which populations are affected by differences among groups in age at death. It is often used to compare the social, economic, or population impact of different causes of mortality for several groups of an entire population. The measure of person-years of life lost gives greater weight to deaths occurring earlier in life than those occurring later. For example, each death due to infant mortality would result in about 70 years of life lost to society, a homicide of a 25-year old would result in 45 years of life lost, and death at 65 would incur 5 person-years of life lost. Interventions to prevent premature death may be related to greater societal benefit from preservation of an individual's economic and personal contribution. Seeking to examine the impact of the age of death on minority/non- minority differences, the Data Development Subcommittee used the measure of excess person-years of life lost as a modification of the more traditional measure. As in other Task Force data, this measure is based on the years lost due to excess deaths, not the total deaths for a given cause. Excess person-years of life lost have been calculated up to age 70 from the Task Force analysis by multiplying the number of excess deaths that occurred at each age by the difference between that age and 70. Results indicated that among Black men 914,688 years of life before age 70 were lost each year in excess of the person-years lost by the White population. Among Black females 573,159 excess person-years were lost annually in excess of the loss among White females. Among Native Americans, 34,087 excess person-years for males and 17,406 for females were lost annually in excess of the White population. These figures 18 give an indication of the personal and societal loss due to mortality disparities between minority and nonminority populations. Life expectancy also was examined by the Data Subcommittee for each minority group. Comparisons of life expectancy in population groups are a standard indicator of differentials in health. Life expectancy takes into account the death rates at different ages in the population at risk and indicates the number of years one is expected to survive from a given age, at birth (age 0) or after reaching a particular age. The person-years of life lived by each cohort are used in computing life expectancy for each minority. Table 2 (males) and Table 3 (females) show life expectancy at selected ages for Blacks and Whites and compare life expectancy changes since 1961-71. Differences over the decade of the 70's for Whites and Blacks indicate gains in life expectancy for Blacks, although the overall disparity in life expectancy continues. Smaller differences in life expectancy between Whites and Blacks after age 65 demonstrate the cross-over effect among older Blacks. Relative risk of death due to a specific cause, as used by the Task Force, is the ratio of the minority death rate to the White rate and indicates the proportional risk to the minority population relative to the White population. Limitations of this measure as an index of overall health status must also be recognized. Diseases that have extremely low, overall mortality rates have few numbers of expected or excess deaths. However, relative to the White population, certain minority groups may show a high relative risk for those diseases. Thus an uncommon cause of death may appear with a high relative risk although it may not appear significant when considering excess deaths because its impact in number of persons affected may be small. Reliance on this measure to set national priorities may be misleading if it diverts attention from those causes of death which have the largest population impact to those which have a high relative risk but low population occurrence. Examination of differences in overall relative risk of death for each age group is, however, useful in suggesting targets for intervention. As with other measures, age-specific data available to the Task Force were most complete to determine mortality rate ratios for Blacks compared with Whites. The Task Force Data Subcommittee used overall ratios to examine the age categories of greatest disparity between Blacks and Whites. Table 4 indicates that in infancy (under age one) and among those age 25-54, Blacks have a death rate twice that of the White population and show greatest Black/White inequities. Combining age groups with selected causes of death, risk for major causes of death are presented in Table 5. The age-specific data reinforce the selection of the six priority areas for minority health activity that can affect minority health status. The cross-over effect 19 is shown again in these data by the ratios that are under 1.0 for o age categories. Morbidity Measures Although the NCHS morbidity database generally includes race and Hispanic identifiers in questionnaires, most data are from national sample surveys with proportional population representation, but in which minorities are not oversampled. Smaller groups, particularly Native Americans and Asian/Pacific Islanders, are represented in these national morbidity data by very few households. Consequently, population-level inferences of health needs are not appropriate because there are not enough responses to draw valid conclusions concerning the health of the group as a whole. The Task Force reviewed available overall morbidity data and urged the six subcommittees to consider morbidity in their individual reports. Four sources of morbidity data were examined by the Data Development Subcommittee to supplement national mortality data. First, several morbidity measures were taken from the National Health Interview Survey (NHIS) of NCHS. The NHIS is an annual survey of approximately 40,000 households, with health information obtained on all household members. In order to collect data on a wider range of chronic diseases in recent years, households sampled for the NHIS were divided into subsamples and each subsample was asked about the occurrence of one of six different lists of chronic diseases. Because of the small numbers in each subsample, the Task Force found it necessary to combine several years of NHIS data together to make reliable estimates of the prevalence of many diseases. Rates for five conditions considered important by the Task Force are summarized in Table 6. Data from NHIS related to use of medical services and self-reported days of disability, Tables 7, 8, 9, and 10, were reviewed by the Task Force. The DHHS Office of Civil Rights made available to the Task Force a survey of hospitals under the Hill-Burton Act. This survey reports the annual number of admissions and emergency room visits to each hospital. Racial and ethnic information is available for all patients. The aggregate nature of the hospital data, however, precluded analyzes by age, sex and diagnosis, or other important characteristics. This database, therefore, was too limited in scope for the Data Development Subcommittee to derive meaningful statements concerning disparities in minority health or to suggest appropriate national policies for their alleviation. A relationship between health and socioeconomic status, widely reported in the literature, suggests that those of lower income and educational levels generally have higher mortality and morbidity rates. Nevertheless, studies reported in the research literature do not contain national level information on this relationship for identified diseases or mortality with minority status taken into account. Consequently, the 20 Survey of Income and Education (SIE) was explored as a possible data source to examine the health status of minorities and socioeconomic status. The SIE, a nationwide household survey conducted in 1976 by the Census Bureau, asked respondents if they had any disabling health problems, chosen from a list including such conditions as heart disease, arthritis, blindness, deafness or hearing impairment, speech problems, or nervous disorders. Income and education of the respondent were obtained. Although this is the largest national health survey of socioeconomic data the disease specific measures in the SIE were not asked in the context of a general health survey. The findings from the_ SIE thus did not indicate the cause-effect relationship of socioeconomic status and health. However, the SIE confirmed a general relationship between poor health and lower socioeconomic status for minority groups. A paper by Haan and Kaplan, commissioned by the Task Force also examines the overall relationship of socioeconomic status to health for minorities. The morbidity database of the Health Care Financing Administration (HCFA) was examined by the Data Subcommittee. These data, however, are limited to information on medicare and medicaid populations and do not represent health needs of an entire population. The limited data reported by HCFA (Table 11) suggest a higher percent of hypertension-related End Stage Renal Disease (ESRD) among Blacks compared with Whites. These ESRD data were reviewed by the Subcommittee on Cardiovascular and Cerebrovascular Diseases and discussed m the report of that subcommittee. The Data Development Subcommittee also examined the DHHS Health Data Inventory to determine the extent to which departmental databases include indicators of the minority status of the population covered by them. Departmental data were found to be primarily administrative and inadequate to compare health status of minority and ^^ity Americans. For example, the Centers for Disease Control (CDC) and Health Care Financing Administration (HCFA) have large data collection activities for administration, but their information is reported in aggregate without race/ethnic categories. Data on Native Americans are provided by the Indian Health Service (IHS) and are collected only for those populations served by the IHS. Consequently, other departmental databases were not included in the Task Force analysis. 21 N5 N3 Hgure 1 Death Rates by Age at Death: White Males and White Females (Average Annual Rates for All Causes of Death, United States, 1979-1981) Rates per KK).(KX) Population 2(),(XX) 15,000 10,000 5,000 4,000 3,000 2,000 1,000 Under 1 / White Females 5-9 15-19 25-29 35-39 45-49 55-59 Age at Death in Years 65-69 75-79 85 + Source: Duke University, Analysis Commissioned by the Task Force on Black and Minority Health, 1984-1985. Figure 2 Death Rates for Black Males Relative to White Males (Average Annual Rates for All Causes of Death, United States, 1979-1981) Rates 0* c cd cd VC c cd o per 100,000 Population 1,600 1,200 - 800 400 White Rate* -400 h 800 — 1,200 — 1,600 ; \\____________________________________________^ i i i i i i i i i i i i i i i Under 1 5-9 15-19 25-29 35-39 Age at Death in Years 45-49 55-59 65-69 *See Fig 1 for White Death Rates Source: Duke University, Analysis Commissioned by the Task Force on Black and Minority Health, 1984-1985. Figure 4 Death Rates for Native American Males Relative to White Males (Average Annual Rates for All Causes of Death, United States, 1979-1981) N3 Rates CD cd - Or- ll ooo<=>«<=>« «a--Or~-00»O00CM-Ma in •«- i— o -O -a -a -o in CM in r-^ ro oo ro ro CM in CM CM ro in ro r-~ r-~ in IO o- Gr-in ro ao -a CM ■»• CO I— r- -o r- -^ o ro *»- o- o- ^- OD M) in o ~o in —• CM CM »»- CM «*■ LU <=> in oo o- in »- ro 1— oo 1— O- in *»- 1— «*- r-- 3C h— =C OO CO «*• or- in o in -a CM r-- ro Q» -a ro ■o oo -o in CO * -o -o -o in in in ^~ «*- ro ro CM CM —' 3,. LU r~- r-~ ro in -a Or- oo ro ar- ■O -a O in ro o i~~ r— cn je -O CM -o CO o- ^~ CM Or- in cs r~ -o r-» o- Or- CM r-» ro CO r-» -a ^~ to ro CM CM CM Q_ =3 cn cn moinoinoinom o-«NNMM»*inin —.in—i i i i i i i i i i i ■ i i oinomomomoin o — in — — N(MMM**inm -a mi i— i— i i i i o in o o oo m ao mi) in i-. o ao in oo deaths due to malignant neoplasms would have been greater by 25 percent in 1970 and 34 percent in 1980. Lung cancer deaths would have been increased by more than 75 percent in both years. Under 1970 rates, Spanish surname males would have had over 25 percent more deaths from major circulatory diseases. By 1980, a fair amount of convergence of death rates from these causes is noted. Due to the faster decline in death rates among non-Spanish surname males, by 1980, additional deaths from major circulatory diseases would have been only about 16 percent of total expected deaths due to these causes. Deaths from major circulatory diseases accounted for about two-thirds of all deficit deaths in 1970 but only one-third in 1980. At the same time, deficit deaths from all cancers rose from 27 percent of the total to 46 percent. The result of these largely offsetting patterns in causes of death is shown in Table 3 as the "net difference" between excess and deficit deaths. Altogether, the net effect of substituting non-Spanish surname death rates is very small because of the nearly equal magnitude of excess and deficit deaths: Spanish surname deatns would have been reduced only 2.5 percent in 1970 and 4 percent in 1980. In those causes in which there is excess or deficit mortality the effects are very clear—i.e., either a cause contributes to a net excess or a net deficit. Only in the case of cerebrovascular diseases do age specific death rates almost exactly offset one another. Effects of Mortality Differences on Life Expectancy. - We have reviewed mortality levels in the Spanish surname and other white male population of Texas and their implications for excess mortality in the former. The results are straightforward, ano in some cases tne excess mortality that might be avoided (or the "deficit" mortality that might be added) is quite large. Numbers of lives that might be "saved" if death rates were equalized have dramatic appeal. But what significance do they have for length of life? Life expectancy provides a practical means for illustrating differences in mortality of populations. For most ages (and over all ages when considering sane causes of death) death rates are so small that, in comparing populations, large percentage differences may be statistically significant, but practically and substantially trivial in terms of the actual effect of those differences on average length of life. In this section we show how differences in age and cause specific death rates contribute to differences in life expectancy. As an overview, life expectancies at various ages appear in Table 4. In 1970, the life expectancy of Anglo males at birth exceeded that of Spanish surname males by .4 (68.1 years as compared to 67.7 years). By 1980, E(0) had increased 2.5 years for both groups of males (to 70.6 and 70.2 years, respectively) so that the differential remained constant at .4. In general, the computations displayed in Table 4 show that, in both 1970 and 1980, Spanish surname males have somewhat lower life expectancy than Anglos at younger ages, while the reverse is true at older ages. In their analysis of 1970 data for California, Schoen and Nelson (1981) report a similar crossover to more favorable Spanish surname life expectancy by age 40. The crossover occurs even earlier—in fact, by age 25—in Texas. The maximun life expectancy advantage (nearly one year of life) of Spanish surname males is reached in middle age, while that of other white males (only .4 years) occurs at age zero. The timing of death by cause must account for these patterns. Although useful for general descriptive purposes, the comparisons in Table 4 are far too broad to permit an adequate assessment of the differentials that separate the two ethnic populations of interest. Consequently, as noted in the earlier section on methods, we prepared life tables for a large number of specific causes of death as a basis for computing cause-adjusted estimates of life expectancy. The adjusted life expectancies 54 are those for Spanish surname males that would result if their age-specific death rates for a specific cause of death were to become the same as the rates for that cause for their Anglo counterparts. Results of the application of this procedure for selected ages appear in Appendix B for the years 1970 and 1980. The essential findings are shown in Figures 1 and 2, which illustrate the difference values in graphic form for selected major causes of death. Note tnat since the Spanish surname life expectancy estimates are taKen as the subtrahend, a negative value in Figures 1 and 2 indicates that Spanish surname males have an actual mortality advantage, and a positive value denotes a current Spanish surname disadvantage. In both 1970 (Figure 1) and 1980 (Figure 2), we see the crossover in life expectancy based on all causes of death occurring by age 25 (after which the all-causes slope falls below zero on the Y-axis). we also see illustrated in botn figures the fact that Spanish surname males are considerably better off than their Anglo counterparts with respect to the major chronic and degenerative diseases. In 1970, Spanish surname males would nave lost about one year of life if their death rates from major circulatory diseases had been the same as tne Anglo rates (as indicated by values near -1.0) up until age 55, followed by a moderate upward inflection of the curve at older ages. This advantage was diminished (to around -.4) and the curve was nearly flat for death from circulatory diseases in 1980. In regard to cancer also, tne Spanish surname group has more favorable death rates, and their relative advantage was greater in 1980 than in 1970. The bulK of tnis advantage is due to lower death rates from lung cancer at botn points in time. The disparity begins to diminish after about age 45, and life expectancy differences related to cancer become extremely small at the oldest ages. In sharp contrast are the patterns of death due to external causes. Spanish surname males nave much higher deatn rates than other white males from both motor vehicle accidents and homicide, beginning at birth and lasting into the young adult years. After age 30, however, a downward trend in the differences in life expectancy is clearly evident, indicating a convergence with Anglo rates at older ages. In 1970, both homicide and motor vehicle accidents contributed very heavily to Spanisn surname mortality, but by 1980, homicide was tne major factor depressing life expectancy relative to other white males. Figures 3 and 4 show 1970 and 1980 intraethnic changes in life expectancy adjusted in a manner analogous to the adjustment of interethnic differences as found in Figures 1 and 2. (Appendix C presents these results in tabular form.) The intraethnic adjusted differences may be interpreted as measuring the effects on life expectancy of age-specific mortality gains or losses in 1980 as compared to 1970. Positive differences indicate the result of lower death rates leading to greater life expectancy and negative differences denote higher mortality rates in 1980 than in 1970. The 1970-1980 differences for Spanish surname males appear as Figure 5, and those for Anglo males are found in Figure 4« Among Spanish surname males, great improvement is apparent with respect to major circulatory diseases (Figure 3). For all ages up to 75, the 1980 death rates, had they occurred in 1970, would have produced a gain in life expectancy of well over one year. By contrast, there was no gain in life expectancy from reduction in the age-specific death rates from external causes among Spanish surname males between 1970 and 1980. Improvements with regard to deatns from motor vehicle accidents and other external causes were competely offset by rising death rates from homicide. 55 Figure 1—Differences between observed life expectancy by age for Spanish surname white males & tnose wnicn would nave occurred if such males had experienced the age-specific death rates of other white males, for selected causes of death—Texas: 19 70 * .13— 0 I F F E R E N C b. I K E E X P E C T A N C Y 1.0- B^^-H kv 0.5-f^—' A, 3 \ -0.5-3%-^—#—#—«*-..#.AH>_^—*—*" T3- *t] ^ -1.5- 11)1111 ITT]-ITrT|Tt TTTTTTT|TmjTTTT^TTrrp-rTrpil T| II Up 1 122334465667788 0S05O505O50S0S0505 LEGENDi CAUSE AGE +-+-*• FILL CAUSES *-■*--* LUNG CANCER e-o-B EXTERNAL v-v-y HOMICIDE #•-*--# ALL CANCERS *-*-* CIRCULATORY -*—»■-+ MOTOR VEHICLE 56 D I F F E R E N C E 1 N L I F £ E X P E C T A N C Y Figure 2—Differences between life expectancy by age for Spanish surname white males & those which would have occurred if such males had experienced the age-specific death rates of other white males, for selected causes of death—Texas: 1980 l.SH 1.0-f N 0.5- ■-O.S-f -a—a—SL-a—-a--^ y \ * i -i -i i.o-i i :?*-*-"«►—•—*---«—•—«--.♦- j 1111 t ft > i-p-n TjTTtf r tt rriTrrrrTTT-r niTiiiitiMiiiini tti 0 5 1 1 2 0 S G 334 455667788 050505050505 AGE LEGEND: CAUSE +-+-+ ALL CAUSES A--A--A- LUNG CANCER e-i^-B EXTERNAL v-^-v H0M0CIDE #--*--# ALL CANCERS *-*-* CIRCULATORY -«-+-»■ MOTOR VEHICLE 57 Figure 3—Differences between observed life expectancy by age for Spanish surname white males in 1970, and those which would have occurred if such males had experiences the age-specific death rates for 1980, for selected causes of death—Texas 3.0- D I F F E R E N C E I N L I F E E X P E C T A N C Y 2.5 2.0 "*—■* 1.0^ 0.5-3 -i H mt -[**—+--+---1---K.^ 0 • Q-d*Kc-£— A--A-~-£-=-*--«~W^fr—^-rVF^t— --*-^^:2:i4U--^a^-. -0.5- B-B ^W—V--¥--V-' ^*" 1.0- -1.5 ■Ip-rrTp-rTTp-TT-rpiT-f-j ti 11 pftrpfrtpn t|t t ttpt-tt pTTt | i ni| in i|itii]Ht i | n I i|ini |' 1 122334455667788 05050 5 050505050505 AGE LEGEND: CAUSE «►-+-+ ALL CAUSES a~a~a LUNG CANCER G-th-3 EXTERNAL v-v-v H0M0CIDE #--*-•* ALL CANCERS *-*-* CIRCULATORY h MOTOR VEHICLE 58 Figure 4—Differences between observed life expectancy by age for non- Spanish surname white males in 1970, and those which would have occurred if such males had experienced the age-specific death rates for 1980, for selected causes of death—Texas 3.0-1 2.5- D I F F E R E N C E I N L I F E E X P E 2.Q-jie--*---*--*-,"~* — X-—H—* 1.5 1.0-3 a.sA 0.0-P* jJto-- ■#—■■ jfr~~T% -- • rfe:— g—-B—-B—-q^. •0.5-3 1.0- -1.5- lp-rrrp-rrrpft i put] 11* tp 11 tprrTp-fTfj- TtiTtrrrt ] I I I Tpt ft-p-ft-fj 1 1 12 2334455667788 0 5050505050505 AGE LEGEND: CAUSE +-+-+ flLL CAUSES A-A-A LUNG CANCER e-CHB EXTERNAL Y--V-Y H0M0CIDE #■-#-# ALL CANCERS *-*-* CIRCULATORY h MOTOR VEHICLE 59 Anglo males in 1970 (Figure 4) would have had approximately two years auueu to their life expectancy at all ages up to late middle age, if the 1980 death rates for circulatory diseases had been in effect. Thus, tne substantial gains experienced by Spanish surname males due to lower mortality fran this cause in 195O were surpassed by the even more remarkable improvement among Anglos. Non-Spanish surname white males suffered some increase in death from lung cancer, and because of this, while Spanish surname males had some minor improvement in mortality due to all malignant neoplasms, Anglos became worse off in this regard, beyond this, there were few changes of note in the Anglo adjusted differences. CONCLUSIONS Spanish surname males in Texas have been shown to have only slightly higher mortality and slightly lower life expectancy at birth than other white males in both 1970 and 19bU. Moreover, a crossover in age-specific death rates is observed so that Spanish surname males achieve higher life expectancy than Anglos in early adulthood and thereafter. These results are consonant with previous wor* on Mexican American male mortality which was reviewed earlier. The most salient conclusion from a policy standpoint is that lower Spanish surname life expectancy at birth is due almost entirely to excess mortality from a limited number of causes of death, sane of which may be controlled or prevented. Excess mortality among Spanish surname males is greatest from external causes of death, accounting for 57 percent of the excess in 1980, up fran 48 percent in 1970. In both years, this excess cost Mexican American males on the average one year of life expectancy at birth. As of 1980, eliminating excess deaths due to motor vehicle accidents would add about .2 year of life expectancy at birth for Spanish surname males—about half the overall difference fran other white males. Given the convergence of age specific death rates from this cause from 1970 to 1980, this seems reasonable. Unfortunately, by 1980, most of the disadvantage from external causes was due to hanicide, a cause of death that may be unresponsive to planned intervention. Excess mortality resulting from other external causes of death, including motor vehicle accidents, declined by over half during the 1970s, but homicide increased by two thirds. Spanish surname males are already considerably better off than other white males with respect to diseases that are common at older ages such as cancer and heart diseases. They have a disadvantage in excess mortality from diabetes mellitus. Better management of diabetes would result in a reduction of mortality directly attributed to that disease and also of mortality due to the vascular illnesses with which diabetes is often associated. This should further increase the advantage in mortality that Spanish surname males have in middle age and beyond. The fact that Spanish surname males are disadvantaged mainly fran deaths due to conditions which may be successfully managed (as diabetes) or to events wnich may be prevented (as motor vehicle accidents) or avoided (as nomicides) suggests their life expectancy at birth might be fairly easily improved. On the other hand, it is certainly possible that if intervention programs succeeded in lowering the Spanish surname death rates from external causes, thereby allowing a greater number and proportion of the population to survive to older ages, deaths from cancer, circulatory diseases or other chronic and degenerative conditions would increase, since these tend to be 60 diseases of old age. Finally, it is also true that it is much more difficult to add significantly to the overall life expectancy of populations whose expectation of life at birth is already great (Arriaga, 1983). REFERENCES Arriaga, Eduardo E. 1984. Measuring and explaining the change in life expectancies. Demography 21(June):83-96. Bradshaw, Benjamin S., and Edwin Fonner, Jr. 1978. The mortality of Spanish-surnamed persons in Texas: 1969-1971. Pp. 261-282 in Frank D. Bean and W. Parker Frisbie (eds.) The Demography of Racial and Ethnic Groups. (New York: Academic Press). Bradshaw, Benjamin S., and Edwin Fonner, Jr. 1980. Survivorship and longevity of Spanish-surnamed and other white persons: Texas, border and nonborder regions, Bexar County, 1969-71, and San Antonio, 1950, 1960. Unpublished report prepared for the National Institute on Aging. Bradshaw, Benjamin S., and W. Parker Frisbie. 1983. The usefulness of census Spanish surname and Spanish origin data with vital statistics data. Paper presented at the annual meeting of the Southern Regional Demographic Group. Chattanooga. Buechley, Robert W. 1976. Generally Useful Ethnic Search System: GUESS. Cancer Research and Treatment Center. (Albuquerque, N.M.: University of New Mexico). Davis, Cary, Carl Haub and JoAnne Willette. 1983. U.S. Hispanics: Changing the face of America. Population Bulletin 38(June):1-43- Ellis John M. 1959. Mortality differences for a Spanish-surname population group. Southwestern Social Science Quarterly 39(March):314-321. --------------. 1962. Spanish-surname mortality differences in San Antonio, Texas. Journal of Health and Human Behavior 3(Summer):125-127. Gillespie, Francis P., Andrew M. Greeley, Michael Hout, and Teresa A. Sullivan. 1983. Public policy, ethnic codes and Hispanic vital statistics. La Red/The Net 70(July):9-13- Hernandez, Jose, Leo Estrada, and David Alvirez. 1973. Census data and the problem of conceptually defining the Mexican American population. Social Science Quarterly 54(March):671-87. Kautz, Judith A., Benjamin S. Bradshaw, and Edwin Fonner, Jr. 1981. Trends in cardiovascular mortality in Spanish-surnamed, other white, and black persons in Texas, 1970-1975. Circulation 64(October):730-735. Keyfitz, Nathan. 1977. What difference would it make if cancer were eradicated? An examination of the Taeuber paradox. Demography 14(November):411-18. 61 ----------------# 1982. Population Change and Social Policy. (Cambridge, MA: Abt Books). tutagawd, Evelyn and Philip M. Hauser. 1973. Differential Mortality in the United States. (Cambridge, MA: Harvard University Press). Lee, Eun Sul, Robert E. Roberts, and Darwin R. Labarthe. 1976. Excess and deficit lung cancer mortality in three ethnic groups in Texas. Cancer 38(December):2551-56. Nelson, Verne and Marion Collins. 1976. Computer replication of census Spanish surnane coding. Mirneo. Berkeley, CA: California department of Health. Roberts, Robert E. and Cornelius Askew, Jr. 1972. A consideration of mortality in three subcultures. Health Services Reports 87(March):262-70. Rosenwaike, Ira. 1983. Mortality among the Puerto Rican born in New York City. Social Science Quarterly 64(June):375-85. Schoen, Robert and Verne E. Nelson. 1981. Mortality by cause among Spanish surnamed Californians, 1969-71. Social Science Quarterly 62(June):259-74. ShryocK, Henry S., Jacob Siegel, and Associates. 1973- The Methods and Materials of Demography. Revised edition. (Washington, DC: U.S. Government Printing Office). Stern, Michael P. and Sharon P. Gaskill. 1978. Secular trends in ischemic heart disease and stroke mortality from 1970 to 1976 in Spanish-surnamed and other wnite individuals in Bexar County, Texas. Circulation 58(September):537-43- Sullivan, Teresa A., Francis P. Gillespie, Michael Hout, and Andrew M. Greeley. 1983. Surnane versus self-identification in the analysis of Hispanic data. American Statistical Association, Proceedings of the Social Statistics Section, pp. 117-22. Sullivan, Teresa A., Francis P. Gillespie, Michael Hout, and Richard G. Rogers. 1984a. Alternative estimates of Mexican-American mortality in Texas, 1980. Social Science Quarterly 65(June):609-17. Sullivan, Teresa A., Francis P. Gillespie, and Richard Rogers. 1984b. Effects of ethnic classification on apparent life expectancy: The case of Texas in 1980. Paper presented at the annual meetings of the Social Statistics Section, Joint Statistical Meetings, Toronto. Tsai, Shan Pou, Eun Sul Lee, and Robert J. Hardy. 1978. The effect of a reduction in leading causes of death: potential gains in life expectancy. American Journal of Public Health 68, No. 10(October):966-971. Tsai, Shan Pou, Eun Sul Lee, and Judith A. Kautz. 1982. Changes in life expectancy in the United States due to declines in mortality, 1968-1975. American Journal of Epidemiology 116, No. 2:376-384. 62 U.S. Bureau of the Census. 19/2. General social and economic characteristics. 1970 Census of Population. PC(1)-C45. Washington, D.C.: U.S. Government Printing Office. U.S. National Center for Health Statistics. 1983. Births of Hispanic parentage, 1980, by Sephanie J. Ventura. Monthly Vital Statistics Report 32, No. 6 Supplement (September). 63 APPtiMblX A Cause of Deatn Categories and Corresponding ICD Codes Primary cause of deatn 8th revision 9th revision Malignant neoplasms, all sites Trachea, bronchus, and lung All other malignant neoplasms Diabetes meilitus Major circulatory diseases Major ischemic heart diseases Acute myocardial iniarction Chronic ischemic heart disease Cerebrovascular disease Pneumonia ana influenza External causes oi deatn Motor venicle accidents homicide and legal intervention All other external causes All other causes of deatn 140-209 162 140-161, 1oj)-209 230 410-413, 4^0-438 410-41:? 410 411-413 430-438 470-48b E800-E999 E810-E823 E980-E978 E800-E809, E824-E959, E979-E999 Residual 140-209 152 140-161, 163-209 250 410-414, 430-4.58 410-414 410 411-414 430-438 480-487 E800-E9y9 E810-E825 E950-E978 E800-E809, E82o-E939, E979-E999 Residual 64 APPENDIX B DIFFERENCES BETWEEN OBSERVED LIFE EXPECTANCV AT SPECIFIED AGES FOR SPANISH SURNAHE WHITE HALES AND THOSE WHICH MOULD HAVE OCCURRED IF SUCH HALES HAD EXPERIENCED THE AGE-SPECIFIC DEATH RATES OF OTHER WHITE HALES, FOR SELECTED CAUSES OF DEATH—TEXAS: 19?0 ft 1980 1969-P1 19?9-81 CAUSE OF DEATH AGECX): 0 30 15 60 ?5 0 30 15 60 ?5 ALL CAUSES 0.11 -0.65 -0.95 -0.61 -0.15 0.13 -0.32 -0.?5 -0.81 -0.68 ALL HALIGNANT NEOPLASHS -0.19 -0.52 -0.19 -0.28 0.02 -0.?1 -0.?5 -0.?3 -0.51 -0.1? LUNG -0.36 -0.39 -0.38 -0.21 0.03 -0.1? -0.50 -0.19 -0.31 -0.01 OTHER -0.13 -0.13 -0.11 -0.0? -0.01 -0.25 -0.26 -0.21 -0.18 -0.12 DIABETES HELLITUS 0.23 0.21 0.23 0.1? 0.05 0.22 0.23 0.21 0.22 0.10 HAJOR CIRCULATORS DISEASES -0.9P -1.06 -1.01 -0.?? -0.69 -0.11 -0.15 -0.11 -0.12 -0.53 HAJOR ISCHEHIC HEART DIS. -1.03 -1.11 -1.05 -0.?1 -0.18 -0.11 -0.1? -0.15 -0.35 -0.31 ACUTE HrOCARDIAL INFARC. -0.?0 -0.?6 -0.P2 -0.51 -0.29 -0.29 -0.31 -0.30 -0.21 -0.21 CHRONIC I.H.D. -0.31 -0.3P -0.35 -0.21 -0.21 -0.15 -0.16 -0.15 -0.11 -0.11 CEREBROVASCULAR DISEASE 0.03 0.03 0.02 -0.05 -0.25 0.01 0.01 0.00 -0.09 -0.21 PNEUHONIA ft INFLUENZA 0.11 0.01 0.01 0.01 0.02 0.05 0.01 0.01 0.03 0.02 ALL EXTERNAL CAUSES 0.91 0.2? 0.02 0.02 -0.01 1.01 0.19 0.13 0.00 -0.03 HOTOR VEHICLE ACCIDENTS 0.36 0.18 0.11 0.01 0.01 0.18 0.13 0.08 0.01 0.01 HOHICIDE 0.50 0.21 0.03 0.02 0.00 0.90 0.13 0.12 0.02 0.00 ALL OTHER EXTERNAL CAUSES 0.05 -0.12 -0.11 -0.01 -0.02 -0.05 -0.0? -0.0? -0.06 -0.05 ALL OTHER CAUSES 0.66 0.13 0.29 0.21 0.19 0.39 0.12 0.00 -0.13 -0.09 APPENOIX C DIFFERENCES BETWEEN OBSERVED LIFE EXPECTANCV AT SPECIFIED AGES FOR SPANISH SURNAHE AND OTHER WHITE HALES IN 19?0, AND THOSE WHICH MOULD HAVE OCCURRED IF SUCH HALES HAD EXPERIENCED THE AGE-SPECIFIC DEATH RATES FOR 1980, FOR SELECTED CAUSES OF DEATH—TEXAS CAUSE OF DEATH AGECX): SPANISH SURNAHE 30 15 60 ?5 30 OTHER WHITE 15 60 ?5 ALL CAUSES ALL HALIGNANT NEOPLASHS LUNG OTHER DIABETES HELLITUS HAJOR CIRCULATORY DISEASES HAJOR ISCHEHIC HEART DIS. ACUTE HVOCARDIAL INFARC. CHRONIC I.H.D. CEREBROVASCULAR DISEASE PNEUHONIA ft INFLUEN2H ALL EXTERNAL CAUSES HOTOR VEHICLE ACCIDENTS HOHICIDE ALL OTHER EXTERNAL CAUSES ALL OTHER CAUSES 2.53 1.60 1.56 1.3? 0.?5 2.55 1.93 l.?5 1.1? 0.52 0.13 0.12 0.08 0.05 0.01 0.01 -0.03 -0.0? -0.11 -0.10 -0.03 -0.01 -0.05 -0.01 -0.02 -0.08 -0.08 -0.10 -0.12 -0.0? 0.16 0.16 0.13 0.09 0.06 0.08 0.06 0.03 0.01 -0.03 0.08 0.09 0.06 0.02 -0.01 0.03 0.03 0.03 0.03 0.02 1.32 1.11 1.13 1.29 0.9? 1.9? 2.0? 2.02 1.6? 1.11 0.88 0.91 0.95 0.82 0.61 1.51 1.62 1.58 1.22 0.?3 0.53 0.5? 0.56 0.12 0.28 0.9? 1.03 0.98 0.69 0.32 0.33 0.36 0.3? 0.38 0.31 0.51 0.51 0.51 0.19 0.38 0.10 0.12 0.12 0.12 0.32 0.36 0.3? 0.3? 0.3? 0.35 0.30 0.11 0.11 0.11 0.08 0.20 0.10 0.10 0.08 0.06 -0.01 -0.03 0.0? 0.11 0.06 0.06 0.15 0.16 0.08 0.03 0.21 0.11 0.11 0.05 0.02 0.01 0.05 0.0? 0.01 0.01 -0.12 -0.21 -0.09 0.00 0.00 -0.0? -0.02 -0.01 0.00 0.00 0.20 0.0? 0.05 0.05 0.05 0.12 0.12 0.10 0.01 0.02 0.53 -0.0? -0.1? -0.1? -0.31 0.3? -0.28 -0.36 -0.12 -0.16 The Contribution of Socioeconomic Position to Minority Health Mary N. Haan George A. Kaplan Human Population Laboratory California Department of Health Services Berkeley, California The Contribution of Socioeconomic Position to Minority Health Introduction Investigation of the differential health experience of minorities and whites cannot help but raise important questions concerning the reasons for these differences. In Part 1 of this report, we will argue that socio- economic position (SEP) represents an important and plausible area of investigation in the search for reasons. It is important because SEP and minority status are clearly intertwined, and examination of both will potentially clarify our understanding of minority health. We say plausible because limitations in the available data advise caution in interpretation and application. However, there is considerable information which points to the critical role of SEP. We will examine this role in several stages. First we will review the strength and consistency of the association between SEP and a variety of disease outcomes. Rather than focusing on specific organ systems, we will use the epidemiologic triad of person, place, and time to organize the massive amount of evidence on SEP and health. Then we will consider the association between SEP and membership in minority groups. Our next step will be to consider available evidence concerning the consistency of the association between SEP and health, between and within specific minority groups. We will then move to evidence which indicates how much of the differences in health between minorities and whites can be attributed to SEP. In Part 2, we will examine evidence which suggests some of the ways in which low SEP may be associated with poorer health. As a first step, we need to consider for a moment what is meant when we refer to SEP. This topic has been addressed recently by Morgenstern (66). Most investigators have viewed SEP as an amalgam of income, educa- tion, and occupation. Various indices have been constructed in an attempt to combine, on empirical or theoretical grounds, information from these three domains. Indices of social status have also been constructed, again on empirical or theoretical grounds, which rank people according to "prestige." Lastly, the construct "social class" has been used to order groups in a number of ways, ranging from broad occupational groupings to orderings based on influence, authority, and power in the economic structure. It is clear from this brief discussion that we can mean many things by SEP. Measures which combine different domains of socioeconomic information can hinder our understanding of the ways in which SEP is associated with health. Although different socioeconomic measures may be related, they have differential utility depending upon the question being asked. For example, in situations where illness is likely to effect occupation and income, education may be the preferred socioeconomic measure. Income level, however, may be more important in obtaining services or meeting needs than education would be. On the other hand, because of secular and group-specific trends in educational attainment, level of education may behave differently as a risk factor for different cohorts. Measures of occupation which group individuals in broad classes such as professional- technical, managers-administrators-proprietors, or semiskilled operatives may obscure large income or educational differences within these classes. On the other hand, occupation may summarize the cumulative effects of education and income or measure other aspects of SEP not tapped by education and income. 69 In what follows, we will report on socioeconomic measures which are derived from income, education, or occupation. Where possible, we will utilize more than one measure in our discussion. In addition, in some cases, we will utilize measures which reflect characteristics of the areas in which individuals live. Census tract characteristics such as median family income, median years of education, or per cent in a particular occupational category are sometimes used as proxy measures of individual levels. Recognizing the potential "ecological fallacy" (65) involved in the use of such measures, we will use them with great caution and only where the pattern of findings is consistent with results using other measures. In Part 2 of this report, we will indicate, however, how such area characteristics may, in themselves, be important to our understanding of the health differentials between minorities and whites. Part 1 All-Cause Mortality: The General Picture Ever since the 12th century, when data were first recorded on this topic, those at the lowest socioeconomic levels in the community have been found to have higher death rates (1, 90). This pattern is reflected in a large number of reports which have examined the association between socio- economic factors and all-cause mortality. An illustrative example comes from Kitagawa and Hauser's study of adult mortality in the United States in 1960 (51). They found a consistent inverse gradient of mortality rates associated with socioeconomic position. Those who had higher SEP had lower mortality rates. This was true whether the measure of socioeconomic position was based on family income, median income of census tract of residence, education, or occupation. In the many studies of which we are aware, this pattern of increased all-cause mortality associated with lower socioeconomic position is found in well over 80 per cent. Furthermore, in many cases, there is an orderly gradient of rates associated with increases or decreases in SEP. In what follows, we will briefly examine the consis- tency of this finding for different age groups, diseases, geographical locations, and time periods. Consistency by Age All-Cause Mortality. Socioeconomic gradients of all-cause mortality are found in most age ranges. There is some evidence that the association between socioeconomic factors and health is somewhat weaker at the older ages. Kitagawa and Hauser (51) found that the gradients associated with income and education were larger for persons 25-64 years of age than they were for those 65 years or older. In analyses (49) of the 18-year mor- tality experience of a large (n=6,928) cohort of individuals representative of Alameda County, California, in 1965 (6), we found that the increased risk associated with low compared to high family income decreases with age, becoming non-significant between 60 and 70 years. Others have reported similar findings (51). However, in interpreting the significance of this apparent dilution of effect, we must take into account the fact that income generally declines with retirement, resulting in a disproportionate lower- ing of the income of those who were not previously in the lower income categories. Thus some portion of those in the lower SEP groups have only recently entered these groups. The absence of lifelong measures of SEP may 70 result in dilution of the association between SEP and mortality in the later years. Similarly, average levels of education have increased in successive birth cohorts, and the educational requirements for most occupa- tions have increased. It is, therefore, reasonable to believe that the health consequences of a low education may have similarly increased. As always, the interpretation of "age" effects are complicated by period and cohort effects. Diseases of the Young. The overall consistency of the association between socioeconomic position and health status can be further seen by examining outcomes which are age-related. A substantial body of evidence exists linking higher rates of infant mortality to socioeconomic position (2, 7, 12, 16, 35, 36, 60, 73, 79, 84, 102). Many studies have shown that perinatal and infant mortality rates are elevated for those with lower income, lower educational attainment, poorer occupational status, or other types of social disadvantage (9, 10, 51, 105). There is also evidence which suggests that higher rates of birth defects are found in the poor (14, 28). The major sources of mortality from unintentional injury in children (housefires, drowning, and suffocation) also show a strong asso- ciation with SEP (4). A similar pattern is found for a wide variety of health outcomes in the young (9, 10, 27, 57, 63, 83, 85, 89, 92, 104). Diseases of the Middle Years. When we turn to diseases of middle-age, we see similar patterns. Vital statistics data confirm the inverse asso- ciation between SEP and the various manifestations of atherosclerotic disease (29, 37, 46, 69, 76, 80). In the United States, both prevalence and incidence of cardiovascular disease are inversely related to SEP (15, 48, 50, 51, 53, 78, 95, 112), although for some groups, there have been changes in the direction of the association between SEP and cardiovascular disease (64). In the 1972 Health Interview Survey (101), those who had family incomes under $5,000 had 33 per cent higher prevalence of heart conditions than those with family incomes of $15,000 or more. The rates of hypertension without heart involvement were over 60 per cent higher in the poorer group. Similar findings have been reported in a number of studies. Findings from the Hypertension Detection and Follow-Up Program show a strong inverse gradient of prevalence of hypertension associated with years of education (41). Mortality from coronary heart disease shows a consistent SEP gradient when SEP is measured by occupational groupings as well (51). In addition, survival from coronary heart disease appears to be inversely associated with SEP (82). An SEP gradient is also found for unintentional injuries in this age group. For example, residents of low income compared to high income counties have about three times the mortality rate for motor vehicle occupants even through they are likely to be driving less (4). Similar patterns are found for other diseases of this age group (11, 31, 40, 91, 112). Diseases of the Later Years. Many cancers which reach their peak prevalence in the later years show inverse gradients with SEP. Among these are cancers of the lung and pleura (56, 81, 109), oral cavity and pharynx (42, 111), esophagus (59, 110), and stomach (43, 93). There are, of course, sites which evidence the opposite gradient such as breast (3, 26) and testicular (68, 77). However, it is notable that the poorer survival associated with lower SEP is found both for sites where there is an inverse association with SEP, for example, prostate (21), and for sites where there is a direct association with SEP, for example, breast (22). Gradients of disease related to SEP are also found for stroke, osteoarthritis, and other diseases and for various measures of disability and impairment (19, 51, 99, 71 100). Mortality rates for those in this group from pedestrian injuries, falls, fires and burns, and exposure to cold also show a strong association with SEP (4). Although not an exhaustive listing, the evidence presented above is quite compelling regarding the consistency of the association between socioeconomic position and health at different age groups. In general, those at lower levels of SEP have higher rates of most diseases, covering a wide range of ages and organ systems. Consistency by Place The association between socioeconomic position and health is consis- tently found throughout the world. Mortality differentials associated with socioeconomic position are found in countries as diverse as the the United States and India. These differentials are found in England and Wales (67, 76), Sweden (30), Finland (69, 80), France (24), Norway (45), Australia (29), New Zealand (74), Latin America (5), Ghana (13), Canada (62), and many others. Within the United States, associations between socioeconomic position and health outcomes have been found in such diverse places as Evans County, Georgia (95), and Alameda County, California (38); Iowa (31) and Hawaii (83); and Chicago (51) and Charleston (50). Consistency over Time Despite the large improvements in health seen during the last 60 years, the gradient of health associated with SEP has changed very little. Hollingsworth (45) has done the most extensive review of changes in the SEP health gradient over time. He examined changes in all cause mortality by occupational class in England and Wales for the period 1891-1971. The standardized mortality ratio for the lowest social class (V) compared to the highest social class (I) in the period 1890-1902 was 1.50. In the period 1970-1972, the same ratio was 1.58. Although there clearly are problems in the comparability of data sources and definition of social class over this 80-year period, the similarity between the two figures is striking. It is especially striking when we consider that these data cover a period in which there were major changes in the leading causes of death. Similarly, Kitagawa and Hauser (51) found very little convergence of the socioeconomic differentials in mortality for all causes, excluding infant mortality, in Chicago during the period 1930-1960. Others have reported similar findings (54, 88). Analyses of mortality between 1960-1970 in Birmingham, Buffalo, and Indianapolis suggest that there was a slight increase in the gradient of mortality associated with socioeconomic posi- tion during that period (112). Some reports have suggested that the association between socioeconomic position and specific diseases has changed over time (58, 64). For example, mortality from cancer of various sites has changed over time (56). Blaxter (8) summarized these changes in England and Wales between 1930-1963 by noting that for sites which have been more common in those of lower socioeconomic position, the gradient associated with SEP has increased, whereas the gradient associated with SEP had decreased for sites more common in those of higher SEP. For sites which are decreasing in mortality, SEP gradients are increasing, and for sites which are increasing, SEP gradients are reversing. Changes in the SEP gradient for coronary heart disease have also been noted. For example, in analyses in Evans County, Georgia (64), and England 72 and Wales (58), there appears to have been a reversal in the SEP gradient for mortality from coronary heart disease (CHD). That is, CHD among low SEP men has increased, and CHD among high SEP men has decreased. However, in both cases, this trend has been seen only for men; low SEP women have consistently had higher rates than high SEP women. It is important to note that although SEP gradients for CHD in men may have reversed in rural Georgia (Evans County) during this period, there is no evidence in the total mortality experience of this cohort which suggests that at any time during the early part of the 20-year follow-up, low SEP individuals had better survival than high SEP individuals (95). To summarize, lower SEP is consistently associated with poorer health. This association is found when considering different ages and diseases, different geographical locales, and has been relatively stable over a considerable period of time. In the next section, we will present evidence which argues for the important role of SEP as a risk factor in the examina- tion of minority and white health differences. Socioeconomic Position and Minority Status This section describes the socioeconomic position of minority groups in the United States. Data on income, education and occupation is presented for blacks, hispanics, Asians, and American Indians. Much of the available data permits only analysis of white compared to "non-white." The "non-white" group is approximately 85 per cent black and 15 per cent other "non-whites." When possible, more detailed groupings will be presented. From a health standpoint, the lack of detailed information on other minority groups is a deficit since what data does exist suggests there are some important differences in SEP and in health status between the various minority groups. Income Table 1 shows the income distribution for hispanics, blacks, and all others including whites (23, 97). This data shows that hispanics and blacks are similar in income and that both have substantially lower incomes than whites. For children under 18, black children are four times more likely to live in poverty than whites. When the family is headed by a woman, black children are 56 per cent more likely to live in poverty than whites. However, the poverty status of blacks is not entirely due to the higher proportion of female-headed households. In fact, the black-white poverty difference decreases when comparing female-headed households only, suggesting that presence of dependents and lower incomes afforded women in general also serve to increase the poverty rates. Comparison of 1970 median incomes earned by non-black minority groups shows that white males earned more than three times that earned by American Indians, 47 per cent more than Japanese males, and twice that earned by Chinese males and Filipino males. The median income differentials were less striking for females, but white females tended to earn 10 per cent more than other females except for American Indian women, who earned two times less than white women (23). Occupation The white labor force participation rate is 7 to 8 per cent higher 73 than the rate for blacks and other minority groups. This picture is further complicated by the fact that black women have a 4 per cent higher participation rate than white women, while black men have an 8 per cent lower rate than white men. Examination of employment status among persons over 16 years of age for other minority groups shows that American Indians are the most disadvantaged (36% of males not in the labor force), while Japanese, Chinese, and Filipino males are similar to whites, with approxi- mately 21 per cent of males over 16 not in the labor force. Rates for women of non-black minority groups follow a similar pattern, except that their non-participation rates tend to be around 50 per cent. About 65 per cent of American Indian women are not in the labor force (23, 97). Table 2 shows the occupational distribution by minority group and sex. These figures were calculated as the relative proportion of whites employed in a category to minorities employed in that category. For example, white females were 39 per cent more likely to be employed in white collar jobs than black females. Consistently, minority groups have proportionately fewer members in white collar jobs and greater numbers in blue collar and service jobs. Blacks, hispanics, and American Indians are most similar in this regard. Asians are more similar to whites, except for employment in service jobs where white females exceed Asian females, and white males are slightly fewer than Asian males. Job tenure also varies by minority status. Thirty per cent of white males have job tenure of 20 years or greater. Females of both groups have the shortest tenure (10%), and black males have job tenure 10 per cent less than that of white males. Job tenure is associated with increased social stability, increased income, and increased post-retirement benefits and therefore affects socioeconomic position. Education Blacks, hispanics, and American Indians have lower educational attain- ment, and lower college enrollment than whites. This is much less true for Asians, whose educational attainment is similar to whites. According to 1978 data, among persons over 18 years of age, 83.9 per cent of whites, 69.8 per cent of blacks, and 55.5 per cent of hispanics were high school graduates. Blacks aged 18-19 had lower college enrollment (25%) compared to whites of the same ages (35%). Hispanics were eight times less likely to have college or greater education than whites and 2.5 times more likely to have a less than eighth-grade education than whites. In 1978, the black-white ratio for college enrollment among persons 14-34 years of age was .13 for men and .15 for women. Tables 3 and 4 show educational attainment levels for whites and minorities. The data presented above abundantly demonstrate that blacks, hispanics, and American Indians are of lower socioeconomic position than whites as masured by income, education, and occupation. Asians appear to be at less of a disadvantage with respect to educational attainment but are also disadvantaged with respect to income and occupation. The recent changes in immigration patterns may have altered the socioeconomic position of Asians, and an examination of more recent data (i.e., 1980 Census) could be useful. It seems apparent that our understanding of minority health status must include examination of SEP. The evidence provided in previous sections on SEP as an independent risk factor and on the close association between minority status and low SEP point to the need for this approach. 74 Minority Status, Socioeconomic Position, and Health This section will review research on minority status and health which also examines socioeconomic position. We will specifically examine all- cause mortality, cardiovascular disease, cancer, infant mortality, and mortality from non-disease causes such as accidents, fires, and drownings. In general, research on minority and health has not simultaneously examined SEP. This is of particular concern because of the close association between minority and socioeconomic position discussed above. Without such an approach in studies of minority health, especially of the more disadvan- taged groups such as blacks, hispanics, or American Indians, it is difficult to conclude whether any results obtained are due to some minority characteristic or due to the socioeconomic conditions prevailing in that group. All-Cause Mortality and General Morbidity Black, hispanic, and American Indian minority groups in the United States generally incur higher mortality rates from all causes and exhibit higher rates of other indicators of morbidity. Table 5 shows some measures of morbidity for blacks, hispanics, and all others (including whites) of all ages for incomes less than $5,000 and greater than $15,000 for 1976 (106). The prevalence of morbid conditions, hospitalizations, and activity limitations are negatively associated with income for all groups. The rates for each of these measures of morbidity are very similar between the different groups at each income level. The number of days of restricted activity varies somewhat but a consistent income gradient is still present within each group. Kitagawa and Hauser (51) have shown that SEP is consistently asso- ciated with mortality and that the association between SEP and all-cause mortality is as consistent within minority groups as it is for whites. A problem found in the Kitagawa and Hauser study and many others is that the great majority of minorities have lower incomes than whites, making adjustment for SEP difficult within minority groups and for purposes of comparison to whites. An opportunity to examine the contribution of SEP to the differential survival experience of whites and blacks presented itself in studies of Alameda County, California, residents. In 1965, the Human Population Laboratory of the California Department of Health Services selected a representative sample of almost 7,000 adults to participate in a longitu- dinal study (6). The mortality experience of this cohort has been ascertained through 1982. Survival of blacks, as expected, was poorer than whites. A proportional hazards model showed that the age-sex adjusted hazard rate was 34 per cent higher for blacks (p=.004). When a measure of income adjusted for family size was introduced, the difference between black and white survival was no longer significant, while the impact of income was significant (p=.0001). Figures la and lb present the differen- tial survival experience of blacks and whites in this cohort without adjustment for income (la) and with such adjustment (lb). Thus in these analyses, differences in SEP appear to account, to a great extent, for the differential survival experience of blacks and whites. The association between SEP, minority status, and health is relatively consistent when specific disease outcomes are examined. The next sections will discuss these associations with respect to cancer, cardiovascular 75 disease, and infant mortality. These outcomes were chosen because they represent major causes of morbidity and mortality and because data are available that permit adjustment for both minority status and SEP. Cancer Several studies have reported associations between increased cancfr incidence and poorer survival with socioeconomic position (17, 55, 71, 72). The association has been observed for cancer incidence and survival for all sites but varies by specific site. Similarly, differences in cancer incidence and mortality vary by minority group. For example, blacks have higher incidence rates and poorer survival from rectal cancers than whites. White women have higher incidence of breast cancer and better survival than black women (71). A study of cancer patient survival among minority groups in the United States reported that survival from all-site cancer was substantially worse for blacks, American Indians, and Chinese than for whites (113). Table 6a shows the ratio of white five-year survival rates to each minority group s survival rates for males. Table 6b shows these ratios for breast and corpus uteri for females. Black males and Hawaiian males have been reported as having higher cancer incidence rates than whites and other minority groups for all sites. (107). White females and black females have similar incidence rates (300/ 100,000) for all cancer sites, while Hawaiian women have much higher rates (400/100,000) than all other groups. Site-specific incidence rates for blacks vary considerably, as do those for other minority groups. For example, black males have higher incidence rates than white males for lung cancer, pancreatic cancer, and prostatic cancer, and black females have higher incidence rates than white females for cervical cancer. Hispanics have notably lower incidence rates than most other groups for all sites and for most site-specific cancer incidence rates. Although site-specific cancer incidence among blacks is not dramatically higher for most sites than white rates, survival from cancer for blacks and some other groups is poorer than whites' for many sites. Blacks1 five-year cancer survival is poorer than whites' for colon, rectum, nasopharynx, larynx, lung, bronchus, skin melanoma, prostate, urinary bladder, kidney, pelvis, brain and other nervous system, thyroid, non-Hodgkin's lymphoma, breast, and corpus and cervix uteri. American Indians also have poorer survival rates than whites for a large number of cancer sites. In general, blacks and American Indians are at greater risk than whites, while other minority groups appear to do better than whites. Notably, Chinese do worse with all-site and stomach cancer than whites. It is useful to note that the more disadvantaged groups (i.e., American Indian, black) have poorer survival rates than whites for many sites (113). SEP may also affect survival from cancer by affecting access to medical care or availability of information on cancer. Unfortunately, data showing incidence and survival for each minority group by SEP are not available. Cancer incidence from some sites may also be associated with SEP. For example, a study of coke plant workers found that blacks had a lung cancer SMR six times greater than whites employed in the same plant (61). Black workers in this study were employed in much greater numbers in jobs where the exposure to benzopyrene and other carcinogens was high. This study demonstrated, in part, that differences in employment opportunities may lead to differences in exposure and disease occurrence. 76 In fact, few studies of cancer survival and incidence among minority groups have also examined SEP. Limitations of available data are part of the reason for this lack: cancer is a rare disease, and a large number of cases is needed for such multivariate analyses; also, accurate information on socioeconomic variables is often not available. However, those studies which have examined SEP, minority status, and health have produced some important results. Dayal (21, 22) has conducted two analyses examining black-white differences in survival from prostate and breast cancer and the contribu- tion of SEP to those differences. In both studies, black-white survival differentials became non-significant with adjustment for SEP. A factor complicating the understanding of minority cancer differences is that some minorities present cancers at a later diagnostic stage than whites. The Dayal Study on breast cancer found that, even with adjustment for diagnos- tic stage, the black-white difference is significant. However, adjustment for SEP rendered the black-white survival difference non-significant. Lung cancer incidence rates for blacks are higher, and survival rates are lower than whites. A study by DeVesa and Diamond (25) reported an SEP gradient for lung cancer incidence in males within both black and white groups. Black rates were higher than white rates, and rates for low SEP persons in both white and black groups were poorer than rates for high SEP persons. The group with the lowest rates was high SEP white males, and the group with the highest rates was low SEP white males. However, the overall black-white difference lost significance when adjusted for age, area of residence, income, and education. Comparison of white to black males at the same educational level suggested that there were no significant differ- ences. A major shortcoming of this study was the lack of data on smoking, a major risk factor in lung cancer. However, the study authors felt that adjustment for smoking would not explain all of the differences and that SEP had an independent effect on lung cancer. Little research has been done on cancers in minority groups other than blacks in the United States. Some of the survival rates experienced by these groups are shown in the tables above. These rates are not adjusted for SEP, so it is not possible to determine what effect SEP may have on cancer incidence and survival in these groups. A study of cancer survival among Asians and Pacific Islanders in Hawaii reported that Caucasians had the lowest median survival time overall and that Chinese, who were at the lowest status economically, survived the longest (107). Hawaiians and Filipinos who were at the lowest SEP level exhibited the shortest median survival time. After adjustment for sex, age at diagnosis, stage of disease, and SEP, many of the white-minority and minority-minority differ- ences were non-significant. The strongest predictor of death in that study was stage at diagnosis, which has been associated with both SEP and minority group status in other studies. Thus there exist substantial differences between whites and minority groups with respect to cancer incidence and survival. Both minority status and SEP are associated with incidence of some cancers and with stage at diagnosis for many cancers. It seems apparent that minority status and SEP are intertwined in the etiology of and survival from cancer. Cardiovascular Disease Blacks in the United States have among the highest rates of cardiovas- cular disease (CHD) in the world (33, 34). Reports of CHD mortality have 77 shown black male rates to be higher than white rates for the past twenty years. Rates among black and white women are similar. Ischemic heart disease and stroke account for 35 per cent of mortality among blacks and other non-whites as a group. However, CHD incidence and mortality rates for other minority groups do not follow the same pattern as for blacks. A study of CHD mortality in Los Angeles County, California, reported blacks as having the highest CHD mortality; whites as second; and hispanics, Japanese, Chinese, and Filipinos in descending order of mortality. Black rates were 10 per cent higher than whites for all major cardiovascular diseases and 24 per cent higher for cerebrovascular diseases (32). Few cardiovascular disease studies have addressed the simultaneous issues of minority group membership and SEP. A study conducted in Evans County, Georgia, between 1960-1977 (34, 95) has attempted one such analy- sis. In this study (as in many studies), virtually all blacks were of low SEP. In fact, they were of lower SEP than most low status whites. This study reported that 20-year survival from all-cause mortality was almost identical for low SEP whites and blacks, and both were higher than high SEP whites. The risk of dying from ischemic heart disease associated with blood pressure, cholesterol, and smoking was similar for low status whites and for blacks, and both were substantially different from high status whites. A study in Charleston, South Carolina (50), comparing CHD incidence among black males and females, white males and females, and high SEP black males for the period 1961-1975 demonstrated that SEP is strongly negatively associated with CHD. This study reported that high SEP black males had the lowest incidence of all categories of CHD compared to all blacks and all whites, except for arteriosclerotic heart disease, for which they had the highest rates. Table 7 shows these results. The lower rates observed in high SEP black males were found for all CHD, non-fatal CHD, fatal CHD, acute myocardial infarction (both fatal and non-fatal), angina, and sudden death. The number of cases of CHD (n=12) among high SEP black males was low, however, and the observations are most valuable for the trend they suggest. CHD Risk Factors The CHD risk factors most frequently measured include hypertension, blood lipids, smoking, diabetes, Type A behavior, overweight, ECG abnor- malities, and, in some studies, heavy alcohol consumption. Blacks are reported (33, 34, 95) as being at greater risk for CHD from hypertension, diabetes mellitus, ECG abnormalities, and overweight (among black women). As a general pattern, these CHD risk factors operate for blacks and other minority groups as they do among whites. Several studies on the distribution of CHD risk factors have suggested that there is an association between SEP and risk of CHD among blacks. Attempts to examine the association between CHD risk factors, minority status, and SEP have included both ecological level and individual level measures. Research (94) by Tyroler and Cassel has reported that ecological measures of social disorganization are strongly associated with mortality from stroke among black males and females. Another study using ecological measures of urban stress (40) reported a positive association with systolic and diastolic blood pressure for blacks but not for whites. Kraus, Borhani, and Franti (52), in their study of CHD risk factors, 78 reported a consistent negative SEP gradient for CHD risk factors and a consistent negative association between SEP and CHD risk within ethnic groups. Comparison of different minority groups at the same SEP levels suggests that (a) low SEP white males are at greater risk than minority males at the same SEP level, and (b) high SEP white males are at lower risk than minority males at a high SEP level, except hispanic males. For example, low SEP white males were at 30 per cent greater risk than low SEP black males, 52 per cent greater risk than low SEP Asian males, and 130 per cent greater risk than low SEP hispanic males. Conversely, high SEP white males were at 28 per cent lower risk than high SEP blacks males, 8 per cent lower than high SEP Asians, but 25 per cent higher risk than high SEP hispanics. A study by Stern et al. (87) on Mexican Americans reported that they had higher CHD risk factors levels than whites with respect to plasma lipids, diet, and adiposity but lower risk from cigarettes, blood pres- sure, and alcohol. The study made the point that Mexican Americans were primarily of low SEP but did not report any SEP-stratified data. A study by Roberts and Lee (75) on Mexican Americans reported that adjustment for health practices reduced but did not completely explain the health differ- ence between whites and Mexican Americans. Data from the Hypertension Detection and Follow-up Program has been reported (47) showing that prevalence of hypertension among blacks and whites of both sexes decreases with increasing education. Furthermore, the difference between whites and blacks generally decreases with increasing education. The evidence presented above suggests that SEP is a powerful risk factor that may help to explain the higher incidence and mortality rates and the poorer survival rates among certain minority groups. Furthermore, it suggests that the association between SEP and minority status and CHD cannot be fully explained by adjustment for risk factors such as smoking, alcohol consumption, or obesity. In short, SEP appears to exert an inde- pendent influence upon CHD and to partially explain the differences between blacks and whites. The association between SEP and CHD risk and occurrence among other minority groups is less clear, primarily because data on these groups is sparse and not generally presented with information on SEP. Infant Mortality As discussed in an earlier section, the association between infant mortality, low birth weight, and SEP is well established. The question to be examined in this section is whether the differences observed between blacks and whites can be explained by SEP. Data from 1976 on low birth weight reported by NCHS (104) show that (a) the percentage of infants weighing 2,500 grams or less at birth decreases with increasing education of mother or father, (b) that the percentage of low birth weight infants is greater at all education and income levels among blacks but declines with increasing SEP for both groups. Education of the mother appears to have little direct effect on low birth weight when prenatal care is totally absent. When prenatal care is present, education of the mother has a strong effect for both groups. The effect of prenatal care is not unrelated to SEP, however, since access to medical care and awareness of the need for prenatal care are probably both associated with SEP (89). The prevalence of low birth weight infants is greater among lower SEP individuals of both black and white groups and 79 is greater for blacks than for whites at all SEP levels. Infant death is strongly associated with SEP for both whites and blacks (103, 105) whether measured by education of father or mother or by family income. Blacks at all income or education levels had higher rates than whites. However, the black-white difference decreases as SEP increases. For example, the black-white difference is 12 per cent when the father's education is eighth grade or less and 4 per cent when the father s education is at high school level (105). In this data, blacks were not present at the highest SEP levels, so it was not possible to determine whether the black-white differential for infant mortality would disappear at higher SEP levels. These data suggest that SEP as measured by parental education or family income can help explain the black-white differences in infant mortality. Data on higher SEP blacks and other minorities are lacking, however. SEP and minority status appear to have direct effects on infant mortality and an indirect effect on neonatal mortality. A 1980 study by Brooks (12) using area measures of racial composition and income reported that racial composition did not affect post neonatal mortality, whereas income and low birth weight together explained 65 per cent of the variance. Low income alone explained 57 per cent of the variance. The addition of racial composition to a model including low birth weight, illegitimate status, and low income increased the explained variance by only 1.2 per cent, a non-significant change, suggesting that income and not minority status contributed most to infant mortality. Neonatal mortality in this study was best explained by a model including low birthweight, low income, and racial composition. Low birth weight and low income were highly correlated (.80) in this study, as were racial composition and low income (.683) (57). Further evidence for the hypothesis that the higher rates of infant death and childhood experienced by blacks may be partly explained by SEP is provided in research by Mare (57). His research reported that for both blacks and whites, mother's education and family income were negatively associated with death for children of all ages under 19 years of age. Furthermore, this association increased in size with increases in the child's age. In general, mortality rates for white males were higher at most ages than rates for black males at both high and low income levels. Among females, the effects of income were less clear. For annual family income less than $10,000, black females generally had lower rates than white females, and this was more true at older age levels. Comparison of white females to black females at family incomes over $10,000 shows that black females suffered substantially higher mortality rates at all age levels. Examination of the association between mother's education and mortality produced somewhat different results. Mortality rates for white male children were lower than black male children at younger ages and higher at older ages for education less than twelve grades. For mother's education at twelve grades or more, the reverse was true, and black males mortality rates were lower than whites. The mortality rates for black females whose mothers had less than a twelfth-grade education were somewhat higher than white females. For those females with mothers educated at twelve grades or better, black females suffered substantially higher mortality than white females. This study demonstrated a clear association between SEP and childhood mortality for both blacks and whites and suggests that, for males at least, the higher childhood death rates suffered by blacks may be due to lower socioeconomic position. 80 These data suggest that neonatal and post-neonatal mortality are affected by SEP for both blacks and whites. Such socioeconomic factors as low income, low education, low status occupation, minority status, teenage pregnancy, and non-married parents are all closely associated risk factors in the etiology of low birth weight, neonatal mortality, and infant mortality. The data available on SEP was insufficient to conclude that all black-white differences could be explained by SEP. However, it is clear that SEP is a powerful risk factor in both infant and neonatal mortality for both blacks and whites and that increasing SEP substantially decreases the black-white differences with respect to low birth weight, infant mortality, and childhood mortality. Non-Disease Causes of Injury and Death: Minority Status and Socioeconomic Position There is substantial variation in injury death rates among ethnic groups. In general, Native Americans and blacks have the highest death rates from such causes, and Asian Americans have the lowest. For a number of non-disease causes of death, the differences between ethnic groups is lessened or eliminated with adjustment for some measure of SEP. This section reviews some available data which demonstrates these points for unintentional injuries, motor vehicle accidents, accidental death from firearms, and deaths from house fires. More than 160,000 Americans died in 1980 from unintentional injuries, including such causes as accidental ingestion of poison, poisoning by faulty heaters (i.e., carbon monoxide), and motor vehicle accidents (nearly 50% of all unintentional injuries), etc. Asian Americans have the lowest rates, and Native Americans have the highest rates, with blacks and whites falling in between. All rates decline substantially with increasing income, although the differences between minority groups are not greatly reduced by such adjustment, except for Native Americans. The rate per 100,000 for Native Americans drops by nearly 300 per cent between per capita income less than $3,000 and per capita income of $5,000. The rate for blacks drops by more than 100 per cent with adjustment for income. The change in rates for Asians and whites is similar to those for blacks with income adjustment. Death rates from unintentional firearm injury for whites and blacks are similar, with the rates for both groups declining precipitously with increasing income. Blacks have much higher death rates from housefires than whites. However, this difference declines substantially with adjust- ment for income. The difference between blacks and whites at per capita area incomes less than $3,000 is three-fold. At incomes greater than $6,000, the difference is less than 100 per cent higher. The black-white death rate difference for occupants of motor vehicles and for pedestrian deaths declines with income adjustment also, although blacks have lower rates than white from the former cause and higher rates than whites from the latter cause. In general, it may be said that differences in non- disease mortality rates between whites and minorities, especially blacks and Native Americans, are diminished with adjustment for income (4). Part 2 In the preceding sections, we have argued that socioeconomic position 81 ought to be considered as a potential explanatory variable when considering minority and white health. We have reviewed the evidence that SEP is consistently related to a variety of health outcomes for different ages, places, and times. We have briefly presented evidence of the strong asso- ciation between SEP and membership in minority groups and have reviewed much of the available evidence that differences in the distribution of SEP may account for the differential health experience of whites and minori- ties. Our intent was to make the argument plausible. The evidence which we have presented, in our opinion, supports such an argument. However, it is important to specify why low SEP is associated with poor health. It has has been argued that such associations reflect the downward drift of less healthy individuals into lower socioeconomic strata. However, there are a number of reasons to believe that this is not what accounts for the association between SEP and health. Although it is undoubtedly true that long-term illness has an impact on income, it is difficult to see how such an explanation might apply to groups of individuals. Given the over- all pattern of lower SEP associated with minority status, it is hard to argue that this lower SEP is the result of poorer health. Indeed, in one analysis (18), income differences between minorities and whites were substantially reduced when there was statistical adjustment for age, educa- tion, occupational prestige, hours worked in the previous week, and average income of the state of residence. This adjustment accounted for 57 per cent of the income differences between whites and blacks. The comparable figures for Mexican Americans, Puerto Ricans, and American Indians were 49 per cent, 93 per cent, and 70 per cent, respectively. In short, the lower SEP of minorities is not due to poorer health, rather it reflects an overall pattern of disadvantage. The argument is also not plausible given the variety of measures of SEP shown to be associated with poorer health. As we have pointed out earlier, although each of the measures of SEP has some interpretive problems, the overall pattern across measures is sufficiently consistent to be compelling. Differential patterns of risk factors are often proposed as explana- tions for SEP gradients of disease. Our review has not turned up consistent patterns of risk factor differences which could account for the disparities between minority and white health. There are few studies which allow us to examine in detail the validity of these explanations. The few studies there are suggest that such explanations do not adequately account for SEP gradients. With respect to cardiovascular disease, there are three studies which have had the opportunity to directly address this issue. In one study of cardiovascular disease among 18,000 male British civil servants, it was possible to examine the contribution of serum cholesterol, smoking, hypertension, and other cardiovascular risk factors to the gradi- ent of cardiovascular disease associated with SEP, measured by broad occupational groupings (76). In these analyses, there was a consistent gradient of cardiovascular mortality associated with SEP; those in adminis- trative classifications had the lowest rates, followed by those in professional/executive positions, clerical positions, and the remainder. Figure 2 presents the results from this study when cardiovascular risk factors were introduced. Taking into account the standard risk factors for cardiovascular disease did not alter the gradient associated with SEP. Similar results were found by Salonen (80) in Finland, and Holme et al. in Oslo (46). 82 Turning to lung cancer, SEP gradients do not seem to be entirely due to higher rates of smoking among lower SEP groups. Although lower SEP groups such as blacks may have higher rates of current smoking, some evidence suggests that they smoke fewer cigarettes and tend to use weaker tobacco products (86, 111). Analyses of data from the Third National Cancer Study (25), the Washington County, Maryland, Study (15), and other studies (109) suggest that adjustment for level of smoking does not elimi- nate the SEP gradients for lung cancer incidence. Differences in access to medical care are also often proposed as an explanation for SEP gradients in health. However, such factors do not adequately account for SEP gradients. The presence of the National Health Service in England and Wales and the equivalent services in the Scandinavian countries would seem to provide reasonable access to care. However, in England and Wales, Sweden, and Finland, there are substantial SEP gradients of health. The evidence in England and Wales is that these gradients did not change substantially following introduction of the National Health Service (45). Similarly, the last 20 years in the United States have seen large changes in the accessibility of medical care to the poor. Between 1964 and 1976, persons in the lowest fifth of the income distribution increased utilization of physician and hospital services by one third (54, 70). Similar changes in health insurance coverage have occurred, particularly for the aged. However, despite these changes, national data do not indicate any major changes in the SEP gradient of prevalence or mortality. This is not to say that such changes have not had important health consequences but only that they do not seem to have resulted in major changes in the association between SEP and health. Further evidence that differences in levels of risk factors or medical care do not account for SEP gradients of health comes from analyses we have recently completed at the Human Population Laboratory in Alameda County, California (38). In these analyses, we examined the nine-year mortality experience of a representative sample of adults in Oakland, California, beginning in 1965. At that time, a portion of Oakland was federally desig- nated as a poverty area, based on rates of unemployment and income reported in the 1960 Census. Table 8 shows some of the characteristics of the poverty area compared to the nonpoverty area (44). Approximately 41 per cent of Oakland's population lived in the designated poverty area. The poverty area exhibited disproportionate levels of unemployment for both men and women, poorer health measured in a variety of ways, and poorer quality of housing. Those in the poverty area had three times the rate of unem- ployment, twice the number with an eighth grade education or less, two and one-half times the rate of inadequate incomes, and almost two and one-half times higher rates of no health insurance compared to residents of the nonpoverty area. We were interested in the extent to which this pervasive pattern of socioeconomic disadvantage would be associated with poorer health among the residents of the poverty area. Furthermore, because data were available for each participant, we were able to ascertain if poorer health among the poverty area residents might be due to differences in age, income, baseline health status, lack of medical care, minority group status, health prac- tices such as smoking and alcohol consumption, or psychological factors such as depression. When we examined the nine-year mortality experience of this cohort, residents of the poverty area were at significantly increased risk of death. Furthermore, when all of the above factors were taken into account statistically, poverty area residents had 46 per cent higher mortality from 83 all causes. In other analyses (49), we have shown that this survival disadvantage persists over 17 years of follow-up. In addition, when adjustment for residence in the poverty or nonpoverty area was carried out, there were no significant differences in mortality for whites and non- whites. These results suggest that we need a broader based approach to our examination of SEP gradients in health and their value in explaining minority health experience. Poverty areas are characterized by a large number of vectors of disadvantage ranging from poorer environmental quality, higher unemployment, lower income and education, higher rates of crime, greater social isolation, poorer services, to higher levels of reported stress. It is of great significance that these are the areas in which a disproportionate number of minority group members live. This clustering of high socio-environmental demands such as pollution, bad housing, and crime, coupled with low resources such as low income, social isolation, and inadequate services, may be what is responsible for SEP gradients of health. Several research efforts, using ecological measures of social area characteristics, have produced results relevant to this approach. Jenkins et al. (48) found census tract SEP indicators such as low occupational status, substandard housing, and low median education to be associated with mortality from hypertensive diseases. They also found significant associations for mortality due to all respiratory diseases, cerebrovascular disease (excluding hypertension), and ischemic heart disease. Dayal et al. (20) has reported that residence in low socio- economic level neighborhoods is associated with mortality from both lung and non-lung cancers, suggesting that both air pollution and socioeconomic variables are associated with poorer health among low SEP groups. This association was not affected by adjustment for race. Harburg et al. (40), using an area measure of social stress, found a significant cross- sectional association between systolic blood pressure and residence in such areas for black males and females. Similarly, area measures of social disorganization were found by Tyroler and Cassel (94) to be positively associated with stroke mortality. Finally, a step toward integrating ecologic and individual level variables has been taken by Hakama et al. (39) in an analysis of cancer of the breast and cervix. The findings in his study suggest that social and physical environmental factors might be relatively more important in the etiology of breast cancer than cervical cancer. Conclusions Studies which have examined minority/white differentials in health have often alluded to differences in culture, lifestyle, or genetics and have generally ignored the role of SEP. However, minority status and SEP are closely associated, and the evidence suggests that a portion of the difference in health between whites and minorities can be explained by differences in SEP. Furthermore, SEP gradients of health cannot, in many cases, be explained by differences in risk factor levels or differences in medical care. Finally, in analyses of all-cause mortality, survival differences in cancer of the breast and prostate, male lung cancer incidence, and mortality from coronary heart disease, minority/white differentials in health decrease significantly when SEP is taken into account. For many other outcomes, the evidence suggests a diminution of minority/white differentials with adjustment for SEP. 84 These results suggest that it is not minority status, itself, which leads to poorer health. Indeed, some minority groups evidence, for some outcomes, better health. Rather, it is the association of low SEP with minority group membership which has consequences for health. It is clear from this review that more research and analysis is needed on the health status of minority groups. Much of the available data only focuses upon white versus black differences and excludes other minority groups or includes them in a non-white grouping. As we have demonstrated in the preceding sections, there are significant differences between the various minority groups with respect to both SEP and to health. However, our understanding of the role of SEP in minority health is compromised by the lack of data on patterns of incidence, survival, and medical care utilization. As has become apparent in the consideration of the declines in CHD mortality, such information may be critical to our understanding of mortality differences. These data may be particularly significant in unraveling the impact of SEP on minority and white health. Similarly, the effort to understand minority health experience would be greatly improved by analyses that also examine the role of SEP. The evidence presented in this report strongly suggests that such analyses would be particularly helpful in clarifying the reasons for the substantial differences between whites and minorities observed for most major disease outcomes and all-cause mortality. As we have discussed previously, the measurement of SEP is prob- lematic. The most commonly used measures — income, occupation, and education — may not adequately assess the effects of SEP on health. For example, a white collar worker and a blue collar worker may have the same income and education but experience a different social and physical environment at work. Similarly, a highly educated person may have a rela- tively low income. Also, different measures may affect health in different ways. For example, income may affect health through the ability to purchase adequate medical care, while occupation may affect health through differential social and physical exposures on the job. Finally, one or two-time measures of SEP may fail to capture the lifetime exposures that individuals actually incur. As we discussed earlier, much research suggests that social and physical risk factors may co-occur in consistent patterns which are not random but are determined by larger socioeconomic forces. Our understanding of the role of SEP in minority health would be enhanced by examination of both ecological and individual-level risk factors. The studies by Jenkins et al., Dayal et al., Tyroler and Cassel, and Harburg et al. and Hakama et al. (20, 39, 40, 48, 94) indicate that SEP involves more than measures of income, education, and occupation can capture. As we have amply demonstrated, a large proportion of minority group members are also low SEP group members. Therefore, our understanding of minority health will be improved if analyses capturing the complex interrelationships between these different levels and types of risk factors can be attempted. An approach which combines environmental and individual level analyses can provide a method for a more coherent description of disease etiology than approaches which focus on only one level of analysis. This approach could be especially important in the investigation of minority health and SEP, factors which are multi-faceted and which exert their effects at both group and individual levels. Without such an approach, it is unlikely that we will be able to understand the reasons for the differential health experience of minorities and whites. 85 TABLE 1 Income by Minority Status for Currently Employed Persons 17 Years of Age and Older (per cent), 1976 Income Hispanic Black All other (including white) $5,000 $5,000 to $9,999 $10,000 to $14,999 $15,000+ 13.5 30.5 26.9 29.1 17.2 30.6 28.0 28.0 7.4 17.6 24.5 50.5 Totals 3,662 7,418 69,463 (number) Source: (23) 86 TABLE 2 Occupational Distribution of Minority Groups Ratio of White to Minority Group Category Black (112) Asian (88) American Indian (88) Hispanics (23) Sex M M M M White collar Blue collar Farm Service Employed 1.39 1.59 1.03 .96 1.50 2.04 4.70 .77 .85 .88 1.42 .68 .77 1.54 1.56 1.08 1.20 1.03 .78 .66 .56 .49 1.03 .35 .57 .57 1.08 1.07 1.90 1.39 2.95 1.82 .13 Source: (23) TABLE 3 Educational Attainment by Sex and Race, 1978 Females Males % % % % White Black White Black to 8th 4.3 4.0 4.3 6.5 grade 4 years of 21.4 12.6 27.6 10.7 college Source: (97) 88 TABLE 4 Median Years of School by Minority Status, 1970 American Indian Japanese Chinese Filipino Women 9.9 12.4 12.2 12.7 Men 9.7 12.6 12.5 11.6 Source: (97) 89 TABLE 5 Age-Adjusted Rates for Selected Conditions by Income and Minority Group Condition Hispanics Blacks All Others $5,000 $15,000+ $5,000 $15,000+ $5,000 $15,000+ Limit of 19.7 12.2 24.9 10.4 23.0 10.8 activity due to chronic condition Hospitalization 11.7 9.1 13.7 9.0 12.6 9.7 in short-term stay hospital Days of 16.3 4.2 12.8 8.5 10.1 5.5 restricted activity Source: (106) 90 TABLE 6A Ratio of White/Minority Group 5-Year Survival Rates for Males Minority Group All Sites Lung Cancer Site Stomach Rectum Prostate Anglo 1.00 1.00 1.00 1.00 1.00 Hispanic .97 1.00 .60 1.09 .92 Black 1.07 1.14 .82 1.40 1.10 American Indian 1.48 2.00 1.50 1.67 1.55 Chinese 1.15 .67 1.13 .81 .94 Japanese .91 .80 .43 .78 .82 Filipino 1.11 .89 .82 1.03 .80 Hawaiian 1.29 .80 .75 .75 .98 TABLE 6B Breast Corpus Uteri Anglo 1.00 1.00 Hispanic .99 .99 Black 1.16 1.65 American Indian 1.25 1.23 Chinese .90 .96 Japanese .79 .95 Filipino .97 .95 Hawaiian .96 1.15 Source: (113) 91 TABLE 7 Age-Adjusted Incidence of Cardiovascular Disease by Sex and Minority Status, and SEP Males Females Blacks Whites Hi-SEP Black Blacks Whites All CHD 131.7 188.4 61.2 161.0 113.8 CHD Deaths 79.8 93.8 38.3 62.2 46.3 Source: (95) 92 TABLE 8 Poverty Area Characteristics 41% of Oakland's Population 66% of unemployed males 14 years or older 61% of unemployed females 14 years or older 85% of Oakland General Assistance recipients 79% of AFDC recipients 79% of aid to disabled 63% of blind receiving aid 65% of police work load 68% of active TB cases 69% of Oakland's deteriorating housing units 75% of Oakland's non-owner occupied units 89% of Oakland's housing units with shared or no bathroom Source: (44) 93 Figure 1 Age and Sex-Adjusted Survival for Blacks and Whites in the Alameda County Study In la, blacks have significantly poorer survival (p < 0.004); when there is adjustment for SEP (lb), this difference is no longer significant (p > 0.05). la Survival Curves by Race Adjusted for Age and Sex Black versus White 1.00 to 0.95 \ -Q Probability o o \. o White Estimated Survival P o 00 bo Black \ > 075 — i i i i i i i I l ) 2 4 6 8 10 12 14 16 18 Years in Study lb Survival Curves by Race Adjusted for Age, Sex, and Income Black versus White 1.00 ^^fi* 0.95 Probability p o XNq, White — Nv ^ Black >a bN A. \ ed Survival o bo Cfl E to w 0.80 0.75 — 1 1 1 1 1 1 1 1 1 ( 3 2 4 6 8 10 12 14 16 18 Years in Study 94 Figure 2 Relative Risk of Death from CHD Compared to Administrative Classification Adapted with permission from reference 76. Relative Risk (Log Scale) r.\t D Cholesterol Smoking Blood Pres. Others Unexplained 1.0 Administrative y/ss. Professional/ Executive 3.2 Clerical 4.0 2.6 Others 95 REFERENCES 1. Antonovsky A. Social class, life expectancy, and overall mortality. Milbank Mem Fund Q 1967; 45:31-73. 2. Antonovsky A, Bernstein J. Social class and infant mortality. Soc Sci Med 1977; 11:453-70. 3. Austin DF, King M, Roe K. Cancer incidence in white women of the San Francisco Bay Area. National Cancer Institute Monograph, 1979; 53:95-101. 4. Baker SP, O'Neill B, Karpf RS. The injury fact book. Lexington, MA: DC Heath, 1984. 5. Behm H. Socioeconomic determinants of mortality in Latin America. In: Proceedings of the meeting on socioeconomic determinants and consequences of mortality, WHO, Geneva, 1980. 6. Berkman LF, Breslow L. Health and ways of living: the Alameda County Study. Cambridge, MA: Oxford University Press, 1983. 7. Black D, Morris JN, Smith C, Townsend P. Inequalities in health: the Black Report. Middlesex, UK: Penguin Books, 1982. 8. Blaxter M. Social class and health inequalities. In: Carter CO, Peel J, eds. Equalities and inequalities in health. London: Academic Press, 1976:111-125. 9. Bradshaw J, Lawton D, Staden F, Weale J, Weekes A. Area variations in infant mortality, 1975-77. J Epidemiol Comm Health, 1982; 36:11- 16. 10. Brennan ME, Lancashire R. Association of childhood mortality with housing status and unemployment. J Epidemiol Comm Health, 1978; 32: 28-33. 11. Bright M. A follow-up study of the Commission on Chronic Illness Morbidity Survey in Baltimore. II. Race and sex differences in mortality. J Chron Dis 1967; 20:717-29. 12. Brooks CH. Social, economic and biologic correlates of infant mor- tality in city neighborhoods. J Health Soc Behav 1980; 21:2-11. 13. Caldwell JC. Education as a factor in mortality decline: an exami- nation of Nigerian data. In: ed. Proceedings of the meeting on socioeconomic determinants and consequences of mortality. WHO Geneva, 1980. 14. Chase HC. The relationship of demographic factors and medical care to adverse pregnancy outcomes. In: Kelly S, Hook EB, Janerich DT Porter IH, eds. Birth defects - risks and consequences. New York: Academic Press, 1976:139-57. 96 15. Cornstock GW, Tonascia JA. Education and mortality in Washington County, Maryland. J Health Soc Behav 1978; 18:54-61. 16. Congressional Research and Service. Infant mortality, a report pre- pared by the Congressional Research Service for the Subcommittee on Health and the Environment and the Subcommittee on Oversight and Investigations of the Committee on Energy and Commerce, US House of Representatives. Washington, DC: US Government Printing Office, 1984. 17. Cooke KR, Skegg DCG, Fraser J. Socio-economic status, indoor and outdoor work, and malignant melanoma. Int J Can 1984; 34:57-62. 18. Cooper R, Steinhauer M, Miller W, David R, Schatzkin, A. Racism, society and disease: an exploration of the social and biological mechanisms of differential mortality. Int J Health Serv 1981; 11(3):39-414. 19. Cunningham LS, Kelsey JL. Epidemiology of musculoskeletal impairments and associated disability. Am J Public Health 1984; 74:574-79. 20. Dayal H, Chiu CY, Sharrar R, Mangen J, Rosenwaike I, Shapiro S, Henley AJ, Goldberg-Alberts R. Ecologic correlates of cancer mor- tality patterns in an industrialized urban population. J Nat Cancer Soc 1984; 73:565-74. 21. Dayal HH, Chiu C. Factors associated with racial differences in survival for prostatic carcinoma. J Chron Dis 1982; 35:553-60. 22. Dayal HH, Power RN, Chiu C. Race and socioeconomic status in survi- val from breast cancer. J Chron Dis 1982; 35(8):675-83. 23. Department of Commerce. A statistical portrait of women in the United States: 1978 special studies. Series P-23, No. 100. Washington, DC: US Government Printing Office, 1980. 24. Derrienic F, Ducimetiere P, Kritsikis S. La mortalite cardiaque des Francais actif d'age moyen selon leur categorie socio-professionelle et leur region de domicile. Revue d'Epidemiologie, Medecine Sociale, et Sante Publique 1977; 25:131. 25. DeVesa SA, Diamond EL. Socioeconomic and racial differences in lung cancer incidence. Am J Epidemiol 1983; 118:818-31. 26. Dorm HF, Cutler SJ. Morbidity from cancer in the United States. Public Health Monograph No. 56. Washington, DC: US Government Printing Office, 1959. 27. Ericson A, Eriksson M, Westerholm P, Zetterstrom R. Pregnancy out- come and social indicators in Sweden. Acta Paediatrica Scandinavica 1984; 73:69-74. 28. Feldman JG, Stein SC, Klein RJ, Kohl S, Casey G. The prevalence of 97 neural tube defects among ethnic groups in Brooklyn, New York. J Chronic Diseases 1982; 35(l):53-60. 29. Fisher S. Relationship of mortality to socioeconomic status and some other factors in Sydney in 1971. J Epidemiol Comm Health 1978; 32:41-46. 30. Forsdahl A. Are poor living conditions in childhood and adolescence an important risk factor for arteriosclerotic heart disease. Br J Prev Soc Med 1977; 31:91-5. 31. Frey RS. The socioeconomic distribution of mortality rates in Des Moines, Iowa. Public Health Rep 1982; 97:545-49. 32. Friis R, Nanjundappa G, Prendergast TJ, Welsh M. Coronary heart disease mortality and risk among hispanics and non-hispanics in Orange County, California. Public Health Rep 1981; 96(5):418-20. 33. Gillum RF. Coronary heart disease in black populations. I. Mortality and morbidity. Am Heart J 1982; 104:839-851. 34. Gillum RF, Liu KC. Coronary heart disease mortality in United States blacks, 1940-1978. Trends and unanswered questions. Am Heart J, 1984; 108:728-32. 35. Goldstein H. Factors related to birth weight and perinatal mor- tality. Br Med Bull 1981; 37(3):259-64. 36. Gortmaker SL. Poverty and infant mortality in the United States. Am Soc Rev 1979; 44:280-97. 37. Gorwitz K, Dennis R. On the decrease in the life expectancy of black males in Michigan. Public Health Rep 1976; 91:141-45. 38. Haan MN, Kaplan GA, Camacho-Dickey T. Poverty and health: a pros- pective study of Alameda County residents. Presented at 17th Annual Meetings of the Society for Epidemiologic Research, Houston, TX, 14 June 1984. 39. Hakama M, Hakulinen T, Pukkala E, Saxen E, Teppo L. Risk indicators of breast and cervical cancer on ecologic and individual levels. Am J Epidemiol 1976; 116(6):990-1000. 40. Harburg E, Erfurt JC, Chape C, Hauenstein LS, Schull WJ, Schork MA. Socioecological stressor areas and black-white blood pressure: Detroit. J Chron Dis 1973; 26:595-611. 41. HDFP Cooperative Group. Race, education and prevalence of hyperten- sion. Am J Epidemiol 1977; 106(5):351-361. 42. Hirayama T. An epidemiologic study of oral and pharyngeal cancer in Central and Southeast Asia. Bulletin WHO; 1969; 34:41-69. 43. Hirayama T. Epidemiology of stomach cancer. Gann 1971; 11:3-19. 98 44. Hochstim JR, Athanasopoulos DA, Larkins JH. Poverty area under the microscope. Am J Public Health 1968; 58(10):1815-27. 45. Hollingsworth JR. Inequality in levels of health in England and Wales, 1891-1971. J Health Soc Behav 1981; 22:268-83. 46. Holme I, Helgeland A, Hermann I, Leren P, Lund-Larsen G. Four-year mortality by some socioeconomic indicators: the Oslo Study. J Epidemiol Comm Health 1980; 34:48-52. 47. Igra A, Stavig RG, Leonard AR. Hypertension and related health problems in California: results from the 1979 California Hypertension Survey. California Department of Health Services, August, 1982. 48. Jenkins CD, Tuthill RW, Tannenbaum SI, Kirby C. Social stressors and excess mortality from hypertensive diseases. J Human Stress 1979; 5: 29-40. 49. Kaplan GA, Seeman TE, Cohen RD, Guralnik J, Knudsen LP. Mortality, morbidity, and functional ability in an aging population: evidence from the Alameda County Study. Presented at the American Public Health Association Annual Meeting, Anaheim, CA, 14 November 1984. 50. Keil JE, Loadholt CB, Weinrich MC, Sandifer SH, Boyle E. Incidence of coronary heart disease in blacks in Charleston South Carolina. Am Heart J 1984; 108:779-86. 51. Kitagawa EM, Hauser PM. Differential mortality in the United States: a study in socioeconomic epidemiology. Cambridge, MA: Harvard University Press, 1973. 52. Kraus JF, Borhani NO, Franti CE. Socioeconomic status, ethnicity and risk of coronary heart disease. Am J Epidemiol 1980; 111(4):407-14. 53. Kuller LH, Cooper M, Perper J, Fisher R. Myocardial infarction and sudden death in an urban community. Bull NY Acad Sci 1973; 49:1973. 54. Lerner M, Stutz RN. Have we narrowed the gap between the poor and the non-poor? Part II. Narrowing the gaps, 1959-1961 to 1969-1971: mortality. Med Care 1977; 15:620-35. 55. Lipworth L, Abelin T, Connelly RR. Socio-economic factors in the prognosis of cancer patients. J Chron Dis 1970; 23:105-16. 56. Logan WPD. Cancer mortality by occupation and social class, 1851- 1971. Office of Population Censuses and Surveys-Studies on Medical and Population Subjects, 1982; No. 44:1-253. 57. Mare RD. Socioeconomic effects on child mortality in the United States. Am J Public Health 1982; 72:539-47. 58. Marmot MG, Adelstein AM, Robinson N, et al. Changing social class distribution of heart disease. Br Med J 1978; 2:1109-12. 99 59. Martinez I. Cancer of the esophagus in Puerto Rico. Mortality and incidence analysis, 1950-1964. Cancer 1964; 17:1279-88. 60. McCormick MC, Shapiro S, Starfield B. High risk young mothers: infant mortality and morbidity in four areas in the United States, 1973-1978. Am J Public Health 1984; 74(1):18-21. 61. Michaels D. Occupational cancer in the black population: the health effects of job discrimination. J Nat Med Assn 1983; 75(10):1014- 1018. 62. Millar WJ. Sex differentials by income level in urban Canada. Canadian J Public Health 1983; 74:329-34. 63. Morgan M, Chinn S. ACORN group, social class and child health. J Epidemiology Community Health 1983; 37(3):196-203. 64. Morgenstern H. The changing association between social status and coronary heart disease in a rural population. Soc Sci Med 1980; 14A: 191-201. 65. Morgenstern H. Uses of ecologic analysis in epidemiologic research. Am J Public Health 1982; 72:1336-44. 66. Morgenstern H. Socioeconomic factors: concepts, measurements. In: Measuring psychosocial variables in epidemiological studies of cardiovascular disease. Galveston, TX: NHLBI, in press. 67. Morris JN. Social inequalities undiminished. Lancet 1979; 1(8107): 87-90. 68. Mustacchi P, Millmore D. Racial and occupational variations in cancer of the testis: San Francisco, 1956-1965. J Nat Can Inst 1976; 56:717-33. 69. Nayha S. Social group and mortality in Finland. Br J Prev Soc Med 1977; 31:231-7. 70. Newacheck PW, Butler LH, Harper AK, Piontkowski DL, Franks PE. Income and illness. Med Care 1980; 18:1165-76. 71. Nomura A, Kolonel L, Rellahan W, Lee J, Wegner E. Racial survival patterns for lung cancer in Hawaii. Cancer 1981; 48:1265-71. 72. Page WF, Kuntz AJ. Racial and socioeconomic factors in cancer survival: a comparison of Veterans Administration results with selected studies. Cancer 1980; 45:1029-40. 73. Paneth N, Wallenstein S, Kiely JL, Susser M. Social class indicators and mortality in low birth weight infants. Am J Epidemiol 1982- 116(2):364-75. 74. Pearce NE, Davis PB, Smith AH, Foster FH. Mortality and social class in New Zealand. II: Male mortality by major disease groupings. 100 New Zealand Med J 1983; 96:711-6. Roberts RE, Lee ES. The health of Mexican Americans: evidence from the Human Population Laboratory studies. Am J Public Health 1980; 70:375-84. Rose G, Marmot MG. Social class and coronary heart disease. Br Heart J 1981; 45-13-9. Ross RK, McCurtis JW, Henderson BF, et al. Descriptive epidemiology of testicular and prostatic cancer in Los Angeles. Br J Cancer 1979; 39-284-92. Rudov MH, Santangelo N. Health status of minorities and low income groups. USDHEW, Public Health Service, 1979. Rush D, Cassano P. Relationship of cigarette smoking and social class to birth weight and perinatal mortality among all births in Britain, 5-11 April 1970. J Epidemiol Comm Health 1983; 37:249-55. Salonen JT. Socioeconomic status and risk of cancer, cerebral stroke, and death due to coronary heart disease and any disease: a longitudinal study in eastern Finland. J Epidemiol Comm Health 1982; 36(4):294-7. Seidman H. Cancer death'rates by site and sex for religious and socioeconomic groups in New York City. Environ Res 1970; 3:234-50. Shapiro S, Weinblatt E, Frank CW, Sager RV. Social factors in the prognosis of men following first myocardial infarction. Milbank Mem Fund Q 1970; 48:37-50. Simpson SP. Variation in infant mortality rates among census tracts in Hawaii. Hawaii Med J 1983; 414-16. Simpson SP. Causal analysis of infant deaths in Hawaii. Am J Epidemiol 1984; 119:1024-9. Stanley MG, Shaw H, McCabe K. Effects of socioeconomic status and race on weight-defined and gestational prematurity in the United States. In: Reed D, Stanley F, eds. Epidemiology of prematurity. Urban and Schwarzenberg, 1977. Sterling TD, Weinkam JJ. Smoking patterns by occupation, industry, sex, and race. Arch Environ Health 1978; 33:313-7. Stern MP, Haskell WL, Wood PDS, Osann KE, King AB, Farquhar JW. Affluence and cardiovascular risk factors in Mexican Americans and other whites in three northern California communities. J Chron Dis 1975; 28:623-36. Stockwell EG, Wicks JW, Adamchak DJ. Research needed on socioeconomic differentials in US mortality. Public Health Rep 1978; 93:666-72. 101 89. Sunderland R. Dying young in traffic. Arch Dis Childhood 1984; 59: 754-7. 90. Syme SL, Berkman LF. Social class, susceptibility and sickness. Am J Epidemiol 1976; 104:1-8. 91. Syme SL, Oakes TW, Friedman GD, Feldman R, Siegelaub AB, Collen M. Social class and racial differences in blood pressure. Am J Public Health 1974; 64(6):619-20. 92. Taylor EM, Emery JL. Family and community factors associated with infant deaths that might be preventable. Br Med J 1983; 287:871-4. 93. Torgersen 0, Petersen, M. The epidemiology of gastric cancer in Oslo: cartographic analysis of census tracts and mortality rates of sub-standard housing areas. Br J Cancer 1956; 10:299-306. 94. Tyroler HA, Cassel J. Health consequences of culture change II. The effect of urbanization on coronary heart mortality in rural residents. J Chron Dis 1964; 17:167-77. 95. Tyroler HA, Knowles MG, Wing SB, Logue EE, Davis CE, Heiss G, Heyden S, Hames CG. Ischemic heart disease risk factors and twenty-year mortality in middle-age Evans County blacks males. Am Heart J 1984; 108:738-46. 96. USDHHS. Mortality from diseases associated with smoking: United States, 1960-1977. Vital Health Stat; Series 20, No. 7. 97. US Dept. of Commerce and Census. Social indicators. III. Selected data on social conditions and trends in the United States. US Dept. of Commerce, Dececember 1980. 98. USDHEW. Prevalence of chronic conditions of the genito-urinary, nervous, endocrine, metabolic, and blood and blood-forming systems and of other selected conditions. Vital Health Stat; Series 10, No. 109. 99. USDHEW. Prevalence of chronic skin and musculoskeletal conditions, United States, 1976. Vital Health Stat; Series 10, No. 124. 100. USDHEW. Disability days, United States, 1980. Vital Health Stat 1983; Series 10, No. 143. 101. USDHEW. Prevalence of selected chronic circulatory conditions, United States, 1972. Vital Health Stat; Series 10, No. 94. 102. USDHEW. A study of infant mortality from linked records by birth weight, period of gestation and other variables, United States, 1960, live birth cohort. Vital Health Stat 1972; Series 20, No. 12. 103. USDHEW. Comparison of neonatal mortality from two cohort studies: a study of infant mortality from linked records. Vital Health Stat 1972; Series 20, No. 13. 102 104. USDHEW. Factors associated with low birth weight, United States, 1976. Vital Health Stat 1976; Series 21, No. 37. 105. USDHEW. Infant mortality rates: socioeconomic factors. Vital Health Stat 1972; Series 22, No. 14. 106. USDHEW. Health characteristics of minority groups, United States, 1976. Advance Data 1976, Rept. No. 27. 107. Wegner EL, Kolonel LN, Nomura AMY, Lee J. Racial and socioeconomic status differences in survival of colorectal cancer patients in Hawaii. Cancer 1982; 49:2208-16. 108. WHO. Proceedings of the meeting on socioeconomic determinants and consequences of mortality. Geneva: WHO, 1980. 109. William RR, Horm JW. Association of cancer sites with tobacco and alcohol consumption and socioeconomic status of patients: interview study from the Third National Cancer Study. J Nat Cancer Inst 1977; 58:525-47. 110. Wynder EL, Bross IJ. A study of etiological factors in cancers of the esophagus. Cancer 1961; 14:389-413. 111. Wynder EL, Stellman, SD. Comparative epidemiology of tobacco related cancers. Cancer Res 1977; 37-4608-22. 112. Yeracaris CA, Kim JH. Socioeconomic differentials in selected causes of death. Am J Public Health 1978; 68:432-51. 113. Young JH, Ries LG, Pollack ES. Cancer patient survival among ethnic groups in the United States. J Nat Cancer Inst 1984; 73(2):341-52. 103 Associations of Health Problems with Ethnic Groups as Reflected in Ambulatory Care Visits M. Alfred Haynes, M.D. President/Dean Charles R. Drew Postgraduate Medical School Los Angeles, California Girma Wolde-Tsadik, Ph.D. Associate Professor Department of Internal Medicine Charles R. Drew Postgraduate Medical School Los Angeles, California Paul Juarez, Ph.D. Health Planner Center for Preventive and Community Medicine Charles R. Drew Postgraduate Medical School Los Angeles, California ASSOCIATIONS OF HEALTH PROBLEMS WITH ETHNIC GROUPS IN AMBULATORY CARE M. Alfred Haynes, M.D., Girma Wolde-Tsadik, Ph.D. and Paul Juarez, Ph.D. INTRODUCTION A comparison of the health status of various ethnic groups must be approached from several different perspectives because no single data source provides a fully satisfactory picture. Since birth and death events are more widely and reliably recorded, vital statistics are frequently quoted. The differences in mortality are easily summarized as differences in life expectancy, and one can easily grasp the significance of wide gaps in longevity. For example, in 1983, life expectancy at birth for the White population was 75.2 years while for the Black population, it was 69.6 years.1 But even this basic information is not always available for all ethnic groups. Trevino has noted that "at present we do not even know how many Hispanics die each year in this country, let alone their health status, use of services or unmet health care needs."^ Another approach to comparing health status would be to use institutional data such as that of the National Hospital Discharge Survey3 or the National Nursing Home Survey.4 This approach reflects only the experience of institutionalized populations and, therefore, has certain limitations. A much broader picture could be obtained from the National Health Interview Survey which is a nationwide survey conducted through personal household interviews. Data are collected on illnesses, injuries, impairments, chronic conditions, the use of health resources and a variety of other health and health-related characteristics.5 These data are subject to many limitations including problems about the accuracy of diagnosis and the fact that respondents cannot address problems of which they are not aware. On the other hand, the National Health and Nutrition Examination Survey6 obtains data by means of direct physical examinations along with clinical and laboratory tests. The sample size is much smaller than that of the National Health Interview Survey. An alternative method is the collection of data from a sample of physicians in office-based practices. This is the approach of the National Ambulatory Medical Care Survey,7 a continuing national probability sample of ambulatory medical encounters. The survey does not include visits to hospital-based physicians or to specialists in anesthesiology, pathology or radiology. Non-office visits and telephone contacts are not incorporated. 107 In order to obtain a true picture of the differences of health status of various groups, it is necessary to examine indicators from all of these surveys and to draw conclusions on the basis of the composite picture. Each analysis adds another dimension of significance. The purpose of this study was to examine the differences from the perspective of ambulatory care and to complement other analyses under consideration by the Task Force on Black and Minority Health. Inasmuch as most of the contacts with the health care system are made in the context of ambulatory care, this study should be considered as one of the important elements of the total picture. It differs from previous studies in ambulatory care which focus on other aspects. For example, there are studies of the patterns of ambulatory care visits to surgeons,° obstetricians9 and family practitioners.10 Also, there have been studies on specific groups such as the Asian Pacific Islanders1 or, in general, the poor.12 This study, on the other hand, is based on a comparison of the major racial/ethnic groups — White, Black, Hispanic, Asian and Native American. METHODS The data source was the 1981 National Ambulatory Medical Care Survey that included 43,123 patient visits. The frequency of patient encounters by the various racial/ethnic, sex and ten most frequent diagnostic groups was analyzed using the multiway contingency table method based on the log linear model. The basic chi-square technique for two-factor analysis was also employed in the preliminary examination of association between race/ethnicity and disease conditions. In addition, the importance of certain disease-ethnicity associations was highlighted by calculating the percentage contribution of these specific associations to the overall significance of the findings. The small number of visits of minority patients did not allow the performance of epidemiologically meaningful analyses by age groups. In fact, the 66 patient encounters by Native Americans were too few to incorporate in any one of the analyses. However, the frequency distribution of Native Americans by diagnostic categories will be presented. In order to maintain the minimum allowable expected cell frequencies, Black and White Hispanics were grouped to form a single category. FINDINGS The preliminary description of the data consists of a cross-tabulation of 18 diagnostic groups with the five racial/ethnic categories. The frequency counts are given in Table 1. More than 20% of the cells do not satisfy the minimum expected cell frequency of 5. The collapse of the relevant diagnostic categories, the exclusion of the Native American group and the subsequent performance of a test of association reveals a highly significant relationship between the health problems and the racial/ethnic groups (p£.0001). 108 In Table 2, the results of the test of association are presented in a reduced form. The diagnostic categories constitute those conditions that show more than equal contribution to the overall association. The entries are the percentage of the total chi-square attributable to the given cell. The plus and minus signs, respectively, indicate the observed frequencies that are more and less than the expected values under independence of the two factors. All ethnic groups showed significant association with some diagnostic category. The White patients were more likely to present themselves at ambulatory care settings with psychiatric problems than the other groups. Blacks and Hispanics, in that order, were far less likely to seek care for these conditions. Asian Americans were more frequently seen for diseases of the nervous and sensory systems, whereas Blacks and Hispanics were less likely to present themselves with these problems than would be expected. Among Blacks, a high frequency of problems of the circulatory system contributes to the overall significance of the relationship between diagnosis and race/ethnicity. Hispanics show a substantially lower frequency of visits for circulatory problems and, to a lesser extent, Asians account for infrequent care of this category. Respiratory problems are frequent causes of visit for Blacks as are digestive diseases for Hispanics. Musculoskeletal diseases, as a cause of visit for Hispanics, are major contributors to the disease and ethnicity relationship whereas Asians are less likely than expected to seek care for this set of conditions. Visits for the diagnostic category classified as "supplementary" are strongly associated with Asians. The classification is, generally, described as factors influencing health status and contact with health services. More specifically, it enumerates health hazards related to communicable diseases and family history as well as familial conditions related to social, mental and economic well-being. Finally, in Table 3, we present a key result of the three-way contingency table analysis. The three factors constitute sex, race/ethnicity and the ten most frequent diagnostic categories. The latter incorporate the conditions identified in Table 2 as well as injury/poisoning, genitourinary and skin diseases. The table entries represent the partial associations due to the indicated factors after the effect of the missing factor is removed. Hence, the rows of primary interest are the fourth, fifth and sixth. As demonstrated by the p-values, all the paired partial associations are significant. The findings presented in Tables 1 and 2 are further substantiated by the significance of the ethnicity-diagnosis partial association. It should be noted, however, that those results were based on more diagnostic categories and the marginal distributions of the frequency of visits. A more detailed scrutiny of the sex-diagnosis association reveals a more or less general pattern across ethnic groups. This pattern encompasses a disproportionately high number of male visits for genitourinary health problems and those of the supplementary classification, and an unexpectedly high number of female visits for injury and poisoning. The common configuration appears in Whites, Blacks and Hispanics. The latter also reveal a relatively high frequency of female visits for diseases of the digestive system. The 109 Asian group did not show any strong association between health problems and sex. The remaining results of Table 3 are of academic interest, ihe first three rows substantiate the disproportionate number of male, White and certain disease categories in the survey sample. The last row or sex-ethnicity-diagnosis with the non-significant result (p=0.1852) merely confirms the appropriateness of the log linear model. DISCUSSIONS The interpretation of the results should be characterized by caution. People with different attitudes towards health present themselves for care at different levels of severity. The high frequency of visits by Whites for psychiatric conditions, and those by Asians for diseases of the supplementary classification might reflect the groups' greater access to care and/or a greater sense of medical need. The supplementary classification includes preventive services but the significance of this finding cannot be established due to the relatively small number of patient visits. On the other hand, if Blacks and Hispanics feel a lesser sense of medical need, then the magnitudes of the identified circulatory, digestive and musculoskeletal problems are underestimated and there might be other problems to which they are at greater risk but whose significance is suppressed. This is of special interest since the morbidity findings are certainly not as striking as the mortality statistics would suggest. Some of the findings suggest the need for more detailed studies with respect to age and sex. For example, the disproportionately high number of female visits for injury and poisoning in Whites, Blacks and Hispanics is not exactly what one would have expected. The absence of association between health problems and sex in Asians also deserves attention due to its possible implications for the other groups. In summary, this study, despite its limitations, has established strong associations among the factors of sex, race/ethnicity and health problems as reflected in ambulatory care settings. A more detailed study of a larger sample of minority populations would provide additional information on these findings by permitting further analysis of more specific age categories. 110 TABLE 1: FREQUENCY OF DIAGNOSIS BY RACE AND ETHNICITY Race/Ethnicity Principal Diagnosis 1 Non-Hispanic Hispanic Asian Native 1 American 1 Total White Black Infective & Parasitic 1 1,006 123 51 13 7 1 1,200 Malignant Neoplasms 1 685 49 20 6 0 1 760 Other Neoplasms 1 385 37 15 2 0 1 439 Endocr./Nutr i./Metabol. 1 1,204 155 73 9 0 1 1,441 Blood/Blood Forming Organs 1 199 26 8 5 0 1 238 Psychoses & Neuroses 2,816 148 86 21 1 1 3,072 Nervous & Sensory Systems 3,419 273 132 51 5 1 3,880 Circulatory 3,679 452 123 17 4 1 4,275 Respiratory 4,037 495 230 41 11 1 4,814 Digestive 1,614 185 110 17 3 1 1,929 Genitourinary 1 2,351 248 137 16 3 1 2,755 Diseases of the Skin 1 1,711 145 75 16 2 1,949 Musculoskeletal 1 2,704 271 212 15 4 3,206 Congenital Anomalies 1 196 17 13 4 0 230 Symptoms 1 1,247 147 81 8 0 1,483 Injury & Poisoning 1 3,016 343 166 29 5 1 3,559 Supplem Class 1 6,433 610 319 90 21 1 7,473 Other Causes 1 142 30 18 7 0 1 197 Unrecorded 1 188 25 8 2 0 1 223 TOTAL 37,032 3,779 1,877 369 66 43,123 111 TABLE 2: STRENGTH OF ASSOCIATION BETWEEN DIAGNOSES AND ETHNICITY (CELL PERCENT CONTRIBUTION*) Race/Ethnicity Principal Diagnosis 1 Non-Hi spanic White 1 Black I Hispanic 1 Asian Psychoses and Neuroses 2.8 1 ( + ) 1 12.8 (-) I 4.0 1 1 (-) Nervous and Sensory Systems 3.1 (-) 1 1.9 1 (-) 2.2 ( + ) Circulatory 3.7 ( + ) I 5.0 1 (-) 2.5 (-) Respiratory 3.0 (+) Digestive 1 1.9 1 (+) Musculoskeletal I 8.8 1 ( + ) 1 1.3 1 (-) Supplementary Class 1 2.5 1 ( + ) * —h. rilne ^nr3 minus sian. re spectivelv . indica te more and less than expected frequencies under independence. -Empty cells have values less than or equal to 1.25 percent. 112 TABLE 3: PARTIAL ASSOCIATION OF FACTORS Degree of Freedom Partial Association Factors Chi-Square P-Value Sex 1 1558.7 0.0000 Ethnicity 3 67576.3 0.0000 Diagnosis 10 6456.2 0.0000 Sex-Ethnicity 3 30.1 0.0000 Sex-Diagnosis 10 1088.0 0.0000 Ethnicity-Diagnosis 30 277.2 0.0000 Sex-Ethnicity-Diagnosis 30 36.7 0.1852 113 REFERENCES National Center for Health Statistics: Health, United States, 1984. DHHS Pub. No. (PHS) 85-1232, Public Health Service, Washington, U.S. Government Printing Office, Dec. 1984. Trevino, F.M.: Vital and Health Statistics for the U.S. Hispanic Population. American Journal of Public Health, Sept. 1982; 7(9): 979-82. National Center for Health Statistics: Utilization of short-stay hospitals, annual summary for the United States, 1982, by E.J. Graves, Vital and Health Statistics, Series 13-No. 78, Public Health Service, Washington, U.S. Government Printing Office. National Center for Health Statistics: The National Nursing Home Survey, 1977 summary for the United States, by J.F. Van Nostrand, A. Zappolo, E. Hing, et. al., Vital and Health Statistics, Series 13-No. 43, DHHS Pub. No. (PHS) 79-1794, Public Health Service, Washington, U.S. Government Printing Office, July 1979. National Center for Health Statistics: Current estimates from the National Health Interview Survey, United States, 1981, by B. Bloom, Vital and Health Statistics, Series 10-No. 141, DHHS Pub. No. (PHS) 82-1569, Public Health Service, Washington, U.S. Government Printing Office, Oct. 1982. National Center for Health Statistics: Plan and Operation of the second National Health and Nutrition Examination Survey, 1976-80, by A. McDowell, A. Engel, J.T. Massey, and K. Maurer, Vital and Health Statistics, Series 1-No. 15, DHHS Pub. No. (PHS) 81-1317, Public Health Service, Washington, U.S. Government Printing Office, July 1981. National Center for Health Statistics: 1981 Summary, National Ambulatory Medical Care Survey, by L. Lawrence and T. McLemore, Advance Data for Vital and Health Statistics, No. 88, DHHS Pub. No. (PHS) 83-1250, Public Health Service, Hyattsville, Md., Mar. 16, 1983. Cypress, B.K.: Patterns of ambulatory care in office visits to general surgeons: The National Ambulatory Medical Care Survey, United States, January 1980 - December 1981. Vital Health Statistics [13], Sept. 1984; 79:i-iv, 1-61. Cypress, B.K.: Patterns of ambulatory care in obstetrics and gynecology: The National Ambulatory Medical Care Survey, January 1980 - December 1981. Vital Health Statistics [13], Feb. 1984; 76:1-62. 114 10. Cypress, B.K.: Patterns of ambulatory care in general and family practice: The National Ambulatory Medical Care Survey, United States, January 1980 - December 1981. Vital Health Statistics [13], Oct. 1983; 75:1-60. 11. Yu, E.S. and Cypress, B.K.: Visits to physicians by Asian/Pacific Americans. Medical Care, Aug. 1982; 20(8):809-20. 12. Kleinman, J.C., Gold, M. and Makuc, D.: Use of Ambulatory Medical Care by the Poor: Another Look at Equality. Medical Care, Oct. 1981; 19(10):1011-29. 115 Nutritional Status and Dietary Pattern of Racial Minorities in the United States Shiriki Kumanyika, Ph.D., R.D. Assistant Professor Department of Epidemiology The Johns Hopkins University School of Hygiene and Public Health Baltimore, Maryland Deborah L. Helitzer Doctoral Candidate Department of International Health The Johns Hopkins University School of Hygiene and Public Health Baltimore, Maryland ABSTRACT Dietary patterns of minority groups may differ from those of the general population according to several factors. These fac^s include the nature of the original diet, the ways the diet has been adapted to or supplanted by dietary patterns from the dominant u.b. culture, availability of preferred foods, and acculturation. The objectives of this review have been to describe general themes ana emphases in the diets of Asian, Black, Hispanic and Native Americans and, where possible, to identify associations between these dietary patterns and excess nutrition-related health risks. The content or the review has been affected by the uneven availability of data for different subgroups and by certain complexities which are associated with cross-cultural interpretation of dietary and nutritional status data. Following a general discussion of relevant nutrition concerns and data issues, specific dietary and nutritional concerns for each minority subgroup are summarized. In general, nutritional factors which may contribute to health disparities between minorities and whites can be grouped into four categories, as follows: 1)excess risks related to the low-income status of a high proportion of minority individuals and families; 2) risks related to the excess prevalence of obesity in some minority groups; 3) carryovers of dietary risks from traditional diets; and 4) increased risks related to movement away from traditional diets that are low in fat and cholesterol and high in complex carbohydrates to an opposite, more typically American, pattern. 118 CONTENTS LIST OF TABLES AND FIGURES 1. INTRODUCTION 2. BACKGROUND 2.1 Sources of Data on the Diets and Nutritional Status of Minority Groups 2.2 Nutrition-Related Health Variables 2.3 Methodological Issues 3. ASIAN AMERICANS 3.1 General dietary patterns 3.2 Nutritional risk 3.3 Conclusions 4. NATIVE AMERICANS 4.1 General dietary patterns 4.2 Nutritional risk 4.3 Conclusions 5. HISPANIC AMERICANS 5.1 General dietary patterns 5.2 Nutritional risk 5.3 Conclusions 6. BLACK AMERICANS 6.1 General dietary patterns 6.2 Nutritional risk 6.3 Conclusions 7. SUMMARY LITERATURE CITED SUPPLEMENTARY REFERENCES APPENDICES Appendix 1: Background Information on Vitamins and Minerals Appendix 2: Recommended Dietary Allowances and Estimated Safe and Adequate Daily Dietary Intakes 119 LIST OF TABLES AND FIGURES Table 1: Current Nutrition Concerns for U.S. Children Table 2: Current Nutrition Concerns for U.S. Adults Table 3: A Summary of Dietary Determinants Related to the Nutritional Risk Status of U.S. Minority Groups Table 4: Characteristic Chinese Foods and Food Choices Table 5: Characteristic Japanese Foods and Food Choices Table 6: Characteristic Filipino Foods and Food Choices Table 7: Examples of Some Southeast Asian Foods Table 8: Examples of Traditional and Contemporary Foods in the Diets of American Indians Table 9: Characteristic Mexican-American Foods and Food Choices Table 10: Characteristic Puerto Rican Foods and Food Choices Table 11: Characteric Black Foods and Food Choices Tabel 12: Examples of Some Caribbean Foods Figure 1: Percent of U.S. adults ages 20-74 who are overweight, by race and sex 1. INTRODUCTION Minority groups in the United States comprise several subgroups whose traditional dietary patterns may differ from those of the general population. The diets of minority group persons are influenced by several, interrelated factors, for example, the nature of the traditional diet, ways in which the diet has been adapted to or supplanted by dietary patterns from the dominant U.S. culture, tne availability of preferred foods, and acculturation. The contribution of culturally-determined dietary patterns to disparities in morbidity and mortality relates both to protective features of the original diets which may have been discontinued as well as to potentially harmful practices which may persist or have been acquired during acculturation. Minority-specific factors operate within the larger context of universal determinants of dietary intake and nutritional need. These universal determinants, which apply in all societies, include factors which are biological (e.g. heredity, gestation, growth, energy expenditure, illness), psychological (mental state, knowledge, attitudes), sreio-cultural (religion, household structure, child rearing practices, family interactions, food beliefs), ana socioeconomic (per capita food availablity, purchasing power, dependency, stability) in nature. The Secretary's Task Force on Black and Minority Health was convened specifically to identify aspects of the environment which predisposeMinority groups to disparities in health and to recommend ?elevantlolicy. Therefore, the Task Force deliberations provide a very appropriate context for reviewing diet and nutrition literature related to minority groups. The objective of this review has been to describe, in very general terms, the themes and emphases in the diets of various U.S. minority sub-groups and, where possible, to identify apparent associations between these dietary patterns and excess Stion-rllated health risks. The specific questions addressed here are whether Black, Hispanic, Asian, or Native Americans are at excess nutritional risk and whether this excess risk can be attributed to culturally- rather than to socioeconomically-based dietary factors. Complex issues are involved. Technical aspects are accompanied by extremely sensitive social and economic concerns. Some examples follow: 1) Food patterns are an integral aspect of culture. Retention of a traditional food pattern is an expression of ethnic identity. Thus, evaluation of minority group diets must be objective and must be sensitive to perceptions that the underlying culture is being criticized or patronized; 121 2) Although diets of members of a particular minority group may be more similar to each other than to those of the general population, there is a great deal of heterogeneity. Insufficient attention to intracultural diversity will lead to incorrect conclusions and may be viewed as stereotyping; 3) Findings of undernutrition among the low-income segment of the minority population may be attributed to inadequacies in the food and nutrition programs designed to attentuate poverty-associated malnutrition. In addition, such findings may be attributed to incompetence of low-income minority individuals and families. From a social policy point of view, both of these attributions are quite sensitive. As recently pointed out by the President's Task Force on Food Assistance (1983), documentation of even a modicum of undernutrition amidst the abundance of food in the U.S. is a sign of societal failure at some level. 4) Findings of obesity and excess risk of some nutrition- related chronic diseases among minorities suggest that adaptation to U.S. dietary patterns may have disproportionately deleterious consequences for minority group persons; In the text which follows, we have noted the nutrition concerns which we consider relevant to current issues of minority health, have listed some inference issues related to assessment of nutrition in non-white populations, and have summarized pertinent literature in this context. We have been especially aware of the tremendous gaps in the nutrition literature in this respect and of the outdatedness of much of the literature which is available. We have exphasized studies dated 1975 or later. However, even studies published within the last decade may not accurately represent those aspects of nutritional status which respond to changes in levels of food and nutrition program funding and outreach. 2. BACKGROUND 2.1 Sources of Data on the Diets and Nutritional Status of Minority Groups The availability of timely, representative dietary and nutritional status data varies greatly for the different U.S. minority sub-populations in rough proportion to the sub-group population size. National probability sample estimates for nutrition-relevant variables are available for blacks from the National Health and Nutrition Examination Surveys, 1971-/b ana 1976-80 (NHANES I and NHANES II) and from the National Food Consumption Survey 1977-78 (NFCS) and will be available for Mexican, Puerto Rican, and Cuban Americans from the Hispanic Health and Nutrition Examination Survey (HHANES) completed in 1985. Special surveys of the NFCS were also conducted in Alaska, Hawaii, and Puerto Rico, permitting—at least theoretically—estimates for minority sub-populations in these areas. NHANES data provide four types of nutritional status measures for each sample person (anthropometric, clinical, biochemical, ana dietary) as well as accompanying health ^^r%J^^^^S't dental and physical health by examination). The NFCS (and some other subdivision^ of the Consumer Food and Economics Institute) provides detailed data on individual and household food consumption patterns accordingto h^Ssehold characteristics. At present, one of the m3or shortcomings of these surveys results from the substantial time lag between data collection and publication. While these data sources have been useful for monitoring over relatively large time intervals, this time lag renders them unsuitable for assessment of acute nutritional status, short-term, or very recent changes. The Ten-State Survey, often cited in the nutrition literature, was conducted in 1968-69 and is now completely out of date. »? Stlte findings should be used only as a baseline for the nutritional SsS vulnerable groups prior to implementation of the expanded base of food and nutrition programs (program expansions were in large part a response to Ten-State Survey findings). With the exception of HHANES, the ability to assess race-specific nutritional vulnerability is not assured by the NHANES sa^fn^d^gn. The sampling objective is to permit repressive ^rimtes for the overall U.S. population rather than to create a Ita^all for sKoup comparisons. Although sub-groups at risk due to Lw^com^, Reductive status (women of childbearing age), and age (preschool children and the elderly) have been o^p3^1*^, rhese surveys, sample sizes for racial subgroups do not necessarily meet criteria foritable estimates-especially if cross-classified by sex,age, and socioceonomic indicators. 123 Separate tabulations for blacks are frequently but not routinely reported in NHANES publications. Accompanying statistical analyses of the significance of black-white differences are sometimes included. Data for minority groups other than blacks are either excluded from sub-group tabulations or included with data for blacks under a general heading of "non-white". Additional information on blacks can often be gleaned from published reports of special analyses by National Center for Health Statistics staff or from published analyses of NHANES data by members of the research community at large. Fewer relevant reports based on NFCS data than on NHANES data were identified during the preparation of this background paper. The apparent amount of NFCS data is larger than the actual data base which can be used for racial comparisons, because tables with racial breakdowns are often not age- and sex-adjusted. Another major source of population data on nutritional status is the Centers for Disease Control Pediatric and Pregnancy Nutrition Surveillance System (CDC-PNSS) (Centers for Disease Control, 1983). This system compiles height, weight, and iron status data for children and women in publicly-supported health programs in more than 20 states. CDC reports include separate tabulations for black, Hispanic, Native American and Asian children (including Southeast Asian refugees), using measurements taken upon entry into a nutrition or health program. These data are suitable for racial comparisons within the data base, but they cannot support generalities about minority children in the general population, or even to low-income minority group children. Participation in the CDC-PNSS is voluntary and varies by state, county, and program. Additionally, minority groups are variably represented in prevalence data derived from baseline screenings for cardiovascular disease studies (e.g.,Framingham, Lipid Research Clinics, Honolulu, Puerto Rico, and San Antonio Heart Studies), in data from evaluations of nutrition programs (such as School Lunch, the Supplemental Feeding Program for Women, Infants, and Children, Head Start, or the Nutrition Program for Older Americans) and in small studies on various nutritional status issues. Unfortunately, for the numerically smaller minority groups, the non-comprehensive and non-representative data sources are often the only basis for describing possible nutritional disparities. This unevenness of coverage is clearly evident in the later sections of this paper. 2.2 Nutrition-Related Health Variables The undernutrition- and chronic disease-related nutritional concerns addressed in this paper are summarized in Tables 1 and 2. For reference, listings of essential vitamins and minerals and recommended nutrient intakes have been appended (see Appendices 1 and 2). Tables 1 and 2 are not comprehensive listings of nutrition and health concerns. Rather, the problems indicated are those most relevant to current public health policy. 124 Table 1: Current Nutrition Concerns for U.S. Children LIFE CYCLE STAGE DIET and NUTRITION VARIABLES ASSOCIATED HEALTH OUTCOMES GESTATION, INFANCY, CHILDHOCO, and ADOLESCENCE inadequate intakes of energy and other essential nutrients overfeeding, feeding of high-calorie-low nutrient-density foods between-meal snacks, sugary snacks - low birth weight, developmental or functional impairment, growth stunting, iron deficiency anemia, low resistance to infection, poor general health - obesity, poor food habits, predisposition to obesity and obesity-related diseases - dental caries SOURCES: Food and Nutrition Board, 1980; National Institutes of Health, 1982 Conceptual and methodological limitations preclude definitive answers to the types of nutrition and health questions that are currently relevant to U.S. population groups—minority group issues included. At present, the occurrence of severe nutritional deficiencies in U.S. populations is limited to certain unusual circumstances. The most pertinent undernutrition issues relate to subtle effects of subclinical nutrient deficiencies on health outcomes at various stages of the life cycle. Most of these outcomes are not nutrition-specific, i.e. they are influenced by other biological and environmental factors. The pertinent ovemutrition issues are defined in terms of consistent, although not firmly established, associations of various dietary excesses with chronic disease risk factors or outcomes—again not nutrition-specific outcomes. 125 Table 2: Current Nutrition Concerns for U.S. Adults LIFE CYCLE STAGE REPRODUCTIVE PERIOD (Girls,Women) DIET and NUTRITION VARIABLES inadequate intakes of energy and other essential nutrients ASSOCIATED HEALTH OUTCOMES poor preconceptual nutrition, poor pregnancy outcomes, poor lactation ADULTHOOD - inadequate intakes of energy and other essential nutrients, drug-induced nutrient deficiencies - excess calories ■ excess sodium - excess smoked or pickled foods excess saturated fat and cholesterol excess saturated fat suboptimal vitamin A* suboptimal vitamin C suboptimal potassium suboptimal fiber suboptimal calcium poor general health and functional status, nutritional anemias - obesity and obesity-related risk of diabetes, heart disease, endometrial and gallbladder cancers - increased hypertension risk - increased hypertension and gastric cancer risk - increased heart disease risk increased breast cancer risk increased lung cancer risk increased esophageal cancer risk increased hypertension risk increased colon cancer risk increased osteoporosis risk * high vitamin intakes have been been associated with prostate cancer (Graham et al., 1983; Enterline, 1984); SOURCES: Food and Nutrition Board, 1980; American Heart Association, 1982; Dietary Guidelines for Americans, 1980; Committee on Diet, Nutrition, and Cancer, 1982; and Willet and MacMahon, 1984; National Institutes of Health, 1984 126 In low-income groups, problems of undernutrition and chronic disease risk may coexist. For example, Trowbridge has pointed out the excess prevalence of both linear growth stunting and obesity among Native American and Hispanic children in the 1982 Centers for Disease Control Surveillance System. The data indicate that the diets of these groups may be quantitatively adequate but qualitatively deficient in some respects. High quality protein or essential vitamins and minerals may be lacking in the diets of these children relative to the level of calories available (Trowbridge, 1983). An additional example of this possible coexistence of dietary inadequacy and dietary excess relates to sodium intake. The addition of salt or other high sodium seasonings (soy sauce, canned chicken broth, bouillon, salt pork) enhance the flavor but not the nutritional value of foods. A dietary pattern may be both nutritionally inadequate and excessive in sodium content. 2.3 Methodological Issues Many diet and nutritional status assessments rely on self-report data or physical measurements which vary widely within each individual from day to day as well as between individuals. Representative measurements of groups may not be strictly repre- sentative of any individual within the group. An additional factor which limits the ability to generalize about nutritional status is the lack of a single dietary or nutrition variable that can be used to indicate risk across all dietary or nutritional dimensions. No one dietary pattern is inherently superior to another. The flexibility, elasticity, and feedback controls inherent in biologic systems permit the attainment of nutritional status equivalence through intakes of a wide variety and combination of foods—as demonstrated by the diversity of nutritionally adequate diets across cultures throughout the world. Nutrition problems identified on the basis of food guides (e.g. the U.S. basic four food groups) or recommended dietary allowances should be interpreted cautiously. Dietary adequacy is best evaluated by a combination of dietary measures and direct physiologic indicators of nutritional status. The appropriateness of applying certain dietary and physiological reference standards which are based on the U.S. population at large to minority populations is also an issue. Some examples of minority-specific nutrition methodology issues follow. a. The precision of dietary assessment methods decreases with increasing dietary diversity or instability; this may pose a particular problem of dietary assessment among minority group members whose food habits are evolving in response to migration or other social and environmental factors. 127 Conventional methods of dietary assessment and evaluation in the United States are based on the combinations of foods and food preparation practices common to the majority. These methods may not reflect important dimensions on which diets of minority groups differ from more typical U.S. diets (Jerome and Pelto, 1981; Sanjur, 1982). Calculated intakes of nutrients do not reflect the foods and combinations of foods from which the nutrients were derived, nor do they reflect certain features of food preparation which can affect biological availability of nutrients or actual nutrient content. For example, the presence of meat or vitamin C in a meal increases overall iron absorption from that meal (Food and Nutrition Board 1980); tea and certain elements in foods may reduce absorption of iron or other minerals; soaking bones in acid solution (a step in some Chinese soup preparation) increases the calcium content of the soup (Suitor and Crowley, 1982); consumption of foods prepared from lime-treated corn adds significantly to dietary calcium and to niacin availability in some Mexican and American Indian cultures (Roe, 1973); use of iron cooking utensils adds to dietary iron; the practice of chewing bones in many cultures provides nutrients which are not reflected in calculations based on fleshy parts of the meat. There is substantial evidence that the distributions of hemoglobin are different for blacks and whites (Owen and Yanochik-Cwen,1977; Meyers et al., 1979; Koh et al. 1980b, Frerichs et al. 1977)—the black distribution is approximately one mg/dl lower—apparently related to the relationship between hemoglobin and transferrin saturation (Meyers et al., 1979). This difference is independent of iron intake. The prevalence of iron deficiency anemia among blacks may be overstated if the "normal" range for hemoglobin levels is calculated from white population data (Owen and Yanochik-Owen, 1977). The applicability of reference data for skeletal and dental development is influenced by racial differences in patterns of skeletal and dental development. As summarized by Knishbacher (1983) black infants tend to be smaller, but developmentally advanced at birth; black boys and girls are taller than their white counterparts during the 2 to 14 year age range, bone density is greater and tooth eruption earlier among black vs. white children (Gam and Clark, 1976; Kramer et al., 1977; Shutte, 1980). Use of white population standards may lead to overestimates of birth weight-associated nutrition risks and underestimates of linear growth stunting. 128 The level of health risk at a given level of exposure to nutrient excess may vary on a genetic basis. For example: blacks may be more sensitive than whites to a given level of sodium exposure due to evolutionary factors favoring renal sodium conservation (Gillum, 1979); a "thrifty gene" favoring the survival of individuals with maximum fat stores during times of food scarcity has been proposed as a factor in the excess prevalence of obesity and diabetes among Pima Indians (Sievers and Fisher, 1981). Some environmental factors and health problems which increase nutritional risk are in excess prevalence among minority groups, e.g., a greater proportion of black than white men are cigarette smokers (National Center for Health Statistics, 1983a; Neaton et al., 1984); alcohol use is more prevalent among some minority sub-groups vs. the general population (Sievers and Fisher, 1981) Strimbu and Sims, 1974; the major chronic diseases which are aggravated by dietary excesses are in excess prevalence among minority groups, e.g., hypertension and diabetes among black Americans; diabetes among Mexican Americans and American Indians (refer to other sections of the Task Force report). Thus, the requirements for "normal" nutrition of minorities may be misspecified by standards designed for populations with different patterns of health-related nutritional risk. Dairy product use, particularly milk drinking, is traditionally lower in non-white populations than white populations after infancy—a pattern thought to be related to the higher prevalence of lactose intolerance in non-white vs. white children and adults after infancy (Paige et al., 1971). However, assumptions about calcium intake based on presumed lactose intolerance should be avoided. Patterns of milk drinking among non-white minority groups indicate that lactose intolerance is not equivalent to milk intolerance; studies among lactose intolerant vs. tolerant children and adults have not always observed significant differences in milk consumption—particularly when the milk is offered in conjunction with subsidized feeding programs programs (Stephenson et al.,1977 Newcomer et al., 1977, Marrs, 1978). In addition, evaluation of the adequacy of calcium intake solely on the basis of dairy product intake should be avoided. Dairy products are the best, but not the only sources of calcium. Traditional calcium sources among non-white populations include tofu (soybean curd) and small bony fish in Asian diets, leafy green vegetables in the diets of southern Blacks, and beans and lime-treated com in Hispanic diets. The density and bioavailablity of calcium in these foods varies; however, since calcium 129 absorption is higher at lower levels of calcium intake and varies with diet composition and other factors (Avioli, 1980), calcium intake may not be a good indicator of calcium status except where dietary deficiencies are extreme. The use of number of "milk group" servings according to the U.S. "basic four" food guide is misleading as a standard for evaluating calcium intake in individuals from traditionally non-milk-drinking cultures. Further, adequacy of a given calcium intake may be different for different ethnic groups. There has been substantial interest in the possible role of chronic low calcium intakes in the incidence of osteoporosis among postmenopausal U.S. women (National Institutes of Health, 1984). However, even if the calcium-osteoporosis association exists in white women, there appear to be other, yet undetermined factors which afford relative protection for some non-white women against osteoporosis. The associated relative risks are substantially higher for white women than for black women, (Farmer et al., 1984) (although not necessarily higher than for all non-white women (Yano et al., 1984)). This is in spite of calcium intakes among black women which are lower than those of white women and which are often below the accepted U.S. standard of dietary calcium adequacy. h. Dietary intake is influenced by socieconomic status. Nutrient intakes are higher at higher levels of disposable income (in the low to middle income range), with the exception of carbohydrate intake, which decreases with increasing income (Adrian and Daniel, 1976). This complicates the separation of culturally- and socioeconomically-based nutritional differences in data which are not income-adjusted. Similarly, income-related differences in access to health care complicate the interpretation of nutritional status measures which are sensitive to medical intervention (e.g. prevalence of low birth weight, dental care). In concluding this background section, it is useful to give a general overview of the factors which determine dietary quality for minority groups within the U.S. environment. This overview (table 3) avoids the redundancy of specifying these factors separately for each minority group, and at the same time draws attention to common pathways of nutritional risk. Also, considering the limited basis for making quantitative judgements on issues of interest, this overview is intended to facilitate qualitative inferences about potential areas of concern. 130 Table 3: A Summary of Dietary Determinants Related to the Nutritional Risk Status of U.S. Minority Groups > culture of origin > length of time in the U.S. > place of birth (native vs. foreign born) > age at migration > family composition > community of residence within U.S. (ethnically similar or dissimilar) > cultural distance from community of origin (frequency of trips to place of upbringing; degree of assimilation into U.S. culture) > changes in social status with migration (relative increase or decrease in income; stability and consistency of income) > changes in activity patterns > changes in social values related to feeding and food (e.g. attitudes towards breast feeding; susceptibility to televised food commercials; abandonment of practices perceived as socially unacceptable (e.g. chewing bones) > practice of geophagia or amylophagia (clay or starch eating), particularly among pregnant women and children in some cultures which may lead to undesirable weight gain (from starch) or other nutritional problems, such as iron-deficiency anemia (Lackey, 1983) > factors related to employment of women before/after migration > nature of traditional food items and meal patterns > similarity between U.S. foods and accustomed food items > shifts in meal frequency or times (number of meals, snacks per day; time of day when heaviest meal is taken) > shifts in proportions of calories from protein, carbohydrate, and fat > shifts in protein sources (animal, dairy, fish, poultry, vegetable sources) > logistical aspects of accustomed methods of food procurement, storage and preparation (daily vs. weekly shopping; requirement for long preparation times) > types of food substitutions made (non-nutritionally equivalent foods within the same food category; manufactured vs. natural forms of food) > flavor and texture preferences and dislikes related to commonly used U.S. foods; > perceived social/health attributes of U.S. foods (health related food beliefs; types of food viewed as "status" foods) > use of vitamin and mineral supplements (discontinued use of traditional supplementary foods; U.S. acquired supplement use practices) > eligibility for and use of food subsidy programs (school lunch, government commodities, elderly meal programs, food stamps, WIC) 131 The greatest quantity and quality of dietary intake and nutritional status data are available for blacks, the ^orl):0£0of whose dietary patterns can be expected to differ least from tnose ~ . the general population. With the exception of sub-groups of the.D^a. population in certain urban areas (e.g., the West Indian conwtun^Y *" New York City, Haitian refugees in Florida) the population of DiacK Americans does not include immigrants from cultures with structurally different dietary patterns. Cultural (rather than socioeconomic; differences between diets of blacks and whites are primarily derived from differences in regional (southeast United States) and religious (more Protestant than Catholic; non-Jewish) influences on food usage and preparation and to a lesser extent from slavery-related dietary adaptations. In contrast, Asians, Hispanics, and Native Americans may retain a carbohydrate-staple structure in their diets which cutters from the meat staple orientation of the diets of U.S. blacks and whites. The literature does not give the impression that minority groups migrating to the United States are from cultures whose dietary patterns are associated with frank vitamin and mineral deficiencies, although quantities of food may have been limiting among certain migrant groups. In general, large differentials between U.S. minority and white population groups can be expected in three areas: 1) differentials related to the low income status of a high proportion of minority individuals and families, including those reflected in lowered growth potential of children from chronically poor cultures; 2) differentials related to the excess prevalence of obesity, particularly among women in certain minority groups; and 3) differentials associated with the lower proportions of refined carbohydrates and saturated fat, and higher sodium content, in the traditional diets of some minority groups, compared to the diets that minority-group individuals may adapt as they become increasingly acculturated. In evaluating diets which include unfamiliar foods or food combinations, care must be taken to differentiate the apparent absence of a food from the absence of a nutrient. Only acute nutritional defiencies or extreme, long-term maladaptations of formerly adequate dietary patterns will result in nutritional risks as they are currently defined and assessed. 132 3. ASIAN AMERICANS 3.1 General Dietary Patterns The Asian American population consists primarily of Japanese and Chinese Americans but also includes Hawaiians, Filipinos, Koreans, Vietnamese, Laotians and Cambodians. Among these are individuals and families who have been in the United States for several generations, first generation immigrants and refugees—the latter in increasing numbers. The dietary staple (i.e., the primary source of dietary calories) for most of these groups is rice, although varieties used and methods of preparation vary. Use of pre-washed or unenriched refined rice poses risks of low B vitamin and mineral intakes, although these risks may be offset by adequate consumption of other foods (e.g. pork and fish). Other starchy foods are also used (wheat and rice noodles, tubers). Seasoned food mixtures, the secondary source of calories, are the major aspect of sub-cultural variation. Intakes of vegetables and fruits, fish and shellfish are higher and the animal protein lower than in typical U.S. diets. Dairy products are used to a much lesser extent among Asians than among the general population. This may pose a potential risk for calcium nutriture in Asian American populations when the traditional sources of dietary calcium in Asian diets (tofu, green leafy vegetables, sardines) are not available or are supplanted by otherwise equivalent but lower calcium density foods. "Status foods", usually foods whose availability or price have been limiting in the traditional diet, may include both nutritious foods (fruits and vegetables, ice cream) and low-nutrient density foods (soft drinks, candy) (Suitor and Crowley, 1984; United States Department of Agriculture, 1980). The short cooking time of many traditional Asian foods is favorable to nutrient retention. Some characteristic food and food choices of several Asian American subgroups are shown in Tables 4 through 7. Adaptations of Asian populations to the U.S. diet inevitably involve increased proportions of calories from animal protein, fat, and refined sugar and decreased calories from complex carbohydrates. These changes also imply decreased intakes of fiber and increased intakes of cholesterol. The higher costs of seafood in the U.S. vs. the home countries of some Asian immigrants may promote shifts toward increased consuption of "meat" protein. The evidence available—most of which relates to Japanese men in Hawaii or California—indicates that dietary changes are reflected in greater relative weight and higher coronary heart disease mortality among Asians in the United States vs. their counterparts in the countries of origin (Tillotson et al., 1973; Hankin et al., 1975; Nomura et al., 1978; Kolonel et al., 1981; and McGee et al., 1982). 133 Table 4: Characteristic Chinese Foods and Food Choices PROTEIN Meat: pork, beef, organ meats FOODS Poultry: chicken, duck Fish: white fish, shrimp, lobster, oyster, sardines Eggs Legumes: soybeans, soybean curd (tofu), black beans Nuts: peanuts, almonds, cashews MILK AND MILK Flavored Milk, milk (cooking), ice cream PRODUCTS GRAIN PRODUCTS Rice, noodles, white bread, barley, millet VEGETABLES Bamboo shoots, beans: green and yellow, bean sprouts bok choy, broccoli, cabbage, carrots, celery, Chinese cabbage, corn, cucumbers, eggplant, greens: collard, Chinese broccoli, mustard, kale, spinach, leeks, lettuce, mushrooms, peppers, potato, scallions, snow peas, sweet potato, taro, tomato, water chestnuts, white radishes, white turnip, winter melon FRUITS Apple, banana, figs, grapes, kumquats, loquats, mango, melons, orange, peach, pear, persimmon, pineapple, plums, tangerine OTHER Soy sauce, sweet and sour sauce, mustard sauce, ginger, plum sauce, red bean paste, tea, coffee SOURCE: California Department of Health, 1975, Table 11 134 Table 5: Characteristic Japanese Foods and Food Choices PROTEIN Meat: beef, pork FOODS Poultry: chicken, turkey Fish: tuna, mackerel, sardines (dried form called mezashi), sea bass, shrimp, abalone, squid, octopus Eggs Legumes: soybean curd (tofu), soybean paste (miso), soybeans, red beans (azuki), lima beans Nuts: chestnuts (kuri) MILK AND MILK Milk (in cooking), ice cream, cheese PRODUCTS GRAIN PRODUCTS Rice, rice crackers, noodles (whole wheat noodles called soba), spaghetti, white bread, oatmeal, dry cereals (Nisei only) VEGETABLES Bamboo shoots, bok choy, broccoli, burdock root, cabbage, carrots, cauliflower, celery, cucumbers, eggplant, green beans, gourd (Kampyo), mushrooms, mustard greens, Napa cabbage, peas, peppers, radishes (white radish called daikon; pickled white radish called takawan), snow peas, spinach, squash, sweet potato, taro (Japanese sweet potato), tomato, turnips, waterchestnuts, yam FRUITS Apple, apricot, banana, cherries, grapefruit, grapes, lemon, lime, melons, orange, peach, pear, persimmon, pineapple, pomegranate, plums (dried, pickled, plums called umeboshi), strawberries, tangerine OTHER Soy sauce, nori paste (used to season rice), bean thread (konyaku), ginger (shoga; dried form called denishoga), tea, coffee SOURCE: California Department of Health, 1975, Table 12 135 Table 6: Characteristic Filipino Foods and Food Choices PROTEIN Meat: pork, beef, goat, deer, rabbit, variety meats FOODS Poultry: chicken Fish: sole, bonito, herring, tuna, mackerel, crab mussels, shrimp, squid, Eggs Legumes: black beans, chick peas, black-eyed peas, lentils, mung beans, lima beans, white kidney beans Nuts: cashews, peanuts, pili nuts MILK AND MILK PRODUCTS Milk: flavored, evaporated Cheese: gouda, cheddar GRAIN PRODUCTS Rice, cooked cereals: farina, oatmeal, dry cereals, pastas, rice noodles, wheat noodles, macaroni, spaghetti VEGETABLES Bamboo shoots, beets, cabbage, carrots, cauliflower, celery, Chinese celery, eggplant, endive, green beans, leeks, lettuce, mushrooms, okra, onion, peppers, potato, pumpkin, radishes, snow peas, spinach, squash, sweet potato, tomato, water chestnuts, watercress, yam FRUITS Apple, banana, grapes, guava, lemon, lime, mango, melons, orange, papaya, pear, pineapple, plums, pomegranate, rhubarb, strawberries, tangerine OTHER Soy sauce, coffee, tea SOURCE: California Department of Health, 1975, Table 13 136 Table 7: Examples of Some Southeast Asian Foods Vietnamese - soup—"pho"—contains rice, noodles, thin slices of beef or chicken, bean sprouts, and greens - fish and/or meat and vegetable dishes - clear soup with vegetables and/or meat - fish sauce—"nuoc mam"—made from fermentation of small fish Cambodian - soup with meat and noodles, and/or rice - fermented fish—"prahoc" - fish sauce—"tuk-trey" - sweets made from palm sugar Laotian - fish—"padek"—and/or meat stew with hot peppers - sweet or glutinous rice (sticky rice) SOURCE: United States Department of Agriculture, 1980 Many Asian-American foods and seasonings are very high in sodium. Examples include bean sauces, dried shrimp, dried salted fish, pickled vegetables, fish sauce, monosodium glutamate, soy sauce, miso, salted eggs (Chew, 1983; Asian/Pacific Islander Task Force on High Blood Pressure Education and Control, 1984; Suitor and Crowley, 1984). Thus, dietary acculturation among Asian Americans may reduce the level of sodium exposure, and thereby lower hypertension risk. Sodium restriction may be particularly beneficial for Asian Americans who are hypertensive, but is difficult to accomplish within the framework of traditional Asian food practices (Chew, 1983; Asian/Pacific Island Task Force of High Blood Pressure Education and Control, 1984). Decreased use of salted fish and pickled vegetables may also reduce the risk of certain cancers (Kolonel, 1981; Yu, 1981). 137 Within the Chinese population, food preparation practices differ according to Mandarin (north), Shanghai (central), or Cantonese (south) influences (Sanjur, 1982). Sanjur (1982) notes two ideolgical themes which underlie Chinese food habits: the Fan-Ts ai principle which balances the proportions of starchy staple foods (Fan) with meat and vegetable components of the diet (Ts'ai) but which considers the Fan component to be the more indispensible. (Sanjur, 1982). An additional principle in Chinese food practices strongly advises against overindulgence (Sanjur, 1982). This implies a relatively low risk of obesity. However, older Chinese may associate excess weight with wealth and prosperity, so that obesity, if it develops, may be considered acceptable (Asian/Pacific Islander Task Force on High Blood Pressure Education and Control, 1984). The Yin-Yang principle is based on beliefs about the inherent physiologic significance of foods and, as Sanjur points out, is analagous to the hot-cold dichotomy in Hispanic food culture (see section 5). Ludman and Newman (1984) give some examples of yin, neutral and yang Foods as follows. Yin foods include bland foods, boiled foods, cold foods (thermal), some types of fruit, some types of fish, pork, most greens, and white foods—including milk. Yang foods include fatty meats, hot foods (thermal), spicy foods, fried foods, chicken, beef, and eggs. Noodles, soft rice, sugar, and sweets are neutral. Health conditions and body organs are also dichotomized according to the Yin-Yang principle. Pregnancy and lactation, two periods of high nutritional risk, are Yin conditions and would theoretically be characterized by de-emphasis of Yang foods (which include high protein foods). However, responses of Chinese-American women to a 1980 survey of health-related food practices indicates that the Yin-Yang principle is not closely followed during pregnancy (Ludman and Newman, 1984). Surveys of dietary transitions among Chinese Americans in New York City (Sanjur, 1982), Lincoln, Nebraska (Yang, 1979), and California (Grivetti and Paquette, 1978) indicate that deviation from charateristic Chinese food patterns increases with increasing years of residence in non-Chinese environments. However, the American foods which are accepted or rejected for reasons of ideology or preference are difficult to predict and may be very specific to interactions between community of origin and community of residence. In addition, patterns of ethnic food use are changing in the home countries of recent Asian immigrants (Grivetti and Paquette, 1978). Several reports provide specific insights into the dietary changes among Vietnamese immigrants. Stoner and Grivetti (1978; cited in Waldman et al., 1979) noted that westernization of traditional food customs in Vietnam began in the 1940's. Thus, some non-traditional dietary practices may precede migration to the U.S. These authors noted that Vietnamese refugees in California adopted a compromise of following U.S. dietary practices during the week and returning to more traditional Vietnamese diets during weekends. 138 In a study of Vietnamese refugees in Florida, 64% of those interviewed reported increases in weight since coming to the United States and had significantly increased their consumption of milk and soft drinks, as well as many other foods (Crane and Green, 1980). Nguyen et al. (1983) compared food habits and preferences of Vietnamese children less than 6 years old who had come to the United States more than or less than a year before being interviewed. The children who had been in the U.S. longer than one year consumed more peanut butter and more sweets (ice cream, milk shakes, and pies) than those who had come more recently, but both groups of children retained a preference for fruits as snacks. 3.2 Nutritional Risk 3.2.1 Infants, Children, and Adolescents Stoner and Grivetti (1978; cited in Waldman et al., 1979) reported that breast feeding was negatively viewed by Vietnamese women in their California study sample, related to the widespread practice of bottle feeding in Vietnam, particularly in urban areas. However, a United States Department of Agriculture review (1980) states that the majority of Southeast Asian infants are breast fed. The CDC pregnancy nutrition surveillance data, which include breast feeding data for a subset of postpartum women participating in the reporting WIC and Maternal and Child Health Service populations, did not include this variable for Asian American women. No other recent data on infant feeding patterns among Asian American women were identified. Thus, there is not a sufficient basis for making any general statement regarding breast vs. bottle feeding patterns or possible associated nutritional risks among Asian Americans. The CDC-PNSS data indicate an excess prevalence of linear growth stunting among the Asian American children in the data base, both in comparison to national standards and in comparison to the prevalence of growth stunting among white children in the CDC population. The latter is a comparison within a population of low-income children receiving publicly-supported services. Among Asian-American children in the CDC data base who were less than 2 years of age, the prevalence of low height for age was 8 to 13% in 1977-78 and 17 to 20% in 1979-81 compared with the 5% expected (i.e., below the 5th percentile) and compared with prevalences of approximately 9% among white children for the entire 5 year period. Among children ages 2 to 5 the excess prevalence of growth stunting was more pronounced—21% in 1977 and 33-37% during 1978-81 vs. 5% expected and 9% in the white children in this data base (Centers for Disease Control, 1983). Neither low weight for height (thinness) nor high weight for height (obesity) were in excess prevalence among Asian American children in the CDC data for these years. No clear pattern of excess risk was evident in the hemoglobin and hematocrit data. 139 3.2.2 Adults As noted earlier, transition from a traditional Asian to a more characteristic U.S. diet may be associated with increases in some aspects of chronic disease risk related to higher saturated fat and lower complex carbohydrate consumption, and with decreases in hypertension risk related to lower sodium consumption. Some reports of inadequate diets among Asian American were noted. Three such studies are summarized below, primarily as examples of the types of data which have been reported concerning the nutritional status of Asian Americans. There are insufficient data on which to base any general conclusions about undernutrition among Asian American adults. Casey and Harrill (1977, cited in Waldman et al., 1979) reported on nutrient intakes calculated from 24 hour recalls of Vietnamese women relocated to Colorado. Diets of women ages 22 to 48 were below the U.S. standards for calcium, iron, and zinc but were otherwise adequate. Only protein and ascorbic acid intakes were adequate for women ages 51 to 65. No nutrient deficiencies were identified by Stoner and Grivetti (1978, cited in Waldman et al., 1979) in their study of Vietnamese refugees in California. The majority of respondents in that study supplemented their food intakes with produce from home vegetable gardens. Kim et al. (1984) reported intakes of calcium, selected other nutrients, and preferences for calcium-rich foods among 40 elderly Koreans in Chicago. The respondents were between the ages of 65 and 81 and were tested to assure mental competence. Calcium intakes were low (below 2/3 of recommended dietary allowances) for a third of the men and two thirds of the women, although slightly higher than reported calcium intakes in Korea. Food preference data indicated that milk and ice cream were liked by more than 60% of the subjects but two thirds of the subjects either disliked or had never tried cheese and yogurt. Calcium-rich Korean rich foods were liked by a majority of subjects but were infrequently consumed due to cost and availability. Other dietary components thought to have an antagonistic effect on calcium nutriture, (e.g., intakes of animal protein) were higher than levels in Korea. Thus, the net calcium status of the Korean Americans may have been worse than if they had been consuming a traditional Korean diet. However, the absence of any data demonstrating calcium-related health problems in this population limits the significance of these findings. Lewis and Glaspy (1975) obtained 3 day food records for 47 college-educated Filipino women in Los Angeles during 1971. Although the diets of many of these women contained Filipino foods, significant dietary changes indicative of U.S. dietary patterns were reported. The calculated nutrient intakes of these women were quite variable; many were below both U.S. and Filipino standards. Intakes of some of the women, including two pregnant subjects, were augmented by appropriate vitamin and mineral supplements. 140 On the basis of the limited data available on the nutritional status of Asian Americans, the most evident area of current concern is the growth status of Asian-American children. This problem presumably relates to the children of recently-migrated Southeast Asian refugees rather than to Asian-American children in general. The extent to which this excess risk is socioeconomically determined, is due to constitutional factors affecting growth potential, derives from transitory, migration-related nutritional maladaptations, or is a function of the growth standards used to specify risk cannot be determined. Additional dietary factors which affect the Asian American population at large are those related to chronic disease. These include elements inherent in certain traditional Asian foods (e.g.,highly salted and pickled foods) and elements acquired as a result of acculturation to conventional U.S. eating patterns (e.g., saturated fat and refined carbohydrate). The coronary heart disease risk factors are of relatively recent acquisition and may be of a lesser concern now than they will be in the future. 141 4. NATIVE AMERICANS 4.1 General Dietary Patterns Native Americans include both American Indians and Alaska Natives. Very little information was identified on the diet and nutritional status of Alaska Natives. An impression of Alaskan Eskimo food patterns is provided in the following paragraph. The remaining material in this section relates to American Indians. The following comments on dietary patterns of Alaskan Eskimos are based on reviews by Gonzalez (1972) and Sanjur (1982). Artie and Subartic food patterns are characterized by extremely high caloric intakes. Primitive Eskimo diets were high in protein, of moderate or high fat content, limited in carbohydrate content, and were low in ascorbic acid (vitamin C) during some seasons. Protein sources varied with region and were either sea mammals, fish, or land mammals. Eskimos live at a very low economic level. Many in rural areas have unstable incomes. Food supplies are most consistent during the winter months when the freezing temperatures permit long-term storage of excess game. Current diets consist of game and store bought items. Important staples include homemade yeast bread, fried breads and pilot crackers. Large quantities of unenriched rice and pancake mixes are used. Wild berries are widely used; wild greens are important in some remote villages. Child feeding practices include premastication of infant foods, use of canned and evaporated milk, increasing use of commercial infant foods; widespread use of dry pre-cooked baby cereals, artificially flavored powdered drinks, and soft drinks. American Indians derive much of their diets from the hunter- gatherer diets of early America. Tribal dietary patterns vary according to geographic region, cultural, and economic traditions. For example, diets of Indians living along the northern coast of California include a variety of seafood, whereas diets of Indians in mountainous areas include more dried meats and fresh water fish (California Department of Health). Dietary patterns of Indians in the Southwest may be more similar to those of Mexican Americans than to those of Indians from other tribal and regional origins. The current diets of the various Indian tribes in the United States are a blend of their traditional foods with the American food culture, except for special occasions. For example, Indians in the Southwest once ate tortillas made from corn, but manufactured wheat tortillas have replaced handmade corn tortillas in recent times. In addition, factors such as income, the availability of foods, particularly U.S. commodity foods, stage of acculturation, and place of residence are important dietary influences. 142 Although there is probably much more heterogeneity than homogeneity in American Indian diets overall, most Indian diets appear to be characterized by a high carbohydrate, high sodium, and high saturated fat content and relatively low content of meat and dairy products. Many Indian dishes are made of starch and meat combinations and many foods are fried and breaded. Consumption of refined sugar products is high. Cholesterol consumption may be low to moderate, depending on the protein sources. Examples of traditional and contemporary American Indian Foods are shown in Table 8. Based on a laboratory analysis of traditional foods of Hopi Indians (southwest) Kuhnlein et al, 1979 concluded that nutrient needs would be met by the Hopi diet if all the quantities of food available were adequate and if all nutrients consumed were absorbed. Toma and Curry (1980) calculated the nutrient content of seven traditional recipes of South Dakota Indian tribes. Five of the seven recipes met the criterion of a nutritious food. The food item of least nutritional value—fried bread—is the only one of the seven which has been retained in the customary diet. Table 8: Examples of Traditional and Contemporary Foods in the Diets of American Indians - hominy (boiled corn) - pinto beans - corn balls - mutton, goat - corn meal - deer - corn tortillas - salmon - wheat tortillas - smoked salmon eggs - oven bread - refried beans - fried bread - lard and bacon fat back - piki bread - small game - cooked cereals - dried meats - dry cereals - fresh water fish - wild rice and nuts - seafood - dried melon and peach strips - wild fruits, berries, and vegetables - corn, beans, squash, melons - potatoes, onions, cabbage - coffee,tea, milk, soft drinks - sweet rolls, cakes, doughnuts SOURCES: Gonzalez, 1972; California Department of Health, 1975; Alford and Nance, 1976; Kuhnlein et al., 1979; Sanjur, 1982 143 Dietary recalls of Eastern Band Cherokee Indian WIC P31^101^^ interviewed in 1978 indicated consumption of several traditional rooa items. Sixteen to 18% of the dietary recalls indicated co^umPt^n8% of cornbread, hominy or boiled corn, pinto beans, and greens; 4 to ■ indicated consumption of fatback, green beans, and fry bread. Participation in WIC was reported to influence both the general family food consumption patterns and the child feeding practices (Slonim et al., 1981). Food intake patterns of Indians on the Standing Rock ^sef^f4°n in North and South Dakota, assessed by Bass and Wakefield (1974) in 1970, did not revolve around a typical menu or meal except tnat coffee consumption was high and some type of breakfast was usually consumed. Consumption of characteristic white middle class foods—including bologna, potato chips and carbonated Averages—was common among those with adequate incomes, while traditional foods were served only on special occasions. Commodity foods were usea oy the majority of families interviewed. Frying and boiling (the traditional method) were used equally. Diets of Indians in the Onondaga Nation (upstate New York) are characterized by corn and legumes as protein sources, high fat-high sodium traditional foods and low intakes of meats, eggs and dairy products (Simpson, 1982). High consumption of refined sugar products (e.g., candy, cakes, pies, soda, Kool-Aid) and a predominance of fried and breaded foods has been noted among Indians of the St. Regis-Mohawk Nation (Rhoades, 1982). 4.2 Nutritional Risk 4.2.1 Infants and Children The Centers for Disease Control Pregnancy Nutrition Surveillance data for 1981 (Centers for Disease Control, 1983) indicated a high prevalence of breast feeding among Native American women at the post-partum follow-up visit: 44 percent of Native American vs. 34% of white mothers under age 20 were breast feeding; 47% of Native American mothers over age 20 were breast feeding, a percentage less than the 53% of white mothers but greater than the 30 and 37% of Hispanic and black mothers who were breast feeding. Several studies in specific Indian populations support the impression that a large proportion of Native American mothers breast feed, and for fairly long periods, although a public health nutritionist serving the St. Regis-Mohawk nation in New York State noted that women in her service population were exceptions in this respect—having a relatively low incidence of breast feeding (Rhoades, 1982). A 1978 study showed a beneficial effect of breast vs. bottle feeding on the risk of gastroenteritis, as observed among all Pima infants under 4 months of age and also among infants of mothers under age 20 (Forman et al., 1984). Thirty five percent of White Mountain Apache preschool children (Arizona) studied in 1976 were reported to 144 have been breast fed—15% for six months or longer and 25% for at least three months. Whereas cow milk was the usual supplementary food for breast fed infants, formula fed infants were more likely to be supplemented with evaporated milk (Yanochik-Owen et al., 1977). Sixty-six percent of Eastern Band Cherokee Indian women participating in a North Carolina WIC program in 1978 reported having breast fed their youngest child— 50% for six months or longer—although the majority of these women supplemented with bottle feeds from the first weeks of life and introduced solid foods within the first three months (Slonim et al.,1981). From the 1982 CDC-PNSS data, Trowbridge (1983) reported excess prevalence of linear growth stunting (12.4% below the 5th percentile of height for age) among Native American one-year-old children but not among children less than one year or 2 to 4 years. Cross- sectional data for Native American children in the CDC data base during 1977 through 1981 indicated a similar pattern but to a lesser degree. The prevalence of growth stunting in Native American children during 1977-1981 was not markedly greater than the prevalence among white children in the CDC data base. Wasting (low weight for height) and low hemoglobin were not in excess among Native American children (Centers for Disease Control, 1983). Yanochik-Owen et al. (1981) reported that three times as many White Mountain Apache children surveyed in 1976 were receiving vitamin supplements as a similar group of children surved in 1969 (15% vs. 5%). Overall energy and nutrient intakes by 24 hour dietary recall were higher in the 1976 vs. the 1969 sample but were below critical thresholds for energy, calcium, vitamin C, and iron for substantial proportions (16 to 38%) of children in 1976. Deficiencies in some areas were probably compensated by contributions from supplements. Although clinical signs of prevalent undernutrition were not found, anthropometric measures did not reflect the apparent improvement in dietary quality, possibly because of the short intervening time interval. Percentages of children below the 10th percentile of height for age were nearly three times greater than expected in both 1976 and 1969 (39 and 37%). There were no statistically significant differences in height or weight measures between 1969 and 1976, although slightly higher thoracic fatfold thicknesses were implied in the 1976 data. The prevalences of low hemoglobin and hematocrit were not high, but the percentages of children in the 1976 cohort with low transferrin saturation and serum ferritin levels (53% and 22% of children for the two measures, respectively) suggested low iron stores—indicative of inadequate long term intakes of dietary iron. Undernutrition—defined by clinical signs, hemoglobin levels, height for age and urinary hydroxyproline to creatinine ratio—was not prevalent among the Chippewa Head-Start children in Wisconsin studied by Homer et al. (1977). Protein intakes were well above the PDA (172%); dietary iron intakes were adequate (80% of the RDA). 145 The prevalence of lactase deficiency among 156 Chippewa children ages 5-17 and adult volunteers on the Leech Lake Reservation in northern Minnesota varied from 62 to 72% and did not vary within tne age ranges studied. Lactase deficiency was almost universal among full-heritage Chippewa and was least prevalent (around 33%) J^g those with less than 50% Indian heritage. However, milk intolerance and milk consumption were only slightly lower among the lactase- deficient vs. the lactase-normal subjects studied, ^yj-0* or children reported milk intolerance (Newcomer et al., 1977). An excess prevalence of overweight was observed among Native American children in the CDC-PNSS in 1982 (13.2, 12.0, and 10.4% of children ages 1,2,and 4, respectively, above the 95th percentile ot weight for height) (Trowbridge, 1983). Percentages of overwe19** children less than one year old and 3 years old were somewhat greater than expected as well (7 and 8.2% respectively). Trends during 1977 to 1981 indicated a prevalence of high weight for height among Native American children which was consistently higher than expected and higher than the prevalence for white children in the sample, particularly in the 2 to 5 (vs. less than 2) year age group. Excess prevalence of obesity was reported among Chippewa preschool children in Wisconsin (Homer et al., 1977). Thirty percent of boys were above the 90th percentile of weight for height; 19% of girls and 37% of boys were above 110% of the NCHS reference standard of weight. Percentages of children with skinfold thicknesses above the 90th percentile were 35% and 37% for girls and boys, respectively. 4.2.2 Adults In a study of pregnant and lactating Navajo women, Butte and colleagues (1981) found the median nutrient intakes of these women to be less than 60% of the RDA for calcium, magnesium, zinc, copper, vitamins A, D, E, B6, biotin, and folacin. Anemia was present in 15-20% of the women. The authors note that the total energy intake of these women was only 74% of the RDA, which is inadequate for pregnancy and lactation. However, it should be noted that 60% of 1200 mg/day, the calcium RDA for pregnant and lactating women, is 720 mg/day, nearly equal to the 800 mg/day RDA for non-pregnant women. Their diets may actually provide adequate calcium and other nutrients for non-pregnant women. A 1970 survey indicated that dietary intakes of Indian people (primarily Sioux) on the Standing Rock Reservation in North and South Dakota were adequate (A 67% of the RDA's) or in excess of the RDA's for most nutrients. Calcium intakes and iron intakes of women of reproductive age were substantially below recommended levels and more than half of the women interviewed reported calcium and vitamin A consumption below 50% of the RDA (Bass and Wakefield, 1974). Actual iron intakes may have been higher than calculated due to the customary use of iron cooking utensils. 146 Although obesity was reportedly uncommon among Native Americans before the 1940's (West, 1974), documentation of the excess prevalence of obesity among Native Americans in all parts of the country is now frequently encountered in the literature. Johnston et al. (1978) have observed that urban Native Americans are at higher risk for obesity than those who remain on reservations. Gillum et al. (1980;1984) have reported an excess prevalence of obesity (vs. white comparison populations) among Native American (primarily Chippewa) school children and adults in Minnesota in 1978 and 1980-81. Studies dating back to the 1960's have noted an excess of obesity among various Indian populations—White Mountain Apaches (Clifford et al., 1963), Navajo (Fulmer and Roberts, 1963; DeStefano et al., 1979, Chase-the-Bear et al., 1979), Seminoles in Oklahoma (Mayberry and Lindeman, 1963), Sioux Indians in North and South Dakota (Bass and Wakefield, 1974) and Seneca Indians in New York State (Doeblin et al., 1969; Judkins, 1978). Obesity among the Pima Indians has received considerable attention in conjunction with the extremely high prevalence of adult onset diabetes mellitus and the apparently lower health risk among Pimas at levels of obesity usually considered severe in the general U.S. population (Pettitt et al., 1982). A higher than average prevalence of adult onset diabetes has also been reported for Indian tribes other than Pimas (Mayberry and Lindeman, 1963; Doeblin et al., 1969; Judkins, 1978; Gillum et al., 1984) Alcohol use is high in many Native American communities. According to a 1960-65 study by Sievers (1968), heavy alcohol use was more common Among Southwestern Indians than among Indians outside of the Southwest but was still higher in both groups of Indians than in a white comparison population. Strimbu and Sims (1974) reported that American Indian high school students in a Georgia survey reported levels of alcohol consumption higher than that of students in most other ethnic groups. Forty percent of Navajo men under age 40 reported using alcohol in a 1977 study, compared with only 22% of the men over 40—a possible sign that alcohol use is increasing among the younger Navajo men (DeStefano, 1979). Alcohol intake was reportedly low among the Chippewa Indians studied by Gillum et al. (1984) in Minneapolis in 1980. 4.3 Conclusions In summary, even with the limited amount of data available, the need for attention to the problems of obesity and obesity-related diabetes in Indian populations is evident. Other dietary risk factors for ischemic heart disease may become important for this population in the future. In addition, the excess prevalence of both growth stunting and obesity among low-income Native American children may indicate poor dietary quality in some elements critical for growth. Alcohol-related nutritional risks are also worthy of attention in this population. 147 5. HISPANIC AMERICANS 5.1 General Dietary Patterns Among the Spanish-speaking U.S. minority groups, the largest proportion are Mexican Americans. Puerto Ricans are the second largest group. There are substantial numbers of Cuban Americans as well. Studies of Hispanic Americans may refer to one or several of the following groups: migrant workers, recent immigrants from Mexico, long term U.S. residents of Mexican descent, Puerto Ricans in Puerto Rico, Puerto Ricans on the U.S. mainland, Cuban immigrants and their descendants, or individuals of other Spanish or Latin American origins (i.e., with Spanish surnames). The socioeonomic status and degree of acculturation vary greatly among these groups, as do the referent cultures. The information in this section relates primarily to Mexican Americans and Puerto Ricans, due to the relatively few available data on Cuban Americans. Some characteristic Mexican American and Puerto Rican foods are shown in Tables 9 and 10. Mexican American and Puerto Rican diets are similar in some respects, but there are underlying differences associated with the Southwestern vs. Caribbean origins of these two Hispanic cultures. Both Mexican American and Puerto Rican diets have carbohydrate staples (tortillas and/or rice). Consumption of green leafy vegetables is limited. Vegetable protein sources are given more emphasis in traditional Hispanic diets than are animal proteins. Consumption of milk and dairy products is low in traditional Hispanic diets, but is not absent. For example, characteristic foods include ice cream, custard (flan), pudding (e.g., arroz con leche), and coffee with milk (cafe con leche, oatmeal cooked with milk (avena) (California Department of Health, 1975; Sanjur, 1982). The more conventional uses of dairy products are increasing in association with participation in subsidized food programs such as WIC and School Lunch (Yanochik-Owen and White, 1977; Lieberman, 1979). The food selections and preparation practices in Mexican American and Puerto Rican diets can potentially provide all essential nutrients. At one time the food patterns of New York City-based Puerto Ricans were publicly criticized as being inherently deficient in folacin and therefore predisposing to a widespread prevalence of megaloblastic anemia (Sanjur, 1982). A study of Puerto Rican foods and food preparation methods indicated that the folacin in the typical diet is adequate (Parker and Bowering, 1976). However, the Hispanic diet is bulky and consists of more moderate nutrient density than high-nutrient-density foods. Adequate quantities of nutrients will be obtained if a sufficient quantity of foods is consumed. If food quantity is limited, e.g. by income, then intakes of essential nutrients may be marginal or low. 148 Table 9: Characteristic Mexican-American Foods and Food Choices PROTEIN FOODS Meat: Poultry: Eggs Legumes: Nuts: beef, pork, lamb, tripe, sausage (chorizo), bologna, bacon chicken pinto beans, pink beans, garbonzo beans, lentils peanuts, peanut butter MILK AND MILK PRODUCTS Milk: fluid, flavored, evaporated, condensed Cheese: American, Monterey jack, Hoop Ice cream GRAIN PRODUCTS Rice, tortillas: com, flour, oatmeal, dry cereals: cornflakes, sugar coated; noodles spaghetti, white bread, sweet bread (pan dulce) VEGETABLES Avocado, cabbage, carrots, chilies, com, green beans, lettuce, onion, peas, potato, prickly pear cactus leaf (nopales), spinach, sweet potato, tomato, zucchini FRUITS Apple, apricots, banana, guava, lemon, mango, melons, orange, peach, prickly pear cactus fruit (tuna), zapote (or sapote) OTHER Salsa (tomato-pepper onion relish), chili sauce, guacamole, lard (manteca), pork cracklings, fruit drinks, Kool-aid, carbonated beverages, beer, coffee SOURCE: California Department of Health, 1975, Table 9 Mexican American and Puerto Rican diets are generally favorable from a chronic disease risk perspective. Fiber content is high. Animal fat content is an important component of Hispanic diets but its proportional contribution to calories is less than in typical U.S. diets. The high carbohydrate content does pose a potential risk of excess caloric intake if the meal pattern shifts from the tradition of taking a heavy meal at midday to the U.S. practice of having the heaviest meal in the evening—particularly if the heavy evening meal is added to, rather than substituted for, the midday meal. 149 Table 10: Characteristic Puerto Rican Foods and Food Choices PROTEIN FOODS Meat: Poultry: Fish: Eggs: Legumes: beef and pork more often than other meats; pig intestines, either fried (cuchifritos) or stewed (sancocho); organ meats; blood sausage (chorizo); ham butts, sausage to flavor certain dishes chicken in limited amounts; salt codfish is most common fried and in cooking white, kidney, pink, beans; pigeon peas, chick peas MILK AND MILK PRODUCTS Milk: in limited amounts, e.g. in coffee; in oatmeal Cheese: native white cheese (resembling farmer cheese; gouda cheese, American cheese GRAIN PRODUCTS Rice, French bread, rolls, crackers, other bread; breakfast cereals: oatmeal (avena), cornmeal, and cornflakes; farina VEGETABLES Yautia, name, plantain, pumpkin, carrots, sweet potatoes, tomatoes, onions, lettuce, cabbage, potatoes; malanga, yuca (root vegetables); white and yellow yams, eggplant FRUITS bananas, pineapples, guava, oranges, papaya, acerolas (West Indian cherry), mango OTHER sofrito (sauce with green pepper, tomato, garlic, lard); lard and salt pork; olive oil and other vegetable oils; sweetened beverages, cakes, pies; guava, orange, and mango pastes, boiled papaya preserves; fruit cocktails, canned fruits; pear, peach, and apricot nectars; canned soups; black malt beer; annato seeds (used for coloring) SOURCE: New York City Department of Health, 1976a; 1981; Sanjur, 1982 150 Sausages and some of the sauces in Mexican American and Puerto Rican diets are high in sodium (Suitor and Crowley, 1984). To the extent that these high sodium foods or seasonings are frequently consumed, sodium-related risks in the traditional Hispanic diet may be similar to those in the general U.S. population. Higher than average sodium consumption among Mexican Americans was suggested by Kerr et al. (1982). These authors found that table salt sales in Hispanic neighborhoods in Houston, Texas were twice as high as those in white neighborhoods. Dietary practices of Hispanic Americans may be influenced by health-related food beliefs derived from the hot-cold dichotomy of body organs, diseases, and other aspects of life. According to Sanjur (1982) the basic premises of this ideology are that a person's normal, healthy state is temperate and that in order to maintain this healthy state the individual must balance heat and cold. The designations assigned to particular food and beverage items as hot, cold, or temperate may vary within Hispanic subcultures. Galli (1975) gives the following categorization for Puerto Ricans: cool foods—whole milk, barley water, chicken, raisins, most fruits, and honey; cold foods—avocado, bananas, coconut, lima beans, and sugar cane; hot foods—evaporated milk, vitamins, iron supplements, calcium pills. The effect of hot-cold food categorizations on weaning practices and on the food intakes of pregnant and lactating women is an important nutritional risk consideration. According to Galli (1975), avoidance of cold foods is advised for 40 days after birth. Similary, Sanjur (1982) cites studies circa 1960 indicating that some Mexican-American pregnant women excluded all fruits and vegetables for 30 to 40 days or for the entire lactation period. The current influence of "hot-cold" food classifications on food intake among Hispanic women cannot be estimated. The impression given by a 1979 report of Lieberman is that such beliefs have been integrated with or replaced by more modern "scientific" beliefs; however, that study population was apparently characterized by more rapid acculturation than is usual among mainland Puerto Rican communities. 5.2 Nutritional Risk 5.2.1. Infancy and Childhood Several relatively recent studies of infant feeding patterns among Hispanic Americans were identified. These studies indicate considerable differences between Hispanics and white or black comparison groups in the incidence and duration of breast feeding, in the determinants of a decision to breast feed, in the types of supplementary or substitute milks used in bottle feeding, in the timing of introduction of solid foods (which is related to breast feeding patterns), and in types of solid foods used (types of commercial foods used and relative proportions of commercial and table foods given at a particular age). 151 Several studies have indicated a low incidence of breast feeding among Hispanic-American women. Among a subset of postpartum women in the CDC-PNSS data base in 1981, the incidence of breast feeding was 21% among women under age 20 and 30% among women over age 20, compared with 34 and 53% among whites. A study of infant feeding practices among 28 Cuban, 28 Puerto Rican and 20 Anglo families in a Dade County, Florida, Maternal and Infant Care program reported that the incidence of breast feeding was higher among Anglos than among Latin women (Bryant, 1982). Only 27 of 76 women breast fed at all; 35% of Anglos breast fed, 10% of Puerto Ricans, and 12% of Cubans. Low incidence of actual or planned breast feeding among Mexican American women has also been reported by Seger et al. (1979), Magnus (1983), Smith (1982), and Rassin et al. (1984). The trend data for 1971-79 reported by Smith et al. (1982) indicated that the increased breast feeding incidence seen among Anglo women is not evident among Mexican Americans. The low breast feeding incidence among Hispanic American women is influenced, but not entirely explained, by socioeconomic factors. Bryant (1982) reported that the Cuban and Puerto Rican women studied in Florida changed to bottle feeding after 2-6 weeks—using condensed, evaporated, and whole cow's milk in addition to commercial formula. Fifty percent of both Hispanic and Anglo women added sucrose to the formula. Ferris et al. (1978b) reported that in a sample of infants in Western Massachusetts, intakes of iron fortified formulas, cereal, fruits and vegetables were lower among Hispanic infants than among black or white infants. Hispanic infants were more likely than black and white infants to be fed fruit juice, as well as vegetable-meat mixtures. Among Puerto Ricans in East Harlem before the introduction of the WIC program, Bowering and coworkers (1978) found that Puerto Ricans used whole cow's milk at 3 months while blacks used formula. Neither group breast fed their infants. Introduction of solid foods was early (at 1 month). Between 9-15 months, Puerto Ricans used baby foods for 30% of energy intake, while blacks used baby foods to provide only 5% of energy intake (table food provided most of the snacks). Puerto Ricans used more juices than blacks. Meat was used by 6% of Puerto Ricans at 3 months; no blacks used meat at 3 months. More Puerto Ricans than blacks used cereal. Vitamin supplements were given to 75% of infants in this sample; iron supplements were given to 90%. These studies suggest a less than optimal prevalence of the currently recommended infant feeding practices in Hispanic communities. The effects of expanded WIC program availability on infant feeding patterns is undoubtedly positive, but some problems in the area of infant feeding may persist. 152 CDC-PNSS data for 1977-81 and 1982 (Centers for Disease Control, 1983; Trowbridge, 1983) for children less than 2 years of age do not show an excess prevalence of low weight for height or low height for age among Hispanic children, compared with reference standards or with white children in the data base. However, an excess prevalence of low height for age (linear growth stunting) is implied in the data for children over 2 years old; for example, the prevalence among Hispanic children was 16% (vs. 5% expected and 6% among white children) in 1982. Findings of an excess prevalence of growth stunting among Hispanic American children have also been reported in several other studies (Yanochik-Cwen and White, 1977; Lowenstein, 1981; Dewey, 1983; Alvarez et al., 1984). The Alvarez et al. (1984) data for Hispanic children in an inner-city neighborhood health center population during 1978 also indicated a 13% prevalence of acute undernutrition (primarily moderate rather than severe levels of underweight, according to the ratio of observed weight to the age-appropriate NCHS median). The prevalence of short stature was greater among immigrant than among U.S. born Hispanic children. A few additional sources have reported low iron intakes, low blood iron status, or low intakes of other nutrients among Hispanic children or adolescents (Yanochik-Cwen and White, 1977; Haider and Wheeler, 1980; Lowenstein, 1981). The 1977-1981 CDC-PNSS hemoglobin data did not indicate an excess prevalence of low values among Hispanic children (Centers for Disease Control, 1983). Data from the National School Lunch Program evaluation did not find major differences between the 24 hour dietary intakes of Hispanic vs. white children, although it was noted that Hispanic children were more likely than white children to have energy and calcium intakes below the RDA (Vermeersch et al., 1984). Overall, these data do not give an impression of any clear pattern of dietary deficiency. The data on stunting are much more consistent. An excess prevalence of overweight has also been a common finding among Hispanic American children (Yanochik-Owen, 1977; Lowenstein, 1981; Centers for Disease Control, 1983; Trowbridge, 1983). According to CDC-PNSS reports for 1977-1982, the prevalence of high weight for height among Hispanic infants was somewhat greater than expected but was similar to that among white children in the data base. Among older Hispanic children the prevalence of high weight for height was higher than expected and higher than for the white children (Centers for Disease Control, 1983; Trowbridge, 1983). 5.2.2 Adults Cardenas et al. (1976) reported that intakes of meats, milk, fruits and vegetables were lower for primigravid Mexican American women than "average" Anglo women. Thirty-nine percent of the Mexican American women were overweight or obese at the first visit. Hunt et al. (1979) reported that 85% of low income Mexican American pregnant women had reported intakes below 2/3 of the 20 mg/day RDA for zinc; 153 protein intakes less than 2/3 of the RDA were reported by 30%. Haider and Wheeler (1979) compared nutrient intakes of Hispanic and black mothers in Brooklyn. Intakes of fiber, calcium, phosphorous, iron and riboflavin were higher among Hispanic than among black women and the diets of the Hispanic women had greater variety than those of black women. However, diets of the Hispanic women averaged only 2/3 of the RDA for energy, were higher in protein and fat than in carbohydrate content, and low in vitamin A. Recent reports of the practice of geophagia were not identified. Sixteen percent of Mexican American migrant women studied by Larson et al. (1974) in 1970-72 reported eating dirt or clay during pregnancy. Based on analyses with data from the National Food Consumption Survey, Windham et al.(1981) reported that nutrient density is relatively consistent across sex and age for all U.S. individuals over age 4. In a later study (1983a) these authors examined the effect of age, ethnicity, and socioeconomic variables, on the nutrient density of diets reported in the NFCS. Multivariate procedures were used to adjust for differences in age, sex, weight and height distributions among racial groups. Since the amount of seasonal variation in reported dietary intakes was low, data from the spring quarter of the survey were used. Intakes of calcium, magnesium, vitamin A and thiamin per 1,000 kcal differed significantly by race. Average calcium density of diets was lowest among Hispanic and black groups, averaging 30-40 mg per 1,000 kcal less than whites. Diets of Hispanic Americans were of lower vitamin A density, by 800 to 1,600 IU per 1,000 kcal, than whites or blacks. The average thiamin density of Hispanic diets was higher than among other ethnic groups. Nutrients for which dietary density did not differ by racial or socioeconomic factors in the model tested were protein, iron, phosphorous, riboflavin, and vitamin B12 (per 1000 kcal). In analyses comparing nutrient density with recommended allowances (i.e., using the Index of Nutritional Quality (INQ)), Windham et al. (1983b) found that the only substandard nutrient intakes which were significantly related to race were vitamin A intakes among Hispanic Americans. Average INQ values were 0.95 (where l=adequate) for Hispanic Americans vs. 1.39 and 1.79 for whites and blacks, respectively. Several authors have raised concern over the prevalence of chronic-disease related ovemutrition among Mexican Americans. Serum cholesterol levels are reportedly similar or lower among Hispanic vs. Anglo Americans (Yanochik-Owen and White, 1977; Williams et al., 1979; Stern et al., 1982). Obesity and diabetes are in excess prevalence among Mexican Americans. Haider and Wheeler (1979) noted that Hispanic mothers in Brooklyn were overweight and had above-standard mean triceps skinfolds values. From the San Antonio Heart Study, Stem et al., (1982) reported an excess of overweight among both low-income and suburban Mexican Americans compared to 154 Anglo Americans of comparable socioeconomic status. Mexican Americans are 2 to 4 times more likely than Anglos to be obese (Stern et al., 1983). Hanis et al. (1983) reported that 50% of Mexican Americans in Starr County, Texas were either diabetic (adult onset) or had a first degree diabetic relative. The excess of obesity does not entirely explain the excess prevalence of diabetes (Stern et al., 1983). Findings related to dietary aspects of coronary heart disease risk have been reported from the Puerto Rican Heart Study. Relative to the men in the Framingham Study, Puerto Rican men consume fewer calories, less cholesterol and proportionately less fat, protein, and alcohol calories (Garcia-Palmieri, 1977). Within Puerto Rico, significant urban-rural differences have been reported for relative weight and serum lipids (rural=lower) (Garcia-Palmieri, 1977). A very small proportion of the lipid differences (approxiamtely 2.5%) could be statistically explained by urban-rural differences in diet. 5.3 Conclusions Data from the Hispanic HANES (HHANES) will permit a detailed description of dietary patterns and nutritional status among Mexican, Puerto Rican and Cuban Americans. Thus, although the dietary and nutritional status picture for Hispanics is based on limited information at this writing, a much more substantial basis for nutritional risk inferences will be available in the near future. The evidence reviewed is consistent in demonstrating a disproportionate risk of both stunting and obesity among Hispanic children. Although the problems of stunting are undoubtedly confined to populations of lower socioeconomic status, and to immigrant and migrant worker subgroups, the problem of obesity in Mexican Americans is seen across socioeconomic levels. Areas of special dietary and nutritional concern which can be pinpointed in the Hispanic population include the low incidence of breast feeding, some associated infant feeding practices which are not in keeping with current pediatric guidelines, a tendency toward suboptimal intakes of vitamin A, and the excess prevalences of obesity and diabetes. Other areas of nutritional concern which apply to the general population should not be overlooked where Hispanics are concerned. The dietary practices of Hispanic Americans are more favorable to low coronary heart disease risks than diets of Anglo Americans, but the differences may decrease as degree of acculturation increases. In addition, the nutritional situation of Hispanic refugees should be given continued attention. A high prevalence of nutritional problems among a group of Cuban refugees in Florida has been noted (Gordon, 1982). 155 6. BLACK AMERICANS 6.1 General dietary patterns If the black American population is to be subdivided on the basis of dietary pattern variations, the subgroups of possible interest are southern- vs. non-southern-born blacks, urban vs. rural, native-born vs. immigrant, and Christian vs. Muslim blacks (among native-born blacks). Fifty-three percent of the black population lives in the South (U.S. Bureau of the Census, 1984) and a substantial proportion of blacks in other regions are southern-born). Black immigrants include persons from Caribbean islands and African countries. However, Spanish-speaking black immigrants are most often classified as Hispanic. The material in this section relates primarily to American-bom blacks. Some comments on Caribbean dietary patterns have been included. Dietary patterns based on Muslim dietary laws are not specifically addresssed. It should be stated at the outset that there is no black American dietary pattern in the same sense that there are Asian, Hispanic, or Native American dietary patterns. Structural dietary influences, if any, of the African origins of the American black population are not evident. The "southern" aspects of black diets are ethnic in the sense that most black Americans live or have origins in the U.S. Southeast. Sanjur (1982) notes the probability of more "eating out" among black Americans due to the high proportion of blacks who reside in urban areas. Only 14.7% of blacks were classified as "rural" in the 1980 census (U.S. Bureau of the Census, 1984). The regional, residence area, and socioeconomic status comparability of black and white survey samples influences interpretation of differences in black and white food consumption patterns. For example, black-white food pattern differences in data for the overall U.S. population may not be evident among blacks and whites in the South or in rural areas. Diets of blacks and whites living in the same regional area and of the same socioeconomic level tend to be more similar to each other than to diets of whites and blacks in different regional areas, and/or at different socioeconomic levels. Some chararacteristic black foods and food choices are shown in Table 11. The unique cultural, rather than regional or socioeonomic, influence on black dietary patterns relates to the types of foods available to blacks during slavery. Sanjur's review (1982) points out the different dietary patterns of field vs. house slaves. Slaves who worked in the fields developed meals requiring minimal preparation which were suitable for preparation in large quantities. Characteristic foods made use of cuts of pork which were inexpensive and undesirable to whites (e.g., tail, feet, chitterlings, ears) (Sanjur, 1982). Those who worked as house servants ate diets more like those of the planation owners. 156 Meal size and frequency were influenced by patterns associated with farming (i.e., heavy breakfast and large midday meal). Food preparation methods and flavoring principles characteristic of soul food include frying, smothering and barbecuing meats, use of black pepper and hot sauce, ham hocks, and salt pork as essential flavorings. Greens prepared in traditional style are boiled for long periods; however, nutrients which leach into cooking water are consumed with the greens as "pot liquor". Table 11: Characteristic Black Foods and Food Choices PROTEIN FOODS Meat: Poultry: Fish: Eggs Legumes: Nuts: beef, pork and ham, sausage, pig's feet, ears, etc., bacon, luncheon meat, organ meats chicken, turkey catfish, perch, red snapper, tuna, salmon, sardines, shrimp kidney beans, red beans, pinto beans, black-eyed peas peanuts, peanut butter MILK AND MILK PRODUCTS GRAIN PRODUCTS Milk: fluid, evaporated in coffee, buttermilk Cheese: cheddar, cottage Ice cream Rice, cornbread, hominy grits, biscuits, white bread, dry cereal, cooked cereal, macaroni, spaghetti, crackers VEGETABLES Broccoli, cabbage, carrots, com, green beans, greens: mustard, collard, kale, spinach, turnips, etc.; lima beans, okra, peas, potato, pumpkin, sweet potato, tomato, yam FRUITS Apple, banana, grapefruit, grapes, nectarine, orange, plums, tangerines, watermelon OTHER Salt pork, fruit drinks, carbonated beverages, gravies SOURCE: California Department of Health, 1975, Table 10 157 Carry-overs of some slavery-related and southern dietary traditions are designated by the general term "soul food". No studies of consumption frequencies for soul food, as such, were identified. Soul food-related preferences and consumption patterns persist to some degree (Wyant and Meiselman, 1979; Cronin et al., 1982; Gite and Perry, 1983). Certain items such as com bread, grits, and greens, which are non-controversial and which lend themselves to daily meal patterns are probably consumed according to habit and individual preference. Items such as chitterlings are probably reserved for special occasions. The consumption of pork products—either in general or with specific respect to "waste" cuts given to slaves—is somewhat controversial among some black Americans, particularly in the younger generations. The controversy apparently stems from a substantial diffusion of Muslim dietary principles among blacks in urban inner cities, intermingled with a deliberate rejection of cultural traditions which are associated with slavery. Relevant positive* features of traditional black food choices include the rich sources of vitamin A (yellow and dark leafy green vegetables) and thiamin (pork), and the relatively high fish and poultry content. Unfavorable features include the high sodium content of many items, the extensive use of frying, the use of less nutritious pork cuts, the overcooking of vegetables, and a tendency toward large, heavy meals. There does not seem to be an ideologically-based food hierarchy in traditional black diets. However, certain health-related food beliefs may persist, particularly among older blacks who were raised in the South (Comely, 1963; Snow, 1976). As described by Snow (1976), southern medical folklore includes an idea that some foods will "make the blood go up or bring it down" in volume, leading to "low or high blood". Foods thought to cause "high blood" or to build up "low blood" include beets, grape juice, red wine, liver, red meat, and other red foods. Foods thought to thin down "high blood" include lemon juice, vinegar, epsom salts, sour oranges, and brine from pickes or olives. Pregnancy was considered to be a condition during which a natural tendency toward "low blood" was assumed. Thus these beliefs favored the intake of several nutrient dense foods by pregnant women. However, failure to distinguish "low and high blood" from low or high blood pressure, may have led to increased intakes of olive or pickle juice, which are *With respect to vitamin A, some limited evidence associating the excess risk of prostate cancer in black men with the level of vitamin A consumption has been noted (Graham et al., 1983; Enterline, 1984). The postulated prostate cancer association with vitamin A consumption is such that risk is higher at higher levels of intake. In contrast, there is a considerable amount of evidence that high vitamin A intakes are associated with lower lung cancer risk. 158 very high in sodium, to "thin the blood" of those with high blood pressure. This is in spite of an equally longstanding belief among blacks that high blood pressure is aggravated by salt (Ward, 1982). The extent to which these beliefs affect current dietary practices of southern blacks is unknown. Data on food consumption patterns or food expenditures of blacks continue to reflect the traditional black American dietary emphases, i.e., relatively high meat diets, but weighted more to fish, poultry, eggs, and pork than beef; high vitamin A consumption, high salt consumption; somewhat lower intakes of dairy products and desserts. Bureau of Labor Statistics Consumer Expenditure Diary Survey data for 1972-1974 indicated that, at equivalent income levels, blacks spent more on beef, pork, poultry, fish and seafood than white households; expenditures for cereal and bakery products, dairy products, sugar and other sweets were less than those of white households (Salathe et al., 1979). Black households in urban areas of the Northeast and North Central regions spent more on non-alcoholic beverages than whites; expenditures on non-alcoholic beverages were lower among blacks than whites in other regions of the country. Market Research Corporation of America data on dairy product purchases of southern households in 1972-74 indicated that per capita purchases of cheese, fluid milk (except buttermilk), and dairy products overall were lower in black than in white households. Congruent with these findings, the levels of calories, protein, and calcium obtained from dairy products were less among blacks than among whites by approximately 50% (Kreidler et al., 1980). A 1973 telephone survey of beverage consumption patterns among approximately 8500 residents of upstate New York and New York City (Cook et al., 1975) indicated susbtantially higher water consumption and substantially lower coffee and tea consumption among blacks than whites. Smaller differences in consumption patterns for milk (blacks less), fruit juice or drinks and soft drinks (blacks more) were seen between blacks and whites in certain sex-age groups. Results of an analysis of data from the National Food Consumption Survey (1977-78) reflect some of the expected regional/racial differences in dietary pattern emphases (Cronin et al., 1982). A larger proportion of blacks and southerners than whites and non-southerners reported consumption of bacon and salt pork; more blacks than whites reported consumption of dark green vegetables, dried beans and peas, and poultry; fewer blacks than whites reported using milk, yogurt, or cheese and beef. A comparison of black and white food frequency data from NHANES I reflected higher intakes of meat, eggs, vitamin A, and salty snack foods among blacks than among whites but did not indicate substantial racial differences in the percentages of persons using milk; use of desserts, fruits and vegetables was lower among blacks than among whites (Villa Dresser et al., 1978) Other reports also imply higher consumption of salty foods or table salt by blacks vs. whites (Karp et al., 1980; Kerr et al., 1982). However, NHANES data do not show a greater use of table 159 salt among blacks compared to whites. Age-specific tabulations of salt use patterns of black and white adults tend to show lower levels among blacks than whites in most age groups (National Center for Health Statistics, 1985). No differences in the proportions of calories from carbohydrate, protein, saturated and unsaturated fat between black and white males or females are apparent in the NHANES data for any age group bewteen 1 and 74 (National Center for Health Statistics, 1985). However, the NHANES data indicate a greater tendency towards meal skipping and less use of vitamin and mineral supplements among blacks than whites (National Center for Health Statistics, 1985). Regarding meal skipping, 20 to 23% of black males and females skip breakfast compared with 12 to 14% of white males and females. Differences in percentages of blacks and whites skipping lunch, dinner, and snacks are of a similar order. Age-sex-race specific tabulations give the same general impression as the overall data. Frank et al. (1978) noted that fewer black than white boys and girls in the Bogalusa Study ate breakfast. Haider and Wheeler (1980) also noted meal skipping by black adolescent girls. NHANES data also indicate higher use of vitamin and mineral supplements among whites than blacks in the U.S. population (National Center for Health Statistics, 1985). Caribbean (West Indian) blacks have traditional diets similar in structure to those described for Hispanics. Puerto Rican diets are one type of Caribbean diet. The cultures and foods of various Caribbean islands include Spanish, French, Chinese, Dutch, African, and Amerindian influences (James, 1978). Examples of some Caribbean foods are shown in Table 12. Table 12: Examples of Some Caribbean Foods LEGUMES chick peas, cow peas, pigeon peas; peas; kidney beans, lima beans, soybeans VEGETABLES apio (root resembles carrot and parsnip); batata (starchy tuber); calabasa (pumpkin and squash); chayote (resembles cucumber and squash); malanga (taro; also called dasheen, tania, tainer, tanyah, yautia malanga); pana (seedless breadfruit); pane de pepita (seed breadfruit); plantain (variety of banana, eaten baked, fried, or boiled); yautia (varieties include belembe, calalu, eddo); yuca (fleshy, tuberous, starchy root) FLAVORING AND basil, sweet chili pepper, sesame seeds, aniseed, SEASONING HERBS coriander, dill, ginger, sweet marjoram, spearmint, oregano, rosemary SOURCE: New York City Department of Health, 1976b 160 6.2 Nutritional Risk 6.2.1 Infancy and Childhood Data from the National Natality Surveys indicate a higher reported incidence of breast feeding in 1980 than in 1969 for infants born to married women in the United States (Fetterly and Graubard, 1984). Breast feeding incidence was higher in white women than black women respondents in both years. Nine percent of black women vs. 19% of white women exclusively breast fed their infants in 1969; in 1980 the percentages were 25% for black women and 51% for white women. A higher incidence of breast feeding among women with higher levels of education was reported for white women in 1969 and 1980 and for black women in 1980. In 1969, the incidence of breast feeding was lower in black women with higher educational attainment. The racial differences are still seen after adjustment for education and parity. Similar findings of race differences which are not explained by education have been reported in studies among small samples of black and white women (including some unmarried women) in different parts of the United States (Rassin et al., 1984; Schaefer and Kumanyika, 1985)• In these studies, the effect of maternal education on breast feeding is more prominent among whites than among blacks. A report of Russo et al. (1981) points out the need to distinguish native-bom from immigrant blacks in this respect. Among several ethnic groups studied, breast feeding incidence was highest among West Indian blacks (64%) and lowest among native-born blacks (15%). Centers for Disease Control Pregnancy Nutrition Surveillance data for 1981 indicated that, among low income postpartum women in their sample for whom breast feeding data were available, breast feeding incidences were 16% and 37% for black women less than 20 years old and over 20, respectively, vs. 34 and 53% for white women in these two age categories (Centers for Disease Control, 1983). Black and white mothers may differ in the types of solid foods they give their infants. Black infants in the Massachussetts survey by Ferris et al. (1978) were more likely than white infants to be fed juices or soup, and less likely to be fed fruit or vegetables, with some differences in these patterns for infants less than three months vs. 3-6 months of age and according to social class. Black mothers surveyed in Philadelphia reported a greater frequency of high sodium feeding practices (earlier introduction of solid foods, adding salt to infant foods, and feeding relatively higher sodium foods as snacks) than white mothers. Some, but not all of these racial differences were statistically explained by black-white differences in maternal education (Schaefer and Kumanyika, 1985). 161 An excess prevalence of linear growth stunting is not evident in the Centers for Disease Control Pediatric Nutrition Surveillance System data (CDC-PNSS) for 1977-1982 for black children ages 2 years and older (although, as noted in section 2.3, inferences on this measure are confounded by differences in growth patterns for black and white children) (Centers for Disease Control, 1983; Trowbridge, 1983). In 1982, the percentage of black children below age 1 with low length for age was 9%, greater than the expectation of 5% below the 5th percentile and was higher than for whites, Hispanics and Native Americans. Prevalence of low length for age among black children in the 1 year age range was similar to that observed for black children less than one, but was lower than the prevalence among children in the other three ethnic groups and was lowest of tne otner three ethnic groups at ages 2,3, and 4 (Asian children were not reported in these data) (Trowbridge, 1983). The same pattern was observed during 1977-1981, except that percentages of Asian children less than two years old with low length for age were highest at all ages and for all years after 1977 compared to the other ethnic groups (Centers for Disease Control, 1983). Alvarez et al. (1984) reported a higher prevalence of short stature (ratio of height to NCHS median height) among black vs. Hispanic children under two years of age (27% vs. 10%) and a lower prevalence among black than among Hispanic children between ages 3 and 12 years (<20% among blacks and >25% among Hispanics). The children studied were from a low income, inner-city health center population in Boston in 1978. Moderate undernutrition was reported for 16% of the black children in this sample. A higher prevalence of low hemoglobin among black children ages two years and over is suggested in the CDC data for 1977-1981 (Centers for Disease Control, 1983). Data from NHANES II indicate that mean hemoglobin levels for blacks are lower than for whites in all age-sex categories and in both poverty and non-poverty categories (National Center for Health Statistics, 1985). However, differentials in the corresponding prevalences of anemia will depend on whether the same criteria are used for both blacks and whites (see section 2.3). Vbors et al. (1981) reported that clinical anemia, where observed in a sample of children in Louisiana, occurred primarily in younger black boys and older black girls. Racial differences in blood levels for several nutrients (e.g., higher mean serum copper, lower mean serum vitamin C, lower serum zinc (males)), are evident in the NHANES II data (National Center for Health Statistics, 1985). However, the significance of these differences for the relative nutritional status of black and white children has yet to be fully interpreted. Tabulations of reported 24-hour dietary intakes for black and white children and adolescents in NHANES II do not show large systematic differences in mean nutrient levels (National Center for Health Statistics, 1985). 162 A 1981 survey of nutrient intakes among female adolescents in the South indicated higher consumption of several nutrients among white girls than among black girls (vitamins E, C, B-12, niacin, and folacin; calcium, phosphorous, magensium and zinc, expressed as individual totals and per 1000 kcal; and protein, vitamin D, and iron) (McCoy et al., 1984). Some of the racial difference was due to higher supplement use among white than among black girls. However, these data do not indicate that the diets of the black girls were inadequate in the nutrients listed. Most diets met the RDA's or were within the recommended ranges for the nutrients ascertained. Possible areas of excess risk among the black girls were intakes of folacin and vitamin D. Answers to a question on table salt use suggested higher salt use by black girls than white girls; however there were no racial differences in stated preferences for salty foods. Gartside et al. (1984) analyzed NHANES II dietary variables related to high density lipoprotein-cholesterol (HDL-C) levels in an attempt to identify determinants of the higher HDL-C levels in blacks compared to whites. Variables considered were the Quetelet Index (QI: weight divided by the square of height), reported 24-hour intake of calories, protein, fat, carbohydrate, saturated fat, oleic acid, linoleic acid, cholesterol, and the ratio of linoleic to oleic acid (I/O), expressed as totals and per kilogram of body weight. Using the data adjusted for body weight, there were no significant differences in intakes of any of the dietary components studied for males ages 6 months through 20 years. Among females ages 6 months to 5 years, total fat, oleic acid, linoleic acid, and the I/O ratio were significantly higher in blacks. Other statistically significant differences which were scattered across sex-age groups were not evident when intakes were expressed per unit weight. In the Gartside et al.(1984) analysis, the only significant racial difference observed in mean Quetelet Index (QI) estimates for children were higher QI values among black than among white girls in the 12 to 20 year age group. Among the low income children in the CDC-PNSS data for 1977-82, the prevalence of high weight for height among black children under age 2 was approximately twice the expected 5% and was somewhat more prevalent than among white children in this data base. High weight for height was closer to the expected prevalence among preschool black children (2 to 5) and was somewhat higher than among white children in 1982 but not in 1977-1981 (Centers for Disease Control, 1983; Trowbridge, 1983). The prevalence of high weight for height among Hispanic and Native American children was higher than for black children at all ages after infancy. High weight for height was not in excess prevalence among the sample of black children studied by Alvarez et al. (1984). 163 6.2.2 Pregnancy Recent studies identifying excess nutritional risk specifically among black pregnant women were not identified. The extent to which possible problems in this area are being addressed may be available in WIC program data. Areas of concern for the nutritional status of black women in general are discussed in section 6.2.3. Although not widely prevalent among black women in the general population, a noteworthy prevalence of pregnancy-associated-pica has been reported among some black women in the rural South and may still be a relevant concern. Vermeer and Frate (1979) reported that 57% of women, 28% of pregnant and postpartum women, and 16% of children of both sexes in a rural Mississippi county practiced geophagia, with an average daily clay consumption of 50 grams. Most of the children practicing geophagia were under age 4. Community norms included the use of clay as a "pacifier" for young children but suppression of the practice as children became older. Geophagia was not practiced by adolescents or adult men, but was common among reproductive-age and pregnant women. Other forms of pica were practiced by an additional 19% of the people surveyed (primarily starch eating, but also eating of dry powdered milk). The authors' impression was that pica was not caused by or associated with dietary or nutritional deficiencies, although it may be indirectly related to appetite through an association with emotional well-being. Although various forms of pica are a hidden aspect of nutritional risk, Vermeer and Frate (1979) found little evidence of deleterious effects in their study. Hematocrit levels of women practicing geophagia were similar to those of non- practitioners. A possible aggravation of hypertension among pregnant women through excess sodium intake (either salt added to clay in preparation or consumed in baking soda) was noted. The authors also cite a 1975 case report (JAMA 1975;234:738) of possible pica-associated hyperkalemia among 5 black patients in Washington D.C. with chronic renal failure. 6.2.3 Adults Using either a "cut-off method" (12 mg/dl cutoff) or a "mixed- distribution" technique, Meyers et al. (1983) reported a higher prevalence of anemia among black than white non-pregnant women in NHANES. Anemia prevalence estimates were low, 0.8 to 1.5 % of the total population for white women and 1.4 to 2.7% for black women; however, the authors noted that these percentages represent a large number of U.S women. NHANES II data indicate lower mean hemoglobin and transferrin saturation levels among blacks than whites in all age-sex groups and lower levels of serum iron among black vs. white women. The serum biochemistries also showed lower mean zinc levels in some age groups of males and females, higher copper levels, and lower vitamin C levels for blacks than whites (National Center for Health Statistics, 1985). 164 Windham et al. (1983a; 1983b) analyzed the nutrient density and nutritional adequacy of diets reported by respondents in the National Food Consumption Survey. Multivariate adjustments for subgroup differences in distributions of age, sex, height, and weight were made to facilitate valid comparisons. Calcium density of reported diets was lowest among black and Hispanic respondents. Blacks had the lowest dietary magnesium density of any racial or ethnic group. Dietary vitamin A and thiamin densities were significantly higher for blacks than whites. Diets of blacks approximated (i.e. were within 80% of) the nutrient density standard for all nutrients studied. Inspection of NHANES II tabulations (National Center for Health Statistics, 1985) of nutrient intakes among black and white adults suggests differences in mean intakes of several nutrients, but often at levels which approximate dietary standards. Apparent lower intakes of calcium and phosphorous, potassium, and iron among blacks in many of the adult sex-age groups may be important. However, statistical tests of the significance of these differences have not been published. The Gartside et al. (1984) NHANES II analysis of possible determinants of black-white HDL-C differences indicated significantly higher calorie, protein, and fat intakes per kilogram of body weight in white vs. black men ages 21 to 65 and among white vs. black women ages 25 and over. Total caloric intakes were also higher among white vs. black women, in contrast to the significantly higher Quetelet Index (QI) of black women at all ages over 21. Significant differences in QI are not seen among black and white men. The Gartside et al. findings regarding higher relative weight among black vs. white women are indicative of what is clearly the most striking nutrition-related disparity between blacks and whites—the marked excess prevalence of obesity among black women vs. white women and vs. white and black men. A graphic presentation of the extent of the problem is shown in Figure 1. Using the implied criterion of a 15% expected prevalence high relative weight, the excess prevalence among black women is approximately 25% vs. an excess of only 10% among white women and men. The corresponding overall prevalences are approximately 40% among black women and 25% among the other race-sex groups. The analyses of Gartside et al. (1984) do not support the hypothesis that this excess prevalence of overweight among black women is caused by higher calorie or fat consumption among black vs. white women, either in terms of total intakes or per unit of body weight. In fact, the opposite appears to be true. 165 Iigure 1: Percent of U.S. adults ages 20-74 who are overweight, by race and sex 50 c o IX) Q- O O 4-> C CD o i- 0) 40 30 - 20 - 10 - BF BM WF WM 1960-62 -----1 BF BM WF 1971-74 WM BF BM WF WM 1------ 1976-80 ------ LEGEND: BF=black females BM=black males WF=white females WM=white males NOTE SOURCE The criterion of overweight is the 85th percentile of w/HP where w-weight in kilograms, H=height in meters, and p=2 for men and 1.5 for women, using the 20-29 year-old U.S. population as the reference group. Data for 1960-62 are from the National Health Examination Survey (NHES). Data for 1971-74 and 1976-80 are from the National Health and Nutrition Examination Surveys, cycles I and II. National Center for Health Statistics 6.3 Conclusions The overall structural similarities between the traditional diets of black and white Americans lead to an expectation of more similarities than differences in nutritional status. The literature and tables reviewed are generally in line with this expectation. Intakes of food in the black population may be somewhat less than among whites leading to lower overall intakes of many nutrients. However, a distinction should be made between intakes that are lower than those of whites vs. those that are lower than standards of dietary adequacy. The dietary and nutritional status data from NHANES II have been published but not yet fully interpreted. No major black-white disparities on these measures could be defined, although concern for the adequacy of iron, calcium, magnesium, and potassium intakes among segments of the black population is suggested by the national data. A discussion of issues related to the interpretation of dietary calcium adequacy is included in section 2.3. Inadequate iron intakes among black women are reflected in their excess prevalence of anemia, even after the appropriate adjustments for differences in hemoglobin distributions of blacks and whites. To the extent that nutritional adequacy among whites is a function of the use of vitamin supplements, blacks—who are less likely than whites to use supplements—will be at higher nutritional risk. Within the limitations of interview data for determination of sodium intake, the NHANES data do not indicate that sodium intakes of blacks are higher than those of whites. However, small studies indicate that a greater preference for and intake of certain highly salted foods among blacks is probable. The quantitative contribution of these foods to overall sodium intakes in blacks is unclear. Breast feeding incidence has increased among blacks but is still much lower than among whites. Depending on the type of milk __ substitutes used and accompanying patterns of feeding solid foods, this lower incidence of breast feeding may put black infants at a relative disadvantage vs. white infants. The nutritional status of reproductive-age black women is of continuing interest relative to the excess prevalence of low birth weight among blacks. There are potential nutritional risks to both mother and fetus associated with the practice of pica among black pregnant women. Pica may also be a concern for some subgroups of black children. A clear picture of nutritional risk among black children was not obtained. Neither linear growth stunting nor obesity was in excess among low-income black children (over age 1 or 2) in the Centers for Disease Control Pediatric Nutrition Surveillance data base. Most of the recent smaller studies identified were among low-income children and could not support generalities to the population of black children as a whole. Excess sodium intake among black girls is implied. Meal skipping by both boys and girls may be problematic. 167 The most outstanding nutritional problem in the black adult population is the excess prevalence of obesity among black ?oine'1' trend which is not necessarily evident in younger black girls duc which appears during adolescence. Data on reported dietary ^axes do not explain this excess obesity. Further study of meal ]^ttej^ and activity levels as well as energy metabolism among black women are indicated. Meal skipping, which appears to be in excess ?™}f blacks of both sexes and at all ages, may contribute to tne ooesity in black women. 168 7. SUMMARY The traditional dietary patterns of Asian, Hispanic, and Native Americans are structurally different from diets characteristic of white Americans in that carbohydrates rather than meats are the major sources of kilocalories. The diets of these minority subgroups also include many unique flavoring principles, food varieties, and food preparation methods. Diets of black Americans are generally similar to those of white Americans, particularly in the southeastern U.S. Specifically-black dietary elements relate to foods and cooking methods used by blacks during slavery. Nutritional risks associated with minority group dietary patterns are generally not inherent in traditional food patterns (except for the high dietary sodium contents of some of the traditional foods). Potential risks are determined by social, economic, or acculturation factors (see Table 3). Pood ideologies which may influence health-related food behaviors are seen in all cultures, e.g., the "hot-cold" dichotomies of Asian and Hispanic cultures. Nutritional risks which may derive from strict adherence to certain food ideologies include unfavorable food restrictions among pregnant women or young children. Although it is doubtful that such food beliefs or associated restrictions persist at levels which have public health significance, certain subgroups may be vulnerable in this respect. Similarly, the traditional practice of consuming non-food substances (pica) among pregnant women and young children may be a concern in certain minority subgroups. The available data indicate that nutritional risk in U.S. minority populations is related to the following factors: the income-related limitations on the quantity and quality of food consumed, the excess prevalences of obesity and obesity-related diseases in some minority subgroups, and the shifts toward dietary patterns which have been epidemiologically associated with high rates of cancer and cardiovascular disease. In addition to these general areas of risk, some specific nutritional concerns evident in the data for each minority group have been identified (see sections 3.3, 4.3, 5.3 and 6.3). With the caveat that the problems identified are a function of the data available, these group-specific nutritional problem areas can be summarized as follows: - among Asian Americans, evident areas of concern are the compromised growth and overall nutritional status of immigrant subgroups; and to carryovers of the high sodium intakes from traditional Asian diets. - among American Indians, evident concerns relate to child growth, to excess alcohol intake, obesity and diabetes; - among Hispanic Americans, evident concerns are growth stunting, obesity, infant feeding patterns, and adequacy of dietary vitamin A; 169 - among black Americans, evident concerns relate to vitamin and mineral intakes (vitamin A, vitamin C, iron, calcium, magnesium, potassium), to high sodium intakes, to infant feeding patterns, and to the excess prevalence of obesity among black women. In addition to the numerous methodological issues in nutritional assessment, there are several cross-cutting issues which apply specifically to the validity of nutritional risk inferences from data on non-white populations. These issues, as outlined in section 2.3, are of three general types: 1) issues related to the applicability of standards and references used to interpret dietary, anthropometric, and biochemical measures of nutritional status; 2) possible differences in sensitivity to a given level of nutrient excess between whites and non-whites due either to heredity factors or to differences in levels of nutrition-related morbidity; and 3) the large socioeconomic status effects within minority group data. Although nutrition, as such, is not a separate area of emphasis by a Task Force subgroup, considerations discussed in this review have implications that may be relevant to the deliberations of several Task Force subcommittees. The importance of standardized data on dietary and nutritional status measures for sufficiently large and representative samples of each minority subgroup is clear. There is also a need to ensure that dietary and nutritional status reference data are applicable multiculturally. Many nutritional effects of potential public health importance are related to subtle variations within characteristic U.S. food intake patterns. Monitoring race-related nutritional risks will require approaches which are sensitive to the often subtle differences between nutrient intakes of minority group Americans and whites. 170 LITERATURE CITED Acosta, PB; Aranda, RG; Lewis, JS; and Read, M; (1974) Nutritional status of Mexican-American preschool children in a border town. Am J Clin Nutr 27(12):1359-1368 Adrian, J; Daniel, R; (1976) Impact of socioeconomic factors on consumption of selected food nutrients in the United States. Am J Agr E Con 58:31-38 Alford, BB; Nance, EB; (1976) Customary foods in the Navajo diet. J Am Diet Assn 69(5):538-539 Alvarez, SR; Herzog, LW; Dietz, W Jr; (1984) Nutritional status of poor black and Hispanic children in an urban neighborhood health Center. Nutr Res 4:583-589 American Heart Association Nutrition Committee; (1982) Rationale of the Diet-Heart Statement of the American Heart Association. Circulation 65:839A-854A Asian/Pacific Islander Task Force on High Blood Pressure Education and Control; (1984) Social and Cultural Considerations in the Treatment of Hypertensives in Selected Asian/PacifJC Islander Populations. 310 8th Stret, Suite 305B. Oakland, CA 94607. Avioli, LV; (1980) Major minerals. A. Calcium and phosphorus. In: Goodhart, RS; Shils, ME; eds. Modern Nutrition in Health and Disease. Lea and Febiger, Philadelphia, PA. Chapter 7, pp 294-309 Bass, MA; Wakefield, IM; (1974) Nutrient intake and food patterns of Indians on Standing Rock Reservation. J Am Diet Assn 64:36-41 Bowering, J; Lowenberg, RL; Morrison, MA; Parker, SL; Tirado, N; (1978) Infant feeding practices in East Harlem. J Am Diet Assn 72(2):148-154 Bryant, CA (1982) Impact of kin, friend, and neighbor networks on infant feeding practices. Soc Sci Med 16:1757-1765 Burroughs, AL; Huenemann, RL; (1970) Iron deficiency in rural infants and children. J Am Diet Assn 57:122-128 Butte, NF; Calloway, DH; and VanDuzen, JL; (1981) Nutritional assessment of pregnant and lactating Navajo women. Am J Clin Nutr 34:2216-2228 California Department of Health Services. Maternal and Child Health Branch. 714/744 P Street, Room 300, Sacramento, CA 95814. (1975) "Ethnic, Social and Economic Influences on Diet Patterns". In: Nutrition Purina Pregnancy and lactation. Chapter 8, pp. 73-93. 171 Cardenas, J; Gibbs, CE; Young, EA; (1976) Ifcitritional beliefs and practices in primigravid Mexican-American women. J Am Diet Assn 69(3):262-265 Casey, P; Harrill, I; (1977) Nutrient intake of Vietnamese women relocated in Colorado. Nutr Rep Int 16:687-693. Centers for Disease Control (1983) Nutrition i^irYft3l1ance'1981* U.S. Department of Health and Human Services. HHS Pub. No. (CDC)I Chase-the-Bear, R; Bonnell, M; Morse, HG; Rate, BG; (1979) Hopis and Navajos not lean. N Eng J Med 13:301(24):1348 (letter) Chew, T; (1983) Sodium values of Chinese condiments and their use in sodium-restricted diets. J Am Diet Assn 82(4):397-401 Clifford, NJ; Kelly, JJ; Leto, TF; Eder, HA; (1963). Coronary heart disease and hypertension in the White Mountain Apache Tribe. Circ 8:926-931. Committee on Diet, Mitrition and Cancer (1982) Diet. Nutrition and Cancer. National Research Council, National Academy of Sciences. Washington, D.C. Cook, CB; Eiler, DA; and Forker, CD; (1975) Beverage consumption patterns in New York State. J Am Diet Assn 67:222-227 Comely, PB; Bigman, SK; Watts, DD; (1963) Nutritional beliefs among a low-income, urban population. J Am Diet Assn 42:131-135 Crane, OT; Green, NR; (1980) Food habits and food preferences of Vietnamese refugees living in northern Florida. J Am Diet Assn 76(6):591-593 Cronin, FJ; Krebs-Smith, SM; Wyse, BW; Light, L; (1982) Characterizing food usage by demographic variables. J Am Diet Assn 81:661-673 DeStefano, F; Coulehan, JL; Wiant, MK; (1979) Blood pressure survey on the Navajo Indian reservation. Am J Epi 109:335-345 Dewey, KG; Chavez, MN; Gauthier, CL; Jones, LB; and Ramirez, RE; (1983) Anthropometry of Mexican-American migrant children in northern California. Am J Clin Nutr 37(5):828-33 Dietary Guidelines for Americans (1980) U.S. Depts. of Agriculture and Health and Human Services. Home and Garden Bulletin No 232. Doeblin, TO; Evans, K; Ingall, GB; Dowling, K; Chilcote, ME; Elsea, W; Bannerman, RW; (1969). Diabetes and hyperglycemia in Seneca Indians. Hum Hered 19:613-627. 172 Enterline, J; (1984) Personal communication regarding preliminary results of a case-control study of prostate cancer among black men. Howard University. Washington, D.C. Farmer, ME; White, LR; Brody, JA; Bailey, KR; (1984) Race and sex differences in hip fracture incidence. Am J Pub Health 74:1374-1380 Ferris, AG; Vilhjalmsdottir, LB; Beal, VA; and Pellett, PL; (1978b) Diets in the first six months of infants in western Massachusetts. II. Semi-solid foods. J Am Dietet Assn 72:160-163. Fetterly, K; Graubard, MS; (1984) Racial and Educational Factors Associated with Breast Feeding - United States, 1969 and 1980. MMWR 33(11):153-154 Food and Nutrition Board (1980) Toward Healthful Diets. National Research Council, National Academy of Sciences. Washington, D.C. Forman, MR; Graubard, BI; Hoffman, HJ; Beren, R; Harley, EE; Bennett, P; (1984) The Pima Infant Feeding Study: Breast feeding and gastroenteritis in the first year of life. Am J Epi 119(3): 335-349 Frank, GC; Berenson, GS; and Webber, LS; (1978) Dietary studies and the relationship of diet to cardiovascular disease risk factor variables in 10 year old children - The Bogalusa Heart Study. Am J Clin Nutr 31:328-340 Frerichs, RR; Webber, LS; Srinivasan, SR; Berenson, GS; (1977) Hemoglobin levels in children from a biracial southern community. Am J Pub Health 67(9):841-845 Frishancho, AR; Leonard, WR; and Bollettino, LA; (1984) Blood Pressure in Blacks and Whites and its relationship to dietary sodium and potassium intake. J Chron Dis 37:7; 515-519 Fulmer, HS; Roberts, EW; (1963) Coronary heart disease among the Navajo Indians. Ann Int Med 59:740-764 Galli, N (1975) The influence of cultural heritage on the health status of Puerto Ricans. J Sen Health: 45(1): 10-16 Garcia-Palmeri, MR; (1973) Precursors of coronary artery disease in Puerto Rico. Am J Clin Nutr 26(10):1133-1137 Garcia-Palmeri, MR; Tillotson, J; Cordero, E; Costas, R; Sorlie, P; Gordon, T; Kannel, WB; Colon, AA; (1977) Nutrient intake and serum lipids in urban and rural Puerto Rican men. Am J Clin Nutr 30(12):2092-2100 173 Gam, SM; Clark, DC; (1976) Problems in the nutritional assessment of black individuals. Am J Pub Health 66:262-267 Gartside, PS; Khoury, P; and Glueck, CJ; (1984) Determinants of HDL cholesterol in blacks and Whites: the second national HANES. Am Heart J 108:#3(2): 641-653 Gillum, RF; (1979) Pathophysiology of hypertension in blacks and whites. A review f the basis for racial blood pressure differences. Hypertension 1:468-475 Gillum, RF; Prineas, RJ; Palta, Ma; Horibe, H; (1980) Blood pressures of urban Native American school children. Hypertension 2:744-749 Gillum, RF; Gillum, BS; Smith, N; (1984) Cardiovascular risk factors among urban American Indians. Blood pressure, serum lipids, smoking, diabetes, health knowledge, and behavior. Am Heart J 107:756-776 Gite, L; Perry, J; (1983) Diet RightI Losing weight soulfully. Essence Magazine 14(4):117-118 Gonzalez, NL; (1972) "Changing dietary patterns of North American Indians". In: Moore WM et al.; eds. Nutritionf Growth, and Development of North American Indian Children. HEW Pub. No. (NIH) 72-26. Washington, D.C. Gordon, AM Jr.; (1982) Nutritional status of Cuban refugees: a field study on the health and nutriture of refugees processed at Opa Locka, Florida. Am J Clin Nutr 35(3):582-590 Graham, S; Haughley, B; Marshall, J; Priore, R; Byers, T; Rzepka, T; Mettlin, C; Pontes, JE. (1983) Diet in the epidemiology of carcinoma of the prostate gland. J Natl Cancer Inst 70:687-692. Grivetti, LE; Paquette, MB; (1978) Nontraditional ethnic food choices among first generation Chinese in California. J Nutr Ed 10(3):109-112 Haider, SQ; Wheeler, SM; (1980) Dietary intake of black and Hispanic teenage girls. J Am Diet Assn 77(6):677-681 Hanes, V;Vermeersch, J; Gale, S; (1984) The National Evaluation of School Nutrition Programs. Program impact on dietary measures. Am J Clin Nutr 40 (supp): 390-413. Hanis, CL; Ferrell, RE; Barton, SA; Aquilar, L; Garzaibarra, A; Tulloch, BR; Garcia, CA; and Schull, WJ; (1983) Diabetes among Mexican-Americans in Starr County, Texas. Am J Epi 118:5:659-672 Hankin, JH; Nomura, A; and Rhoads, GG; (1975) Dietary patterns among men of Japanese ancestry in Hawaii. Cancer Res 35:3259-3264 174 Horner, MR; Olson, C; and Pringle, DJ; (1977) Nutritional status of Chippewa Head Start children in Wisconsin. Am J Pub Health 67(2):185-186 Hunt, IF; Murphy, NJ; Gomez, J; and Smith, JC Jr.; (1979) Dietary zinc intake of low-income pregnant women of Mexican descent. Am J Clin Nutr 32(7):1511-1518 James, SM; (1978) When your patient is black West Indian. Am J Nursing 78(11):1908-1909. Jerome, NW; Pelto, GH; (1981) Integrating ethnographic research with nutrition studies. Fed Proc 40:2601-2605 Johnston, FE; McKigney, JI; Hopwood, S; and Smelker, J; (1978) Physical growth and development of urban Native Americans; a study in urbanization and its implications for nutritional status. Am J Clin Nutr 31(6):1017-1027 Judkins, RA; (1978) American Indian medicine and contemporary health problems. IV. Diabetes and perception of diabetes among Seneca Indians. NY State J Med 78:1320-1323 Karp, RL; Williams, C; and Grant, JO; (1980) Increased utilization of salty food with age among preteenage black girls. J Natl Med Assn 72:3;197-200 Kerr, GR; Amante, P; Decker, M; and Callen, PW; (1982) Ethnic patterns of salt purchase in Houston, Texas. Am J Epi 115:906-916 Kim, KK; Kohrs, MB; Twork, R; Grier, MR; (1984) Dietary calcium intakes of elderly Korean Americans. J Am Diet Assn 84(2):164-169 Knishbacher, S; Porter, DV ; (1983) Nutrition and Diet-Related Diseases. Congressional Research Service. Library of Congress. Washington, D.C. (1981 document revised and updated by Porter in 1983) Koh, ET; (1980a) Selected blood components and urinary B vitamins as related to age and sex of the black population in southwest Mississippi. Am J Clin Nutr 33:670-676 Koh, ET; Chi, MS; and Lowenstein, FW; (1980b) Comparison of selected blood components by race, sex, and age. Am J Clin ttitr 33:1828-1835 Kolonel, LN; Nomura, AMY; Hirohata, T; Hankin, JH; and Hinds, MW; (1981) Association of diet and place of birth with stomach cancer incidence in Hawaii Japanese and Caucasians. Am J Clin Nutr 34:2478-2485 175 Kramer, MJ; Hunter, R; Rosse, C; Smith, NJ; (1977) Race related differences in peripheral blood and in bone marrow cell populations of American black and American white infants. J Natl Med Assn 69:327-331 Kreidler, PL; Boehm, WT; Lentner, MN; and Driskell, JA; (1980) Dairy product purchase by southern households. J Am Diet Assn 77:41-46 Kuhnlein, HV; Calloway, DH; and Harland, BF; (1979) Composition of traditional Hopi foods. J Am Diet Assn 75(1):37-41 Lackey, CJ; (1983) Pica during pregnancy. Contemporary Nutrition. 8 (11) Larson, LB; Dodds, JM; Massoth, DM; and Chase, HP; (1974) Nutritional status of children of Mexican-American migrant families. J Am Diet Assn 64:29-39 Lewis, JS; Glaspy, MF; (1975) Food habits and nutrient intakes of Filipino women in Los Angeles. J Am Dietet Assn 67:122-125. Lieberman, LS (1979) Medico-nutritional practices among Puerto-Ricans in a small urban north-eastern community in the US. Soc Sci Med: 13B:191-198 Lowenstein, FW; (1981) Review of nutritional status of Spanish Americans based on published and unpublished reports between 1968 and 1978. World Rev Nutr Diet 37:1-37 Ludman, EK; Newman, JM; (1984) Yin and Yang in the health-related food practices of three Chinese groups. J Nutr Ed 16(1):3-5 Magnus, PD (1983) Breastfeeding among Hispanics. Am J Public Health 73(5):597 (letter) Marrs. DC; (1978) Milk drinking by the elderly of three races. J Am Diet Assn 72(5):495-498 Mayberry, RH; Lindeman, RD; (1963) A survey of chronic disease and diet in Seminole Indians in Oklahoma. Am J Clin Mitr 13:127-134 McCoy, H; Kenney, MA; Kirby, A; Disney, G; Ercanli, FG; Glover, E; Korslund, M; Lewis, H; Liebman, M; Livant, E; Moak, S; Stallings, S; Wakefield, T; Schilling, P; Ritchey, SJ; (1984) Nutrient intakes of female adolescents from eight southern states. J Am Dietet Assn 84:1453-1460 McGee, D; Rhoads, G; Hankin, J; Yano, K; and Tillotson, J; (1982) Within-person variability of nutrient intake in a group of Hawaiian men of Japanese ancestry. Am J Clin Nutr 36(4):657-663 176 Meyers, LD; Habicht, JP; Johnson, CL; (1979) Components of the differencein hemoglobin concentration in blood between black and white women in the United States. Am J Epi 109(5):539-549 Meyers, ID; Habicht, JP; Johnson, CL; Brownie C; (1983) Prevalences of anemia and iron deficiency anemia in black and white women in the United States estimated by two methods. Am J Pub Health 73:1042-1049 National Center for Health Statistics (1982) Diet and dental health, A study of relationships. United States, 1971-1974. Vital and Health Statistics, series 11. Data from the National Health Survey (No. 225) (DHHS Pub (PHS) 82-1675). National Center for Health Statistics; (1983a) Health United States, 1983. Tables 30 (pp 129-130) and 31 (pp 131-132). DHHS Pub. No. (PHS) 84-1232. Public Health Service. Washington. U.S. Government Printing Office, Dec. 1983 National Center for Health Statistics. S. Abraham: 1983b; Obese and overweight adults in the united States. Vital and Health Statistics. Series 11, No. 230. DHHS Pub. No. 83-1680. Public Health Service. Washington. U.S. Government Printing Office. National Center for Health Statistics (1985) Caloric, selected nutrient, and food intake of persons 1-74 years of age, by race and sex. United States 1971-1974 and 1976-1980. Vital and Health Statistics, series 11 (unpublished draft) National Institutes of Health; (1984) Osteoporosis. Consensus Development Conference Statement. Vol 5, #3; US Government Printing Office, Washington, D.C. Stock #1984-421-132:4652 National Research Council (1980) Recommended Dietary Allowances. Ninth Edition, National Academy of Sciences, Washington, D.C. Neaton, JD; Kuller, LH; Wentworth, D; Borhani, NO; (1984) Total and cardiovascular mortality in relation to cigarette smoking, serum cholesterol concentration, and diastolic blood pressure among black and white males followed up for five years. Am Heart J 108:759-770 New York City Department of Health. Bureau of Nutrition Services. 93 Worth Street, Room 714, New York, NY 10013; (1981) Characteristics of the Puerto Rican Food Habits New York City Department of Health. Bureau of Nutrition Services. 93 Worth Street, Room 714, New York, NY 10013; (1976a) Description of Puerto Rican Foods. New York City Department of Health. Bureau of Nutrition Services. 93 Worth Street, Room 714, New York, NY 10013; (1976b) Glossary of Caribbean Foods. 177 Newcomer, AD; Thomas, PJ; McGill, DB; Hofmann, AF; (1977) Lactase deficiency: a common genetic trait of the American Indian. Gastroenterology 72 (2): 234-237 Nguyen, TT; Do, TH; Craig, WJ; Zimmerman, G; (1983) Food habits and preferences of Vietnamese children. J School Health 53(2):144-147 Nomura, A; Henderson, BE; and Lee, J; (1978) Breast cancer and diet among the Japanese in Hawaii. Am J Clin Nutr 31(11):2020-2025 Owen, GM; Yanochik-Owen, A; (1977) Should there be a different definition of anemia in black and white children? Am J Pub Health 67:865-866 Owen, GM; Garry, PJ; Seymoure, RD; Harrison, GG; Acosta, PB; (1981) Nutrition studies with White Mountain Apache preschool children in 1976 and 1969. Am J Clin Nutr 34(2):266-277 Paige, DM; Bayless, TM; Ferry, GD; Graham, GG; (1971) Lactose malabsorption and milk rejection in Negro children. Johns Hopkins Med J 119:163-169 Parker, SL; Bower ing, J; (1976) Folacin in diets of Puerto Rican and black women in relation to food practices. J Nutr Ed 8(2), cited in Sanjur (1982), p. 246 Pettitt, DJ; Lisse, JR; Knowler, WC; and Bennett, PH; (1982) Mortality as a function of obesity and diabetes mellitus. Am J Epi 115(3):359-366 Rassin, DK; Richardson, CH; Baranowski, T; Nader, PR; Guenther, N; Bee, DE; Brown, JP; (1984) Incidence of breast feeding in a low socioeconomic group of mothers in the United States. Ethnic patterns. Pediatrics 73:132-137 Report of the President's Task Force on Food Assistance (1984) Washington, D.C. Rhoades, SJ; (1982) St Regis-Mohawk Health Services Needs Assessment of Nutrition Services, pg. 5 (New York State Public Health Nutritionist^ report included in briefing book for the New York State Nutrition Watch Commission Roe, DA; (1973) A Plaque of Com. The Social History of Pellagra. Ithaca, New York. Cornell University Press Russo, RM; Patel, R; Laude, TA; Rajkuraar, SV; Gururaj, VJ; (1981) Infant feeding practices by ethno-cultural grouping. J Med Soc NJ 78:737-740. 178 Salathe, LE; Gallo, AE; Boehm, WT; (1979) The impact of race on consumer food purchases. USDA ESCS-68 Washington, DC Sanjur, D (1982) Social and Cultural Perspectives in Nutrition. Prentice-Hall. Inc. NJ. Schaefer, LJ; Kumanyika, SK; (1985) Maternal variables related to potentially high sodium infant feeding practices. J Am Dietet Assn 85:433-438. Seger, MT; Gibbs, CE; Young, EA; (1979) Attitudes about breastfeeding in a group of Mexican-American primigravidas. Texas Med 75(l):78-80 Shutte, JE; (1980) Growth standards for blacks. Current status. J Natl Med Assn 72:973-978 Sievers, ML; (1968) Cigarette smoking and alcohol usage by southwestern American Indians. Am J Pub Health 58:77-82 Sievers, ML; Fisher, JR; (1981) Diseases of North Americans Indians. In: Rothchild, HR; ed. Biocultural Aspects of Disease. Academic Press, Inc. Chapter 8 pp 191-213 Simpson, M; (1982) Annual Report of Pilot Nutrition Programs to Onondaga Indian Nation, pp 1-2. Cited in briefing materials for New York State Nutrition Watch Committee, 1982) Slonim, A; Kolassa, KM; and Bass, MA; (1981) The cultural appropriateness of the WIC program in Cherokee, NC. J Am Diet Assn 79:164-168 Smith, JC; Mhango, GG; Warren, CW; Rochat, RW; Huffman, SL; (1982) Trends in the incidence of breastfeeding for Hispanics of Mexican origin and Anglos on the US - Mexico border. Am J Pub Hlth 72(1):59-61 Snow, LF; (1976) "High Blood" is not high blood pressure. Urban Health 6:54-55 Stem, MP; Pugh, JA; Gaskill, SP; and Hazuda, HP; (1982) Knowledge, attitudes, and behavior related to obesity and dieting in Mexican Americans and Anglos: the San Antonio Heart Study. Am J Epi 115(6):917-928 Stem, MP; Gaskell, SP; Hazuda, HP; Gardner, LI; and Haffner, SM; (1983) Does obesity explain excess prevalence of diabetes among Mexican-Americans? Results of the San Antonio Heart Study. Diabetologia 24:272-277 Stephenson, LS; Latham, MC; Jones, DV; (1977) Milk consumption by black and white pupils in two primary schools. J Am Diet Assn 71:258-262 179 Stoner, MK; Grivetti, LE; (1978) Vietnamese refugee. Adjustment, diet and food behavior. Read before the Society for Applied Anthropology, Mexico. Strimbu, JL; Sims, OS Jr.; (1974) A university system drug profile. Int J Addiction 9:569-583 Suitor, LW; Crowley, MF; (1984) "Promoting sound eating habits in different sociocultural situations". In: Nutrition■ Principles and Applications in Health Programs. Second Edition. J.B. Lippincott, Philadelphia, PA Chapter 9, pp 91-105 Tillotson, JL; Kato, H; Nichaman, MZ; Miller, DC; Gay, M; (1973) Heart disease and stroke in Japanese men living in Japan, Hawaii, and California: Methodology for comparison of diet. Am J Clin Nutr 26:177-184 Toma, RB; Curry, ML; (1980) North Dakota Indians' traditional foods. J Am Diet Assn 76(6):589-590 Trowbridge, FL; (1983) Prevalence of growth stunting and obesity: Pediatric Nutrition Surveillance System, 1982. MMWR 32(4SS): 23SS-26SS U.S. Bureau of the Census; (1984). Statistical Abstract of the United States, 1985 (105th Edition). Washington, D.C. U.S. Dept of Agriculture, Food and Nutrition Service; (1980) Southeast Asian American Food Habits. FNS-225 Washington, DC U.S. Dept of Health, Education and Welfare, Children's Bureau; (1965) The nutritional status of American Negroes. Nutr Rev 23(6):161-164 Vermeer, DE; Frate, DA; (1979) Geophagia in rural Mississippi. Environmental and cultural contexts and nutritional implications. Am J Clin Nutr 31:2129-2135 Vermeersch J.; Hanes, S; Gale, S; (1984) The National Evaluation of School Nutrition Programs. Program impact on anthropometric measures. Am J Clin Nutr 40 (supp): 414-424. Villa Dresser, CM; Carroll, MD; and Abraham, S; (1978) Selected Findings: Food Consumption Profiles of White and Black Persons 1-74 years of age in the US 1971-1974. Advance data from the Vital and Health Statistics of the NCHS #21 Voors, AW; Frank, GC; Srinivasan, SR; Webber, LS; and Berenson, GS; (1981) Hemoglobin levels and dietary iron in pubescent children in a biracial community. Pub Health Rep 96(1): 45-49 180 Waldman, EB; Lege, B; Oseid, B; Carter, JP; (1979) Health and nutritional status of Vietnamese refugees. South Med J 72(10): 1300-1303 Ward, GW; (1982) Improving public understanding of high blood pressure. Urban Health 11(4)-.117-118 West, KM; (1974) Diabetes in American Indians and other native populations of the New World. Diabetes 23(10):841-855 Willet, WC; MacMahon, B; (1984) Diet and cancer. An overview. N Eng J Med 310:633-638 and 697-703. Williams, CL; Carter, BJ; Arnold, CB; Wynder, EL; (1979) Chronic disease risk factors among children: the xKnow Your Body' Study. J Chron Dis 32: 505-513 Windham, CT; Wyse, BW; Hurst, RL; Hansen, RG; (1981) Consistency of nutrient consumption patterns in the United States. J Am Diet Assn 78(6):587-594 Windham, CT; wyse, BW; Hansen, RG; and Hurst, RL; (1983a) Nutrient density of diets in the USDA Nationwide Food Consumption Survey, 1977-1978. I. Impact of socio-economic status on dietary density. J Am Diet Assn 82(1):28-34 Windham, CT; wyse, BW; Hansen, RG; (1983b) Nutrient density of diets in the USDA Nationwide Food Consumption Survey, 1977-1978. II. Adequacy of nutrient density consumption practices. J Am Diet Assn 82(1):34-43 Wyant, KW and Meiselman, HL; (1979) USAF Food Habits Study: Part II. Food Preferences of Whites and Blacks and Males and Females. US Army Food Sciences Lab. Tech Rep #TRr79/042 Natick, Mass. Yang, GIP; Fox, HM; (1979) Food habit changes of Chinese persons living in Lincoln, Nebraska. J Am Diet Assn 75(4):420-424 Yano, K; Rhoads, GG; Kagan, A; and Tillotson, J; (1978) Dietary intake and the risk of coronary heart disease in Japanese men living in Hawaii. Am J Clin Nutr 31(7):1270-1279 Yano, K; Wasnick, RD; Vogel, JM; and Heilbrun, LK; (1984) Bone mineral measurements among middle-aged and elderly Japanese residents in Hawaii. Am J Epi 119(5):751-765 Yanochik-Owen, A; White, M; (1977) Nutrition surveillance in Arizona: selected anthropometric and laboratory observations among Mexican-American children. Am J Pub Health 67(2):151-154 Yu, MC; Ho, JHC; Ross, RK; Henderson, BE; (1981) Nasopharyngeal carcinoma in Chinese - salted fish or inhaled smoke? Prev Med 10:15-24 181 SUPPLEMENTARY REFERENCES Armstrong, H. (1975) Nutritional status of black preschool children in Mississippi. J Am Diet Assn 66(5) :488-493 Brittin, HC; Zinn, DW; (1977) Meat-buying practices of Caucasians, Mexican-Americans, and Negroes. J Am Diet Assn 71:623-628 Caliendo, MA; (1981) Nutrition and Preventive Health Care. Chapter 5. Mac Millan Pub Co, NY pp.133-169 Caster, WD; (1980) The core diet of lower-economic class women in Georgia. Food and Nutrition 9:241-246 Cerqueira, MT; Fry, MM; Connor, WE; (1979) The food and nutrient intakes of the Tarahumara Indians of Mexico. Am J Clin Nutr 32(4):905-915 Chang, B; (1974) Some dietary beliefs in Chinese folk culture. J Am Diet Assn 65(4):436-438 Chase, HP; Kremar, U; Dodds, JM; Suberlick, HE; Hunter, RM; Burton, RS; and Spalding, V; (1971) Nutritional status of preschool Mexican-American migrant farm children. J Dis Child 122:316-324 Costas, R; Garcia-Palmeri, MR; Nazario, E; Sorlie, FD; (1978) Relation of lipids, weight, and physical activity to incidence of coronary heart disease: The Puerto Rico Study. Am J Cardiol 42:653-658 Dales, LG; Friedman, GD; Ury, HK; Grossman, S; and Williams, SR; (1979) A case-control study of relationships of diet and other traits to colorectal cancer in American blacks. Am J Epi 109(2):132-144 Desai, ID; Lee, M; (1974) Nutritional status of Canadian Indians: I. Biochemical studies at Upper Laird and Ross River, the Yukon Territory. Can J Pub Health 65(5):369-374 Dilling, LA; Ellestad-Sayed, J; Coodin, FJ; Haworth, JC; (1978) Growth and nutrition of preschool Indian children in Manitoba; I: Vitamin D deficiency. Can J Pub Health 69(3):248-252 Drabman, RS; Hammer, D; Jarvie, GJ; (1977) Eating styles of obese and non-obese black and White children in a naturalistic setting. Addictive Behavior 2(2-3):83-86 182 Farris, RP; Cresanta, JL; Frank, GC; Webber, LS; and Berenson, GS; (1984) Dietary studies of children from a biracial population: Intakes of fat and fatty acids in 10 and 13 year olds (Bogalusa study). Am J Clin Nutr 39:114-128 Ferris, AG; Vilhjalmsdottir, LB; Beal, VA; and Pellett, PL; (1978a) Diets in the first six months of infants in western Massachusetts. I. Energy yielding nutrients. J Am Dietet Assn 72:155-159. Gillum, RF; Grant, CT; (1982) Coronary heart disease in black populations. II. Risk Factors. Am Heart J 104(4 Pt 1): 852-864 Guthrie, H.A. (1979) Introductory Nutrition. 4th edition. The C.V. Mosby Co., St. Louis, Mo. Haider, SQ; Wheeler, SM; (1979) Nutritive intake of black and Hispanic mothers in a Brooklyn ghetto. J Am Diet Assn 75(6):670-673 Hoppner, K; McLaughlan, JM; Shah, BG; Thompson, JN; Beare-Rogers, J; Ellestad-Sayed, F; and Schaefer, 0; (1978) Nutrient levels of some foods of Eskimos from Artie Bay, NWT, Canada. J Am Diet Assn 73:257-260 Jerome, NW; (1982) Dietary patterning and change: A continuous process. Contemporary Nutrition 7(6) Johnston, JL; Williams, CN; Weldon, KM; (1977) Nutrient intake and meal patterns of Micmac Indian and Caucasian women in Shubenacadie, N.S. Canadian Med Assn J 116(12):1356-1359 Koh, ET; Caples, V; (1979a) Frequency of selection of food groups by low-income families in southwestern Mississippi. J Am Diet Assn 74:660-664 Koh, ET; Caples, V; (1979b) Nutrient intake of low-income black families in southwestern Mississippi. J Am Diet Assn 75:665-673 Lowenberg, ME; Lucas, BL; (1976) Feeding families and children: 1776-1976. A bicentennial study. J Am Diet Assn 68(3):207-215 McMurry, MP; Connor, WE; Cerqueira, MT; (1982) Dietary cholesterol and the plasma lipids and lipoproteins in the Tarahumara Indians: a people habituated to a low cholesterol diet after weaning. Am J Clin Nutr 35(4):741-744 Mettlin, C (1980) Nutritional habits of blacks and whites. Preventive Medicine 9:601-606 Rush, D; Davis, H; Susser, M; (1972) Antecedents of low birth weight in Harlem, New York City. Int J Epi 1: 375-377 183 Stavig, GR; Igra, A; Leonard, AR; (1984) Hypertension among Asians and Pacific Islanders in California. Am J Epi 119(5):677-691 Stemmermann, G; Haenszel, W; Locke, F; (1977) Epidemiologic pathology of gastric ulcer and gastric carcinoma among Japanese in Hawaii. J Natl Cancer Inst 58(1):13-19 Wagner, PA; Krista, ML; Bailey, LB; Chistakis, GJ, Jernigan, JA; Aranjo, PE; Appeldorf, H; Davis, OG; and Dinning, JS; (1980) Zinc status of elderly black Americans from urban-low-income households. Am J Clin Mitr 33:1771-1777 Wenkam, NS; Wblff, RJ; (1970) A half century of changing food habits among Japanese in Hawaii. J Am Diet Assn 57(1):29-32 Yohai, F; (1977) Dietary patterns of spanish-speaking people living in the Boston area. J Am Diet Assn 71(3):273-275 184 Appendix 1: Background Information on Vitamins and Minerals VITAMINS Vitamin A Vitamin BI (Thiamine) Vitamin B2 (Riboflavin) Vitamin B6 (Fyridoxine) Vitamin B12 (Cyanoco- balamin) Vitamin C (Ascorbic acid) FUNCTIONS Aids in normal growth, helps to prevent infections, promotes healthy skin, and aids in night and color vision Assists in carbohy- drate metabolism. Maintains reserves of energy. Helps to release energy from food. Helps main- tain healthy skin. Assists in the formation of red blood cells. Assists in the formation of the neurotransmitters important in the brain function and in protein metabolism. Assists in the develop- ment of red blood cells. Helps maintain nerve tissue. Forms collagen, a substance that holds body cells together. Hastens the healing of wounds. Enhances iron absorption. IMPORTANT FOOD SOURCES Liver, butter, carrots, cantaloupes, dark green leafy vegetables, apricots, broccoli, fortified margarine, sweet potatoes, cheese, cream, & milk fortified with vitamin A. Liver, meat, poultry, whole-grain flours and cereals, wheat germ, seeds like sunflower and sesame, legumes, nuts, leafy vegetables. Liver, kidney, lean meat cheese, milk, yogurt, eggs, leafy vegetables, beans & peas, fortified grain products. Muscle meats, liver, whole-grain bread, flours and cereals, soybeans, bananas, peanuts, potatoes, beans, brown rice. Vitamin B12 can only be found in animal products: meat, poultry, fish, shell fish, eggs, milk & milk products. Citrus fruits, tomatoes, cantaloupe & other melons, berries, green leafy vegetables, peppers broccoli, cauliflower, fresh potatoes. continued . . . 185 Appendix 1: (continued) Background Information on Vitamin and Minerals VITAMINS Vitamin D Vitamin E FUNCTIONS Promotes healthy teeth & bones. Prevents cell membrane damage. IMPORTANT FOOD SOURCES Fortified milk products, fish oils, egg yolks, butter, liver, fatty fish (like sardines, salmon, and tuna) Vegetable oil, margarine, shortening, green leafy vegetables, asparagus, rice germ, wheat germ, butter, liver. Vitamin K Niacin Promotes blood clotting. Essential in releasing energy from carbohydrates, fats & protein. Green leafy vegetables, liver: Note: Main source is synthesis by normal bacteria, in the intestine. Organ meats, lean meats, poultry, fish, fortified grain products. Nuts, seeds, beans & peas. MINERALS Calcium Chloride FUNCTIONS Present in the body in greater amounts than any other mineral; needed for hard bones & teeth, for normal behavior of nerves, muscle tone & irritability, & blood clotting. Part of hydrochloric acid, which is found in high concentration in gastric juice and is important in digestion of food in the stomach. IMPORTANT FOCD SOURCES Milk & milk products, dark green leafy veget- ables (except spinach & chard), citrus fruits, dried peas & beans, small fish eaten with bones, tofu (soybean curd) sodium chloride (table salt) 186 continued . • . Appendix 1 (continued) Background Information on Vitamins and Minerals MINERALS Magnesium Phosphorous Potassium Sodium Sulfur Iron FUNCTIONS Found in all body tissues, principally in the bones; is an essential part of of many enzyme systems responsible for energy conversions in the body Is present with calcium, in almost equal amounts in bones and teeth and is an important part of every tissue in the body. Major constituent of fluid inside individ- ual body cells. With sodium, helps to regulate body fluid balance & volume Found mainly in blood plasma and in the fluids outside body cells. Helps to maintain normal water balance inside and outside of cells. Present in all body tissues, related to pro- tein nutrition (is a component of several amino acids); also part of two vitamins: thiamine and biotin. An essential part of hemoglobin, to carry oxygen to body via the red blood cells. Lack of iron can cause anemia. IMPORTANT FOOD SOURCES Nuts and beans, whole grains, green leafy vegetables. Meats, poultry, fish, eggs, whole-grain foods, beans & peas. bananas, oranges, dates, cantaloupes, tomatoes, & baked potatoes, vege- tables, meats, poultry , fish, milk Salt, salted foods, MSG, soy sauce, baking powder, cheese, processed foods such as breads, cereals, ham, bacon, crackers. Eggs, meat, milk & cheese, nuts, legumes. Liver, red meats, dried beans and peas, enriched or whole grain breads and cereals, prunes, raisins. continued. . . 187 Appendix 1: (continued) Background Information on Vitamins and Minerals MINERALS Iodine Zinc Fluoride Copper Manganese Chromium FUNCTIONS Forms part of hormones of thyroid gland, which helps regulate body metabolism. Lack can cause goiter. Needed for tissue repair and normal growth of skeleton. Part of several horomones, in- cluding insulin. In- volved in cell metabolism. Helps prevent dental caries. Helps stabilize bones, teeth. Deficiency can cause tooth decay, osteoporosis. Vital to enzyme system and in manufacturing red blood cells. Need for utilization of iron. Anemia possible, but deficiences are rare. Vital to various enzyme systems involved in pro- tein and energy metabo- lism. Essential for normal bone structure and functioning of central nervous system. Essential for normal glucose metabolism. Helps regulate insulin levels. IMPORTANT FOOD SOURCES Iodized salt, seafood, plants grown near sea. Red meat, milk, liver, seafood, eggs, whole grain or fortified cereal Fluoridated drinking water best source. Oysters, nuts, liver, kidney, whole grain breads and cereals, mushrooms. Nuts, whole grain breads and cereals, tea, vegeta- bles, fruits. Brewer's yeast, cheese, whole grains, meats. continued . . . 188 Appendix 1: (continued) Background Information on Vitamins and Minerals MINERALS Selenium FUNCTIONS Essential role in enzyme systems of animals and proper functioning of blood. IMPORTANT POOD SOURCES Varied diet provides adequate amounts - fish, meat, breads, cereals. Molybedenum Essential to function of enzymes involved in production of uric acid and in oxidation of sul- fites and aldehydes. Meats, grains, legumes. SOURCES: National Health Information Clearinghouse "Health Finder: Vitamins." Office of Disease Prevention and Health Promotion. Public Health Service; U.S. Dept. of Health and Human Services. February 1985. A Primer on Dietary Minerals. FDA Consumer. September 1974. HHS Publication No (FDA)77-2070. Lecos, C. Tracking Trace Minerals. FDA Consumer. July/August 1983. HHS Publication No (FDA) 83-2176. Food and Nutrition Board. Recommended Dietary Allowances. Ninth Revised Edition. National Academy of Sciences. Washington, D.C. 1980. Mayer J. A Diet for Living. New York. David McKay Company, Inc. 1975. Appendix 5. 189 Appendix 2: FOOD AND NUTRITION BOARD. NATIONAL ACADKMY OF SCIENCES-NATIONAL RESEARCH COUNCIL RECOMMENDED DAILY DIETARY ALLOWANCES." Revised 1980 Designed for the maintenance of good nulntum of practically all healthy people in the U.S.A. Fat-Soluble Vitamins Water-Soluble Vitamins Minerals Vita- Vita- Vita- Vila- Thia- Ribo- Vita- Fola- Vitamin (.al- Pirns- Mag Weight Me.gbt prote|n mmA mmD mmF mmC mm (lavm Niatin minB-6 cr/ b-12 in.m phoius nes,um Iron /m« "dm, (years) (kg) (lb) (cm) (in) (g) 120 7 10 2H 62 132 52 34 700 10 7 45 1.2 1.4 16 1.6 300 3.0 8 - 30 500 10 6 45 0.9 1.0 11 1.3 7 10 28 62 .32 52 34 700 10 7 45 1.2 1.4 16 1.6 ~~ ~~ i5 ^ Ma'« " I4 « *> >57 62 45 ,0°° 10 8 5J 7 H 20 4M 30 00 .200 .00 .8 .5 150 15 18 66 145 176 69 56 1000 10 10 60 1.4 1.7 18 2.0 400 3.0 i£ '«■»•« ™ ,M ,?7 7° 56 10,)° I"5 I' S M 16 ! 2 4)0 3 .0 15 .50 23-50 70 154 178 70 56 1000 5 10 60 14 1.6 18 11 4UU 3.0 5'+ 7° 'H ,?87 ?0 ffi *Z 10 *8 50 ii Is ' 00 ,200 300 1" .5 ,50 Females ,1 .4 46 101 157 62 46 800 0 8 50 . .3 5 I ^ ^ ^ S 15-18 55 120 163 64 46 800 10 8 60 . .3 I AO 40 O 19 22 55 120 163 64 44 800 7.5 8 60 1.1 1.3 4 2.0 400 3.0 » .»* w» » ion ifi* fi4 44 800 5 8 60 1.0 1.2 13 2.0 400 3.0 800 800 MM in f II 20 63 64 44 Z 5 8 60 1.0 1.2 ,3 2.0 400 3.0 800 800 300 10 ,5 .50 , 5,f 55 ,2° ,63 ^ Jo +200 +5 +2 4-20 +0.4+0.3 +2 4-0.6 4-400 4-..0 4-400+400 + .50 +5+25 \ZZZ ___________+20 +400 +5 +3 +40 +0.5 +0.5 +5 +0-5 +.00 +1.0 4-400 +400 +150 * +10 +50 • The allowan, es are intended to provide for individual variations among most normal persons treatment with enzymes (conjugases) to make polyglutamyl .onus of the vitamin available to as thev live in the United States under usual environmental stresses. Diets should be based the test organism. on a variriy of common foods in order to provide other nutrients for which human require- • The recommended dietary allowance for vitamin B-12 m m.ants » based on average concen- men.s have .Ken less well defined. See text for detailed discussion of allowances and of tration of the vitamin in human m.lk. The allowances alter weaning are based on energy mm ients not tabulated See Table 1 (p. 20) for weights and heights by individual year of age. intake (as recommended by the American Academy oi Pediatrics) and consideration ol other See Fable 3 (p. 23) for suggested average energy intakes. factors, such as intestinal absorption; see text. • Ret.nol equivalents 1 retinol equivalent = 1 Mg retinol or 6 Mg 0 carotene. See text for "The increased requirement during pregnancy cannot be met by the iron content of habit cal< ulat.on ol vitamin A activity of diets as retinol equivalents. American diets nor by the ex.st.ng iron stores ol many women; therelore the use ol 30- <" As i holcialc ilerol. 10 fig cholecalciferol = 400 iu of vitamin D. mg of supplemental iron is recommended. Iron needs during lactation are not subst * a-totopherol equivalents. I mgective ccmmunities. Depending on the circumstances involved at the time of the prog ran development, the sponsoring organization could be private non-profit or local government. Chinatown Health Clinic, located in the heart of Chinatown in New York City, is an exanple of such a clinic. It is supported by an Urban Health Initiative Grant, National Health Service Corps, Medicaid, and a small amount of third party and patient fees. Their total budget in 1983 was $549,000. Responding to the great need for bilingual services, the Clinic provides bilingual primary care services and health education to the Chinatown residents and the surrounding community. 196 The demographic profile of tne clients seen in the Clinic annually are as follows: -Over 90% of the Clinic's patients are belcw 150% of national poverty line. -Cver 90% are of Chinese descent, including ethnic Chinese refugees from Southeast Asian. -Over 56% do rot have any sources of support for health care, including Medicaid and Medicare. -Many are workers in the garment industry's sweatshops. Because of the extremely low economic status of the client population, the Clinic is heavily dependent on government funding and will remain so for a long period of time. The unit cost for a physician visit in 1983 was $30.95. While Chinatown Health Clinic is primarily supported by Federal grant funds, the Asian Health Services (AHS), established in 1974 in Oakland, California, is supported by state and local funds to a significant degree. Because of the presence of a greater variety of Asian groups in the area, the clinic is utilized by several Asian groups. For this reason, the clinic is staffed by bilingual staff who speak Cantonese, Mandarin, Vietnamese, Korean and Tagalog. Among the 2500 patient families who are users of the clinic services, their demographic profile includes: -60% who are Chinese; 10.6% Pilipinos; 8.7% who are Koreans; and an increasing number of Southeast Asian refugees. -Cver 83% are non-English speaking immigrants. -70% are on Medicaid or have no insurance. -The number of average visits to the Clinic per user was 4.4. Like the New York clinic, the Oakland clinic is also funded by Urban Health Initiative. In adition, they receive stote *^loca?- f^1?*' SL well as other third party insurers and client fees. Their budget in 1984 was approved at $680,000, whhich was to serve the special needs of the Asian residents of Oakland and surrounding ccmmunities, which have a rapidly increasing Asian population, estimated to be about 100,000. In addition to the primary care services, other services provided at the Clinic include lab tests, immunization, x-rays, and optometry services. The staff of 6.43 inlcudes 2.04 physicians, an optometrist, health educator, social worker, and 1.39 health aides. Possibly because of tne Civil Service restrictions involved in personnel hiring, requiring stricter professional and educational qualifications, there are fewer government sponsored programs. However, one example of such a program may be found in San Francisco, at the Northeast Health Center, #4, located in Chinatown. It is funded and operated by the San Francisco Health Department. It provides free or low- 197 cost bilingual preventive health services, which include: a well-baby clinic, prenatal care, family planning, nutrition counseling, WIC, dental care and supplemental food screening. The Center's budget of $1,670,839 (FY 1983-84) is supported by combined funding frcrn federal, state, and city (San Francisco) government. The 39 health care professionals ard paraprofessionals available at the Center include physicians, nurses, nutritionists, health educator, ard health aides. The majority of the clients of the Center are Chinese, with utilization ranging among various programs from 60% to 100%. 2. Cbmmufiity Based Comprehensive Health Center: Although there are many advantages to having a comprehensive health center designed specifically for Asian Americans, it requires the presence of a large user population ard tne availability of sufficient funds. Even then, because of tne complexity involved in organizing a multi-service system, only a handful of communities have succeeded in establishing such a center. Northeast Medical Services (NEMS) in San Francisco is considered the largest community health center serving a predcminatly Asian population in tne country. Funded by federal, state ard local governments, NEMS has been in operation since 1971, when the federal government was making an effort to provide services in Medically Underserved Areas (MUA). Chinatown in San Francisco was identified at that time as one of the MUAs. One of the unique features of the center is its use of both clinic ard community physicians. For primary care, a registrant has the freedom to choose a family physician from either the NEMS clinic or a group of ,contracted private physicians frcrn the ccmmunity. The proportion of patients receiving primary care frcrn NEMS staff vs. contract physicians is 80% vs. 20%. The FTE count of NEMS health care staff is 17.5 FTE currently, ard includes 4 primary care physicians, part- time specialty physicians, dentists, an optometrist, pharmacist, nutritionist, and other allied health workers. In addition to the primary care services, services available at the center include: 4 specialty services (acupunture, urology, podiatry, ard allergy services), basic laboratory tests, radiology services, nutrition counseling, nursing, pharmaceutical services, dental care, optcmetric services, social services, health education and outreach services. Through the 50 participating contract physicians, NEMS also provides a host of other specialty services off site. Although NEMS does not provide hospital care, they maintain a coordinating and referral relationship with several hospitals, including the private Chinese Hospital ard the county-operated general hospital. The annual budget of NEMS for 1983 was $3.1 million, of which $2.1 million were frcrn the combined funding of local, state, ard federal ccmmunity health center grants. Additional revenues of $520,000 were generated frcrn third-party sources, including private insurance, Medicaid ard Medicare. The patient fees, based on a sliding-fee schedule, brought in an additional $480,000. The NEMS patient population served during the year was approximately 23,000, bringing the annual cost per patient to $130. 198 The characteristics of the patients seen at NEMS are as follows: -87% are non-English speaking -98% are Chinese ard ethnic Chinese refugees -19% are elderly -Over 60% are at or below poverty level The service area population of the center, as defined by the City of San Francisco, includes the Chinatown area and has approximately 70,000 residents. Of those 70,000, about 50% are Chinese. However, because of the bilingual, and bicultural capacity of the center, other Chinese in the City are also attracted to these services. The potential user pool, for this reason, could be as high as 50,000 to 60,000. According to the 1980 census figures, the Asian American population in San Francisco is 147,426, which is 21.4% of tne city's population. Although other Asian groups (Pilipino, Southeast Asian refugees, Japanese, Koreans) are also known to have an acute need for specialized bilingual health services, because of the limited resources available, development of services for other Asian groups, with the exception of refugee services, has been fragmented and limited. 3. Health Maintenance Organization: Chinese Hospital, San Francisco While most Asian health service agencies serve a large number of poor clients and require a significant amount of public funds for basic support, a group of Chinese physicians, health care providers and community leaders recently made a bold move to establish a privately financed health maintenance organization. Chinese Hospital, in San Francisco, was initially established in 1925 to deal with persistent discriminatory practices against Chinese in the general health care delivery system. It was founded by broadly based community groups involving a diversity of political, religious, ard social organizations. Initially requiring funding support from the ccmmunity, the hospital now has a solid fiscal and professional base in tne community, with a large roster of specialists (Loo, 1984). The hospital, located in the heart of Chinatown, has 110 affiliated physicians ard over 200 ancially staff. The number of patients admitted during the past year (1984) was 2,350. Cver 95% of the patients are Chinese, of wham nearly 80% are Medicaid ard Medicare eligible, and the majority are non-English speaking. The annual operating budget of the hospital is approximately $10 million dollars; the daily bed rate is $530. Recently, tne Chinese community leaers have been concerned about the growing trerd amoung employers, including those of the working class Chinese, to purchase economically more advantageous health maintenance organizations (HMO) or preferred provider type health plans for their workers. This could mean that many of the working monolingual Chinese will be deprived of access to local bilingual health care services. In order to forestall further erosion of the services in the Chinese community, the health care leaders in the community have established an exclusive health provider organization of their own. The Chinese Ccmmunity Health Plan, established in 1984, is a collaborative effort of Chinese Hospital, the California Physicians Insurance Corporation, and over 70 physicians. 199 The developmental cost for the Plan has been over $250,000. In order to become competitive with the larger, more established HMOs in the area, each of tne partners in the venture has agreed to accept lower reimbursement rates. The monthly membership rates are: individual - $62.80; two party - $125.10; family - $179.90. The eligibility for the plan is limited to the working employees of policy holder organizations. Although the plan will rot be able to assist the unemployed poor, it is targeted toward lcwer income working employees by keeping its premium at the lowest available rate. This is a bold and new attempt at meeting the needs of working Asians whose resources are limited and their needs somewhat different frcrn those of the general population. Multi-Service Model: While other minority groups nave touted the value of locating a variety of human ard health services under one agency or one location (Padilia et al., 1975), only a few Asian American ccmmunities have succeeded in creating such a system. Boston's South Cove Community Health Center is a successful exception (Lee, 1979). Also located in Chinatown, tne Center was established in 1972. It provides not only primary care services, dental, eye, nutrition, and health education, but it also has a well organized mental health component. This strategy was based on the belief that the scmatizing tendency of Asian Americans cannot be properly treated without coordination between health and mental health. In order to provide one stop assistance to multi-problem Chinatown residents, the Center works closely with other service agencies under one roof, i.e., programs offering housing assistance, vocational training, legal services, social services, day care and youth programs. Funding for the health center is largely public, coming from federal, state, ard city government. The Center's budget for 1984 was approximately 2 million dollars. The user population is well over 10,000 and the encounter visits over 40,000. Cver 90% of the patients are Chinese and over 70% are non-English speaking. In addition to the 3.75 FTE physicians, there are approximately 14 FTE allied health professionals and ancillary service staff within the health unit. Services for Special Groups: Although special needs were identified for tne elderly, women, or youth groups, it is more difficult to develop categorical health programs designed especially for these groups because of the small numbers in target populations. However, possibly because of the cultural value placed on the care of the elderly, there seems to be support for service development for this groups in various Asian communities. Among the existing services, On Lok Senior Health Services is the most ambitious and comprehensive system in the country (Lew, 1984). Established in 1972 in San Francisco as a non-profit, free-standing, it is a cemmunity-based long-term organization serving the frail elderly. With a client population of approximately 300 elderly, about 75% of whem are Chinese, the agency is supported by the budget of approximately $1.7 mixiion. Although they were previously supported by the Administration on Aging (AOA) and the Health Care Financing Administration (HCFA) as a 200 research and demonstration project, the agency took a new financial position in 1984 as a demonstration project. It assumed the total financing risk of providing for all health services for the registered patients. This demonstration project will assess the feasibility and desirability of provider assumption of risk as a means of improving the quality and controlling the costs of long-term care. Cn Lok has obtained service waivers frcrn Medicare ard Medicaid with research ard development funding frcrn a consortium of foundations: Robert Wood Johnson, Hartford, Kaiser Family ard Retirement Research. The new venture attempted by Qri Lok is quite unique in that it is the only organization in the country to assume financial risk, under a fixed capitation rate, for an all-inclusive range of services provided to an elderly population. On Lok's capitation rate is $1400 per month. For Medicare-Medicaid eligible participants, Medicare's monthly share is $650, and Medicaid's is $750. For non-Medicaid participants, a copayment system involves participants and family members in financing the cost of care. The services offered by the agency include acute hospitalization, nursing heme, pharmacy, contracted professional services such as dentistry, optometry and all the subspeeially medical services. On Lok operates three adult day health centers ard was the first such program in 1978 to be designated as eligible for Medicaid benefits. In addition, they manage On Lok House, a HUD development of fifty-four apartments housing 66 individuals, and 6 communal living facilities housing 3 to 6 elderly residents person each. Also at the site, there is a 2-bed respite unit for the clients of On Lok programs so that those needing temporary attendant care may be accomodated. The unit is atterded by health workers 24 hours a day, with easy access to nursing and medical staff. Heme health services, transportation, ard nutrition programs complete the range of services provided. Although the example cited is a program targeted for the elderly population, other specialty services ne^/e also been developed in various cemminities. Although not as extensive as the ax lok model, they are designed to meet the special needs of particular groups, such a women's health, substance abuse programs ard youth programs. Depending on the resources available in tne community, it may be more feasible to develop such specialty services than to develop a more cemprehesive program. PROBLEMS IDENTIFIED BY EXISTING SERVICES: Many of the Asian health services initiated during the past several years have been successful in overcoming a number of problems identified in earlier pages as barriers to Asian service utilization. However, all of the programs cited have irdicated their concern that a greater effort must be made to deal with a variety of additional obstacles. Following are seme of the more common issues: 1 While the number of Asian immigrants continues to grow, increasing both in size ard the number of ethnic groups represented, the resource allocation has not shifted according to the changing nature of the population. The programs, therefore, are forced to extend already cver-stretched resources to meet the need of the new populations or to ignore their needs. 201 2. The current system of funding and reimbursement, which is leaning more toward capitated costs, does not allcw for the cost of translation or outreach to work with these difficult, multi-problem populations. 3. Current Immigration ard Naturalization Service regulations, penalizing immigrants frcrn seeking any kind of public assistance, including medical care, are ccmpourding the dilemma of the struggling immigrant groups. 4. While it is more feasible to develop bilingual services for the more dominant, larger Asian groups, it is much difficult to develop resources for smaller Asian groups, whose individual needs may be just as great. 5. Although there is a definite trend toward prepaid health care systems in the health care industry, specialized ethnic services are lost in the process by larger organizations. At the same time, because of their small size, it is difficult for most Asian health care organizations to develop a competitive prepaid system. 6. While many of the existing Asian health care providers are finding the advantages of preventive programs in reducing the utilization of more expensive services, the reimbursement policies of government and private industries is to eliminate such costs. 7. While there is a significant concentration of bilingual Asian health care providers in certain metropolitan cities, it is very difficult to recruit such personnel in areas with fewer amenities and a smaller number of Asian residents. 8. While an area may have an adequate number of non-Asian physicians, the lack of appropriate bilingual physicians could qualify that are an MUA for the Asian population. However, because of the simplistic application of the federal regulations, the non-English sp)eaking patient groups are under served. RECOMMENDATIONS: Based on the experience gained from various Asian communities, we are new beginning to understand seme of the essential ingredients for the successful health services within our communities. Because of the limited resources available in the Asian American ccmmunity, much support is still needed from federal, state ard local governments. As this paper was developed for the purpose of review by the Task Force on Asian Health under the auspices of the Department of Health and Human Services, the following recommendations are being made to the Task Force for consideration: 1. Federal support should be sought for the allocation of funds to implement Asian health services which are culturally ard linguistically accessible to the ccmmunity. 2. Federal mandates should be developed to require bilingual personnel in health care agencies serving Asian clients with limited English speaking ability. 202 The Federal Goverrment, particularly DHHS, should provide support and incentives to the innovative service delivery models currently in the Asi< American ccmmunity. The federal goverrment should develop mandates to the states so that federal furds being administered by the states will be appropriately allocated for program development in the Asian American community. Federal grant support should be provided to conduct research into the feasibility of innovative Asian health services. Federal support should be provided to encourage tne training of Asian health care professionals in both Eastern ard Western health care practices. Demonstration Centers should be established by federal funds to provide technical assistance to ccmmunities with urderdevelopjed Asian health resources. Federal regulation should be revised to allow flexibility for the special needs of Asian population, e.g., MUA description criteria, reimbursement rates, allowable service modalities. 203 REFERENCES Asian Health Services. Progress Report: Oakland, CA: 1984. Chan, C. and Chang, J. K. "The Role of Chinese Medicine in New York City's Chinatown," Am. J. of Chinese Medicine, 4, 1, 31^15, 1975. Chinatown Health Clinic. Needs-Demand Assessment. New York: 1983. Commission on Civil Rights. Success of Asian Americans: Fact or Fiction. Washington, D.C: 1980. Easton, J. ard Weil, R. J. Culture ard Mental Disorders. Glencoe, 111.: The Free Press, 1955. Fraumeni, J. F. ard Mason, T. J. "Cancer Mortiality Among Chinese Americans, 1950-69," J. of the National Cancer Institute. 52, 3, 659-665, 1974. Gordon, T. "Mortality Experience Among the Japanese in the U.S., Hawaii ard Japan," Public Health Reports. 72, 543-553, June 1957. Hessler, R. M., Nolan, M. F., Ggbrue, B. ard New, P. K. "Intraethnic Diversity: Health Care of the Chinese Americans," Human Organization. 34, 3, 253-262, Fall 1975. Huang, C. ard Grachow, F. The Dilemma of Health Services in Chinatown, New York City. New York City Department of Health. 1974. King, H. and Haenszel, W. "Cancer Mortality Among Foreign and Native Born Chinese in tne U.S.," J. of Chronic Disease. 26, 623-646, 1973. Lee, E. "Mental Health Services for the Asian Americans: Problems and Alternatives." In Commission Civil Rights. Issues of Asian and Pacific Americans. 734-757, 1979. Li,F. P. "Chinese Community Health Task Force Study," Am. J. of Public Health. 62, 4, 536-539, April 1972. Loo, C. M. and Yu, C. Y. "Pulse on San Francisco's Chinatown: Health Service Utilization ard Health Status." Amerasia Journal. In Press. Low, H. On-Lok Senior Health Services - Community-Based Long-Term Care. Unpublished Paper. 1984. North East Medical Services. Health Currents. San Francisco: 1984 Oriental Service Center. Health Surveys. Los Angeles. April 1970. Pad ilia, A. M., Ruiz, R. A., and Alvarez, R. Ccmmuriity Mental Health Services for the Spanish Speaking/Surnamed Population. American Psychologist, 30, 892-905, 1975. 204 REFERENCES (continued) Selltiz, C. Jahcda, M., Deutch, M., and Cook, S. Research Methods in Social Relations. New York: Holt, Rinehart, and Winston, 1959. Starr, P. The Social Transformation of American Medicine. New York: Basic Books, 1982. Tseng, W. S. "The Nature of Somatic Complaints Among Psychiatric Patients: The Chinese Case," Comprehensive Psychiatry. 16, 3, 237-245, 1975. U. S. Depot, of H. E. W. Towards a Comprehensive Health Policy for the 1970 's. A White Paper. Washington, D.C: U. S. Government Printing Office. 1971. Weaver, J. L. National Health Policy ard the Underserved: Ethnic Minorities, Women, and the Elderly. St. Louis: Mosby, 1976. Wong, M. A Survey Report on Medical Care in Oakland's Asian Ccmmunity. Oakland, CA: Asian Health Services, May 1975. 205 Asian-White Mortality Differences: Are There Excess Deaths Elena S. H. Yu, Ph.D. Associate Professor of Sociology in Psychiatry and Research Associate Ching-Fu Chang, M.S. Graduate Research Assistant Department of Sociology William T. Liu, Ph.D. Professor of Sociology and Director Stephen H. Kan, Ph.D. Research Associate Pacific/Asian American Mental Health Research Center University of Illinois at Chicago Chicago, Illinois ACKNOWLEDGMENT The authors are grateful to Phyllis Flattery for her helpful suggestions and encouragement, and to Aiko Igarashi for her timely assistance. Paul Kurzeja's and Medy Masibay's tremendous patience in typing the tables in this report is also appreciated. 208 ASIAN-WHITE MORTALITY DIFFERENTIALS: ARE THERE EXCESS DEATHS? Asian Americans are the fastest growing segment of the U.S. population today. According to the 1980 Census, the number of persons who originate from the Asia/Pacific Triangle has ballooned by 120 percent over the past decade—to 3.5 million—while the number of whites has grown by 6.4 percent, blacks by 17.4 percent and Hispanics by 60.8 percent. Factors accounting for most of this increase are: immigration, births, and inclusion of new groups in the census definition. The term Asians comprises a number of diverse groups; in many ways, they are as different from one another as they are from other races. More than 20 Asian populations were reported in the 1980 Census. Table 1 shows the number and percent distribution of some of these groups in 1980 compared with 1970. Chinese, historically the first Asians to enter the United States and the first group to be legally barred from becoming U.S. citizens (by the Exclusion Act of 1882), emerged as the largest Asian population in 1980. The Japanese, who were the largest group in 1970, fell to third in 1980, surpassed by Pilipinos, who were the second largest group in both 1970 and 1980. Table 1 With the passage of a law in 1981 which allotted a separate immigration quota of 20,000 persons per year for Taiwan, in addition to the 20,000 assigned to Mainland China by virtue of the 1965 Amendments to the Immigration and Naturalization Act, Chinese population growth in the coming decades will accelerate more rapidly than that of any other Asian group. Yet, except for an overview of the occupational transition made by Chinese Americans over the last one hundred years (King, 1981) and a few pieces of work based on analysis of 1960 data (Kitagawa and Hauser, 1973; King, 1974), little has been published on the health of Chinese Americans or, for that matter, Japanese and Pilipino Americans. Indeed, an examination of the three most current bibliographies on Pacific/Asian Americans (Yu, Murata and Lin, 1982; Doi, Lin and Vohra-Sahu, 1981; Vohra-Sahu, 1983) reveals a dearth of research on this subject. I. Objectives The purpose of this paper is to provide a description of the mortality patterns of Chinese, Japanese, and Pilipinos in America, by examining available data extracted from death certificate records submitted by each of the 50 states to the National Center for Health Statistics (NCHS). Created in 1960, the National Center for Health Statistics is mandated to collect, analyze, and disseminate statistical and epidemiologic data on the health of the nation. However, because the 209 size of the Asian American population remained numerically insignificant until recently, national data for this special population are difficult to analyze and interpret even though they have existed for some time at NCHS in two of its vital statistics systems, U.S. birth and death files. Furthermore, since analysis of such data depends heavily upon the availability of population denominators supplied by the Bureau of the Census, the absence of intercensal estimates for Asian Americans in general and Chinese, Japanese, and Pilipinos in particular, has severely limited the use of these records for research purposes. This paper focuses exclusively on the mortality patterns of the three largest Asian groups in 1980 and compare their differences with the majority white population, as well as examines the intra-ethnic differences by age, sex, and nativity. In what follows, the term Asians refers only to these three groups, unless otherwise specified. II. Geographic Distribution and Sociodemographic Profile Although the Asian population was more geographically dispersed in 1980 than in 1970, it remained highly concentrated in the West. In 1980, 56 percent of the Asian population lived in the West compared with 70 percent in 1970. However, the degree of concentration in the West varied among the groups. For instance, about 8 out of 10 Japanese, 7 out of 10 Pilipinos, but only 1 out of every 2 Chinese were residing in the West in 1980. (In contrast, only about 4 out of 10 Koreans and 2 out of every 10 Asian Indians resided in the West.) Regardless of their regional distribution, Asians as a whole lead a predominantly urban existence. Ninety-seven percent of the Chinese live in urban areas, followed by Pilipinos and Japanese, both 92 percent, compared to 71 percent for the majority whites. Perhaps the most impressive sociodemographic characteristic of the Asian populations is their level of education. Fully one-third of Asian adults have finished college, compared with 17.5 percent of majority whites. Further breakdowns by age and sex (Table 2) show that, in the male 25 to 35 age group, 57 percent of the Chinese have four years of college education or more, compared with 25 percent for whites. The corresponding figures for Japanese and Pilipinos are 49 percent and 33 percent respectively. Among females in the same age group, 46 percent of both Chinese and Pilipinos, as well as 40 percent of Japanese, have such a high level of education, compared with 22 percent of whites. The educational advantage is sustained in the next higher age groups, 35-44 and 45-64 years. Among males, almost twice as many Asians aged 35-44 years (55 percent of Chinese, 49 percent of Japanese, and 48 percent of Pilipinos) as whites (26 percent) have at least a college education. For females, more than three times as many Pilipinos (54 percent) and twice as many Chinese (34 percent) have such a level of education compared to whites (16 percent). The corresponding figure for Japanese women (25 percent) is not as strikingly different as those found for the other Asian groups. In the 45-64 age range, only Japanese females have lower percentage (8.9) of population completing 4 years or more of college, compared to white Americans (9.6 percent). 210 Table 2 However, the unparalleled educational attainment of Asians does not necessarily assure the advantage that one would expect in the U.S. occupational structure. The 1980 census sample data show that, for both sexes, 19 percent of the Chinese, 15 percent of Japanese, and 14 percent of Pilipinos have a professional occupation compared to 13 percent for whites. At the other end of the occupational ladder, 19 percent of the Chinese, 17 percent of Pilipinos, and 13 percent of Japanese are in service occupations, compared to only 11 percent of whites (Table 3). Within each Asian group, proportionally more foreign-born persons than native-born ones are in service occupations (21 percent for Chinese, 20 percent for Japanese, 17 percent for Pilipinos, compared to 11 percent among white Americans). Although proportionally fewer Asians than whites are employed as operators, fabricators, and laborers, more foreign-born persons than native-born among the Chinese and Japanese populations are engaged in such types of work. Pilipinos seem to be an exception insofar as nativity is concerned. The proportion of foreign-born persons engaged in professional occupations is more than twice that found for the native born. This striking contrast is undoubtedly due to the large influx of physicians, pharmacists, nurses, and other professionals from the Philippines for several decades. Table 3 Income data from the 1980 census reveal that regardless of nativity, the median family income for Asians is higher than that found for whites, the percentage of families earning $25,000 or more in 1979 being higher for Asians than for white Americans (Table 4). However, caution is warranted in interpreting such types of data because they mask the true poverty situation of Asian Americans. Differences in the family structure and kinship concept between Asians and whites enable large numbers of Asian adults to tolerate the sharing of a household for reasons of exigency, thereby inflating the reported "family" income. Using the 5-percent sample from the 1980 Census Public Use Microdata, Kan and Liu (1984) found that besides the Vietnamese, Chinese and Korean Americans actually have poverty prevalence well above the national level. Table 4 Nevertheless, it is clear that changes in the immigration laws and sociopolitical developments in the international scene in the last 40 years have led to the influx of a new generation of Asian immigrants—one recruited from a different social stratum in their home country than those who came in the 19th century or the first half of the 20th century. Such marked shifts in the characteristics of the immigrant population would 211 suggest a different mortality (and morbidity) pattern compared with the earlier cohorts of immigrants. In the next section, we will examine the health data compiled from the birth and death files maintained by NCHS. We begin first by reviewing the most recent findings on inter-ethnic differentials in infant mortality, followed by general mortality. Next, gender differences are compared between groups and within each Asian group, using the 1979-1981 mortality files. Finally, because three-fifths of the Asian/Pacific Islander population as a whole are foreign-born, we will take a look at the significance of nativity, if any, in the mortality patterns of Asian Americans. III. Infant Deaths Infant mortality rates, that is, deaths under 1 year of age per 1,000 live births, have often been used as an indicator to compare the health of different populations. Data on infant deaths by specific nonwhite races have been available at the National Center for Health Statistics for a number of years. Perhaps the most recent analysis of inter-ethnic infant mortality data which include information on Asian Americans is the one conducted by Yu (1982). Using both published and unpublished data prior to 1980 on Chinese, Japanese and whites, the author compared four different types of rates: fetal (at least 20 weeks of gestation), neonatal (within 28 days of birth), postneonatal (within 28 to 365 days of birth), and infant mortality (Table 5). She found large inter-ethnic differences in fetal mortality (5.9 per 1000 total births for Chinese, 8.5 for Japanese, and 9.7 for whites). Likewise, substantial differences exist between Asians and whites in the neonatal death rates (3.7 per 1000 live births, 5.5, and 10.3, respectively) and in the infant mortality rates (5.5 per 1000 live births for Chinese, 7.4 for Japanese, and 14.1 for whites). Just why these rates are so different is difficult to interpret because of the limited information on the death certificates. We may, however, evaluate some possible record-keeping errors such as underreporting, misreporting of race, or misclassification of time of death, and maternal factors such as age and education. Table 5 A. The Possibility of Underreporting Since infant mortality rate is defined as: the number of deaths in a year of children __________less than 1 year of age_________ x 1000* the number of live births in the same year two possible sources of reporting error exist. The first lies in the denominator which reflects the number of live births; the second may be found in the numerator which shows the number of infant deaths. Generally speaking, underregistration is relatively common among births occurring outside of hospitals. However, findings from unpublished NCHS data for 212 the entire United States during 1973-77 indicate that only 0.7 percent of all reported live births to Chinese as well as to Japanese mothers had occurred outside of the hospital setting. The rate for Pilipinos, though somewhat higher—1.5 percent—is still low. If the underreporting error lies in the numerator, then a systematic study is warranted of the death registration procedures—especially in those States where these three Asian groups are found in large numbers—in order to determine the magnitude of underregistration. Unfortunately, such a study has yet to be conducted. Relevant to the above issues is the possibility that a vital event may not have been registered in both the numerator and the denominator. McCarthy et al. (1980), for instance, demonstrated from their research in Georgia that such blatant omissions can indeed happen, although they seem to occur disproportionately in the rural areas, among unmarried mothers, and for black American infants. In contrast, we know that Asian vital events are more likely to occur in urban areas because of the nature of Asian settlements in the United States. Compared to either black or white Americans, the proportion of unwed mothers thus far is rather low among Asians in general and Chinese in particular. Nevertheless, taking the McCarthy findings into consideration, we must bear in mind that for underregistration of deaths and/or births to be a major source of bias in the observed Asian infant mortality rate, we not only need to provide evidence that underreporting exists, but we also need to demonstrate that the magnitude of underreportage is greater for Asian than for white American infants. In the absence of any systematic study in this respect, we can neither establish nor dismiss it as a source of error. We can, however, momentarily sidestep this issue so that we may proceed to evaluate other possible explanatory factors. B. Misreporting of Race To date, two studies have documented the existence of racial misclassification in births and deaths for Asian Americans. The first uses the 1965-67 linked birth and infant death records in California (Norris and Shipley, 1971) to determine if the race of infants born to Asian parents are recorded consistently at birth and at death. Norris and Shipley reported that the death rates for both the Chinese and the Japanese appeared to be substantially lower than the rates for the white population, mainly because 39.2 percent of the Japanese infants and 13.7 percent of the Chinese infants born during that period were mistakenly classified as white upon their deaths. Adjusting for this error, they found that the Japanese, instead of having a more favorable pregnancy outcome than white Americans, actually had higher infant mortality rates. However, the Chinese cohort rate, especially for the neonatal period, remained to be the lowest of any racial group, despite the adjustment for misreportage. The second study on racial misclassification uses the 1968-77 data from Washington State (Frost and Shy, 1980). Misclassification errors in Asian births and deaths were found to occur. However, the number of 213 Asians living in the State of Washington, especially Chinese Americans, is extremely small. Patterson's advice (1980) to take caution in the use and interpretation of vital records for numerically small racial groups is, therefore, very well taken. One finding that emerged from Frost and Shy's study is that offspring of interracial marriage appear to have a significantly higher rate of discordance between birth and death certificates. However, interracial marriage as a factor in the discordance between birth and death certificates, though relevant, may not be a truly significant factor in accounting for the low Chinese infant mortality rates, although it may be significant for Japanese and Pilipinos where proportionately more interracial marriages have been reported, both in the 1980 and the 1970 census. For these reasons, the analyses that follow will focus exclusively on the Chinese-white differences. C. Misclassification of Time of Death A third possible source of error which might account for the low Chinese infant mortality rates is that the time of death may have been misreported. For example, neonatal deaths, especially in the first day of life, may have been misreported as fetal deaths. If so, we would expect to find an unusually high fetal death rate for Chinese compared to white Americans. Bearing in mind the uncorrected error of possible misreporting discussed above (i.e., some differences in racial classification between birth and death records which we cannot establish nor rule out), we recall that in Table 5, the Chinese fetal death rate (5.9 per 1000 total births) is actually lower than that found for whites (9.7). Indeed, data published elsewhere (Yu, 1982) show that for both fetal and postneonatal periods, the rates for Chinese are the lowest of all racial/ethnic groups. D. The Importance of Record Linkage Frost and Shy's recent publication (1980) and Norris and Shipley's earlier work (1971) demonstrate the importance of matching birth and infant death certificates in order to approximate the true infant mortality rate for different ethnic groups. Since birth and death records in the U.S. are not linked nationwide, Yu (1982) examined the data from California—where record linkages exist and where 40 percent of live births to Chinese mothers occurred during 1973-77. To minimize annual chance fluctuations, the linked data we obtained from California are cumulated over 5 years (1973-77). In addition, data for white Spanish minorities are differentiated from those for white non-Spanish majority. Consequently the observed infant mortality rates for white Americans are not artificially elevated by the higher rates found for white Spanish minorities. The results (Table 6) show that only in the fetal period is the death rate for Chinese significantly lower than that found for white non-Spanish Americans. Insofar as the neonatal death rate is concerned, the observed difference between Chinese and white Americans is not 214 statistically significant. In the postneonatal period, the inter-ethnic difference disappears. If we take these findings to reflect the true rates, then, it seems that some unknown advantage in fetal survival exists for Chinese Americans, but this possible advantage is not strong enough to be sustained through the first 28 days of life, and it disappears thereafter. Table 6 E. Mother's Age and Fetal Deaths Insofar as fetal death rates are concerned, Table 7 demonstrates that the Chinese advantage is evident in every age group, from 20 to 39 years. From the magnitude of the observed differences in rates by age, it appears that factors other than mere reporting or classification errors, are operating. Just what these factors are remains to be explored. Table 7 F. Mother's Age and the Distribution of Total Births Another way of gaining insight into the nature of the differences in the fetal death rate between Chinese and whites is to look at the age distribution of the mothers at the time of the expulsion of the fetus, i.e., birth. An interesting finding from the California data is the differences in the distribution of total births (fetal deaths plus live births) between Chinese and white non-Spanish American women (Table 8). Teenagers accounted for 14 percent of the total births among white non-Spanish women, but only 2 percent among the Chinese. Likewise, a much smaller percentage of Chinese women (19 percent) gave birth between the ages of 20-24 years, compared to white non-Spanish women (35 percent). By far the largest proportion of Chinese women—47 percent—gave birth while they were between the ages of 25-29 years, or slightly older, 30-34 years (23 percent). It is noteworthy that the proportion of births to Chinese women 35 years of age or over (8 percent) is twice that found for whites (4 percent). Given the knowledge that older women have a very high risk of fetal death, the age distribution for Chinese is in fact less favorable for fetal survival than that for whites—and yet their fetal death rate is much lower than that of white Americans. Table 8 215 G. Mother's Education Given the high level of education among Chinese women in the United States, might not the observed low perinatal rates for Chinese be accounted for by the mother's high educational level? Unfortunately, California's birth certificates do not contain information on mother's education. Certificates used in other states with large Chinese populations (e.g., New York and Hawaii) do have information on mother's education but tabulations of linked data for Chinese are not routinely available. However, since the California data show that the advantage, if any, of Chinese appears to exist most distinctly in the fetal period, the author decided to examine the 1973-77 U.S. fetal death files using the race of mother alone to determine the race of child. The advantage of using the fetal death records is that it eliminates the possibility for racial discordance between birth and death records, since fetal deaths are reported on only one record. Table 9 shows the fetal death rate by mother's education. Chinese women have a lower rate at every level of education. Thus, the higher education of Chinese mothers vis-a-vis that of white mothers does not explain the low fetal death rates observed among Chinese. Rather, these findings suggest that the explanatory factors probably lie in the maternal intra-uterine environment which is, of course, highly influenced by the mother's condition and health habits, besides as yet unknown genetic factors. Table 9 Current research on spontaneous abortion and perinatal development in the biomedical disciplines have succeeded in pinpointing certain key factors as critical variables in explaining the fetal death differentials observed in non-Chinese populations. In a number of well-designed and carefully controlled studies, both maternal cigarette smoking and alcohol consumption, for instance, have each been shown to have a dose-response relationship to fetal development. Over 200 published works on the perinatal effects of maternal cigarette smoking have been reviewed in the Surgeon's General Report on Smoking and Health. Therefore, we will only recapitulate the important points below. First, smoking during pregnancy is a risk factor for spontaneous abortions or fetal deaths and perinatal mortality. The highest perinatal mortality risk ratios (smokers versus nonsmokers) reported in the literature is 2.42. Second, the proportion of pre-term live births increases directly with the quantity or frequency of maternal smoking. An estimated 11 to 14 percent of all pre-term deliveries in the United States may be attributable to maternal smoking. Third, full-term babies born to women who smoke during pregnancy are on the average 200 grams lighter than babies born to comparable women who do not smoke. The whole distribution of birth weights of smokers' babies is shifted downward, and twice as many of these babies weigh less than 2500 grams, compared with babies of nonsmokers. This finding has been confirmed by over 45 studies of more 216 than half a million births. Birth weight is affected by maternal smoking independently and to a uniform extent, regardless of other determinants of birth weight. The more the mother smokes, the greater the reduction in birth weight of the baby. This dose-response relationship has been demonstrated in at least 20 studies. What is not clear from the Surgeon General's Report is the circumstances under which maternal smoking produces one type of negative outcome (perinatal death, pre-term birth, or low birth weight) instead of another. Nonetheless, when we juxtapose these findings with reports on the prevalence of smoking among American women, we cannot help but speculate on some possible hypotheses which may explain the observed U.S. white-Chinese differentials in fetal mortality. Might the differentials be attributable to the lower prevalence of smoking among Chinese women compared with white women? If only data on the prevalence of smoking among Chinese women in the United States were available, it would have been possible at least to test the plausibility of this hypothesis on an aggregate level. Unfortunately, our literature search revealed the absence of any recent data base which contains a sample of Chinese American women large enough for us to examine this issue. We, therefore, need to defer the testing of this hypothesis. Smoking, however, is not the only risk factor which might turn out to differentiate between white and Chinese women. Alcohol consumption may be another factor, although it has not been as thoroughly investigated as smoking has been. About the best evidence thus far comes from a case-control study on drinking during pregnancy and spontaneous abortions. Using maximum-likelihood logistic regression analysis, Klein et al. (1977) found that the adjusted odds ratio for this association was 2.62—higher than the highest perinatal mortality risk ratio ever reported for smoking. With non-drinkers as the comparison group, the adjusted odds ratio for drinking daily (2.58) was significantly higher than that for drinking twice weekly or more (2.33), indicating a dose-response effect. Consideration of wine, beer, and spirits separately suggested that the minimum harmful dose was one ounce of absolute alcohol. Several potentially confounding variables, including maternal age, gestation, prior spontaneous abortions, smoking and nausea or vomiting, were controlled for in the analysis. The association between drinking during pregnancy and spontaneous abortions did not vary with these factors. Even moderate consumption of alcohol during pregnancy is a risk factor for, and may be a cause of, spontaneous abortion. Among the possible mechanisms, acute fetal poisoning seems the most likely, although chronic poisoning is also possible. Klein et al. hypothesized that because their karyotype analyses indicate that drinking during pregnancy raised the risk of aborting euploid—rather than aneuploid—conceptions, it is quite likely that spontaneous abortion is a manifestation of the human reproductive sensitivity to alcohol. In this sense, newborns with congenital malformations or fetal alcohol syndromes may well be the rare survivors of 'poisoned' conceptions. In view of the evidence provided by these sociomedical researchers, we are tempted to posit that perhaps the Chinese American women who gave births during 1973-77, being 86 percent foreign-born (non-U.S. residents 217 excluded)—compared to about 8 percent foreign-born in the white population—may have retained a traditional set of sociocultural health habits or lifestyles which contribute to their advantage in fetal survival. At least up until very recently, rigid sex-role socializatio in Taiwan gives men the freedome to smoke and drink but strongly discourages women from cultivating such habits. IV. Ethnic Differences in General Mortality Analysis of the general mortality data reveals that the Chinese advantage in survival is not limited only to the perinatal period. Table 10 shows both crude death rates and age-adjusted death rates. Age adjustment, using the direct method, is the application of the age-specific death rates in a population of interest to a standardized age distribution in order to correct for differences in age structure between and among the racial groups being compared. Without the age adjustment, differences in the observed rates due to age differences in population composition can distort comparisons of overall mortality. Therefore, the age-adjusted death rates are what mortality levels would be if age distributions were identical for all the racial groups being compared. Table 10 For 1980, the ranking from the highest to the lowest mortality rates per 1,000 population are: blacks (8.3), American Indian and Alaskan Native (5.8), whites (5.6), Chinese (3.5), and Japanese (2.9). The crude death rate of blacks is actually lower than that of the white population, demonstrating the importance of age adjustment. This ranking of age-adjusted rates for 1980 is identical to that found for 1970. Japanese Americans have maintained the lowest mortality rates throughout the past 30-year period, while Chinese Americans have maintained the second lowest rates for the past 20 years. Data for Pilipino Americans (not tabulated for 1970) reveal an overall mortality rate for 1980 that is even lower than that found for Japanese Americans. A body of literature exists (Gordon, 1967; Kitagawa and Hauser, 1973; NCHS, 1975; King, 1974, 1981) which suggests that socioeconomic status influences one's chance of staying alive. However, direct evidence of the generalizability of this conclusion to Asian-white differentials is impossible to obtain at present because death rates needed to make comparisons between these races by socioeconomic status are lacking. Indirect evidence is available from the 1980 census which suggests that Asians appear to have high socioeconomic status. In aggregate, They are proportionately more educated than white or black Americans (see Table 2 earlier). Furthermore, notwithstanding the bimodal distribution of occupations, more Asians are employed in high-prestige occupations, especially in the professional and technical fields. Data on income distributions reveal that a higher proportion of Asian families than white reported having incomes in 1979 of more than $25,000 (see Table 4 218 earlier). At the other extreme, we find that whereas 5.6 percent of the white families earned $5,000 or less in 1979, only between 3 to 4 percent of the Japanese and Pilipinos are in that situation. The figure for Chinese (5.9 percent) is comparable to that for white. Therefore, by and large, the evidence suggests the presence—on an aggregate level—of an inverse relationship between socioeconomic status and mortality which may account for some of the observed inter-ethnic mortality differentials. Indeed, the following data strongly indicate that the effects of socioeconomic status may be operating in age-specific deaths, and in mortality from specific causes of death. A. Age-specific Differentials Examination of the age-specific mortality rates (Table 11) indicates that the age pattern of deaths is similar across racial/ethnic groups. At the youngest age group (0-5 years), death rates are high, but they drop to a minimum in early childhood (5-14 years), only to increase steadily to a maximum at the oldest ages. One notes, further, that in every age group, the rates for Asians are lower than those reported for white Americans. Table 11 B. Race-Mortality Ratios By dividing the age-specific death rates for each Asian subgroup with the corresponding white rates, we may obtain race-mortality ratios for the different ethnic groups. A ratio of 1.0 means that the death rate for the minority population is no different from that found for the majority white population. On the other hand, a ratio of greater than 1.0 suggests the number of times by which the minority death rates exceed the majority rates. Table 12 shows that for every age group, the mortality ratios are less than 1.0—which signifies that, for All Causes of Death, white death rates exceeded Asian rates. Table 12 Sufficient evidence exists to indicate that there is a positive association between social class and overall mortality and that the class differentials are largest in the middle years of life. Table 12 shows that the race-mortality ratios are lowest in the 15-24 and 25-34 age groups for Chinese—implying that the differences between white and Chinese are greatest during these years. For Japanese, the ratios are lowest in the 15-24, 25-34, and 35-44 age groups, increasing slightly in the 45-54 age group, but declining markedly in the 55-64 age group—a pattern which was replicated by the Pilipinos, except for the 45-54 age group rise. 219 These findings lend confidence to the tentative conclusion that perhaps, beyond the issue of errors in filing death certificates, there exists a strong socioeconomic factor which accounts for the observed white-Asian differentials, especially in the early and middle years of life. C. Leading Causes of Death Heart disease, cancer, cerebrovascular disease, and accidents have been the leading causes of death in the United States since around 1950 (Fingerhut, Wilson, and Feldman, 1980). Table 13 shows the 10 leading causes of death for the United States and the rank order of these causes for whites and Asians. Insofar as the first 4 leading causes of death are concerned, all four groups have identical rankings. However, they differ in proportional mortality. In 1980, heart disease accounted for nearly two-fifths (39 percent) of all deaths for white Americans, compared to about one-third of the Asian deaths (32 percent for Chinese, 30 percent for Japanese, and 34 percent for Pilipinos). On the other hand, cancer accounted for a larger proportion of deaths among Chinese (27 percent) and Japanese (25 percent) than among whites or Pilipinos (each 21 percent), while cerebrovascular diseases made up 11 percent of all deaths for Japanese and 10 percent for Pilipinos, and about 9 percent for whites and Chinese. Table 13 In terms of ranking, we note that pneumonia and influenza rank fifth for the Asian groups, compared to sixth rank in the total U.S. population, while suicide ranks higher as a cause of death in both the Chinese and Japanese populations (seventh and sixth, respectively) than in the Filipino group or in the total U.S. population. Table 14 shows the age-adjusted race-mortality ratios for specific causes of death. Here, we note that in the area of accidents, the age-adjusted mortality ratios for Chinese (.34), Japanese (.44), and Pilipinos (.39) are extremely low. From a public health point of view, deaths due to accidents is of considerable interest because it is for this particular cause of death that the differential consequences of social class will be most visible and detectable. After all, it is only when the cause of death is preventable that one's social class is predictive of one's ability to command resources to engage in prevention behavior. Having said that, we note that for such a preventable cause of death as accidents, the Asian mortality rates are only about one-third to two-fifth the size of the white majority rate. This finding gives a strong hint of the possible combined effects of culture and socioeconomic status on deaths due to accidents. Table 14 220 On the other hand, heart disease, cancer, and cerebrovascular disease—the top three leading causes of death in the country and the leading ones as well for the ethnic groups under comparison—show a different pattern. In general, the ratios for heart disease, where the largest number of deaths have been reported, range from 0.4 to 0.5 (if we round off the figures). This finding suggests the possibility that perhaps, social, cultural, dietary, as well as environmental factors are at play in heart disease—more than we ever realized in the past. The near similarities in race-mortality ratios experienced by the three Asian groups vis-a-vis white Americans pinpoints new avenues for prevention research in heart disease. Deaths from atherosclerosis ranks 9th in the white population's leading causes of death, but it ranks only 14th for Pilipinos, 13 for Chinese, and 10th for Japanese. Table 14 indicates that the mortality ratios for atherosclerosis and heart disease are fairly similar for Chinese (0.57 and 0.54, respectively) and Japanese (0.41 and 0.42, respectively), but not for Pilipinos, a group which—for historical and cultural reasons—behaves differently from Japanese and Chinese. Likewise, low mortality ratios are found for COPD among Chinese, Japanese and Pilipinos (0.50, 0.34, and 0.31, respectively). It is possible that this may be attributable to proportionally more non-smokers and/or possibly ex-smokers among these Asian groups compared to whites. However, among Asians, behavior such as abstention from smoking may very well be confounded with social class. Therefore, to the extent that smoking can be a good predictor of social class, to that extent high socioeconomic status will be inversely associated with such diseases as COPD. That lifestyle may be an important factor in the differential mortality rates of Asians can be surmised from another cause of death for which the age-adjusted mortality ratios are nearly as low as—if not lower than—those found for accidents. In Table 14, the ratios for chronic liver disease and cirrhosis of the liver, for which heavy alcohol use and abuse is known to be a definite risk factor, are striking. Indeed, this particular cause of death represents the lowest mortality ratios—that is, the greatest Asian-white differential—for Pilipinos (0.29) and Japanese (0.34, which is identical to that found for COPD), and the second lowest ratio observed for Chinese (0.42), next only to accidents (0.34). Insofor as cancer is concerned, the age-adjusted mortality differences between white and specific Asian groups are small (Table 14). They range from 0.4 to 0.8 depending on which Asian group one looks at. Given that cancer is a disease which includes malignant neoplasms of various parts of the body, little can be said about the variation in the race-mortality ratios for this specific cause of death without a lengthy discussion into the genetic vulnerability and socioenvironmental risk factors for each type of cancer across the racial groups being compared. Moreover, in order to do justice to inter-ethnic comparisons of cancer death rates, attention must be given to the proportional mortality of each type of cancer in the overall cancer deaths for each group. Chinese, for instance, have unusually high morbidity and mortality rates from 221 nasopharyngeal cancer (NSP), a very rare disease. In her review of the literature, Yu (1982) found that for all countries of the world in which cancer registries exist, the average annual incidence rate for nasopharyngeal cancer is less than 1 per 100,000 population (age-adjusted to world population). However, the only areas or populations which have reported average annual incidence rates of greater than 5 per 100,000 are: San Francisco Bay Area Chinese (19.1 among males; 6.4 for females); Singapore Chinese (18.7 for males; 7.1 for females), and Hawaii Chinese (10.3 in males; 5.1 in females). These figures excluded data from China and Taiwan, where NSP rates are known to be very high but where data were unavailable for review at the time. The Chinese proclivity for nasopharyngeal cancer may well boost this group's cancer death rate vis-a-vis other Asian groups. Research now shows that Chinese have a genetic susceptibility for this type of cancer, in addition to a high exposure to chemical agents formed from ingestants that popularly consumed in the folk diet (Ho, 1979). For cerebrovascular disease, the ratios hover in the range of 0.7 to 0.8, with Japanese and Chinese showing a ratio of 0.76, and Pilipinos, 0.66. These rates are sufficiently close to suggest greater similarity within the Asian group and some amount of differences with white death rates. On the other hand, the category of pneumonia and influenza, shows as high—if not higher—mortality ratios for two of the Asian groups (0.81 for Chinese and 0.73 for Japanese). For diabetes mellitus, the mortality ratio for Chinese is again relatively high (0.81), but for Japanese, it is somewhat lower (0.64), and for Pilipinos, even lower (0.49). The reasons for these differential ratios by cause of death and across ethnic groups are far less clear. Additional review of the literature would be necessary to evaluate the relative importance of socioenvironmental risk factors for these different diseases and across race. One cause of death for which biological factors can be safely ruled out as an important one is suicide. Here, we note that the ratios for Chinese and Japanese are fairly similar—their rates being three-fifths that found for white Americans. The ratio for Pilipinos is even lower, 0.30. These dissimilarities in deaths from suicide within the Asian subgroups strong suggest the suicide is closely interrelated with culture, besides age and sex. A separate study on suicide between Asians in Asia and those in the United States is necessary to search for clues as to what accounts most for these observed differentials. D. Sex Differences In this section, gender comparisons are made within groups and between groups. Table 15 shows the within-group differences in mortality rates for All Causes of Death. In every ethnic group, the age-adjusted male death rate in 1980 was slightly higher than the female rate, resulting in a sex-mortality ratio of greater than 1.0. Table 15 222 to OJ a 13 13 O 13 M Ml h M H- O 13 Hi O O 3 13 fD fD 0) 01 0) 0 »-( 1-5 3 3 0) M O fD ft rt fD rt M fD 3 h- 3 3 O M)Ml^ O 0) ft fD O O u> 3 O far as malignant neoplas al mortality is higher f 6 percent for women vers r female versus 22 perce r female versus 27 perce s were observed; the pro 13 3 3 C 0 3 i-3 0 rt rt oi h 0) Oi 1-5 0" (t Hi Hi W ^ 0) t-> H- 0 0 O 0 H ro 0 H 1-5 3 fD 3 13 ro h-1 0i 3 3 01 3 O ««j H 0i 01 H 0 HH(5 rt 3 3 fD (D fD 3* O 0 01 0) 3 0i fD H ^^ ft 3 M rt • - 3 D) Hi i-h fD H* Oi 0 0 Pi H- > 3 H H rt 3 O. ^ ^< O 3 3 01 3 £ o> oi ro l-h vO 3* 3 3 0) 0 H- -' 1-5 C_| rt » H- KJ 0) fD 3 0i 3 13 D cr fD 0i > H- rt M 3 3 3 !-• 3" CD fD fD H- H Ol 0) H 13 0) H 3 fD H- H- fD ~J a " O 3 ^- 01 0 «£> - £ 3 3 0) H 0 0 0) 0 rt 3 — C 3* 0) 0) -^ to 13 ro 3 C to Ol 0) O VD H- 3" 0) Ol Hi 01 Hi H- 13 OJ 3 O 3 O 3 O ,an ttn h H 13 O U) £ O O 13 H 3* £ 3" H- 0 H- 0) H- H 13 rt CL 3 h- 0 0) 01 13 H D* 0) H- n > ►< 0) 3 H 3 , ° " ro c_i ^> cn M 01 *» h- u o —» o oi *> Oi 3 U OJ 3 3 ro oi 0) 0) h 13 o oi 01 h ui^3 n m w rt ro h 3 ft Hi ft^ o M Hi Hi Hi "* o o 3 H H (Jl OJ. 3 ro oi c ft 3 M 3 3 *- 0) rt ro O 0i M ft Hi 01 I- o hi «• ro HO 01 0) rt 0) -ft fl) H 01 to 3 rt 13 — oi H ft Oi O H- 3 fD O Qi 3 3 rt q> ai w M CT O Hi C O rt H 0) rt 0) cr c o rt 3 0) fD XJ O 3 C H rt oi fD 13 fD 13 h ro G H ro n 0 H rt i-1 Oi H O H- C rt Hi «• 0 H O 0 3 3 Oi 13 to 0 h u fD 01 O O H ifl Oi G H fD fD M 01 U Di3 MO) O O 0) Hi uQ 3 fD Oi fD i_j. < ifl O fD H H H O *< c O 13 B) ft - Qi O H H- fD 0 Du ifl < 0) 0) 0) 0i H 0) ft Oj G 3* M C 0) fD h-1 0) 01 0) 0) H 3 O O a 3 hi H- uQ s 01 oj. O H G 0) H Oj H- O < 01 01 G c H 0) H O H- 01 fD 01 01 fD 01 Ol 3 O 3 01 Hi ifi O H H O c U 13 H- I H" I h- s: 13 3* h- ro 3 H O ro oi *< 01 3 a 01 cr O rt C 3" rt H to Oi • rt o ro 01 Oj > fD 0) 0> O c H el- s' O 0 O 01 fD H 01 - 0) fD H a ro < < ro fD H H 0) 01 c c 0) 01 rt h- 3" rt rt ^< O *- < H Oj 01 Hi 0) OJ J> OJ O OJ ID 3" 0) H HJ 0) ft Oj 01 3* O 0) fD 0) fD oi •<: — o c G 3 O o3 id 3 0) 3 rt 3 fD O 13 D* «o oi ro <*> Hi K O fD UI Oj 13 13 h- h- rt 13 13 01 13 01 3 3" ro oi h oi ro o oi H H O H 0i 0) rt o o ro o oi ro ro 3 o> ro 3 3 ft 3 0) ft rt rt Hi 0i fD Ml Hi O Hi H X O O h O ro h h h Hi H - Hi Hi 01 Hi 01 Hi 3 fD fD 3 fD H Hi 0) 3 3 0i 3 <0 fD 01 01 0) (—' Ol fD H 0) H H fD H 0) fD fD fD 0) fD ft 3 '— 0) 0) w 0) G U) «— >-^ •» ^ 0) I-1 • 01 Hi O 0i O. 13 0) 3 13 0i H fD a 0 ro oi oj o a -j c H 13 Oi 01 01 Hi H- 0) 3 3 0i ft H1 3" 0) fD a s ro >< oi o ft c 3* 3 •JJ Oj 01 Hi rj- oi fD 13 Pi H Oi o cr Oi ft 3 cr 3- o O 3 C 01 lO ft 01 x s: M fD 3 01 H iO 01 3 cr 0) ^< 01 O H a 01 s 01 I I cn on ft (I) 0) K H 0 Hi 0) rt oO 3" H- 0i ft ft 3" H O H- 01 3 rt -J | H- i» ifl O H 0i O OiO c Hi fD 13 Ui to 13 3 0) H- 0) 0i uQ 0) H H 0) O fD 0) ft 0) 01 3* 3 cr ro rt *~. fD HJ >—' to ft 0) H X ai £ 01 Ol ro o H- Z, 13 ro o H H- 01 3 3 H- rt H a 3 3" G H uQ 0) Ul t-1 Cj, 3 0i cr 0i rt Ol H 01 13 •—' 3 ifl 3" Oi > a ro H- 3 01 3 01 Hi to ft O- 01 O >o 0) M • 3 Oi H 0i 0 ft O OJ. 0 3" s: H O H- fD fD 31 H a O ro 01 I-1 O 3 O cr i: rt « k: H c 01 ro 0 oi 01 HJ 1 01 rt H- 0 01 H H O .—. H- rt Hi 01 Ul 13 H- 3 H- o a 13 3 3 ro H- ro O 0i 01 3 H 01 H ft 13 o 3" O fD ,—. 3 H 3 H" O 0) rt ft OH H 3 Oi »-' ft o 3 • 13 0) 3 rt fD |_i iQ H H- O O rt 3 0i ro •< 0) C 3 3 01 ft H- fD "—» o) o c O 0) 3* H Hi 3 p. p. P, i£ a 3" iO ro s: 01 Oi s- 0) ft rt H- ft 3" 3" ft 0) > 0) Hi 0) O < 3* H C 0 ,—. I-1 3 H n 3 H- OJ 3" fD o H- H H- 3 0i a ro cr ro 01 H ro ro 13 OJ H- Hi Hi 13 o- rt 01 ft 03 H- 0) o 3 0> 01 0) H- t, o 01 H 01 H 0) 3 3 H Hi a> 0 TT ro O rt 0) Qi O Hi 3 H- 0) 0i M 0) 3" h-1 H-1 01 0> ft a rt on 3 £ H- 0) 0) 3 H H- ro H- ro 1 ft 3" 01 0) 0) ft 0) ro 3 rt H to > fD •— 3 01 rt 01 >» 13 Hi M 01 • ^ Hi rt ^ 01 rt 13 O ro ifl 0 H- ET 3" *< 0i H ro 01 H 01 0 01 H- 0i ro H >» ►-3 3 H1 0 a ITJ 3 Oi 0) «: OJ lQ S- O Hi 0) 3 n 3" H 3 3" H (t> 0) 13 0 ro rt 01 ft H- 13 0 HJ 3 3 0i c 0) 3" 1 I-1 ft 01 c 0) H- 0i 0i H 3 rt 0) 1 ^ ro H 13 X h-1 h-1 ^< ro rt € ^ 0 0 H- ro a rt Hi 3* Hi ro H- 0) 13 0) cr Hi 0 ro ro rt 0) 3 01 0) H- •— 0) rt 0 H 3 H 3^ 3 rt o 3 ^ O H H 01 0) ro 01 O H- 0 O M M O 01 3 01 Cj, cr Ul o> H- ro rt 3 ro Hi H uQ 01 0) o H" 3 3" 0 0) H I-1 ^-, 13 ro H 0) 0) o H' ^< on 0i H 13 13 rt 0i rt ^-» ro 0) .» 3 < ro 0) 3" rt 3 ■>3 0) a 3" 0) 0) H H H- ro 0) < Oi rt H- 13 0) a n o 0) H c cr 3" H- iO 01 0) 01 01 <—» 0) M \-> 01 3 3* H Hi 3 3 01 -^ 3 01 ft O «-« O rt rt iP 01 a 0) H- 0 13 ro t» H ro cr ro H 3 01 H 3 >J Hi 0 M 0i 01 O ft 0 rt n O Hi H 0) rt cr rt O 3" 13 13 3" H 01 3" H c 0) 0 Hi ro H- 3 3 H-1 0) 0) ro H 0 H 3 Hi 01 id on H 3" 3 rt H o 0) ro H1 0) —■ 0) 01 O rt ro H- 0) 01 3 0) • rt iO C O X O 3 3 0) 0i «—■ ro ro 3 rt 3 Oi rt I-1 a )-3 0) H .—» ro 01 01 H3 H- Hi O- 0 01 0 01 Hi to • Oi cr 3" 0) o C O aa Hi 01 O vc rt H H- H fD H- ro H J 0) 01 ft o 0) < 13 > 0) s: 3 ft ro ua 0 0) 3 fD H H1 H- 0 01 O 3 H 0 H Oi H 3 H- cn 0) ^ H ft 0 H- 01 H O 0 3 *— 0) 01 01 c a c 0) ro 01 • cr H- 01 13 X 13 ro 01 01 3 ft rt ro Hi 0i n 3 ft J n H- n a ft >» < o 0) Oi 3 3 ro 0 3 0) OJ ro Hi 0 ES* c 0 3 H- H O 3 H 3 01 ft 0) h-1 13 o 01 13 01 H 13 01 0 01 01 H- 01 1 3 01 c Oi ft 3 ft O 3 H to 0 H 01 3 H 0) lQ 01 cr 3* H- 0) o 3 0 0i 0) o H ft o 01 in 0) OJ H" cr ft C n ro 0) Ol 3 vr> ro H ro 3" H- 0) rt rt rt rt £ rt 01 ro ro H- a ^ 3" H- 0 O 3* 13 3 rt ro 3 O H- Hi 0) < 01 01 0) 3 rt H- 3 to 0) ft O H 0) ro 01 ft H- o — cn 0) 0) H O H X ft > 3 • H 0) 01 0 a fD «o H 3 3 c ro 01 H •< 0i ro 01 ft Hi 0 H 01 to H 0) 01 H I-1 0) rt 31 01 0 c 13 13 cr h- < < 01 oi oi 13 Hi rt Oi a CT ft Cj, 0) Qi H Hi X 0 rt H 0) 01 " ro c O 3" 01 H- ro 3" 01 X 01 • 0) 0 Hi 3" 0 o H 0) O 0) 0) H fjj 01 0 ro 13 O 01 O £ 3 ro 13 Oi Oi ►3 ^< o 0 01 3 OJ 0) 01 01 01 ft *« 01 a 0 C 3 cr ro 3 ft H- HJ • 01 c > 3 13 3" 0i < ro > H 0) ro >< 13 o ro cr H- UI cn 01 0) 01 ro ft •s ua 0) Oi 01 rt (D H- Z O rt H 3 H- h-1 -J 0) ro H- 01 H- H 3T O H rt h- cn H- •« 3 Oi z C H- rt 13 M H- •< oi ro 0 01 H- 3* oi ro 0 13 ft 01 3 cr rt h- 13 fD Ol Oi 0 3 3 rt o ua 3 01 3 0) 0 H- ft iQ h- ro ro rt H- 01 3 3 Hi fD 3* H 0) H 0) H < H- • M hi a •< 3 M a a O o" 0 0 3 rt ua o 0 ft H- < H- Hi O 0) H- < H Hi H- 0) C 0 3* H 3* Hi 0 Oi ft H- rt fD ft O OJ o rt ro a 01 H- 13 3 ro 0 H- Hi 3 *< rt k: O O Hi 3 O • H- 0) H 01 ua oi ua C 3 Hi ft ". >< ft 01 Hi o H O c ua 3 01 13 ua 0 ft Oi 3* 0) M o H H- €, H H- H H- c rt ro 01 H 3* < 0 ^3 0) Hi Ol H fD 01 X 31 3 3* a O ro hi 3" 3" X Hi ro fD 01 H 01 0) 0 N M 01 u3 0 O H- = H- 0) H- 0) fD i s: o H- H O 0 H £U O 0) H 0) 0 rt a ft 01 H 3 H- M ua \-> H- fi- rt n H H H- •» 0) H- 3 ro hi ro fD 01 fD H 0 rt 3 ve 0) 3* O H- rt a oi 3 0) 01 0) ro H O H 3" cr 1 00 D" fD H Oi 3* OJ 13 > H- a 3 01 rt > ft ro 0" O l-J ft fD 01 H- rt • ro 0) 0) 01 3 ft 3" 0) H 01 i. rt 0 fD 13 0) 3 3* to o Hi rt 01 H 0 3- 01 H- 0) H 3" K H O M a 0 ua oi oo rt O ro rt ro 3 0i ft 01 01 H- H- (I 3 0) H- 01 m c a ft H- H 3 3* 01 uQ 3 3 a ft ft 0) 3 3 O c a ro ro Hi < rt Hi 01 H- >< ro 3 13 0) 01 H- fD 0) o> 01 O 0) H- H —. n rt 0 3 0) 1 0) C Oi rt 0) 0) ft &) H M 3 Oi 0) --J 3" 3* H 3* ua H ^ ua H 0) 3 :£ 3" rt 3" 3 •<; 01 id rt on H- fD Oi 0i H 0) * •<: 3* rt ro o ft 0) 0) 3 ft < O ft oi 0 0 0) 3" ifl 3" H 3* Hi 0) K ro C 3" fD 01 H- ro c 3 > 0i H 01 H fj (D 0l H- 0 3 13 H 01 01 3* 0i c 0 X 13 01 01 3 0) 13 0 ft i£3 ua 3 H O Oi fD 0) 0) H- ft 01 0) 0) to H- Q) 0) C - H- ro ro 0) rt 3 H ft ro hi ua N3 .-^ 0i l-J rt ft 13 01 0) n ^ rt 01 0) Hi ro 13 o) 01 Oi 0 0) £~ Ln 3 ►< 3" H- 13 ft ua 0) 3* 0) H- ro oi ro H H 3 00 0) 0) fD rt 0) ro H H H- ft H 3 0i 3 a h x 0 0) a H- H- 3 H 0i 0 Hi 3 3* 3 0) 3 Oi ro rt i Hi oi ro 13 oi oi a 01 a 3" rt C fD 0) 0) a H Oi H- 3 0 0) H ro 01 01 01 H- 13 3 01 ^s 0) 0) ft O 0 a 0 X H 0 o 0 0) 13 O . fjj ro H 0) 3 0 01 3" c H ro 3 a n 01 Hi 0) H O 0) 3 r-> Oi ro 0i < i-" rt 0J 01 »-v H- ro a 3 - fD Oi o ro h ro Hi 1-5 01 01 rt H- 0) Mi 3 -e > ro cr ro a £ 0) 3 ro h-1 H H 0i H M 3" 0) ro hi rt 3* 0) 3 H- ft c H- a i > — 0 rt h- > ft ro (D *— 0 h- rt H ?r 3* fD rt oo 3 13 Oi - 3 0) Qi rt 0) H H-j qi H- H- H- 3" on rt 0 13 3 . ua ^< 0 3 > ro 0 ro 3 -e Oi 3 0) rt 0 3" H ro a rt 3 01 3 rt 13 3 Hi 01 H a o C ^< 0) rt 3 3" H- 1 H 0) I-1 13 O 3-3 0) •< H- rt fD 0i a a ro a 2! 0) 0i ^< 0) 01 01 0) 0 0 3 H 0i HJ H H- ro a 0 O rt 3 3 01 H < h cr 0 Hi O 01 a h H- H- X Hi 3 H 3 X H- 0J H a ^ H fD rt O H- « ° 3 H- 0) H ft *— H- 01 ro 0) 0 ^< 0) H- JD H H ua p a 0) H- ►< • ro ua i ft UJ 01 0) x z ro CTH3 H 0) 0) 0 3 0i 13 01 3 01 3" H 0 0) «— ro rt •< H- CT 3 H- 3 H- 3 H- H H- aa fD O 0 n • 3* rt — 0 0) 0) oi a 3 0i *1 ua rt 01 1-" C Hi 0 rt 0 3 hi •< a 01 rt rt Hi 0) 0 ft 0 ro c O 13 0J 3" 0 H- Q) x . H- H- fD M H- 0 H" H 3 a 0 o) ■« ua 0) 0) ^ 3 n h cn 0) O 3 C 3 < 3 0 ro o> 3 0) H H- 3 0 13 13 • O O ua ro Oi 0) 0 H £ - rt ro 0 H 01 ua ft H Ol h- rt H H 3" Oi o 01 3* 01 3 H H H 3T H- O 3 01 C rt - 0) 01 H l-1 Hi H C ro rt 0) ■ > 0) 0 fD 3 H 0 0) 0i 3* 0) H 0 H- a H- fD H o H c o 3 rt 01 01 0) 3 N fD 01 ft > H H • M 0) 13 (-. H- 01 UJ H H- 3 H H- 01 X 3" 0) ►< 0) 0) Oi 13 CT fD O a 0) Ol O ft G 01 H- ft n o 1-5 01 H ft Hi 3 01 0) h- ro 3 rt Qi 3" Hi 0) 3* H- O M 0 3* I-1 0) j> 01 H O 0) a 3* 3 0i H- c 3 ro h-> oi oi ^< a ft -^ 0) 0) -O ft £ o 3-n < 0) 3 0) ro 3 3 a 0) 0 ro hi 0) ■« fD 3" H 3 01 01 a H D" H i-i 13 Oi s: o M OJ 0) 13 € 3 ft a H" 3 H- 01 (D 01 0i 0i ft 0 3" H- • ua c H- 0) 3" ro oo a 01 rt ft ft ft rt 3* 3 H < r-> 0) M rt ua Qi 0i ,—' 0) H- s: H- 01 rt 0) ro 0 0) OJ i 3 3" 3 rt rt ^ H- Oi 0 ro 0 0) H 3 3 H 0i 0 H- H- •y 3 0) ro Hi H 0 ^ H- Hi a 3 3 rt 3 •» K Hi fD 0 3 0 0 3 01 n a 0 OJ. 0) C 01 0) ►r] H Hi Hi H 0) 0 n H- 1-5 C H ft a \-> rt 0 C ,-» ft H o M 0) 0) ^< cr ro 0) 3" 0 0 0) to ua 3" to 3 0) 0) HJ H- 0) 01 rt 0) 01 01 0) Cr C 3 • H fD • 0) 0) H a < 01 c 0) a 0) 0) N 00 0) to 3 0 ro 0) H- a H- Cj, 0 H Qi 01 H- Ol 00 01 0) to n 01 i-i 1-5 ro o 01 01 Ml 0i ua H 3 *—' rt 0) • H- 3* H- 0) r-J- 3 0) 13 ft ro < ua »—« > ro X 01 H- 0) rt a 01 a 01 0i Oi ft 01 i 01 to H ro 3 X 0) H- 3 ro H 01 3 3" 01 01 a d . 0) 01 M ua 0) 0) 01 a ■* 0) 0) 0) 0) a 0 o c rt 3 H 01 ro 0) 0) 0i OJ. rt s Ul H- 3" H- ro 0) X OJ 01 o to a ,—. ro fD ua c 3" *-' o 0J 3 O Oi X Oi • 01 H- • 01 8 X 0) 01 0i 0 > H- 3 0 rt 0 -j 01 H KD 01 ua o H- rt rt 3 a 3 3 ro ro rt j> H OJ rt HJ H 0) 3 0) Oi ro to 0) rt H X M 01 3" 3* D 0 0) a o 3 • 3 H o ^< rt 3 0 Hi *•—' c 01 0) cr Oi a «—» o K 0i ft ro H- a 01 0 H 13 C 0) ^ c to 0) 3* 13 rt 3 H- H 01 H- H H 01 0) 0 ■ H- 0 ft 0i rt C ro o 01 ft 3 rt 0i 13 X 01 0) 3" -J 0) Hi * 3 H- 01 H- ^ 01 (D 0 3 0) 1 3 1 H o Hi 0 H o ft ua 0) 3 a 01 0 "~' Hi ft HJ rt 0 01 3* H 0) o Ui cr 0i ro 0) 0 TJ 3 > 0 3" H- 3" H H- 3" 3 H- c 0) 0) ro H H1 0) H- C fD H Oi 0i ua 0 a a H Hi a rt 01 o n n 3 H- rt n 0) 3- 0) 0) 13 01 3" H QJ H H- o a H" 13 o ro H- to 3 Qi Oi H- 0 Hi H ua Hl M HJ 0) H- Hi HJ 3" H 01 • rt 0) 3 ua 3 ro H- ro H- H- D Hi 01 3 0 C3 H- ■»• o 01 0) ro 3- 3 rt o < 0 a 0 H ua 01 oo ^ 0) 0 Oi ►< 0) ^-« H H- 01 01 3* 3" H 0) 0) 0i 3 H Hi £, H to 3 K 3 • ro ro Hi to rt 01 (-■ 0) H 0 H- • 01 ua 3" 0 a 0) 0 • rt cr >< Qi H ft a UI rt 0) 3 H J> M y£> 3" 3 OO H ft 3" H- to 3" o < ro 3" h3 rt • to 0) H- H Oi H- fD H- 01 *— 0) 01 0) 3 ro 3* m cr 0 00 ft 0 0i 3 ro * H c > 0) 0) a ID 01 Hi Hi M O ft 0) 0 1 01 0 01 ft H H- 0i 0 ro ua J> 01 3* ua 01 0i 0) 01 3" H rt Cj. 01 rt H H 3 H Hi 0) M ro 0 0 01 0) ro 01 ro H- ft Oi 0 0 Hi H- 1-5 0 0 O H 0) a i: Ol 3 n r-1 c H 0 3 01 Hi C 01 H- 01 0 H c H- 0) 01 fD a o 0) D H 13 3 a H Hi 01 M 01 3 ro 0) H- HJ 0 fD ua rt a ro 0 1-5 ft 01 0) 01 D a cr 0 01 H cr oi 3 01 a u3 H- 0) 01 to 3* 0) "» ro H ro 0 (D fD ro 0 ft H- ro ro 3 0) 01 • H- Oi 01 *i 0 H cr Ol X H- 0) 01 01 0) ua 0) H ua 0) to rt 3 3 ro rt (-■ H rt rt • 0) 00 3" 0) • 3" 0 H- cr fD fD o a H *~* *->■ 3* I-1- 01 H ua 0 cr H Hi 0 • • H- *j 3 H- 01 0 < 3 ft 0) Hi 0) to cn Hi Ol 0 3 Hi Jj> Hi rt s: cr 0) H- cr Ol 0) H- cn o > Hi H ft 0) 0 3 0 fD 3* 0) 0) 0 0) a h 0) "— *-* ro 0) 0 01 H 0 H H- rt 0 3 H- fD •• •• 01 H fD 01 3 cr ft i-i c ua 3 3 *->■ fD 01 X c 3* 3" 3 ua 0 ^< fD c H i-i H ua o to D 3" X 3 1 H- 0i H- 0) 3" 0) 0 0J 0 0) • 3 fD 1 0 3 n < ua 3 o H Qi ^ ft H c O o oi o 01 Ol 3 0) 0 H- 0) 3" 3" 0 01 H- 13 1 Oi "~^ c H 0 01 H a 01 ft H- 3 ft < a . | c ~ • 3 rt H ft ro 01 H 3* 3 H- 3 cr 01 H- 3 0) i-3 0 ft 0) H- x 01 fD n c 0) 0) 0) ro oi 3 0J H rt 3 3 01 o 0) i-3 0i oi cr H- H H- 3" H- 01 ro 31 01 cr 3 H Oi ft rt 0) 0) 3 a Hi 0 ■* 0) 01 01 ro 3 0 ro Hi K H H 01 01 ua Hi rt 01 3 3 H- rt 0 3 rt 01 Ml -s rt 0 HJ 3 H O n z 3 0 0 cr rt 3 % z rt H- ft O a Ol 3" 0 Oi 3 3* CT 0J cr 3 H 0 0) 3 H- Qi 3* cr 0) H- 0 0 0 ►< 3 0 01 cr M c 0 c 0 0) H H ft 0) fD 0) ft 0) a 0 H H H M 3 0) H- 3 ft 3 3 o 0) 0) 0) ft 01 H H 3" 0) rt 01 ft H- H- ro H- 3 ?«r 3 ^< H- a ^< 3 ft 3" 1 cr c rt rt H H- C 3 0 0 3* rt QJ < ua 3 H < cr rt 0) 0) 13 0i 0) 0 H- 3 H- 13 3" Ml H- 0 a 0) ■• H ft H 0) H- H H- 3 0i < H- c cr 0i M H- H- M ft 0) 3 3 01 3 0) ro H- 0 0 3 3 rt ft 0 01 0 H- rt H- rt ro ft H ua 0) cr 0) 0 3 3 H- ro ro ^ 3 0i ft H- M rt 0i 01 H- M 3 0) Z H- I-1 rt •< Hi H- a ►< 3 0) H- 0 Ml 0 3 3 0) a ft H- M 0) 3" 3 M rt- 0J 01 it- 0 3 ft 0 >< H- < 1 H c «: rt 0) a H- Cj 01 H- 0 ft fD 1 Ol ft ^< 0 s' 3" Hi a 0 H- ro 3 H3 H- rt 0) 0 (l > 0) 3 C 0) H 0 3 ro 1 M cr O a 0) Hi 0) a 3* H H- 0) ft 0) 0 01 3 Hi 1-5 0 M 13 rt 3 01 3 H- 3 0) a ua 0) 0 H ■» cr 0 fD 0 OJ 0) Hl 3 *< H H cr Hi Oi M 0) n 3 ft ro 0 H- 3 Qi 3 c 0) H H- cn 01 Hi 0) c Oi Hi rt rt Hi ft H ft H H 3 c 01 cr 0) iO ua M 0 ^» 0 3 0 3 ua rt ft < 0 0) cr H 01 0) • a H- 01 0) c H- o 01 13 0) TT cr H a H 0) Ml 3 0) 0) 3 H- 0i cr fD M H rt ft > H 0) H H- 01 I-1 X M oo 3 0i 0 0) 0) 01 0 13 0 rt M H-1 01 3 rt ro fD 0) ■» cr 01 fD 0) Hl H H- M H- X 0 rt Mi a H 3 3 H 0) ft 0) H- 01 ^< 3 0) ro a H- O H- Ol Hi 3 M Hi ft >X> 0 ro 3 H- H- 0) K ft ua 0) 0i 0) ft 0 ft a 0) o H- ua ft rt cr 01 0) 0 o 3 0) Hi K ^ Hi 0) Ml ua 0 ua Hi * ro 01 a TT M M 3 H- 3" > c 0) 3 H- 3" (0 3 ft H 0) H 0 ^ 0 3 ro Hi H HJ H- 0) H ^< 0 0) i< 0 Hi H ft 1 < ro cr 0) • ■-3 0) H H 0 13 M rt 0) 3 0) H- H 01 H c H H- ,_. 3 < 3 ft H H H- cr ro 0) 0) to Hi 01 3 01 3 H Hi M cr 0) H Hi i Hi 0 01 01 ft 01 0 0) 0) 0 ro 3 0 ro < H- • 0 CT 0 ft rt ro 0 0 0 ro «-s M ro 0) ft < 0 0) M H- H K 3 3 n H t, H- 3 o H Hd h-1 01 cr H- Hi H -£ K 01 3 3 a ro H- 0 0 H H ^ 3 0) H- 0) ft rt 3 01 a 0 X 0) 0 01 0) 0) 0 3 0) 3 fD Ml C rt ft 0 c H 3 Hi Hi H- < 3 0 0) ft H- < H- H J> H- H 01 0 01 ro a H- ^ H- ft 0 O 01 H- M H- M 0) 3* 13 < 13 01 Hi > 3 H- ua to 01 0) ft 0) a H 0 O 01 H ro ft 0) M 0 0) ua c 3 13 0) 0 ro 0) 0) Hi fD Hi 3 0) o H 01 3 0) H- cr 0) H 3 0 CT ft IX) 3d 0 fD it- H a 1 01 13 H 0) 3 H 1 < ^ 0) 3 0i 0) 0 3 ^< ft U3 3" H- ro cn H- 0) c 01 0) s' Oi H- 0) 0) rt C cr 0) 0 H 0 H- 3 CT 0) 0) ►< 3 01 00 .—. 0 H a on 3 rt 3 o 0) 3 3 a 0) c H ro 0 Hi 0) H o 0 0 H * 0 rt 0i H- 0) ft O rt M H- > 3 3" a M ft 0) c ro H 0) 0) 3 3 D" 01 rt H ^ 0) 0 0) cr 3" H 3 0 '—' cr rt 3 13 n 0 0) 0) ^ 0) 3 0 a 3 ft H 0) > ft H- ^ 3 3 rt 0) 0i 0) M 0) . 01 0) ua M 0) cr c H ro c 1 ft 0) • H- ft 0) H- a H 0 CT 01 3" ft 1 »< c H- rt c 0 3 H- ft 3 it- o M 1 01 ft 0 0 Cf H- H- 0> H- Oi 01 ro 3 0) H- 3 3 1 cr 3 0 01 Ml 0 3 3" er cr 01 n H fD Hi 3 3 01 r-> ft ft H X 0) ua 01 H- 0 Hi 3 01 0) 3 1^ H- Hi H ro rt 01 a 3 O 01 ^< a i-3 3" 3 *< rt ua n H- ft 0 0 0 Hi rt H- 3" 0) 0) 3d *— 0 H- 0 o> *- H »• cr ft 0 01 it- 13 ro ro fD 0 3 0) H H 01 ft Ol ua 0) * H- 0 0) ro • Mi 0) H 0) H 0 ro 3" Hi cr s' H- 01 13 3 H 0) Hi 3 ua 0) ft 0) 3 H 3 H H- H H H- 0 0) 0 01 H- 0) H- fD 3 ft Qi rt a a rt ft H H o H H rt o> HJ ft 3 0 it- 0) 3 H Hi a 3 0 H 01 rt 3 It rt H- ro ft 0 0 ua 0) ^< H- rt ro H- M- 3 3 -* DO er o ua 0) 3 3 0 ua 0) G H- 0) rt 0 a 3* 01 c ro H ft 0i 0) 3 3 ua c Ol 01 0) ro rt- c c ro M 0i fD 0 (I ro 3 ua H 0) 3 H3 ro Oi 3 13 H ft 13 3* H 0) 0) H H H < n Oi H- 3 a a 3 Hi 3 a 0i H- H 3 H cr it- Ii a 0i cr ro rt H- 01 ft 0) 0) 0) HJ 01 a c 01 a 3 Hi rt 3 ft 01 0) 3 3 0) er r-> i: «—-. -' 0) 01 ^ 01 0i rt cr H X < H- ro M 0) ua 01 0 3 01 ua M 0) H- rt < rt H- a ^ a 7T C 0) H 0) H- rt 3 ua ro 3 C on H "S Mi H- 3 H a • H- - cr 0) H- M1 H- H rt 3 ^ H- 0) o 0) 0 H- H- ua rt 0) ua 0 0) 0 ft a 01 H- 3 01 ro M 3 H- M cr 0) 0i Cj, cr rt It • 3" H H < 3 cr cr ro c 3 H 3 0) rt 01 0) ua 01 0) c ft 3 01 ro H- rt * H- ft ro 0 o 0) a CT 3 13 a M > a 13 H- rt > 01 01 H 3 ft 0) rt H- ua 13 H cr 13 rt OJ 3 Oi H 01 H- c a 0) a H M 0 H- 01 01 H 0i ro Hi H- rt 3 c 01 z ro 13 t-J H *< I-1 3 3 c 3 0) rt 0) X 3 3 cn 01 3^ 3 3 ro H 0) H 0 0i 0i ua 0) 3 H- h-1 01 01 0 H- ft ►< 1 0) 0i 3 H C -s *< rt 3 ua Ml ro ua H 0 3 C H c ua 0) 0) V£> M ft o a CT ft 0 1 01 ua H- rt H 0) 0i ft ^ c 0 3 H- ro 3 0) H 01 H- fD 0) ro to 0) ft 3" c ►< ^< Hi 0 01 3 0) cr a Cf 0) H 0) rt 01 H- H ■. a rt 0J 0 0) • 0) H- J> 0) ro o rt C 3 H- 0) ^< H- 0) ro x 3 ro ft CT 0) ro H- 3* it- <—> 3 3* ^ ro M n ro 3 0 cr ft 0 ft 0) Hi 0) a 0) H- c cr ft »c 3 3 fD s' 13 ft cr Ol Hi C H- 3 0 01 c 0) C 01 Hi c a 01 Hi -£ 1 ft * 0 rt 01 3* M ft Hi ro M 0) >< cr hfj H 01 3" ^ 0) cr a ft ft H < a 3 ro H- 1 cr z 3 cr 0) ro H- H 0 H- *— H 0 c M 0) ■ 0 H- H- o 3 0) 0) a ro 0i a H rt 0J 0) 0) 0) 01 H a Hi C H- 01 3 ■» rt 01 M ro 01 3 0 H Hi < 0 Oi H ro 0) ro 31 Hi 3 < rt 0 H 0) 0 H- X rt M 0) 3" 3 0i rt Ml H- 3 ft 0) 01 0i 0 H 3 H- ft ^< Oi H 0) 3 01 ►< 0) H 01 ro 3 n ft ft 3 i-3 ^ ro ft ro H- ft ua it- Hi 01 n 3 0) O 3 M 0 3 H- cr 0 01 H- 3 H- 0 3" \-> cr 0) 0 H ^ • < c 0) er H 0) 1 H 0i ft Mi 0 3 H Hi 0 H a H rt H- >< fD H- 0) H- a i ft ua 01 ua 3 3 H- 0) H- rt 0i 0) Ml ►< 3 •» ft 3 3 0> rt •< i cr ro H H ro ro 3 M 3 3 • 0i 0 fD 0 ft H "< 0 Hi cr •< rt 3" 01 0) 0 c 13 1X5 cn On 3 ua 3 rt 3 H 01 ro cr c rt 0) s: 01 Ol 3 0) > H- 3 3 0) rt 3 ft 0 a ft ro cr 3 H- cn H H- ua Qi < on (-•■ 3 ua H- O 1 it- fD rt 13 Oi ua s' 1 ^< (D a H 0) 0i 1 H 0 a 3 o c Ml OJ. 0 0) i? 13 0 C H 3 3 H 0) rt ft 0 0) ro ft 0i *• 3 0) H- 0) I-1 ua X ua a H- 01 i 3 it 3 HJ a 1 3 ►< a H- H- cr 01 M Hi 0 rt H Cj. H- Hi M H- 01 01 13 ro 3 < rt 13 H- H H- H- 0i 3 0) < ft 0 3 0 3 H- ^< 0) 0) 0) 0 0) 1 0) ^ 0) 1 3 Ml 0) 0) 01 0 O It 1 H H OO cr 0 < ft H 0) cr H- 01 > 0) 0) H H 13 Hi ro H- H 0) H- H 3 It H ua < 0i ^< n O c 0) ft 01 ro H a H- M c 3 0) < 01 0) ft H- ro ft 0) ■ H- 3 i cr 0 01 0) H3 0 0) 0 i-3 0) 0i H Ml 0i 0) Cr 3 0) cr rt H cr D H H- 0) ft 0 0) 0) 3 S- € 01 0) H 01 it M it on rt Oi cr 0i cn ua 0) • K 0) 0) 13 H 0 0) Mi to It H rt 0 3" o 3- a ft a> ro 0i ro 3" 3 3 0i rt rt rt ^a ft s- 3" 3" 0) n fD H 3* 0) O H- rt H 3 fD C ro 0) a 0) 01 0) ro ro > H- K H 0 cr > 0) -s rt 13 O 13 13 ft > 01 3 3 3" Qi 3 Qi 0) C 3* 0 0 cr 0) 0) 0 01 0 0) a H- H < H- H H- 13 H- H H M- n M H- O ro H- ro 0) < o cr C 3 0) H- 3 Oi H- H- o cr 3 3 H- 3* K H ro H- 3 0 3 0 O 0i 13 Hi cr < 13 0) 0) 01 Oi H 01 ro Oi rt 3" S3 H Ol 0) 01 0i 13 0 ft 0) rt it ft o M 3 01 Oi H- 0 3 01 a m H rt H- ro 0i 01 H- 0 3 0) 0) 0) rt rjr ft 3 c • 0i zr 0 "£ a H H" 3 3 • cr l-J h-1 •< ro 3 0 H- 0 rt rt >< 0) 3 01 rt 0) I 3 Ml 3 ft ft 3 3 3" ro ft 3 0 3" o) cn 0) H- 1 0) 0 0 ^< 3 H- ro 03 0i rt 3" H O ro N 0) 3 M M 0) 0 0 rt 0) H- It * Hi 0 3 0) 0) it H- 0 ft rt 3 01 3 O 0 01 H 0 -s Qi 0) G 01 0i 3" > 0 ro r-1 H 3 OJ. rt a Mi ro H a ft 01 0) X > M- rt a O 3" 0 H H- CT cn 01 ro CT H- cr 0 Ui ft O -s - H 0i 3 o H ft 0 a rt H- 0) 0) 0) H- ^< cr rt cr ^< H C 0 0 3 13 01 CT h- rt M 01 H- 0 01 01 Ml 13 01 13 rt 3 01 0) O cr 0 H- 3 H < H ft o « 3* C H- 3 cr ro H It O a 0) 01 ro c 0) K 0) cr O O a 3 0) TT i M rt H M n a 01 H- H 3 c 01 ft H H- ro cr 0) ro c c H- 3 H- H -s OJ. cr o H- 0) ro 01 0) a n H H 0 13 01 O ft H- O fD K 3 o 0) 0 ft ft M H 3* Hi C rt H 3 M- M £) Hi cr 3 C 0 ^< ro H 3" 3 H it 0 ua H- C 0 *< Hi M 3 ua a ft 0) H H- DJ Cr 01 0) 3 Oi M H- 0) 0 H- a * 3" 0) rt H- 3 rt ro rt M ua M Ui a > H 0 0 0) a 3* 0) H- fD ro ft > 0) ro rt Oi rt 0) ?r 3 0) H- H rt Oi 0) 01 < 3 0) 0i H 3 cr 3 0) 3 H- it ua M H- ro n 3 H 0 ro 0 ?r Hi H O rt 7T ro 01 01 M ro a H- 01 rt 0) 3 Qi CT 01 ro 0) i 0 3 01 rt 3 cr ft O 0 O 0) M M i Hi 01 H H- M ^< a a c s: It h ro 1 ^< 0 ro * 3 H- * H H H H 0 M 0) M 0) 01 o Hi o H- ?r 0) 0) H 3 < ft rt 13 O 0 rt 01 -e c 3 n a 01 0 ro 3" 3d a ro 0 cr 3 cr 0) ro H ?T O ro ua a a 3 ro H o 0) H- 01 0) ft ft Hi 3 ro H- 0) H- H- H- 3 3 H- ►< cr C 0) rt 0) H Ml o Oi 3 01 It 0) 0) n H 0J M 01 M rt 0i 0 cr TT I-1 c 0) ft 0 ro < 0> ro 01 3" 0 ft H a 0) M H ro 0) 3 01 ro I-1 3 01 ft 0i 0 fD H- rt Oi K ^< 3 rt c ro c rt 3 O Hi ro H ^ 0) rt H H- a Hi o 0) a 0) C rt Ml 3 n It 13 H- 3 H- Oi 31 H ■< 0i H- oi cr ro ft 0 cr 0) 0) Hl CT 01 n o H 0) 0) 0) H H- 3" ro 3 O H- C rt 0 rt 3 cr 0 O It H 0) 0 0 a H- 3 it 3" 0 0 0 0 O 0) O 3 3 H Hi Hi a H- ro < H H O 3 cr 3 Ml O ft H- 0 rt H- H- 0 0) 0) 3 a O it 0) H- Mi o H 3d O 3 3 3" M 01 c M ft c 01 H 0) fD 0) ua 0) 0) ro 3" 01 Ol n H- -£ O 01 H 0 cr M- H 0) 0) a Oi M rt 0) H- cr ro 0) 3 ro ua TT H •• 0 H < Hi O ro 3 ft Ol 0) < 3 3 ^< Hi rt rt 0) M O a s* o H- it ro 0 •*• Hd 3" CT ro cr H- o o 3 cr M cr £ H- ft fD C £ 0 H- 0) 0 O Hi 0) 01 0 I-1 rt H 3" Cr O c H 3 c H- 01 ua H ro cr 0) 0) fD H- fD 0) H- 0) H oi a 3 0) 0) 3 a 01 ft 0) cr 3 3 3 01 t-3 s ro Mi 0) ua rt n 0) Ol 13 rt 3 01 H Hi a oi Oi 0) * 0) < O I-1 H- 0) H- rt 3 c s It Mi H- M < -s H 3 oi cr It cr H rt 01 3" 01 0 ft 01 Ol a H 01 ro 3* 0) fD 0 H Ol 0i 3 ft H- O 01 rt Oi Qi ft 3 01 H- 3 O 3 Q) a ft H 0 0 O 01 < jO 3- 0 3 0 M Hi 01 3 fD 01 0 3 O c C 0) a H ^ i-i ua fD 0) 0) 0 Ml Cj. 0) cn o It 0) 0 ro n Qi M 0 M it z t cr 3 01 •£ cn O € ft £ Ml rt 01 c rt 3" 3 M Hi M H M 0) c ro cr Oi 0 0) 0 c 3" rt cr O 3" 3" 0 0 13 Ml Qi M- 01 1-5 M 01 n rt H n (I 0) 3 H M M H- 0i 0) M 0i H- M 0) Hi rt 3 3 0 H- 0 01 0 0) M ro 01 H- ft Ml ft rt ro H rt It QJ 3 H- ro 0) 0) M 3 M 01 cr 0) 0) 3 H- H 3 ro 01 01 0) 0) M H- 0) 01 a 0) n i 01 ro 01 Ml 01 H- 0 3 ft 3 0) < H n 0) < 01 it c 01 H- i ro H * 01 Hi H 0) rt rt O 01 £ ro H- 0) -£ ro 0 3" a 3 M O 0) 0) n < 3 O H- H a ro c rt cr rt O ^^ 3 0) ro 0) ft C 1 3 0) < 0) ro rt 0) 0 01 0) 0) 0) H H M H- rt *< H- 3 0) Ml ro 0) 0i oj. cr M -s ft M ro 01 < 3 Ol cr 0 rt 1 H- o 3 3 ro ro M a ro rt 0 8 cr H H- M ro 3 3 rt c H n 0) ro 0) O V 0) a 3 0) 0 3" M 01 H- ^< Hi 0) rt 3 0 0 H- 01 0 H- rt Ml (t ft 01 ft a 3 ro M- a 3 ft 0 c 0 M rt 0 ro -£ 0 0 3" Mi ro a rt cr > > X H ft 0) ro 01 M 0) H- ft 3 3 3 3 0 a ro ft Oi ro u 0) CT o 0i K cr I-1 0) 3 13 Hi 01 0* 01 Ml 0) H- M 01 cr 3 1 O (D rt Oi >< a M H- € M M 0) 0) M 0 a > 0 ft 0) ro ro Ml j> O ro 01 c Ui 0) 0 < 01 M- H- 01 3" 0 It M- M 0) M H- 0) O H H O On M 3 ft rt Ui 0) Ml ro ro Ml 0 1 H- a 3 rt ro M o H- a a H < 0) M 1 ft a c M X << XX 01 cr Oi 0 M 0 • 01 H- 01 OJ. 0 Mi ^< 13 IX) H- *-» ro z ^< 0) H- Ml 01 01 ro cr H O 0) H to ro ■8 Ml 13 cn 0i It O (I h-1 3 01 0) It X 0 0) H- M ro H- 3 M ft c 0) ua ►< H- Ml O H roT 0 ro rt n cr ffi 3 H- rt H 3 ft ro ro H- O M 0i l 0) 01 3 0) 0i a 3" ro It 0 0 0) 0) 3* ►< 0 0 (l M 3 3 M M ^j a It M c 0) 0) 0 -2 c It < 0 01 H M- ft ft ft 3 01 M ua 01 It cr •< 3 Hi 01 rt ro 3 ^ it a H- 3 c 0) 0 H- cr 0) M rt 3 ro O O 0) 0 > H- H- rt < 0J cr H- It it 01 It 01 01 C 0) M ft H 0 0) 0) O it c 01 01 3 3 cr 0) ua a ro 01 01 M ro c 3 (-■ 13 0 0) c rt O 3 •< H- Hi 01 M 3 0 3 h-1 H- 0) M 01 H H- < 13 Ml 3* ro 3 o a 0 H- 01 H- H rt ^ H- M H- rjr ro 3 01 IX) 3 cr ro ^ O H- 3 01 O Ml ro 3 3 0i * ft 3 H- 01 M c O X O H- H- -O < ro 0i M ua H- 3 Mi H 1 Hi ft QJ c 0 rt 0) ro it Mi H 3 ua ro i-1 0) 3" M 01 0) O a a s: 01 0) Ml < a it 0) ro ua H- «: ft 3 1 01 01 01 X H ro It H- M ro 3" 3 * 0 01 ro 1-1 a H- 3 a ro 01 3 M- 00 rt a 0) IX) 01 0) ro 01 0i 0i H- rt M H- 01 0J 0) ua ro M 0 Hi H H- 0 0 -J c a 3 rt o rt rt H- H 0 13 ft ,—. It 01 cr H- M H- ua 0 ^~. ft IX) H- H- rt 0) ro 3* 01 0) 3 Hi Cj. 0i Ml 13 c 3 M H rt Oi ft rt O a 0i cr t-3 H1 1 o 3 3 H a c 0 01 CT 0) M 0 01 H- cr < << 0i 01 0) it 0 0i ►< oo H- ua 13 ft ft to- 3 H 3 13 h~- it 0) 01 M rt 0) ro ■* H 3 rt H- M cr H a 0) Ol ol 0 ft O 01 0) cr M H- ?T Ml H- ft 0i 0 ft H It ro n rt 0) H- > ft H Oi rt 3 ro 01 Ml 0 0 a 01 ft a 3 01 ro 01 0) ft O 3 M 0i It H ro a ft 3 H- 3 Hi M H- 3 •<: O a 0) 3 cr M c 0) H ft 3 01 H- 3 0) 0) 0 0 a H 01 01 a H Hi 3 O 01 01 H 0 Oi cr rt h-J ft H- ro It cr 3 3 o £ n 13 C Hi 0 ^< -s rt ro 3 0) ua 0 H- H- ^ 01 ^ Oi 01 cr rt H 0i ft cr 0 cr 01 ro c ro ft ro 01 01 0) 0) c 01 rt • H 01 ro cr ^< ►< 0 H- < Oi rt 01 3 0 01 rt O i ua H1 •< ro M Mi H ro H o ro 0) rt Ml rt 0) M cr O 0 Hi Oi 01 3" 13 ro H < 0) ro ro 01 cr H ro ro it 0 3 01 0i Hi Hi ua ro 01 a H 13 0 ro c a 01 X • ro a H 0 H H- 0 rt ro ro X 01 3 H- ro 0 3 a Ml 0i lt H- o a 3 ft 3 C H n a X 0 O Ml 01 M Ml rt 01 01 H- 0 0) 13 c 3 01 rt 3" ro n 0 0 3 c Ml ro ft 0J a H- Qi cr ft ■a ua rt 3 a 0 3 01 *— rt cr H- 0i a> M 0) ro 3 ro 3 H- rt H- o H > cr 0) cr 0) 01 c a 0) ro 3 rt Ui ft rt M ft 3 Hi H- H- H- H- H Oi Oi 0) ft H- H 0) cr Ui 01 01 ro fD ^ ua H Hi ro 3 0) 01 H ro rt G H- ro o ro 0) c ua X 0 3 ro 0) 3 a cr c ro rt M 0) CT r-> H- i-3 rt X 0) H- rt rr c 0 Mi rt ft ro OJ M ft H- ro o rt > ro H- M "< a 3" cr O 3 cr ua M 0) H- cr M 01 0) 0 a 3" M ro 0) < 3 0) ^ ro H- OJ 0) -£ 01 H H- 0) 0) 0) ro H ft 3 H- Qi -— ro H- H- ua 01 3 0) O It rt O ro 01 X H 0 o 3 ft a ft 1 0) 01 0) Hi ft QJ 0) 3 Or C 01 0 0) H 0 0) ro ►< 0 3 3 -s rt Oi 3" 3 H -2 0) ro 0 13 0 0) 0) 3 0) 01 13 H 0 cr O H- 0 0) 3 cr 3 3 01 Ml 0) Ml 0) 3 rt M fD ft TT 3 0 ro ft 01 3 OJ 0) ro G 0) it- 01 »-5 Oi Hi a cr 3 0) H- H- 0 Hi 0 ua a ft ua < 3 H- er > o H 0 01 3 3 M 3 H M 01 H- H- H- 3 0) 0) 0) a 3 ro H 0 rt rt 0 ua ua ft It O 01 ro n ft 3 01 It 0) H- ro Hi cr Hi 01 3" 01 13 cr c 3" 01 1 (I ^ Ml 0) 0) H- 01 fD 0 Oi H ro rt cr 0 H- 0) 0) Ol a 01 3 rt 3 c M 3 H- ro H 3 (t 1 0 3 Ml o 0 a rt 3 3" ft ro rt O < rt 0 0) 01 B- 0i 0) ro •< 0 0) ft 3 a 01 ro O 0) < ro H 3 O implicated in the low prevalence of alcoholism in this ethnic group. Wolff (1972) compared Asians (Japanese, Taiwanese, and Koreans) with Caucasian adults in terms of alcohol reactions and found that 83 percent of the Asian adults showed a marked flushing response and increased optical density of the earlobe shortly after alcohol ingestion, compared to less than 2 percent of Caucasians who exhibited similar sensitivity. In addition, Asian adults experienced more symptoms of discomfort such as dizziness, muscle weakness, pounding in the head, and palpitations. These physiologic responses were also documented in Asian infants and in American-born Asians who were raised on western diets, thereby suggesting a strong genetic predisposition for alcohol abstention. From a social-anthropological perspective, the concept of "Happy Hour" or "cocktail party" is foreign to indigenous Asian cultures; Drinking is usually accepted on festive occasions but, then, only if taken with other types of food. That one may find drunken Asians at parties may not be a reflection of alcoholism but, rather, an indication of their inability to tolerate alcohol—what little it takes for them to get drunk compared to white Americans. From the data presented in the aforementioned pages, it is clear that additional studies, focusing systematically on the lifestyle differences between Asians and white, would be necessary to test the hypotheses concerning smoking and drinking behavior in these populations. 227 FOOTNOTE *The above conventional infant mortality rate approximates the probability of death among infants in a given year. Since the numerator and the denominator do not refer to exactly the same cohort, possible bias may emerge. The accuracy of the approximation varies from one situation to another but depends in general on the annual fluctuations in the number of births. This may have some effects on the infant mortality rates of small populations with high level of immigration, such as the Asian Americans. Therefore, when data are available, such as for the case of California, the cohort rate is used. The cohort infant mortality rate is based on the linked records of births and infant deaths; it describes the true probability of infant death. 228 Table 1. Asian Population, 1980 and 1970 United States Number 1980 1970 Percent 1980 1970 Total Asian Population Chinese Pilipino Japanese Asian Indian Korean Vietnamese Other Asians Laotian Thai Cambodian Pakistani Indonesian Hmong All other 3,466,421 1,426,148 812,178 431,583 781,894 336,731 716,331 588,324 387,223 NA 357,393 69,510 245,025 NA 166,377 NA 47,683 NA 45,279 NA 16,044 NA 15,792 NA 9,618 NA 5,204 NA 26,757 NA 100.0 100.0 23.4 30.3 22.6 23.6 20.7 11.3 11.2 ► • • • 10.3 4.9 7.1 4.8 1.4 1.3 0.5 0.5 0.3 0.2 0.8 iData based on sample. 2the 1970 data on the Korean population excluded the State of Alaska. Source: Bureau of the Census (1983). 229 Table 2. Percent of Population Completing 4 Years or More of College By Specified Race, Age and Sex: United States, 1980 M a 1 e Female Race and Age Number Percent Number Percent White1 25-34 years 15,400,161 24.5 15,394,841 21.7 35-44 years 10,711,364 25.9 10,930,907 15.6 45-64 years 18,618,917 18.2 20,292,624 9.6 65 years and over 9,210,721 10.5 13,730,849 7.6 Chinese2 25-34 years 4,453 57.1 4,758 45.6 35-44 years 2,601 55.0 2,619 34.4 45-64 years 3,742 30.7 3,552 15.0 65 years and over 1,391 18.5 1,450 6.8 Japanese2 25-34 years 35-44 years 45-64 years 65 years and over 3,287 1,939 3,878 1,164 49.3 48.9 23.7 7.9 3,517 40.4 2,861 24.7 5,827 8.9 1,442 4.6 Pilipino2 25-34 years 35-44 years 45-64 years 65 years and ove 3,374 33.2 2,740 47.6 2,015 31.9 1,880 8.1 4,832 46.3 3,412 53.5 2,911 27.9 982 11.2 1 Compiled from published census reports. 2Data are from the 1980 Census Public Use Microdata A (5%) sample. 230 Table 3. Percent distribution of employed persons according to major occupational groups for three Asian American groups by nativity, and for the white population U.S. born Chinese Foreign born Total U.S. born Japanese Foreign born Total U.S. born Pilipino Foreign born Total White ho oo Professional Executive, administrative, and managerial Technical Administrative support, including clerical Sales Precision production, craft and repair Operators, fabricators, and laborers Farming, forestry, and fishing Service Private household occupations 19.9 0.3 18.3 0.7 18.7 15.5 13.7 15.0 7.4 18.4 0.6 0.6 1.0 0.8 0.3 0.7 16.3 0.6 12.8 12.6 6.3 11.5 6.0 11.8 6.1 10.7 4.2 13.6 3.8 11.5 4.1 6.7 3.5 7.2 5.8 7.1 5.3 11.1 3.1 23.1 11.9 13.0 8.3 15.7 9.2 21.9 10.7 13.9 10.2 19.7 10.5 21.6 10.0 20.8 5.3 20.9 6.3 17.3 10.7 6.4 5.1 5.5 10.7 6.4 9.5 11.4 7.3 8.1 13.4 7.4 15.5 13.3 10.2 13.7 11.1 16.7 14.2 14.7 17.1 1.0 11.2 0.6 21.1 0.7 18.5 5.0 10.6 3.5 20.1 4.6 13.2 3.6 18.8 3.7 16.8 3.7 17.2 2.9 11.3 0.4 (N) (6,607) (18,062) (24,669) (1,6,810) (6,353) (23,163) (4,219) (17,481) (21,700) (84,027,375) Source: data for Asian Americans were computed by S. Kan from the 1980 Census Public Use Microdata A (5%) sample; data for the white population are compiled from the published census reports. Table 4. Family income for three Asian American groups by nativity, and for the white population, 1979 ________Chinese________ ____ U.S. Foreign U.S. born born Total born Japanese_______ ____ Foreign U.S. born Total born Pilipino_______ Foreign born Total White U) Median family income Percent of families with income Less than $5,000 $25,000 or more $28,955 $21,010 $22,910 $29,573 $21,195 $27,475 $21,310 $24,010 $23,585 $20,835 3.1 6.6 5.9 1.9 6.4 3.1 5.3 3.3 3.7 5.6 60.1 41.5 45.4 62.9 40.7 57.0 38.6 48.1 46.6 37.8 (N) (1,981) (7,368) (9,349) (6,176) (2,209) (8,385) (1,349) (7,032) (8,381) (50,644,862) Source: data for Asian Americans were computed by S. Kan from the 1980 Census Public Use Microdata A (5%) sample; data for the white population are compiled from the published census reports. Table 5. Fetal, neonatal, postneonatal and infant deaths and death rates for specified race: United States, 1973-1977 ________Race of Child__________ White Chinese Japanese Fetal death rate1 Number Neonatal death rate2 Number Postneonatal death rate- Number Infant death rates4 Number Live births Ratio of neonatal to po death rates iFetal death rate is defined in this paper as the number of deaths prior to the complete expulsion or extraction from its mother of a product of conception (which has had at least 20 weeks of gestation) per 1000 total births (number of live births and fetal deaths combined). 2Neonatal mortality rate is the number of deaths for infants within 28 days of birth per 1000 live births. 3postneonatal mortality rate is the number of deaths for infants within 28 days to 365 days of birth per 1000 live births. 4Infant mortality rate is the number of deaths for infants under 1 year of age per 1000 live births. Source: National Center for Health Statistics. Figures on fetal deaths were calculated by E. Yu from unpublished data. All other figures are based on data published annually in the Vital Statistics of the United States, 1973-1975, and unpublished working tables on file in the Division of Vital Statistics, NCHS. 9.7 5.9 8.5 126,098 258 332 10.3 3.7 5.5 133,709 163 213 3.7 1.7 1.9 48,314 76 74 14.1 5.5 7.4 182,023 239 287 12,937,502 43,747 38,9 tneonatal 2.8 2.2 2.9 233 Table 6. Fetal, neonatal, and postneonatal mortality rates for specified race of mother based on linked birth and death records: Single births to California residents, 1973-1977 ______Rate per 1.000 births______ Non-Spanish Chinese White Ratio Fetal death rate 5.5 7.9 .70 Neonatal mortality rate 6.3 7.4 .85 Postneonatal mortality rate 3.4 3.5 .97 Source: Unpublished data from the California State Department of Health, calculated by E. Yu. 234 Table 7. Single fetal 1973-1977, by a Whi Age of Mother Number Less than 15 years 32 15-19 years 1050 20-24 years 2131 25-29 years 2066 30-34 years 1077 35-39 years 431 40-44 years 120 45 and over 7 Age not repor :ted 47 All age groups 6961 Source: Unpublished data from the calculated by E. Yu. deaths and death rates in California, e of mother and specified race non-span______ ______Chinese________ Rate per 1000 Rate per 1000 live births Number live births 25.95 0 NA 8.90 3 * 6.98 15 4.64 6.97 40 5.09 8.60 23 5.83 13.77 9 7.04 20.87 2 * 23.97 1 * 228.16 0 NA 7.88 93 5.50 California State Department of Health, 235 Table 8. Percent distribution of total births by age and race of mother, Single births to California residents: 1973-1977 Chinese Non-Spanish White Less than 20 years 20-24 years 25-29 years 30-34 years 35-39 years 40 and over 2 19 47 23 8 1 14 35 33 14 4 1 Source: Unpublished data from the California State Department of Health, calculated by E. Yu. 236 Table 9. Fetal death rate by mother's education for specified race: United States, 1973-1977 Mother's education Fetal deaths White___________ _______Chinese________ FD FD rate rate per per 1000 1000 Live total Fetal Live total births births deaths births births Total 122,729 13,082,234 Not applicable 36,478 3,421,614 0-8 years 5,549 498,608 9-11 years 16,034 1,842,690 12 years 34,899 4,374,402 13-15 years 10,427 1,533,067 16 years and over 7,486 1,202,055 Not stated 11,856 209,798 9.29 10.55 11.01 8.63 7.91 6.76 6.19 53.49 226 88 17 7 27 15 30 42 40,155 16,706 2,385 1,557 6,619 3,206 8,961 721 5.60 4.24 7.08 4.48 4.06 4.66 3.34 55.05 Source: Calculated by E. Yu from unpublished data provided by the National Center for Health Statistics. Single and multiple births combined. 2Total births is the sum of live births and fetal deaths. 237 Table 10 Crude and Age-Adjusted Death Rates1, According to Specified Race: United States, 1980 and 19702 19 8 0 Crude Age-Adjusted4 Death Rate Death Rate All Races Black White American Indian or Native Alaskan Chinese-American Japanese-Amer ican Pilipino American 8.8 5.9 8.6 8.3 9.1 5.6 5.0 5.8 3.7 3.5 4.0 2.9 2.4 2.5 __________19 7 0_______ Crude Age-Adjusted4 Death Rate Death Rate 9.5 7.1 10.0 10.4 9.5 6.8 7.2 8.2 4.7 4.9 4.2 3.3 1 Excludes deaths by nonresidents of the United States. 2 Data for "All Races" are published in the NCHS Monthly Vital Statistics Report 32, 3 (Supplement), August 11, 1983. Rates for specified race are computed by the author from unpublished data. Because some of the nonwhite races are extremely small in number compared to white deaths, deaths for the 3-year period between 1979 and 1981 were averaged. The denominator is obtained from published reports of the 1980 Census. 3 National Center for Health Statistics, published data. 4 Age-adjusted by the direct method using as the standard population the age distribution of the total population of the United States as enumerated in 1940. Adjustment is based on 10 age groups for 1980 and 11 age groups for 1970. 238 Table 11 Average Annual Age-specific1 and Age-adjusted2 Death Rates (per 1,000 Resident Population) for All Causes of Death by Specified Race: United States, 19803 Age White Chinese Japanese Pilipino All ages, crude 9.1 3.7 4.0 2.4 Age-adjusted 5.6 3.5 2.9 2.5 Under 5 years 3.0 1.5 1.4 1.2 5-14 years .3 .1 .2 .1 15-24 years 1.1 .4 .5 .4 25-34 years 1.2 .5 .5 .5 35-44 years 2.0 1.1 .9 .8 45-54 years 5.4 2.8 2.5 2.0 55-64 years 12.8 7.1 5.3 4.0 65-74 years 29.1 18.9 14.5 13.7 75-84 years 66.2 52.3 39.8 39.3 85 years and over 159.9 112.8 131.6 77.6 ^he numerator consists of 1979-81 cumulative number of deaths, excluding those of foreign residents; the denominator is based on the 1980 Census. 2Age-adjusted by the direct method using the 1940 U.S. population as the standard. ^Source: Unpublished data from the National Center for Health Statistics, calculated by the authors. 239 Table 12 Race-Mortality Ratios1 for All Causes of Death According to Specified Race: United States, 1980 Age Chinese Japanese Pilipino All ages, crude 0.41 0.44 0.26 Age-adjusted2 0.26 0.52 0.44 Under 5 years 0.49 0.47 0.40 5-14 years 0.49 0.56 0.43 15-24 years 0.33 0.43 0.37 25-34 years 0.42 0.44 0.37 35-44 years 0.54 0.44 0.37 45-54 years 0.51 0.46 0.36 55-64 years 0.55 0.41 0.31 65-74 years 0.65 0.50 0.47 75-84 years 0.79 0.60 0.59 85 years and over 0.71 0.82 0.49 iExcludes deaths of nonresidents of the United States. Ratios are computed by dividing the age-specific death rate of a specified ethnic group by the death rate of the white population in that age group. 2Age-adjusted by the direct method, using as the standard population the age distribution of the total population of the United States in 1940. Adjustment is based on 10 age groups. Source: Unpublished data from the National Center for Health Statistics, computed by the authors. 240 Table 13 Rank Order and Proportional Mortality (P.M.) of the 10 Leading Causes of Death, According to Specified Race: United States, 1980 10 Leading Causes, United States ICD-9 Codes White Chinese Japanese Pilipino Rank P.M. Rank P.M. Rank P.M. Rank P.M. 1. Heart Disease 390-398,402,404-429 1 39.3 1 31.8 1 30.4 1 33.5 2. Cancer 140-208 2 21.3 2 27.4 2 25.4 2 20.5 3. Cerebrovascular Disease 430-438 3 8.6 3 8.6 3 11.2 3 10.1 4. Accidents E800-E949 4 5.2 4 4.2 4 5.4 4 6.7 5. Chronic Obstructive Pulm. Disease 490-496 5 3.0 6 2.4 8 2.0 6 2.0 6. Pneumonia and Influenza 480-487 6 2.6 5 3.0 5 3.5 5 2.8 7. Diabetes Mellitus 250 7 1.7 8 2.1 7 2.0 7 1.8 8. Chronic Liver & Cirrhosis Disease 571 8 1.4 9 1.2 9 1.2 9 1.2 9. Atherosclerosis 440 9 1.5 10 0.9 10 1.0 10 0.6 10. Suicide & Self-inflicted injury E950-E959 10 1.5 7 2.2 6 2.3 8 1.5 Source: National Center for Health Statistics, published and unpublished data computed by the authors. Table 14 Age-Adjusted Race-Mortality Ratios for Specific Cause of Death United States, 1980 Causes of Death Chinese Japanese Filipino ■)> ho Heart Disease 0.54 Cancer 0.76 Cerebrovascular Disease 0.76 Accidents 0.34 Chronic Obstructive Pulmonary Disease 0.50 Pneumonia and Influenza 0.81 Diabetes Mellitus 0.81 Chronic Liver Disease and Cirrhosis 0.42 Atherosclerosis 0.57 Suicide and self-inflicted injury 0.64 0.42 0.60 0.76 0.44 0.34 0.73 0.64 0.34 0.41 0.62 0.42 0.40 0.66 0.39 0.31 0.59 0.49 0.29 0.25 0.30 Note: Ratios are calculated for each specific cause of death by dividing the age-adjusted death rate of a specified ethnic group by the age-adjusted death rate of the white population. Source: Unpublished data from the National Center for Health Statistics Table 15 Within-Group Sex-Mortality Ratios1 for All Causes of Death: United States, 1980 Age White Chinese Japanese Pilipino All ages, crude 1.23 1.63 1.33 3.25 Age-adjusted2 1.82 1.75 1.65 1.96 Under 5 years 1.28 1.13 1.37 1.01 5-14 years 1.56 0.73 1.31 1.39 15-24 years 3.00 2.39 2.05 2.80 25-34 years 2.61 1.60 1.84 1.87 35-44 years 1.86 1.27 1.17 1.38 45-54 years 1.88 1.70 1.64 1.34 55-64 years 1.98 1.94 1.72 2.18 65-74 years 1.96 1.97 1.95 2.58 75-84 years 1.65 1.86 1.60 2.07 85 years and over 1.28 1.31 1.38 1.71 iExcludes deaths of nonresidents of the United States. Ratios are computed for each ethnic group by dividing the age specific death rate of males by the death rates of females in that age group. 2Age-adjusted by the direct method, using as the standard population the age distribution of the total population of the United States in 1940. Adjustment is based on 10 age groups. Source: Unpublished data from the National Center for Health Statistics, computed by the authors. 243 N3 4N Table 16 Proportional Mortality for Selected Deaths in the 15-24 Years Age Group: United States, 1980 Selected Causes ________White_________ ________Chinese________ _______Japanese_______ _______Pilipino______ ____________________Total Male Female Total Male Female Total Male Female Total Male Female All accidents 58.3 61.0 50.1 27.0 29.4 20.8 44.1 46.7 38.6 51.0 54.0 42.9 Motor Vehicle 43.6 44.4 41.3 16.2 16.0 16.7 31.8 32.0 31.6 34.8 33.6 38.1 Other accidents 14.7 16.6 8.8 10.9 13.4 4.2 12.3 14.8 7.0 16.1 20.4 4.8 Homicide and legal 8.8 9.1 8.1 21.0 26.1 8.3 7.3 4.9 12.3 14.2 15.0 11.9 Suicide 11.9 12.9 8.8 16.8 15.1 20.8 19.0 21.3 14.0 11.6 12.4 9.5 Source: Unpublished data from the National Center for Health Statistics, computed by the authors. Table 17 Proportional Mortality for Selected Deaths in the 65-74 Years Age Group: United Selected Causes White Chinese Japanese— Total Male Female Total Male Female Total Male Female Malignant neoplasms 27.8 Major cardiovascular diseases 50.6 Diseases of heart 41.0 Ischemic heart disease 32.9 26.7 29.4 31.7 29.8 51.6 49.1 36.6 40.1 42.8 38.4 34.0 37.2 34.3 30.8 25.7 28.9 35.5 31.2 31.4 30.8 30.0 34.8 36.9 30.8 27.7 31.7 33.2 28.8 19.3 24.4 26.2 21.1 Source: Unpublished data from the National Center for Health Statistics, computed by the autho Table 18 Within-group Age-Adjusted Sex-mortality ratios for 10 Leading causes of death: United States, 1980 10 Leading Causes of Death 1. Heart Disease 2. Cancer 3. Cerebrovascular Disease 4. Accidents 5. Chronic Obstructive Pulmonary Disease 6. Pneumonia and Influenza 7. Diabetes Mellitus 8. Chronic Liver & Cirrhosis Disease 9. Atherosclerosis 10. Suicide & Self-inflicted injury White Chinese Japanese Filipino 2.08 2.18 2.00 2.88 1.49 1.58 1.53 1.41 1.19 1.19 1.04 1.44 2.93 1.98 1.97 2.26 2.92 3.74 3.81 2.52 1.76 1.87 1.78 2.05 1.15 1.28 0.95 1.39 2.24 4.69 1.66 2.00 1.33 2.00 1.76 3.90 3.13 0.99 2.22 2.70 Note: Ratios are computed for each specific cause of death by dividing the age-adjusted death rate for males by the age-adjusted rate for females in the same ethnic group. Source: National Center for Health Statistics published and unpublished data computed by the authors. Table 19 Nativity-Mortality1 Ratios2 for All Causes of Death: United States, 1980 Age Chinese Japanese Pilipino All ages, crude 3.58 4.78 5.73 Age-adjusted3 2.30 3.04 2.34 Under 5 years 2.52 2.46 4.13 5-14 years 2.27 5.03 3.68 15-24 years 2.82 2.86 1.34 25-34 years 1.52 1.63 1.27 35-44 years 1.69 1.59 0.93 45-54 years 1.50 1.91 1.18 55-64 years 1.68 4.78 0.89 65-74 years 5.78 3.69 2.86 75 years + 1.88 2.52 16.83 iExcludes deaths of nonresidents of the United States. 2Ratios are computed by dividing the age specific death rate of foreign born by the death rate of the native born population in that age group. 3Age-adjusted by the direct method, using as the standard population the age distribution of the total population of the United States in 1940. Adjustment is based on 9 age groups. Source: Unpublished data from the National Center for Health Statistics, computed by the authors. 247 Table 20 Age-Adjusted Nativity-Mortality Ratios for 10 Leading Causes of Death: United States, 1980 10 Leading Causes of Death Chinese Japanese Filipino N3 00 1. Heart Disease 2. Cancer 3. Cerebrovascular Disease 4. Accidents 5. Chronic Obstructive Pulmonary Disease 6. Pneumonia and Influenza 7. Diabetes Mellitus 8. Chronic Liver & Cirrhosis Disease 9. Atherosclerosis 10. Suicide & Self-inflicted injury 1.98 2.41 2.93 2.13 3.31 2.29 1.52 2.50 2.75 2.71 3.04 2.67 2.82 1.63 4.21 2.76 2.94 2.29 2.88 1.75 4.37 2.92 2.09 2.42 4.95 2.06 6.38 0.93 2.69 1.15 Note: Ratios are calculated, for each specific cause of death, by dividing the age-adjusted death rate of the foreign-born by the age-adjusted rate of the native born population from the same ethnic group. Source: National Center for Health Statistics, published and unpublished data computed by the authors. REFERENCES Breslow, L. and B. Klein 1971 "Health and Race in California." American Journal of Public Health 61, 4 (April): 763-75. Doi, Mary L., Chien Lin, and Indu Vohra-Sahu 1981 Pacific/Asian American Research: An Annotated Bibliography. No. 1, Bibliography Series. Chicago: Pacific/Asian American Mental Health Research Center. Fingerhut, L. A., R. W. Wilson, and J. J. Feldman 1980 "Health and Disease in the United States." Annual Review of Public Health 1: 1-36. Frost, F. and K. Shy 1980 "Racial differences between linked birth and infant death records in Washington State." American Journal of Public Health 70: 974. Gordon, T. 1967 "Further mortality experience among Japanese-Americans." Public Health Reports 82: 973-984. Ho, J. H. 1979 "Some epidemiologic observations on cancer in Hong Kong." National Cancer Institute Monograph 53 (Nov), Second Symposium on Epidemiology and Cancer Registries in the Pacific Basin, NIH Publication no. 79-1864: 35-47. Kan, Stephen H. and William T. Liu 1984 "The poverty class of Asian Americans." Paper presented at the Eastern Sociological Society Annual Meetings in Boston, March 8-10. King, Haitung 1974 "Selected epidemiologic aspects of major diseases and causes of death among Chinese in the United States and Asia." Pp. 487-550 in Arthur Kleinman et al. (eds.), Medicine in Chinese Cultures. U.S. Department of Health, Education, and Welfare Public Health Service. National Institutes of Health DHEW Publication No. (NIH) 75-653. King, Haitung and Frances B. Locke 1981 "Chinese in the United States: A Century of Occupational Transition." International Migration Review 14, 1: 15-42. Kitagawa, Evelyn M. and Philip M. Hauser 1973 "Differential mortality in the United States: A study of socioeconomic epidemiology." Cambridge: Harvard University Press. 249 Kline, J. et al., 1977 "Drinking during pregnancy and spontaneous abortion." Lancet 176. McCarthy, B. J. et al., 1980 "The underregistration of neonatal deaths: Georgia, 1974-77." American Journal of Public Health 70: 977 National Center for Health Statistics 1980 Health, United States 1979. Washington, D. C.: U. S. Government Printing Office. National Center for Health Statistics 1975 Selected Vital Statistics in Poverty and Nonpoverty Areas of 19 Cities, United States, 1969-71. By S. Ventura, S. Taffel, and E. Spratley, Vital and Health Statistics. Series 21-No. 26. DHEW Publication No. (HRA) 76-1904. Health Resources Administration. Washington, D. C.: U. S. Government Printing Office. Norris, F. D. and P. W. A. Shipley 1971 "A closer look at race differentials in California's infant mortality 1965-67. HSMHA Health Reports 86: 810. Office of Smoking and Health 1979 Smoking and Health: A Report of the Surgeon General. Government Printing Office. Washington, D. C. Office of Smoking and Health 1980 The Health Consequences of Smoking on Women. A Report of the Surgeon General. Washington, D. C: Government Printing Office. Patterson, J. E. 1980 "Assessing the quality of vital statistics." American Journal of Public Health 70: 944. Vohra-Sahu, Indu 1983 The Pacific/Asian Americans: A selected and annotated bibliography of recent materials. Chicago: Pacific/Asian American Mental Health Research Center. Wolff, P. H. 1973 "Vasomotor sensitivity to alcohol in diverse Mongoloid population." American Journal of Human Genetics 25: 193. Yu, Elena S. H. 1980 "Filipino migration and community organizations in the United States." California Sociologist 3, 2 (Summer): 76-102. Yu, Elena S. H. 1982 "The epidemiology of nasopharyngeal cancer in Chinese populations." Unpublished paper, Division of Epidemiology, Columbia University. 250 Yu, Elena 1982 "The low mortality rates of Chinese infants: Some plausible explanatory factors." Social Science and Medicine 16: 253-265. Yu, Elena S. H., Alice K. Murata, and Chien Lin 1982 Bibliography of Pacific/Asian American Materials in the Library of Congress. Chicago: Pacific/Asian American Mental Health Research Center. 251 Physical and Mental Health Status Indicators for Asian-American Communities Elena S. H. Yu, Ph.D. Associate Professor of Sociology in Psychiatry and Research Associate William T. Liu, Ph.D. Professor of Sociology and Director Paul Kurzeja, M.A. Graduate Research Assistant with the assistance of Phyllis Flattery Pacific/Asian American Mental Health Research Center University of Illinois at Chicago Chicago, Illinois ACKNOWLEDGMENT The authors are grateful to Thomas F. Drury for his guidance and assistance during an earlier study conducted in collaboration with Elena Yu and William T. Liu, portions of which are summarized in this report. In addition, the critical comments of Mary Doi who read an earlier draft of this report are very much appreciated. 254 PHYSICAL AND MENTAL HEALTH STATUS INDICATORS FOR ASIAN/PACIFIC AMERICANS In the last decade, the number of Asians and Pacific Islanders in the United States more than doubled, increasing from 1.5 million in 1970 to 3.5 million in 1980 (Bureau of the Census, 1981). This dramatic growth came about largely through legal immigration made possible by the 1965 Immigration Act (Keely, 1971), which took effect in 1968, and the 1975 Refugee Act following the fall of South Vietnam and other parts of IndoChina. The surge in the Asian/Pacific Islander population during the 1970s exceeded that for all other ethnic groups including persons of Spanish origin. If the present trend continues, an even greater increase in the Asian/Pacific American population is expected by the 1990s. Yet, despite increased understanding of minority health concerns, the health status of minority persons of Asian and Pacific Island backgrounds is perhaps the least understood (Weaver, 1976; Lieu et al., 1976; True, 1980). A major reason for this has been the unavailability of the necessary statistics for this population. Special studies, particularly at the local level, would be a major contribution in filling this gap. But, lacking such studies and in a time of scarce resources, alternative procedures for developing information should be explored. One such approach is secondary analysis of unpublished data available from the National Center for Health Statistics (NCHS). As the federal agency mandated to collect, analyze, and disseminate national health statistics and epidemiologic data, the National Center for Health Statistics maintains three independent sets of records which provide a modicum of information on the health of Asian/Pacific Islanders: The National Health Interview Survey (NHIS), the National Ambulatory Medical Care Survey (NAMCS), and the birth and death files for the entire United States. With the exception of an article by Yu, Drury, and Liu (1981), and another paper by Yu and Cypress (1982), data on Asian/Pacific Americans based on the first two resources have, for the most part, remained unpublished. On the other hand, data on Asian/Pacific Americans from the birth and death files related to infant mortality have been analyzed and published (Yu, 1982), while data on adult mortality experiences, with specific emphasis on race, sex, age, and nativity differences, have been presented in a companion piece (Yu et al., 1984). I. Objectives The purpose of this paper is to provide some indicators of the physical and mental health status of Asian/Pacific Americans. We will review the unpublished data at NCHS in order to provide some indicators of their physical health status. Data on the mental health of Asian/Pacific American are unavailable on a national basis, and small sample, local-area studies are few and far between. Nonetheless, a few available studies, some of which are still in progress, will be reviewed to provide some indicators of the mental health status of this small yet growing minority group. 255 II. National Health Interview Survey (NHIS) Data The National Health Interview Survey, which has been conducted since July 1957, consists of a continuous sampling and interviewing of households nationwide. Although some changes in its design were made in 1959, 1963, and 1973 (NCHS, 1975), a basic concept has persisted throughout. Each week a probability sample of households in the civilian noninstitutionalized population of the United States is interviewed by personnel of the U.S. Bureau of the Census. Consolidation of samples over a time period, e.g., a given year, produces estimates of average characteristics of the U.S. population for that year. Since the late 1970s, the usual NHIS annual sample has consisted of approximately 41,000 eligible households, which yields a probability sample of about 120,000 persons (NCHS, 1978). Based on verbal responses to an interview schedule, information is collected on the health and other characteristics of each member of the household. Until 1976, persons of Asian/Pacific Islander backgrounds were identified in the NHIS as persons of "other race" based on interviewer observations (Drury, Moy, and Poe, 1980). Since then, as a byproduct of efforts to improve the quality of the racial classification used in the sample, respondents of Asian or Pacific Islander backgrounds have been allowed to identify themselves in the interview as "Asian or Pacific Islander." Despite this improvement, the NHIS in any particular year cannot be expected to draw a large enough sample to produce estimates for persons of Asian/Pacific Islander backgrounds. Certain kinds of health information, referred to as "core" items, are collected annually by the NHIS. These include items on the incidence of acute conditions, disability days, limitations of activity, perceived health status, prevalence of selected chronic conditions, and use of physicians, dentists, and of short-stay hospitals. For each of these types of information, it is possible to produce tabulations for Asian/Pacific Americans based on multiple years, but not single years. This is because Asian/Pacific Americans, representing no more than 1.5 percent of the population counted in the 1980 census are, statistically speaking, "rare elements" in the sampling universe of the U.S. population. Any national sample drawn on a probability proportional to population size basis, as this one is, will yield too few cases of Asians and Pacific Islanders to provide statistically reliable estimates. In fact, only about 1,557 NHIS sample persons have been identified as Asian/Pacific Americans in an average year since 1976. However, pooling these data yields a subsample size of 3,009 sample cases for the two-year period 1976-77; 4,465 cases for the three-year period 1976-78; and 6,228 cases for the four-year period 1976-79. Other kinds of health information, referred to as "supplementary" items or supplements, are collected conjointly with the core items, either periodically or on a one-time basis. Between 1976 and 1979, supplementary information collected included: diabetes (1976), health practices (1976 and 1977) , health insurance coverage (1976 and 1978), problems getting medical care (1977), needs of the disabled for special services (1977), characteristics of usual places of care and blood donorship (1978) , and home care due to a disability or other health problem (1979). Much of this information has been collected for the total sample, although on some topics these data were elicited only from a one-third subsample of persons 20 years 256 and older. Still other information is available only for persons who met certain screening criteria in the interview. In what follows, we will examine the most recent NHIS pooled data for Asian/Pacific Americans on five commonly-used health status indicators, and report on some other unpublished NHIS data from single years. Whenever possible, comparable indicators available from published sources for the United States as a whole or for white Americans are noted, even though complete tables are not presented here. A. Estimates of Selected Health Characteristics Table 1 shows the average annual estimates of selected health characteristics of Asian/Pacific Americans by age and sex for the period 1976-78. It further gives the standard errors for these estimates and the sample sizes upon which these estimates are based. The first column shows the proportion of the population with limitation of activity, which is a measure of the long-term impact of chronic disease. This measure includes people who, because of such diseases, are unable to perform their usual activity (such as working, keeping house, or going to school), or who are limited in the kind or amount of that activity, or are limited in other activities. The 1976-78 average annual estimate of 7.0 percent for Asian/Pacific Americans is considerably lower than the 13.5 percent for the entire U.S. population in 1977 (NCHS, 1978). Lumping NHIS data over the three years 1976-78 proves useful in that the standard errors for these estimates are reduced. In no age group is the standard error more than 4.3 percent. There are marked differences between age groups in regard to limitation of activity, with the elderly having the highest rate (36.3 percent). This rate for Asian/Pacific Americans, however, is still lower than the 43.0 percent reported for all persons 65 years or older in the U.S. In all four age groups described, the proportion with activity limitation is slightly larger for males than for females (2.3 percent compared to 1.4, 5.6 to 4.2, 15.7 to 12.0, and 39.5 to 32.5, respectively). Such a pattern is consistent with that found for the country as a whole, although in every age group, the rates for Asian/Pacific Americans are lower than those for the total United States. Looking now at the second column of Table 1, only about 9 percent of the Asian/Pacific American sample perceived their health as either "fair" or "poor" in response to the question: "Compared to other persons your age, would you say your health is excellent, good, fair, or poor?" This figure is relatively low compared to about 12 percent annually for Americans generally from 1976 to 1981. Sex differences among Asian/Pacific Americans in self-perception of health are minimal, both overall and across age groups. Two other measures displayed in Table 1 (physician and dental visits) give some indication of the frequency of visits made by Asian/Pacific Americans to specific health practitioners. A physician visit is defined as consultation with a physician, in person or by telephone, for examination, 257 diagnosis, treatment, or advice. Such visits include services provided by a nurse or other person acting under a physician's supervision. For the purpose of this definition "physician" includes doctors of medicine and osteopathic physicians. Like other Americans, more Asian/Pacific Americans made a visit to a physician's office in the past year (73 percent) than to a dentist's (48.2 percent). The corresponding rates for the country as a whole are 75.1 percent for physician visit and 49.8 percent for dental visit. In general, men made fewer visits than women. Among Asian/Pacific Americans, women across age groups reported uniformly high proportions of physician visits within the past year, while the distribution for males by age is less consistent. Nearly equal percentages of men under 17 years of age and those between 45-65 years (76.5 percent and 73.6 percent, respectively) visited a physician in the past year, compared to a lower rate (62.1 percent) of those in the 17-44 age group (62.1 percent) and a higher rate (81.3 percent) of those in the highest age group. This pattern differs from that found for the U.S. (see NCHS, 1978: 31). A dental visit is defined as any visit to a dentist's office for treatment or advice, including services by a technician or hygienist acting under a dentist's supervision. Table 1 indicates that while physician visits made by Asian/Pacific Americans seem to increase with age (except for men in the 17-44 age group), dental visits show just the opposite pattern, with the older age groups making proportionally fewer visits than the younger ones. This pattern occurs for men as well as for women, and is consistent with findings for the total U.S. Next, we examine the proportion of Asian/Pacific Americans who reported having one or more short-stay hospital episodes in the past year. Short-stay hospitals are: general hospitals; maternity hospitals; eye, ear, nose, and throat hospitals; children's hospitals; osteopathic hospitals; and the hospital departments of institutions. Overall, the proportion of Asian/Pacific Americans who had such hospital episodes increases with age. A breakdown by sex, however, reveals some divergent age patterns. In the 17-44 years age group, the larger proportion (12.4 percent) of women compared to men (4.3 percent) who had one or more short-stay hospital episodes in the past year is most likely attributable to maternity services. However, in the oldest age group, there is a large disparity between the proportion of short-stay hospital episodes for men (17.5 percent) compared to women (5.4 percent). We suspect that the inability to care for oneself or to be taken care of by others (because of living alone, being without social support, or having a complicated chronic illness), and the ability to pay (as reflected in income, retirement benefits, or medical insurance coverage) are the critical explanatory factors of this difference. B. Data on Health Practices of Asian/Pacific Americans (1977) Data on the health practices of Asian/Pacific Americans (Table 2) was obtained on a one-third subsample of persons 20 years and over in the 1977 National Health Interview Survey. Close to 57 percent of Asian/Pacific Americans ate breakfast everyday, 21 percent of them sometimes did, and only 258 22 percent of them rarely or never did. One-third of those interviewed snacked everyday, 38 percent snacked sometimes, and 29 percent rarely or never snacked. Only one out of five persons reported needing 6 hours of sleep or less, seven out of ten persons usually slept 7-8 hours daily, and another one out of ten slept longer hours. About half believed they were as active as others, and another one-third believed they were more active than others. Insofar as smoking is concerned, 61 percent of Asian/Pacific Americans had never smoked, another 11 percent were ex-smokers, and close to three out of ten (or 28.7 percent) were smokers. Some 32 percent of Asian/Pacific Americans had never used alcohol, while 52 percent reported taking alcoholic beverages sometimes. The proportion of persons who drank alcohol once or twice a week (8.6 percent) is quite similar to that for persons who drank three or more times a week. Because of the small sample size, further analysis of the data by sex or other variables is not advisable. C. Characteristics of Sources of Health Care In the 1978 National Health Interview Survey, data were collected on the characteristics of the regular sources of health care for Asian/Pacific Americans (Table 3). Such information was obtained only from 81 percent of the total sample of 1,456 persons. Of those who had a regular source of care, the place of care in 83 percent of the cases was in a doctor's office, and in about 13 percent of the cases, a hospital clinic, emergency room or health center. Three-quarters of those with a regular source of care had one particular physician whom they consulted for medical problems. For some 19 percent of those with a regular source, the time required to travel from their home to that source was only 10 minutes. For some 38 percent, it was between 10 and 20 minutes. About 27 percent required a travel time of 30 minutes or longer. For those who did not have a regular source of care, the most common reason given was that they "haven't needed a doctor" (55 percent). One out of eight (12.3 percent) tended to see different doctors for different reasons and, hence, did not have a regular source of care. A similar proportion of persons attributed the lack of such care to the fact that they had just changed their residence. About one out of nine reported that they had not found the right doctor or their former doctor was unavailable. D. Problems of Meaning and Measurement in NHIS Data The NHIS data for Asian/Pacific Americans have three types of concept-measurement problems which limit their usefulness: (1) conceptualization and empirical identification of this population; (2) the conceptualization and measurement of their health characteristics in the context of an interview survey; and (3) interpretation of the correlates of health characteristics. 259 (1) Problems in Conceptualization and Identification of Asians From the standpoint of secondary data analysis, perhaps the most basic conceptual point to realize is that existing NHIS data cannot be disaggregated to identify the diversity of peoples included under the rubric Asian/Pacific American. Unlike Hispanic Americans who share a common linguistic root, Asian/Pacific Islanders do not share a common language or descent, either in the U. S. or abroad. This heterogeneity is amply illustrated by the fact that the 1980 Census enumeration codes for Asian Americans and Pacific Islanders included no less than 20 categories. For most purposes, the term designates residents of the United States from the following countries and territories: East Asia: China (including Taiwan and Hong Kong), Japan, and Korea. Southeast Asia: Philippines, Vietnam, Cambodia, Laos, Thailand, Malaysia, Singapore, and Indonesia. Indian Subcontinent or South Asia: India, Pakistan, Bangladesh, Sri Lanka, and Burma. Pacific Islands: Hawaii, Guam, Samoa, Tonga, Fiji, and other Micronesia Islands. In practice, however, only a few of these groups are identified in federal health and other information systems: Chinese, Japanese, Filipinos, and Vietnamese or IndoChinese. Despite the fact that a wide spectrum of linguistic, cultural and racial diversity exists even among these four major groups, the practice of lumping them together into a single ethnic group is accepted in many federal, state, and local programs. With respect to sample selection, since the probability of selection is proportional to population size, persons of Asian/Pacific Islander backgrounds are presumably included in the sample in proportion to their representation in the civilian, noninstitutionalized population. The composition of the NHIS sample of Asian/Pacific Americans most probably reflects, therefore, the numerical frequency of particular Asian/Pacific American subgroups within the population as counted in the 1970 Census. With respect to sample coverage, despite the 96 percent household response rate in the NHIS, it should be recognized that health problems among Asian/Pacific Americans may be understated by the available data. Ill health and other health problems may be unreported or not reported at all because of language and other cultural barriers or unwillingness to participate in the NHIS. (2) Problems in the Conceptualization and Measurement of Health A similar point can be made with respect to the measurement of health characteristics per se. Information obtained through interview surveys is subject to a variety of known limitations (Feldman, 1960; Finkner and Nisselson, 1978; Suchman, 1967). Whether these standard problems have unique features in surveys of persons of Asian/Pacific American backgrounds, however 260 is not known. It has been suggested, for example, that health interview surveys of minority subpopulations are likely to present unique problems due to cultural biases in the wording of questions, inaccuracy of proxy responses in sensitive topic areas, and other response errors because of unfavorable interviewer-respondent interactions (Salber and Beza, 1980; Rice, Drury and Mugge, 1980). But aside from several case studies of methodological problems in field work among particular Asian/Pacific American subpopulations (e.g., Hurh and Kim, 1982; Yu, in press), the systematic exploration of these kinds of methodological issues in surveys and other types of health research on Asian/Pacific Americans has not yet begun. (3) Problems in Interpretation of Correlates of Health Characteristics A substantive issue in using the NHIS data is the concern that differences in health characteristics among age strata may reflect either aging, cohort, or period effects, or possibly differences in the social composition of the respective age strata. Even a cursory analysis of the history of Asian immigration to the United States clarifies the scope and significance of this interpretive problem in analyses of Asian/Pacific American health data. As a result of differences in the waves of Asian/Pacific American immigrants, the interpretation of age variations is often confounded since age is associated with two other variables frequently used in data analysis: period and cohort (Schaie, 1965; Riley, 1973). Awareness of the possible effects of age, period and cohort underlying variations in Asian American data is extremely important. Koreans of a specific age group, for instance, are likely to consist mostly of foreign-born persons while Japanese of the same age group will have a high proportion of native-born individuals. Because these two ethnic groups arrived in the United States at different periods of U.S. history, the age variable alone also implies differential occupational attainments between these groups. For Chinese and Filipinos, the ratio of foreign-born to native-born is slightly greater than 1.0, whereas for Koreans and Vietnamese the foreign-born far outnumber the native-born. Among Japanese, the proportion of foreign-born is relatively small compared to those of the other Asian American subgroups. Secondary analyses of NHIS data on Asian/Pacific Americans need to be cognizant of these complexities so as to avoid spurious interpretations of within group variations. For example, insofar as particular age categories in the NHIS subsample of Asian/Pacific Americans differ in their ethnic group composition, the analyst has to be wary of the kinds of historical and social crosscurrents which age-specific variations in health indicators might reflect. A careful assaying of national data, based on census and/or health statistics, in the context of a sociohistorical and sociodemographic understanding of specific Asian/Pacific American ethnic groups is much needed. Also needed is a specially designed study of the health of specific subgroups of Asian/Pacific Americans. In many ways, the methodological and analytical issues inherent in the NHIS data may also be found in another national data set maintained by the National Center for Health Statistics, i.e., the National Ambulatory Medical Care Survey. Findings from that data-collection system are presented below. 261 III. National Ambulatory Medical Care Survey (NAMCS) Data The NAMCS is a national probability sample of office-based physicians selected from master files of the American Medical Association and the American Osteopathic Association. Sampled physicians maintain a listing of all patient visits in their office during a randomly assigned 7-day period. The strength of these data is in the precision and depth of the medical information that it provides. Reliable data on diagnosis, reason for visit, diagnostic procedure, treatments, and medication therapy are reported by the physicians themselves. A. Limitations of the NAMCS Data Statistics from the National Ambulatory Medical Care Survey were derived by a multistage estimation procedure, which produces essentially unbiased national estimates and has three basic components: 1) inflation by reciprocals of the probabilities of selection, 2) adjustment for nonresponse, and 3) a ratio adjustment to fixed totals. Notwithstanding these safeguards, caution is warranted in the interpretation of the NAMCS data for Asian/Pacific Americans. First, ethnic identification is not made by the patients themselves but by the sampled physicians or their assistants, based on their prior observations or knowledge of the patients. Second, differences exist within the Asian/Pacific American population and between subgroups, as mentioned earlier. However, insofar as Asian/Pacific Americans tend to be viewed as "alike" and are treated similarly in the health care setting, it is legitimate for some data analysis purposes to examine their health care utilization behavior as a group. Third, due to the nature of the sampling design and the resultant disparity in sample sizes between specified race/ethnic groups, the estimates of the volume and characteristics of visits to office-based physicians are more precise for the majority white population than those obtained for Asian/Pacific Americans. Despite this, the NAMCS data for Asians and Pacific Islanders are sufficiently precise as to be useful in two important senses: 1) by way of contrast with similar data obtained from the white majority, and 2) by providing a preliminary estimate of the characteristics of the minorities sampled. B. Findings from the 1979 NAMCS Data for Asian/Pacific Americans The most recent analysis of the NAMCS data for Asian/Pacific Americans was made by Yu and Cypress (1982). Their findings showed that in 1979, the majority of the physician visits made by Asian/Pacific Americans occurred in metropolitan areas (93.8 per cent). In contrast, less than three quarters of white American patients visits occurred in such areas. This pattern for Asians and Pacific Islanders is consistent with their geographic distribution in the United States. For instance, an independent estimate based on the 1978 NHIS indicates that nearly 92 percent of Asian/Pacific Americans reside in standard metropolitan statistical areas. In the NAMCS data, no discernible sex differences were observed between Asian/Pacific Americans and the white majority with regard to visits. In both groups, proportionately more females (60 percent) than males (40 percent) visited physicians. In general, 45 per cent of the visits by Asian/Pacific 262 Islanders and 38 per cent of those by white Americans were for new problems. The ratio of return visits to visits for new problems was 1.2 for Asian/Pacific Americans and 1.6 for white Americans, a difference not statistically significant. Likewise, in terms of the kind of diagnostic and therapeutic services provided by the sampled physicians, no significant differences across specified race/ethnic groups were observed (Tables 4 and 5). Table 4 shows that the proportion of Asians and Pacific Islanders who received immunizations or desensitizations (6.3 percent) was not significantly different from that for white Americans (5.4 percent). Despite the stereotype that Asian women have a strong sense of modesty, the same proportion of Asian and Pacific Islander as white American women (5 percent) received Pap tests (Table 5). Insofar as blood pressure checks are concerned, slightly fewer Asian/Pacific Americans (33 percent) received this diagnostic service than did white Americans (35 per cent). There are, however, significant differences between the groups as far as the age distributions of the patients are concerned (Table 6). For Asian/Pacific Americans, as many as 30.4 percent (roughly 1.7 million divided by 5.6 million) of the estimated visits to physicians' offices were made by children under 15 years of age, which is significantly different from that observed for whites (18.2 percent). The visit rates calculated for the two groups are also shown in Table 6. First, compared with white Americans, Asian/Pacific Americans of all ages have a lower rate. Second, while the visit rate for whites shows an almost linear increase with age, that for Asian/Pacific Americans displays a more U-shaped relationship with age. Third, within-group comparisons indicate that among Asian/Pacific Americans, in contrast to whites, the youngest age group had the highest visit rate. In terms of medical specialty, about one-third of the visits for both white and Asian/Pacific Americans were made to the offices of general and family practitioners (Table 7). However, significantly more visits by Asians and Pacific Islanders (19 percent) were made to the offices of pediatricians than were those of whites (10 percent). These data are, of course, consistent with the earlier finding that a sizable proportion of the physician visits by Asian/Pacific Islanders were made by patients under 15 years of age. Dermatology is another specialty where a statistically significant difference exists between Asians and white Americans. In addition, a significantly smaller percentage of Asian/Pacific Islanders made visits to the offices of general or specialty surgeons or psychiatrists. The cultural underpinnings of these findings will be discussed later. Table 8 presents data on physician visits by major ICD-9 Diagnostic Codes. Only 1.9 percent of the visits made by Asian/Pacific Americans were for mental disorders compared with 4.5 percent for whites. The difference is statistically significant. In contrast, significantly more visits were made by Asian/Pacific Islanders (10.2 percent) than by white Americans (5.3 percent) for problems related to diseases of the skin and subcutaneous tissue. Examination of the more detailed ICD-9 codes for this category reveals that contact dermatitis and other eczemas, as well as diseases of the sebaceous glands, are two diagnoses in which the visits are proportionally higher among Asian/Pacific Americans than whites. Since these diagnoses include diaper rash and acne, their relative preponderance is most likely an artifact of the overrepresentation of visits of children under 15 years old among Asians and Pacific Islanders. 263 Table 8 also shows that 23 per cent of the Asian and Pacific American visits, compared with only 16 percent for the whites, were made for the category "special conditions and examinations without sickness." This diagnostic group includes general medical examinations, routine normal pregnancy (prenatal) examinations, and supervision of healthy children (well-baby examinations), which together constitute the largest proportion ol visits for specific diagnoses regardless of race. Of further interest is the significantly smaller proportion of visits for injury and poisoning observed for Asian and Pacific Islanders (5.3 percent) than for whites (9.3 percent). C. Cultural Diversity Reflected in NAMCS Data In many ways, the results of our analyses are both heartening and disheartening. On the one hand, the visit rate for Asian/Pacific Americans is rather low in every age group except for those under 15 years compared with that found for white Americans. A persusal of NHIS unpublished data lends an interesting interpretation to the NAMCS finding. From 1976 through 1978, visit rates to emergency room/outpatient clinics were significantly higher for Asian/Pacific Americans (798.1 per 1,000 population) than for white Americans in 1978 (566.0 per 1,000). This NHIS finding suggests that the NAMCS data give only a partial picture of the use of health services by Asian/Pacific Americans. It is most likely that language and cultural barriers have prevented many immigrant Asians and Pacific Islanders from having a regular family doctor, as reported earlier in a few localized studies (Lieu et al., 1972; Weaver, 1976). Lack of health insurance, ignorance about the service delivery system in the United States, and a sizable number of unattached adults and elderly in the Asian populations are three other possible factors in accounting for their heavy reliance on emergency room and/or outpatient clinics. On the other hand, at least one reason for optimism exists. Taken at face value, the NAMCS data fail to show gross disparities, across ethnic group membership, in the kinds of diagnostic and therapeutic services provided by the sampled physicians. Instead, some interesting cultural diversity in medical care is reflected in the NAMCS data. For instance, within the Asian/Pacific Islander population, compared with the other age groups, children under 15 years of age had the highest rate of visit to office-based physicians. We suspect that this may be due to the presence of a sizable proportion of persons in the under 15 age group, many of whom are children of immigrants. As a result of the recency of their arrivals, we can expect that the majority of Asian Americans will continue to retain many of their traditional values and health practices in the old countries, where health is viewed as a state of equilibrium ("homeostasis") between man, society, and the cosmic forces of the universe. Maintaining a balance between the "hot" and "cold" elements of the body is the cornerstone of good health. Disease is a disturbance of that relationship. Blood is a source of human vitality and is difficult to replenish. Consequently, surgery is avoided to the extent possible since it increases the risk of losing one's blood. Indeed, in the 1978 National Health Intevview Survey, 93 percent of 919 Asian/Pacific Americans in the 17-64 age range had not given or sold blood in the past 264 year. Seventy-seven percent of those interviewed had never done so. These concepts, of course, are so deeply rooted in the traditional cultures of several Asian societies and they may cross ethnic boundaries. Some of the similarities in herbal prescriptions (e.g., the use of Ginseng for restoring vitality or stress endurance) and behavioral proscriptions (e.g., never sleeping with wet hair; "doing the month" after delivering a baby) are rather striking (Popov and Goldwag, 1973; McKenzie and Chrisman, 1977; Gould-Martin, 1978; Pillsbury, 1978; Sich 1981) . Taking into consideration the composition of the major Asian/Pacific American groups, i.e., Chinese, Japanese, Korean, IndoChinese, and Filipino, it is apparent that historically at least three of these groups have been heavily influenced by China—in language, religion, philosopy, social structure, and medicine. The fact that the Chinese presence in the Philippines began as early as the 11th century suggests the possibility of its influence in Filipino society—though not to the same extent as that found in the other countries. Therefore, despite the apparent cultural diversity among Asians, unifying threads of social values and health beliefs are interwoven into the indigenous cultures. The persistence of these health-related beliefs and practices in Asian American subcultures is evident from the few studies in the United States based on limited samples and sometimes impressionistic observations (Campbell and Chang, 1973; Ling, King, and Leung, 1975; Chan and Chang, 1976). Obviously, precise empirical evidence about the prevalence and sociodemographic variations of these ethnic practices are lacking. In the NAMCS data, demographic factors probably played an important role in the high physician-visit rate of Asian/Pacific Americans for "special conditions and examinations without sickness," while cultural factors may be salient in their unusually low proportion of visits for injury and poisoning. The former is essentially preventive health care, while the latter, though usually unanticipated, is preventable in most cases. Thus, the juxtaposition of these two findings would tend to suggest that although proportionally fewer Asian/Pacific Americans rely upon office-based physicians for their regular source of medical care, many of those who do visit physicians are apparently prevention-oriented. It is further possible that this consciousness for the importance of prevention may extend beyond the supervision of infants. IV. Mental Health Status Indicators In general two types of statistics, treated and true prevalence rates, are used conventionally to provide information on the mental health status of a given population. The 1980 Census provides some information on treated prevalence for selected ethnic groups based on estimates from a sample, whereas true prevalence data are obtained from psychiatric epidemiologic surveys conducted in the community by means of either a standardized diagnostic instrument or at least a symptom-rating scale. In what follows, we will review the census data on inmates of institutions in the United States in order to provide estimates of the treated prevalence rates. Additionally, findings from recent small-scale studies will be reported to provide as much information as possible on the true prevalence rates for Asian/Pacific Americans. 265 A. Treated Prevalence Rates Among the institutions relevant to mental health issues, only mental hospitals, correctional institutions, and homes for the aged have been tabulated in published census reports. In a recent analyis, Liu and Yu (in press) found that in every case, for males as well as for females, the rates for Asian/Pacific Americans are lower than that found for white Americans. For inmates of mental hospitals, the age-adjusted rate for white American males is 1.20 per thousand, for females, 0.70; for Asian/Pacific American males, 0.45 per thousand, for females 0.24. The white-to-Asian ratio among males is 2.67 and among females, 2.92, indicating that at least two-and-a-half times more whites than Asian/Pacific Americans are committed to mental hospitals in this country. For correctional institutions, the age-adjusted rate for white males (2.33 per 1,000) is twice that for Asian/Pacific Americans (1.11 per 1,000). Among females, the difference is smaller, the rate for whites being 0.14 and that for Asian/Pacific Americans 0.08, yielding a ratio of 1.75. In homes for the aged, the Asian rates are low as well. For males, the white rate is 2.72 per thousand and for Asian/Pacific Americans, 1.49 per thousand, the ratio being 1.83. For females, the white rate is 3.72 per thousand and the Asian/Pacific American rate 1.38 per thousand, yielding a ratio of 2.70. The low institutionalization rates for Asian/Pacific Americans should be interpreted with caution, because more than half of them are immigrants. Selective factors which determine the types of persons who are allowed to immigrate to the United States mitigate against the likelihood of their commitment to institutions. Furthermore, some evidence is available that Asian/Pacific Americans tend to return to their parent country for treatment of mental illness rather than face confinement in the U.S. where service providers are less familiar with their cultural conflicts and life stresses (Yeh et al., 1979) . B. True Prevalence Rates Data based on institutionalized populations do not provide useful information on the true prevalence of mental disorders in the community because selection factors determine who among those suffering from a particular type of psychiatric disorder actually receive treatment (Dohrenwend and Dohrenwend, 1969: 5-7). The problem is compounded by the lack of a standardized method of case-finding that can be used in a uniform and consistent fashion in population surveys to detect persons with mental disorders (Kramer, 1976: 188). But the development of case-finding techniques depends on the existence of a consensus, which was lacking, among members of the psychiatric profession as to what constitutes "psychopathology," "mental illness," or "psychiatric disorder." The publication in 1980 of the Diagnostic and Statistical Manual, Version III (or DSM-III for short) by the American Psychiatric Association represents one of several attempts which started in the 1960s to develop a consistent definition of mental disorders. The release in 1981 of the Diagnostic Interview Schedule (called DIS), Version 266 Ill (Robins et al., 1982) represents a new case-finding technique based on the DSM-III criteria that can be used in large-scale community surveys. This technique has been used to collect information in such surveys on white, black, and Mexican Americans in certain parts of this country (Regier et al., 1984). Overseas, the DIS has been used in several independently organized surveys in Taiwan, Mainland China, Hong Kong, Korea, and Peru. The results of these studies, when analyzed and published, should provide the best possible estimates to date on the true prevalence of mental disorders in several countries, as well as establish the reliability and validity of the DIS-Version III in community surveys both in the United States and across cultures. Until then, we are forced to use data collected in smaller studies to ascertain the mental health status of Asian/Pacific Americans. Psychiatric epidemiologic studies designed to provide data on the true prevalence of mental disorders in community settings fall into two broad categories: (1) those that yield information by means of clinical diagnoses based on the use of a diagnostic instrument or clinical judgments rendered by professionals (usually psychiatrists) on symptom data elicited through survey interviews, and (2) those that yield psychiatric symptomatology or psychological distress data but not clinical diagnoses; these studies employ actuarially calibrated symptom scales to screen persons who are "cases," i.e., those who score above a cut-off point which was either pre-specified by previous validity studies with clinic patients or determined from newly studied samples of human subjects. Both sources of psychiatric epidemiologic data continue to suffer from serious methodological problems, but they provide the best information yet available for understanding the rates of mental disorders outside the institutionalized context. (1). Community Surveys using Clinical Judgments Unfortunately, diagnostic data on Asian/Pacific American populations using the most recently developed instrument in the United States, the Diagnostic Interview Survey-Version III (DIS-III), have not yet been collected on a community-wide basis. However, a pilot study to investigate the feasibility of using such an instrument on Chinese Americans has been conducted recently as part of a collaborative project between the Pacific/Asian American Mental Health Research Center and a primary health care clinic in New York City. Our experience indicates that the DIS, with some modifications, can be applied to Chinese Americans in an interview situation. In this primary health care setting, the response rate was better than 95 percent. Men as well as women showed little hesitation in talking about mental health problems when interviewed in a health help-seeking context. Table 9 summarizes the preliminary findings from that study. Of the 342 patients interviewed at the clinic, the proportions that manifested symptoms that yield the clinical diagnoses of Anxiety Disorder using the DIS/DSM-III Criteria were: 7.7 percent of those 18-24 years old, 15.9 percent of those in the 25-44 age group, and 20.4 percent of in the 45-64 age group. Among young adults, more men (11.4 percent) than women (4.6 percent) exhibited clinical anxiety. After age 25, there is a sex cross-over such that more women than men expressed clinically severe anxiety symptoms. 267 About 14 percent of those in the 25-44 age group reported symptoms that are clinically recognizable as Somatization Disorder. The next highest proportion (9.1 percent) is to be found in the 45-64 age group, followed by 3.8 percent among the young adults. For each age group, the rates for women are higher than those for men. Panic disorders, like anxiety disorders, are reported by more persons (9.1 percent) in the oldest age group, 45-64 years, than in the other groups. More women (12.9 percent) than men (7.1 percent) in that age group had the disorder. In younger ages, more men than women reported clinically severe panic symptoms. But the number of cases is far too small for us to have confidence in the meaning of these figures. The percent of persons having had a Major Depressive Episode also increases with age. In the two younger age groups, more women have reported having such a mental health problem, but in the 45-64 age group, sex differences were reversed. (2) Screening Scale Studies Using Cut-Off Scores to Identify Cases Although there have been numerous studies using screening scales to identify likely "cases" of mentally disordered persons, not all have collected information on race or ethnicity (e.g., Manis et al., 1964; Phillips, 1966). Of those that have, findings on non-white populations in the sample are not always analyzed separately. In the 30-year period between 1950 and 1980, only 15 screening scale studies can be found in the literature that report findings on race or ethnic differences in "caseness." None contains data on Asian/Pacific Americans. At present, however, a widely-used screening scale for depression, the Center for Epidemiologic Studies-Depression Scale (CES-D), is being pilot tested on Chinese Americans in our New York clinic study. The findings are shown in Table 9. Using a cut-off of 16 and over to determine caseness, the data indicate a gradual increase of symptomatology with age. In the young adult group, there are hardly any sex differences in depressive symptomatology. But with increasing age, women report more depression symptoms than men. These symptoms may not be clinically severe, however. Chinese American responses to the Demoralization Scale recently developed by a University of Columbia team of researchers are similar. Sex differences are practically non-existent in the young adult group, but are found in the older groups. Further analyses of these data are being conducted to determine some of the socio-demographic factors associated with the various types of clinical disorders and depressive symptomatology. One study has reported on the use of the CES-D in a community sample of Asian Americans (Kuo, 1984). The mean CES-D score for Kuo's Asian American sample is higher than the means of the white samples reported in other studies (e.g., Radloff, 1977; Frerichs et al., 1981). A greater proportion of the 268 Asian sample (19.1 percent) had a depression score of 16 or above, which is higher than the rates previously found for whites. Among Asian Americans, statistically significant differences were found for Koreans, Filipinos, Japanese, and Chinese, even after holding constant several demographic variables (sex, marital status, age, nativity). Kuo's study remains to be replicated in other locales, using a more refined method for identifying and sampling Asian ethnics. C. Mental Health Services Utilization To date, the only published data on services utilization for Asian/Pacific Americans on a national level are those reported by Yu and Cypress (1982). Table 7, shown earlier, reveals the significantly small percentage of visits made by Asian/Pacific Americans to the office of a psychiatrist. Moreover, regardless of the medical specialty of the physican they visited, only a very small percentage of Asian/Pacific American patients received the principal diagnosis of "mental disorders." Thus, the NAMCS data lend confidence to previous observations made by practicing psychiatrists and clinical psychologists. To illustrate, in a study of 17 community mental health services in the greater Seattle area, Sue (1977) found that over a period of three years, only 3.1 percent of the Asian American clients saw a psychiatrist at intake, and not one Asian saw a psychiatrist during therapy. A number of researchers believe that cultural factors play an important role in the Asian reluctance to consult with psychiatrists, even when they admit to having psychological problems. For the most part, Asians and Pacific Islanders appear to have a tendency to somaticize their psychiatric symptoms (Tseng et al., 1974). Hence, the consensus among practitioners is that members of this minority group generally perceive "talk" therapy to be ineffective (Tung, 1980; Chien and Yamamoto, 1981). Instead, self-medication, such as herbal tonics, is the expected treatment even for mental problems, thereby causing an extensive delay in help-seeking until the disorder has clearly become unbearable or unmanageable. Lin and Lin (1978) reported from their experience that the delay among Chinese patients in seeking psychaitric help can be as long as twenty-five years. Intrafamilial resources are first utilized before outside help is sought and, even then, respected intermediaries who have the family's trust are consulted before assistance is sought from a psychiatric professional, who is a total stranger to the family. These findings, consistent as they are, are not based on large-scale epidemiologic studies. Hence, until prevalence and incidence data on both "treated" and "untreated" mental disorders among Asian/Pacific Americans are available, the extremely low percentage of visits to psychiatrists by this growing minority must be interpreted with caution. V. Summary In summary, we have presented the major findings on the physical and mental health status indicators of Asian/Pacific Americans taking into consideration the limitations of the data. Physical health status indicators in the NHIS data indicate that, compared to the total U.S. population, a smaller proportion of Asian/Pacific 269 Americans reported being in "fair" or "poor" health, or having physical health limitations. Even so, approximately one out of five reported having had no regular source of care, and the visit rates of Asian/Pacific Americans to emergency room/outpatient clinics were significantly higher than that reported for white Americans. Data on office visits collected by NAMCS suggest that a significantly small percentage of Asians or Pacific Islanders had visited a physician for injury and poisoning; a substantial proportion of their visits were apparently made for preventive care: general medical examinations, routine normal pregnancy examinations, and supervision of healthy children. Significantly fewer visits were made by Asian and Pacific Americans to the office of a surgeon or a psychiatrist. A cultural resistance to the utilization of these two types of medical specialty is suggested. Some traditional concepts of health rooted in the Asian culture appear to persist. This is evidenced by the extremely low percentage of Asian/Pacific Americans who have ever donated blood. Other interesting cultural influences are discernible in the health practices of Asian/Pacific Islanders. Seven out of ten Asian/Pacific American adults had never smoked or had quit smoking. More than four-fifths of this population either never drank alcoholic beverages or do so only occasionally. Although much of the data reported here are based on small samples, we are impressed by the consistency of findings from independent sources of data-collection systems. By and large, these findings support intuitive impressions that such risk factors as smoking and drinking are indeed less prevalent among Asians than among the majority Americans. A review of mental health status indicators reveals that, compared to white Americans, Asian/Pacific Americans have lower rates of commitment to mental hospitals, correctional institutions, and homes for the aged. However, a pilot study based on patients entering a primary health care setting show one Asian subgroup, Chinese Americans, to have high rates of major depressive disorder, generalized anxiety, depression symptoms, and demoralization. In addition, a community study using the depression symptom rating scale suggests that Asians have higher mean depression scores than reported for whites in previous studies. While the validity of these diagnostic instruments and screening scales has yet to be firmly established for Asian/Pacific Americans, and their reliabilities remain to be tested, there is sufficient evidence to indicate promise in the development of standardized instruments applicable to studies of white and Asian Americans. The extent to which the disclosure of mental disorders or depressive symptomatology among Asian/Pacific Americans is influenced by cultural attitudes toward mental illness has not been studied. Such attitudes may well turn out to influence definitions of behavior disorders with consequent disparities in statistics on mental illness. Certainly, more systematic large-scale studies are warranted on the comparative health and mental health behavior of Asian/Pacific Americans and other ethnic groups, especially white Americans, in order to increase our understanding of intra- and inter-ethnic differences in health and mental health. 270 Table 1. Average annual estimates of selected health characteristics of Asian/Pacific Americans by age and sex: United States, 1976-78 N3 Age and sex Limitation of activity Perceived fair or poor health Physician visit in the past year Dental visit in past year One or more short-stay hospital episodes in past year Sample size 1976-78 *a S.E.b *a S.E.b ta S.E.b *a S.E.b *a S.E.b Both sexes All ages.................... 7.0 Under 17...................... 1.8 17-44......................... 4.8 45-64......................... 13.8 65 years and over--........... 36.3 Male All ages.................... 8.1 Under 17...................... 2.3 17-44......................... 5.6 45-64......................... 15.7 65 years and over------------ 39.5 Female All ages.................... 5.9 Under 17...................... 1.4 17-44......................... 4.2 45-64......................... 12.0 65 years and over------------ 32.5 0.3 0.5 0.4 9.3 0.6 9.7 73.0 0.7 48.2 0.8 7.0 0.3 8.9 0.8 69.6 0.9 46.3 1.3 5.4 0.5 0.7 76.3 1.1 50.0 0.8 4,465 0.3 4.1 0.9 76.3 1.3 49.4 1.5 3.8 0.4 1,448 0.6 8.5 0.8 69.1 0.8 48.5 1.3 8.5 0.6 2,111 1.6 17.5 1.0 76.7 1.7 51.4 2.3 6.7 1.0 670 3.3 23.3 3.4 79.5 2.7 29.4 4.1 12.0 1.9 236 2,150 0.6 4.1 1.3 76.5 1.6 48.7 2.0 3.8 0.9 732 0.8 7.2 1.1 62.1 1.6 45.0 2.3 4.3 0.6 978 2.1 18.1 1.5 73.6 2.4 51.3 3.2 7.3 1.5 315 4.2 25.0 3.4 81.3 3.1 31.4 4.3 17.5 3.3 125 8.6 0.5 2,315 0.2 4.0 0.8 76.0 1.6 50.2 1.5 3.9 0.8 716 0.5 9.7 1.0 75.4 1.3 51.7 1.1 12.4 1.0 1,133 1.9 16.8 1.7 79.8 2.3 51.5 2.1 6.1 1.6 355 4.3 21.3 4.9 77.4 4.7 27.0 5.4 5.4 1.6 111 Source: National Center for Health Statistics, 1976-1978 National Health Interview Survey, unpublished data. aA 95-percent confidence interval for any percentage may be obtained by multiplying the standard error (S.E.) for that percentage by 1.96 and then adding the product to the respective S.E Standard errors (S.E.) for these three-year, average annual estimates were computed by a Taylor-series linearization method• Table 2. Selected health practices of Asian/Pacific Americans 20 years and over: unweighted data Practice _________Number__________^Percent Asian/Pacific Americans 20 years and over.....---- 256 100.0 Frequency of breakfasting Every day-.....------.....------------------- 144 56.7 Sometimes...........------------------------- 54 tt'Z. Rarely or never----------------------------- 56 zz,u Frequency of snacking Every day----------------------------------- 85 33.2 Sometimes----------------------------------- 96 37.5 Rarely or never----------------------------- 75 29.3 Number of hours usually sleeps 6 hours or less----------------------------- 52 20.4 7-8 hours----......-...................-------- 178 69.8 9 hours or more----------------------------- 25 9.8 Perceived physical activeness More active than others---------------------- 88 34.8 As active----------------------------------- 136 53.8 Less active--------------------------------- 29 11.5 Smoking status Never smoked-------------------------------- 154 60.6 Present smoker-----------------■------------- 73 28.7 Former smoker------------------------------- 27 10.6 Frequency of most frequently used alcoholic beverage"" Never use alcohol---------------------------- 81 31.6 Occasionally-------------------------------- 132 51.6 Once or twice a week--------------------...... 22 8.6 Three or more times a week-------------------- 21 8.2 Obesity status Obese3--------...........-----........---..... 32 12.7 Non-obese----.......—...............------..... 220 87.3 Source: National Center for Health Statistics, 1977 National Health Interview Survey, 1/3 subsample of persons 20 years and over, unpublished data. a0besity is defined here as a body mass index (weight/height ) of 27 or greater for males; 25 or greater, for females. Index was computed with weight in kilograms and height in meters. 272 Table 3. Characteristics of regular sources of health care among Asian/Pacific Americans: unweighted data Characteristic Number Percent All Asian/Pacific Americans---------------------- 1,456 100.0 Regular source of care Yes—...............-.............------........ No.......................-—.................... Unknown---------------------------------------- Usual place of carea All with a regular source---------------------- Doctor's office-------------------------------- Hospital clinic or emergency room-------------- Health center---------------------------------- Other-----............---------------......----- One particular doctor5 All with a regular source---------------------- Yes............................------.........— No—.....-....................-.......-.....— Unknown---------------------------------------- Time to get to regular source5 All with a regular source---------------------- Less than 10 minutes--------------------------- 10-14 minutes---------------------------------- 15-19 minutes---------------------------------- 20-29 minutes---------------------------------- 30-44 minutes---------------------------------- 45-59 minutes---------------------------------- 60 minutes or more----------------------------- Unknown---------------------------------------- Reason for not having a regular source All without a regular source---.....----------- Haven't needed a doctor------------------------ See different doctors for different problems---- Former doctor not available-------------------- Haven't found the right doctor-----.......------ Just moved here-------------------------------- Other reason----------------------------------- Unknown---------------------------------------- 1,175 80.7 269 18.5 12 0.8 1,175 100.0 968 82.4 124 10.6 29 2.5 54 4.6 1,175 100.0 876 74.6 252 21.4 47 4.0 1,175 100.0 219 18.6 224 19.1 220 18.7 191 16.3 166 14.1 49 4.2 62 5.3 44 3.7 269 100.0 147 54.6 33 12.3 15 5.6 16 5.9 32 11.9 23 8.6 3 1.1 Source: National Center for Health Statistics, 1978 National Health Interview Survey, unpublished data. ^is information was obtained in the interview only for persons with a regular source of care. 273 ho Table 4. Percent physician visits by type of therapeutic services rendered for specified race: United States, 1979 •i Asian/ Therapeutic Totalx U.S. White Pacific services rendered Americans Total Estimate^ of visits--------- 556,313.431 502,926,839 5,559,524 Percent ------------------- 100.0 100.0 100.0 None------------------------ 19.8 20.2 18.0 Drugs (prescription and non- prescription)-------------- 51.3 50.1 54.5 Injection------------------- 9.6 9.4 13.1 Immunization/ Desensitization------------ 5.2 5.4 6.3 Diet counseling-------------- 6.0 5.8 6.9 Family Planning-------------- 1.4 1.4 1.8 Medical counseling----------- 22.2 22.3 19.1 Physiotherapy---------------- 3.1 3.0 2.9 Office surgery--------------- 7.4 7.6 7.6 Psychotherapy/Therapeutic Listening------------------ 4.4 4.6 2.5 Other therapy---------------- 3.5 3.6 2.8 Includes data for Indians and Alaskans in the United States Percent do not add up to total because more than one service may be rendered per physician visit. Table 5. Percent physician visits by type of diagnostic services rendered for specified race: United States, 1979 Asian/ Diagnostic services Total U.S. White Pacific rendered Americans Total Estimate of visits-------- Percent None- Limited exam/history------- General exam/history------- n> Pap test- Clinical lab test— X-ray-------------- Blood pressure check- EKG---------------- Vision test- Endoscopy— Mental status exam- Other diagnosis---- 556,313,431 502,926,839 5,559,524 100.0 100.0 100.0 10.1 10.6 8.2 63.0 63.0 64.8 16.8 16.6 20.3 4.9 5.0 5.0 23.2 23.0 21.8 8.2 8.2 5.2 36.0 35.3 33.2 2.7 2.7 0.5 6.0 6.2 6.6 1.3 1.4 1.8 1.5 1.5 0.4 3.5 3.6 3.9 Includes data for Indiens and Alaskans in the coterminous United States. Percent do not add up to total because more than one service may be rendered per physician visit. Table 6. Estimates and rates of visits to physicians by age for specified race: United States, 1979 Age group All races Visit Estimates Visit Rate* White Visit Estimates* Visit Rate* Asian/Pacific Visit Visit Estimates Rate* ho All ages- Under 15 years- 15-24 years- 25-44 years- 45-64 years- 65 years and over- -------- 556,313,431 2.58 502,926,839 2.70 101,352,298 2.02 91,267,680 2.21 82,289,782 2.05 74,295,285 2.18 151,713,912 2.59 133,337,485 2.61 128,594,299 2.96 117,859,164 3.04 92,363,140 3.96 86,167,225 4.08 5,559,524 1.55 1,690,800 1.98 675,243 0.99 1,590,530 1.49 1,061,339 1.51 541,612 1.94 * Based on estimates of the civilian, noninstitutionalized population of the United States excluding Hawaii and Alaska, furnished by the Health Interview Survey, 1979. Race for black and white Americans was based on interviewer's report. Race of Asian/Pacific Americans was based on respondent's self-report, Table 7. Percent physician visits by medical specialty for specified race: United States, 1979 Medical Specialty Total U.S. White Asian/ Pacific Americans ho Total Estimate of visits---------- 556,313,431 Percent-------------------- 100.0 General and Family Practice---- 34.2 Internal Medicine------------- 12.0 Pediatr ics------------------- 10.5 Obstetrics and gynecology----- 9.1 Other medical specialties----- 7.0 General surgery-------------- 6.1 Other surgical specialty------ 16.0 Psychiatry------------------- 3.1 Other----------------------- 2.1 502,926,839 5,559,524 100.0 100.0 33.4 32.0 12.3 9.3 10.3 19.3 9.0 13.5 7.3 9.0 6.1 3.8 16.3 11.7 3.2 0.6 2.1 0.8 Includes data for Indians and Alaskans in the coterminous United States. Table 8. Percent physician visits for specified race by principal diagnoses ICDA code: United States, 1979 Principal diagnosis and ICDA codes Total U.S. White Asian/ Pacific Americans All classes Estimate of visits- Percent----------- Infective & Parasitic Diseases- Neoplasms-------------------- Endocrine, nutritional, and metabolic diseases---------- Mental disorders--------- Diseases of the nervous system and sense organs- Diseases of the circulatory system------------------ Diseases of the respiratory system------------------ Diseases of the digestive system---------------- Diseases of the genitourinary system-------------------- Diseases of skin and sub- cutaneous tissue----------- Diseases of the musculoskeletal system--------------------- Symptoms and ill-defined conditions----------------- Accidents, poisoning, and violence-------------- Special conditions & examinations without sickness------------ Other diagnoses None or unknown- 556,313,431 100.0 3.5 2.6 4.1 4.4 9.1 8.9 13.2 4.4 6.6 5.2 6.7 3.1 9.3 15.8 1.5 1.6 502,926,839 100.0 3.5 2.7 4.0 4.5 9.4 8.8 13.1 4.5 6.3 5.3 6.5 3.1 9.3 15.9 1.5 1.6 5,559,524 100.0 4.1 I-4 3.2 1.9 9.1 7.3 15.1 6.6 5.0 10.2 4.0 1.8 5.3 23.3 0.6 1.1 includes data for Indians and Alaskans in the coterminous United States. 2 Includes diseases of the blood and blood-forming organs, complications of pregnancy, childbirth and the puerperium, congenital anomalies, certain causes of perinatal morbidity and mortality, blank diagnosis; noncodable diagnosis; and illegible diagnosis. 278 Table 9. Mental Health Prevalence Rates in Per Cent in Chinatown Health Clinic by Sex. Age groups 18-24 years 25-44 years 45-64 years Male Female Total Male Female Total Male Female Total (N=35) (N=43) (N=78) (N=40) (N=92) CO X o 4-> JZ CD •H G -P CD "D G 14- E o -P CD 4-J G X 4-J C-I E c CD o G 4-> CD 4-J X 2 CO o E i—l 3 O 4-J x: CO CO X >,*»- G CD o O C-i CO CO CO 4-> 4-> G G r—l co u- r—1 C C4- CL4-J a.-p >> CO o p »N ZJ X G G XI o o E CL G CO CD JZ 2 o >^ >> CO 4-) G G 4-> O ■H O CO CD G G -P 4-J (4- r—1 -P X X cr 4-J C-i 4-J G G CO Q.U- G •H U- G ■H CO o 4-> G C O- n CP G c CO o CD 2 CD CD 4-> c •H 4-J - E G CO 4-) cr E G CO CD JZ G 3 TJ CO G E c ■a ■H CD G G 4-1 CO U- to 4-J ■a i—I E CO cn G G CO C G c 4-J C-i 4-J CD as G X — O CD c Q. E G c •H i—l C-i •H G CD CO CO X E G co c CD CD E O Z ■H U- r—l Z CD XI E cr r-H >S«»- G O 4-J 00 CD •H 4-J O G CO J* ■H •H X! 4-) G ■H r—l co O CD CO c On 3 2 CD 4-J G CO CD 4-1 X E c CO •o 3 O o r—1 CO 4J P c G ■H G C G •H o >> U- t*- G cr C-i CO G 1 CO P CD CD 4-1 X CL G < G G CD X o C-i o a G "C •H CD a E o 4-J O CO ■o C-i 2 G CD CD •H CD- C-i CO •H 4-> CD C c 2 •H G 4-J r-H ■H -^ G 4-J CD o •H ZZ CD cr CO O 4-> 4-J * X TZ CO CD X G C CD r-H X x •H P 4-J CD c LH 4-J G o • 2 CD o JZ C-i ■a a. .p CD 4-1 o ■H G G o CD CO G CO CO i—l 4-J G G X X >>-P > >> G X ■H ON CO CO G CD G CD r-H TD CO G c 4-> cr r-H C P 4-J CO i—l CO XI X 4-J X Ch G 4-J CD •H CO ■H ■H CD CD CD CO CD CO 1 4-J CD 3 X CD G CO JZ C-i X c 4-) r—1 CD G l*- •H C G CJ i—l 4-> G 4-J 4-J ■a X cr 1 G 4-J •H 4-J CD XJ 0) o E ■H C-i cr CD LO C-i CD o cr Q. •H CO i—1 •H u TO X E CD Q G CD C ■a cm G G o co 2 CD CD Q. CO CD c 4-J Q. o >> G X i—l G c cr X G •H Q. C-i c-i E CO G CD i—l ■H CJ i—l CO CD i—l CD •H 4-> C-l 4-1 CO CD CD O o CD X C-i TO - i—l C-i X CO c CO CO G t_J 4-J G ax: 4J G CO 4-J CO G G CD G 4-1 c CO •H c 4-J r—l •H G c E 4-1 c X cr c G 4-1 C-i C-i > i—l o C-i CD •H c ZJ X •H X ■H o •H i—l CD o G CD CD G G CD •H o 40 X G CLr— > 4-> CO CD 4-J o •H E c CO G ■a CO G CO 4-J cr 4-J G o C-i r—\ to CO CD P 4-J 4-J >> o c ■H G X c CD ■H •H G Q. G CD CD CO ■H JZ o CD CD ■H G TJ l_i_ C-I G G CD 2 C-I . CO CD 2 •H Q. CO r—1 0_ 4-J a G o o E • iH E CD CO X CD 4-J CD E CO G CO E s: U- U- 4-> •H "O G G 4-1 G C CD ■H X ■H G C-I G C-i O o X) c C-i G X G t— C-i o CD Z 4-1 G C-i > 4-1 G U- 1—1 Ui 1 •H O C 4-J -Ni •H CD 4-> C-i CO 4-) G G G •H CO cr o G G Ui •H C4- G CO CD G CO CD c > CD 4-J 4-J ■H 3 G "O G 4-J U- r—1 U- 4-1 C-i CO 4-> ■H X ■H CD C G C >> G O G r—l o >^ D o 4-J X a CO x 4-J X P •H G •H u- 4-J •H E G O 4-) JZ o CO 4-J 4-J CD cn CD CO o o_ E o ■H CL CO E >^ ■H C-i CO 4-J cr i—l CO CJ z X -P G CO U- G CO TO i—l C-i CO 4-J r-H 3 3 o CD G CJ o I— cr cn 4-> X O < c •H G U- CO G •H 4-J G c G G C-i < p CD c CO co t— C-i o XJ 4-1 CO o - •H XI O G X G X) E CD o G Q_ CO •H CD C CO 4-J C-i •H 4-J l*- G CD 4-1 >s a C X G G G 4-> •H G CO a. CD CO c XI 4-1 O C 4-> CD CO •H G c a G G CO > G to r—l G > c G G CD 4-J ■H ■H > CD CD G G CO X •H G G G G o C-i C CO ■o C-l CD G 4-J TO P 3 r—1 4-> > Z G G > > c CO CO- C-i o CO CO -o o X ■H C ■H CD cr G C-i crx x: C-i G o CO r: X ^ CO G c cr > o Q. G G > <4- G 4-J 4-J G r—l G U- G G X! G 2 •H o CO (-1 G CO CO G o CD "O CO o C-i a CO G G 2: •H — CD cr X c i—l G c c cn G G CO G -cz G C 4-J o CO 4-> 4-> o c TO 4-J G G CO c X cr X X E G G CO G CJ CD CO i—l G r—l CO c cr 4-) G CO ■H 4-J o 4-J r— G •H C-i C-i Z ON CD O a CD CD E CD G CD XI X f-i •H ■H r-H > G - TO i-H t-i CJ >^ G to C 4-1 •H G D C4- 4-J 2 JZ C-i a CO "O o CD l—l X o o i—l CO C-i P CO X CO o CO a o G E G c u CD CJ tj_ G o ■H < G o x; CO G co C-i CO CO C-i CO co ■H CO 4_) x o 4-J c 4-J 4-J G 1 4-> X ■H G G G G 4-J Q. X 4-1 c r— JZ G •H CO CD 4-1 C-i CO X 21 CO "O •H c X ■H G (—1 4-J CD l—l CD p C C-i 3 CO TJ X 2 u ■H 4-) G G 4-J c C-i cn i—1 Z Q.-P G CD O G l—l G CD CO •H E CO i—1 G o c CD Ui E C TO TO Z CO G z ■H TD G c 4-J G G CD 4-1 c CD . CD Ui o o C C 4-J G G o X C G o c r—1 X G U- •H X 1---1 _C •H G G 3 CD CO or CD G 2 co X3 G G G 4-1 X o E G CO G <4- CO ■H E G ZJ O -¥ z x: O 4-J X o G 4-J r— CO •H 4-> CO _J >s CO G CD CD >^ G 4-1 - cn r—l 4-1 ■a a^ l»- 4-1 •H >N •H C CO . CO CO . c G o •H i-H JZ) C-i O CO CO ZJ C-i C-i G o ■H ZJ c cr X >> o CD G o C-i CO u c c 4-> o o 4-1 4-J C-i G E c C-i G o ■H CO -P •H r-H ■H co O •H ■o CO 4-J •H S c c -P G CD C4 X CO E cr G G 4-> G CO G G O G -P CO >^ ZJ CD TO G ZJ r-H C-i X ,,-3 G r—1 G <4- ■o T3 cr jz cr ZJ o co cr CD CO O r-H G c c co 4-> G Q.L- c E G G •H CO X o ZJ C-i O •H •H X CD G •H 4-> G o j* cr <♦- Q. J* G C-i C-i G C-i u_ G •H CO >s4-l G o C G C-I G CD G >s CD 4-> •H Q.4-1 r-H G 4-1 o C-i c rH CO u 4-> r— •H X X r—1 •H CD U- CO o cn G O •H C-i s G CO ■H C-i E CJ CO cn E C-i CO CO CO 2 G ■ >s c c C o - CO- G >> G Q. CO cm G r-H X o CD XJ C >s CO X 4-1 G 4-J X G c > CD o X C •H C-i •H 4-> •H G •H G G •H C-i G r~3 G G CD E CO ■o C-i CO ■H 4-J ZJ •H G 4-J ■o O TZ G > CO CO X) TZ C-i 4-J CO >^ G G c c G C X C-i •H G O o C r-H X3 C-i X CO •H XI CO 4-J G X 2 s: O L_ CO CD G 4-J E •H CO G G S G C T3 G CO Ch ■Q 4-> •H o J* •H O C CO r-H cr •s G G ZJ G >^ > 4-J 4-J CO U- -H CO ■H o G CO ■D o C-i XJ C-l C-l CO •H CO G CO •H ^ G G X) CO CO G CO G I— cr co c-i x: 4-J 4-1 X TJ G CD G G -P CO XI cn g O 4-1 G CD •H cr C-i c G O CO . ■H <4- o TZ rH CO u -a X r-H r—l G cr - C-i co o Q.U- ZJ CO G G G 4-J CD G >> G C-i o C Q. o X C-i C-i r-H X < G C-i G C Q. G •H O a-u co G CO C 4-J X CO C G x: 4-j Q. G 2 G o i-H C 4-J 4-1 CO G X 4-> Ci ■a 4-1 cn CD CO X co o G CJ Ui 4-1 o G >, c O 4-J c >^ G C-i C-i rH ~ Q.JO 4-J G 4-J CD CD cn r-H X >s o G cr G CO >N G •H •H 4-J ■a c cr G G t4- G o > G r-H cr C-i C-i X cr ZJ •H > r-H > CO CO X C-i G O G G u > •H CO P co G X G G CO c J* G c •H CO cr ZJ G G G U- T3 X G ■H G CO -P •H G ZJ o CO 4-1 x: 4-1 co o <—* 4-> "O E X 4-J 4-1 CO G r-H a. •H CD 4-1 4-J G G 4-> G E G G CO 4-1 r-H •H •H "O C •s CO C X XI G C-i X G CO CD G >s E CD O G CO CO G O G C-i G G 2 cr O XI E G -H X G X G -P ■H G ■a C-i o CO O G 4-> 4-> G 2 X C-i X cr CO ZJ G •H CO TO G T3 ZJ C X 4-J as 2 G G CD 2 o X 4-1 et G XJ XI >s CD 4-J G G X •H CO CO 4-> CO r—1 ZJ O -H XJ > r-H X3 cr CO 2 X cn C-i z TO •H CO 2 P-i "O CD •H C-i 2 c TZ G E o cr CL 4-J 4-> CO CO G CO E o CO •H G O 00 CN CD E CO CO G X G CD 4-1 CO CO -^ C-i C-i C-i ON 00 O - 4-J -H G G ZJ X CO G c G G X U- r-4 O CO G X CO TTJ ■H CO >, c G G C-i o G Q.4-) 1 r-H •H 4-1 CO > G 4-> •H ■H CD ■H CO CO CD 00 C-i CD r-H CLTTJ c-i x: •H -M i—l "D 4-J CO- c 4-J r^ g C-i 4-J CO C G 4-1 C-i C i—l cr G CO c o CD ON X CO CD G G CD CO o o •H •H 4-> G •H CO Q r-H t— X Q X X G c G 2 O ZJ •H >> i— ^x: 4-) ■H G c CT CO X r-H r—\ 4-> -H :z X C-i X o_ I— ■H • • • Q-i—I Q. G G 4-J r—1 >> D > Q. CD CO "O •H CO- I--1 C O CD ZJ G G c X CO (4- •H CO C-i G CO X Q CD 2 Q. o 2 E cn X oo CN CO •H CO CO ' 4- CO p TO cr r-i G G Q. G G G o G CO r-i P > z X X G G -P TZ TO •H G •H o 4-J 4-1 P CO N G 1—1 ■r-i E cr 2 X 4-J p z G •H X Ui G CO as r— co en O it- o •H C 4-1 c -P G P 2 -P CD CO 4-J G CO . o CO CO cm g G -P c G CO G cn<4- CO o >, G CO CO- TO • P CO I-H -H CO P •H C p o G CO P CO. CD G CO O G G P CD o -P X z CL -P G P X >, •H G G P CO z G >> CO cn CO 4-J 4-J • p -P r-H -P Z o c XI ■r-i C P CD C CO i-H -H co i-H -P CO XJ G CO 4-J o •H X O G G •H -H g x: p P CO c •H CD CO c ■H P 2 •H > E 2 X P 2 CO o C G P r-H P •H CO CO 4-J -P -r-i O -P 4-J O G •H co -p c 4-J C CD r-i cr r-H CO O CD c ■H r—| 4-J CO X as co_ G CD G i-H G o r-H CO CO -P G o CO CO G > G cr E E CO Z TO •H O CO G ■H •H c z CD co o G •H CO P a. -P P- CD G TO TO CO G 4-J P >s •H 4-J TO -P G O G CD G G G C CO E G CD p XJ -P -P ■I-i TZ a. p N G G CO P •H z •H CO X o CD CO >^ C CO ■H X CO CO G TO G TO G c Cn< >s G i-H -P P- CL G CO 1 TO c C Z p CO C XJ ■i-i <4- O E CO •H •H C >^ o CO TZ G E o TO X -P c o O G TO 4-J CD XJ ■H G a E C TO 4-> Z o G G X r-i -P av CO G CO CO G r-H CO Cn 4-1 G Z >^TO z G •N ClX -P CD X G G C CO X E 4-> G X G G -P CO >> CO G cm P -H CO c P 4-J •H G •H cr E Ui c CO P X z Z 4-1 P O O CO r-H C p G O ■r-i CO o TO 4-> o cm co •H CO ■H -H G 4-J •H G X -P ■r-i < CO G p •H -i-i CD 4-J >s4-l XJ Z CO cr C p- CO -P Z X X p- p CO c •H G CO r-i •H G •H G G >s-H 4-1 4-J G X r-H CO E P P TZ > cm G XJ r-H G 4-1 4-J Z E •H CO •> c C >s . cm c •H P- TO X G •H Q. i-H G CO •H TO o i-H CO z c TO o G r— CO 2 O P X CO cr G •H G CO o G CO 4-> P CL o X G x: g CO CD E TO G a. CD CO ro CO >s p G CD 4-J c ■H TZ CD z X Ch o G G "-» X i-H CJ •H O r—i cn rP C <4- 4-J G G r-i CO -H •r-i r-H G G -H c X P CO c 4-J G CD ■H •H -P G CO 4-J TO v_x > c o 2 co. G -H ►—1 > CO -P X C > G CO C . (D •H CO •H Q. x: r-i G CO CD -P 2 ■r-i G G -H •H c x: CO Q. -UJ CO et r-H r—1 G CO TO G P O -P C CO CD cr O -H a •H X r-H G CO G CO •H G -H E CD TO 4-> 2 E G -P G r-H G X CO 4-J CD 4-J CO 4-J X P G cr G X CO Z CO TO 4-J CO rH G G >> X o ■H CO CO CO -P -P ■H CO Ch O G CO CO CO CO CD G TJ p P CO - CO Ch Z E P G ■H P CO P- P C c G c CD "—I >> CO CD G -P O G -P CO G CO P r-H CD ■H E o G CO CO CO 4-1 P i-H CO CO C X G •H CO CO et •H G Z G CD z G G G G 4-J G J* G 4-J CO TO >>x: CO G G cr E t-i -P CO CO * TO G r-H C ■H •H p G P 1 ■r-i O O O TO CO r-H CD XJ CO cr P > G z CO cm O X CO P D- C TZ G CO f-4 co •H •H O -H > CO X cr •H -P CO G -P TO G CD p CO P TO •H G •H G ■H TO i-H P N C O G CO < G 4-1 C i-H E -* O 2 c G O CO •H G E C-i P co. Q Q.-H G G Ch G CO CO TO G rH •H ZJ o E TO G TZ 4-J G G CO a G -H CO 4-> CO O P- o C G C CO TZ CO TZ 4-J G E r-H -P G CD -H co G CO • a C CD G >> ■H cr CO z XI CO P CLX CO CO O P CO > cr CO TZ CO CO CO -r-i 4-J G P P CO CO G CO CO o o CO X P P P- CO o o c G G G >> p CO G cm ■r-i c O G O O C P G 4-J CO G G XI p Q. p cn c o P -P O G X CO J* G G X -H X G G co ■r-i Ui CO. G r-H G TO 1— G G ■H ef r— P 4-1 > •-. Q. p 4-J Ui CO CO CO G TO -H •H CO Ch G r-H ■H -P G CO G 2 X t-i z P G CO "D r-H G a CO CO r-i CO G > •H G CO -P CO 4-J cm CO C C E I-H G G G o X CO X G Ch O X G G CO O •H CO et t—1 TO X TO G 4-J G G CO TO XJ G CO TO XJ G G cm CO G p >> P c G X G 4-J CO •H 4-J ■H -P 4-J >, r-H CD cm cr > G G ■i-i CO X ■H o i-H cr •H 2 t-> G G cn c 4-> G o X O CD i-H 4-> ■H z > •H 4-1 x: > TO X) - G •H X rH E 4-1 ■r-i c cn o C CO CO 4-1 X 4-J X E G 4-J ■H G c p »• G CO •H c ■r-i 2: r-H cn O G CO co p o >^x P o X CO ■H G ■•—> i-H 4-J XJ CO CO 4-J >> CO ■H 3= TO G X X) G cr z i-H CO 2 4-1 G > -P C X G Z P o CO i-H cr U_ ■I-i P CO G CO >*x CO G G p TO o Ui CD E r-H X X XI 4-J cr p 4-> G G p G G z -P >>4-J CO o 4-> o -P 4-J CO G 4-J a 4-J •H CO TO ■H o P 4-1 ■r-i c > X CO o X •H 2 G C 2 c r-H O CO E o ■H 4-J >s CO. 4-J P G CD X CO co -P CO. ■r-i •H TO rH CO •H o C C G p o c CO r-i 4J CO G 2 c O G CO z TO TZ o G i-H CO G G TO CO •H •r-i P ^H CO ■H •H co CD p X X c G CO E 4-> G CD >^ cr cm to c 4-J 4-J CD X G G P Z CO G o G C o C CO o -P P P CO P TO CD X P •H •H i—i z TZ 4-> i-H CO O p- •H •H 4-> TO cm rH cr CO 2 Q. CO TO > TO G i-H >s G r-i CD CO z o E CO •H ■i-i c X ■H CD C p a ■i-i CO TZ X O G 4-J G TO CO cm r-H G co cr >s G •H G G CO CO C z G > as o Ui p G > TO C CO G •H 2 o P G p TO •H CD -P G ■r-i c C G X G X P o G 4-J c CD TO CD CO ■H TZ 4-J G ■H 1— CO p p G CO z C •H P c X > G o G •H G o x: •H CO 4-J G CO c -P P ■H 4-J G -P TO i-H G 4-1 cr G a. o G . TO •H c CO o_ ■r-i G o CO X p P > CO Z X 4-J z •H E G 2 X •H P G o 4-1 o o ■r-i -P cm CO o P co X 4-J -P CO X Ui CO z ■<-i G Z X 4-J TO CO ■r-i G G cm CO i-H >> o 4-1 c ••—i G G 4-J r-H 4-J X CO ■H G CO i-H G c CO G c 4-J CO Z CD 4-J z r-H ■H cr CO X G 4-J CD rH •H CO X CO. CO 4-J o c 4-1 CO CO CD X •H 2 o CO p TO CO •H CO G c cr C 4-J G CO-TO G ■r-i G •H 4-J •s cm G o ■r-i o •H o CO c -P G X p CD G CO p cm ■i-i X ■H 2 CO CO >^ CO CD X CO G C 4-1 Q. CO CD 4-J 4-J -P CO z 4-J •H -P 4-J CD z r-H p CO ■r-i CO TO CO G •i-i c -P o G CD cm o G > 2 c G CO P o c o 4-J CO G p CO > P 1 P CO ■H o ■H G 4-1 P CO p cn G CD G TO XJ G G TO c -P P CO CO 4-J cm •r-i CO G z 4-1 z •H CO G cm -p X • cr cm G 4-1 i-H XI G CO c CO r-i E N P c Q. G G G CD i-H ■H CO o C z CO r-H ■H P •H E TO CO Q. p cr G G o •H •H G i—1 ■i-i TZ G G •H G p E o o P ■H CO G 2 N ■H TO P 4-J X CO P o CO c •H O P z c G ■H 4-J o 4-1 4-J G o. CO •H 4-J G G X p P z 4-J G CD P CO E CO G CO- r-H 4-J G G r-i G c p G r-i cr cm X l—i CO Q.-P c CD 4-J ■H >> G ■i-i CO U- c G ■H G P- CO G ■H I-H G CO- 4-J c O < CD 2 G O CO- CO • CO CO X CO ■r-i o P CO C P G c cr G 4-J CO-r-H •H Ui G P G >> CO CO G o o a G ■r-i 4-J G . G G G •H 4-1 •H X C •H ■r-i CO c X G X CO r-H CO X C CO- P ■r-i X G G 4-J 4-J G G G X co c a. c 4-1 G CO G X 1— •H CO CD G P •H p- p E p P a a. G X P r-H r-i •H G X G G CD G 4-J G X rH G G z z > P 2 C cm CO 4-J z P CO G CO. •H Q- CL Q_ P P. i—1 p XI 4-J o P ■r-i E X X O O G •H C z CO z CO XI ■r-i X P o 2 G D. CL CO TO •H > r-i CO a CO TO -P O G ACCESS TO HEALTH CARE AND ITS ATTRIBUTES POTENTIAL ACCESS REALIZED ACCESS Individual Predisposing Age 6 or less Age 65 or over Enabling Financing: Income, Employment, Type and Extent of Insurance Coverage, Out-of-Pocket Expense Organization: Regular source of Care, Type of Care Provider, Particular Provider, Travel Time, Waiting Time Need Perceived Health Disability Days Limits on Activity Recent Medical Emergencies Serious Illness in Family Community Availability Physician/Population Ratio Location Region Rural Residence Central City Residence Objective Use Physician Visits Site of Visit Preventive Activities Dental Visits Hospitalization Difficulty in Receiving Care Use Relative to Need Subjective User Satisfaction with Travel Time Waiting Time Time with Physician Information Received Out-of-Pocket Expense Quality of Overall Visit or Stay 289 III. Potential Access Potential access indicators measure the characteristics of the population under consideration and of the health care delivery system that influence whether care will be received. Andersen has grouped individual determinants into three classes: ---Predisposing; ---Enabling and —Need. (2) Predisposing characteristics are those factors which suggest the likelihood of using services and which exist in the individual before the onset of illness. They include characteristics which cannot be changed such as age, sex, race or ethnicity and those which can such as educational level, knowledge of good health practices and general health care attitudes and beliefs. Enabling factors describe the resources available to a person should he decide to seek care both individually and within the community. Need attempts to measure health status and immediate causes for seeking care. Predisposing Factors Table 1 illustrates information on some predisposing access indicators which are demographic in nature. As the table shows, a larger percentage of the minority population than the White population is young, suggesting a proportionately higher need for preventive services The overall level of child bearing for minority women is higher than for White women. The fertility rate for Black women is 2.3 children compared to 1.7 for Whites. The average number of children in U.S. Hispanic families is 2.3. The American Indian birth rate is almost twice that of all U.S. races. Asian Americans have larger households and average family size than the U.S. population as a whole.(3) Smaller percentages of minority populations are elderly. With the exception of Asian Americans, minority populations achieve lower levels of education than the White population. This has important consequences for their income, ability to pay for care, and insurance status. It can also affect knowledge of services available and how to purchase services. Enabling Factors Table 2 illustrates variations in income and employment among racial and ethnic groups. 290 TABLE 1 All Races* % Age 5 or Less 7.7 POPULATION AND EDUCATION INDICATORS FOR WHITES, BLACKS, AND HISPANICS, 1983 White 7.3 Black 10.3 Hispanic 11.1 % Age 65 or over 11.2 11.9 7.8 4.1 N3 V£3 Years of School Completed (Persons 25 years of age or over 0-8 years 15.1 14.1 23.3 15.0 4 Years College 18.6 or more 19.5 9.5 18.8 SOURCES: U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 142; and P-20, No. 388 and unpublished data. Labor force data are published by U.S. Bureau of Labor Statistics Employment and Earnings, January 1984. U.S. Bureau of the Census, Current Population Reports, Series P-20, No. 368 and Series P-60, No. 144 and unpublished data. TABLE 1A Years of School Completed (Persons 25 years of age ko or over POPULATION AND EDUCATION INDICATORS FOR AMERICAN INDIANS AND ASIAN AMERICANS 1980 All Races American Indians Asian Americans ?o Age 5 or Less 7.2 10.1 8.6 ?o Age 65 or over 11.3 5.2 5.9 0-8 years 18.3 25.0 16.4 4 Years College 16.2 or more 7.7 32.9 SOURCE: U.S. Bureau of the Census, 1980 Census of the Population, Vol. 1, Chapter C (PC80 1-C) TABLE 2 INCOME AND EMPLOYMENT INDICATORS FOR WHITES, BLACKS, AND HISPANICS 1983 All Races White Black Hispanic Median Family Income $23,433 Persons below poverty line 15 $24,603 12 $13,599 35.6 $16,228 29.9 ?o Persons 16 and over below poverty nj line 64.0 64.3 61.5 62.9 Unemployment Rate or more 9.6 8.4 19.5 16.3 SOURCES: U.S. Bureau of the Census, Current Population Reports, Series P-60, No. 142; and P-20, No. 388 and unpublished data. Labor force data are published by U.S. Bureau of Labor Statistics Employment and Earnings, January 1984. U.S. Bureau of the Census, Current Population Reports,,Series P-20, No. 368 and Series P-60, No. 144 and unpublished data. TABLE 2A INCOME AND EMPLOYMENT INDICATORS FOR AMERICAN INDIANS AND ASIAN AMERICANS 1980 All Races American Indian Asian American Median Family Income $19,917 $13,724 $22,713 % Persons below poverty line 12.4 27.5 13.1 % Persons 16 and ^ over below poverty ^ line 62.0 58.6 66.6 Unemployment Rate or more 6.5 13.2 4.7 SOURCE: U.S. Bureau of the Census, 1980 Census of the Population, Vol. 1, Chapter C (PC80 1-C) Black, Hispanic, and American Indian families have substantially lower incomes than the White population. Black family income is only 55 percent of White family income, while Hispanic family income is 66 percent of the majority group. Within the Hispanic population income varies with Puerto Ricans reporting the lowest family income.(4) Median family income for American Indians is also low. Relatively high income levels are found in the Asian American population, many of whom live in households containing several working adults. Minority groups, with the exception of Asian Americans, also experience higher unemployment rates, with the consequence that fewer minorities have access to employment-based third-party health insurance the prevailing national mode for the financing of health care. Lower incomes result in less disposable income available for the direct or indirect purchase of health care. A prerequisite for access to care is a method of paying for services. This usually takes the form of some type of third-party insurance, either public or private. The Robert Wood Johnson survey found that in 1982 approximately 9 percent of the American adult population under 65 reported no health insurance coverage of any kind. This percentage had held relatively constant since 1976.(5) In 1983 the Census Bureau looking at all ages of the American population found 15.2 percent of all persons to be not covered by either private or public health insurance. Fourteen percent of Whites lacked health insurance, 21.8 percent of Blacks and 29.1 percent of Spanish origin persons.(6) Table 3 illustrates insurance coverage by type for White, Black and Hispanic populations under 65 years of age. Minority populations are two to three times as likely to be uninsured as the White population. Inability to pay is the most commonly cited reason for not having health insurance.(7) Minority populations are also more likely to be covered by Medicaid than the White population. The especially high Medicaid coverage rates for Blacks and Puerto Ricans can probably be attributed to two factors. Greater percentages of these families are headed by single women than other population groups, thereby increasing the likelihood of categorical Medicaid coverage through Aid to Families with Dependent Children. Also, it is probable that many of these population groups live in states, such as New York, with relatively generous optional Medicaid coverage.(8) Although Medicaid has provided health insurance coverage to many minority families, not all States offer the full range of services allowed under the program and eligibility requirements vary from State to State. 295 TABLE 3 INSURANCE COVERAGE OF WHITES, BLACKS AND HISPANICS UNDER 65 YEARS OF AGE 1978-1980 White Black Hispanic Mexican Puerto Rican Cuban % Insurance Coverage: Persons Under 65 Private 86.1 61.3 58.6 59.0 47.7 72.2 Medicaid Only 2.7 20.0 13.4 10.7 31.9 5.4 on Not Covered 8.7 17.8 25.7 29.9 19.7 16.6 Other 2.5 1.0 2.3 .4 .7 3.8 (1970-1980) SOURCE: Trevino, F.M. and Moss, A, "Health Insurance Coverage and Physician Visits Among Hispanic and Non-Hispanic People." Mexican Americans are the population subgroup least likely to report health insurance. They are also the group least likely to report unemployment as a reason for non-coverage. This suggests the likelihood that members of this group are often employed in organizations that do not provide health insurance.(9) 1977 data show that employment is no guarantee of health insurance. Across population groups, 22 percent of the working poor lacked any form of health insurance throughout the year.(10) Nearly one fourth of all agricultural workers are uninsured during part of the year, with 16 percent uninsured for the entire year. Insurance coverage also is low among blue collar and service workers.(11) The Robert Wood Johnson Foundation survey found that during 1982 about 21 percent of Americans experienced changes in their insurance coverage. On balance, Americans experienced a net positive change, but the poor, and especially poor minorities were more apt to have their coverage dropped or reduced.(12) The 1977 National Medical Care Expenditure Survey (NMCES) examined continuity of insurance coverage, finding 84 percent of its respondents always having insurance coverage, 7 percent insured part of the time and 9 percent always uninsured. For both Whites and Blacks, continuity of insurance is directly related to income. For Hispanics the relationship between continuity of insurance coverage and income is less clear, but generally tends towards an association between the two factors (Table 4). This same survey found the largest number of ambulatory visits (4.0) among the always insured and the lowest number (2.2) among the always uninsured. Whites have higher numbers of ambulatory visits at nearly every income level.(13) Nearly all the U.S. population over age 65 is covered by Medicare. Because Medicare does not provide full insurance coverage for all health needs, in many cases beneficiaries supplement their coverage with private insurance. Overall 65.2 percent of the over 65 population have both Medicare and private insurance coverage, 20.4 percent have Medicare only and 10.6 percent have both Medicare and Medicaid. The White population supplements Medicare with private insurance more than twice as frequently as Blacks (69 percent versus 31 percent).(14) The 1977 National Medical Care Expenditure Survey collected information on out-of-pocket expenditures for personal health services. Expenses per person for those persons with out-of-pocket expenses were relatively constant across ethnic/racial lines: $196 per capita for Whites, $180 for Blacks and $191 for Hispanics, although the lower incomes for minority groups makes the burden of these expenditures greater. Blacks and Hispanics were considerably less likely than Whites to have high annual out-of-pocket expenses and while 3.3 percent of all White families 297 TABLE 4 MEAN NUMBER OF AMBULATORY VISITS BY HEALTH INSURANCE STATUS, AND ETHNIC/RACIAL BACKGROUND AND INCOME LEVEL (NMCES household data: United States, 1977) CHARACTERISTICS ALWAYS INSURED INSURED SOME OF THE TIME ALWAYS UNINSURED All persons 4.0 3.6 2.2 White NJ 00 Poor Other low income Middle income High income 4.8 4.6 4.1 4.3 4.1 3.9 4.3 3.4 3.4 2.8 2.5 3.0 Black Poor Other low income Middle income High income 3.8 3.1 3.1 3.5 2.4 2.4 2.7 * 1.8 1.3 1.4 Hispanic Poor Other low income Middle income High income 5.3 3.5 3.4 2.9 Adjusted for family size E x: CO co G >s 4-J >s J* G G X X rH O iH G 4-J •H p -P G G av c rH CO CO x: G G P J* CD ^ CD rH j^ 2 P 4-J cr O ■H P vc p G CO CO CO CO CD G G X 4-> •H E i—i CO -H CJ N^ 0 •H TO P- X+J P rH G G >> • p G o p c CO it- cm CO Q. P r-i P G G E G G G o G Z CO G O P X 4-> x: > E 4-J CD X o -P P- CO 4-J G c G ■r-i G CO G p G ro G G X G Z r-i rH P G CO TO cn G c E X -P G CO G G -P P -P G 4-J o 0 u- 2 P TO CO Z CO > cr ■r-i •H 0 0 P. G O r-i z O G O o X 4-J p O O CO. C G co rH g cm P •H 3: P >s 4-J P cr r-H cm Q.4-J o >^4-> X CN TZ CD o ■r-i rH z E CO as Q_ G -H TO -P rH G G •H co co •H H G O c :o G P 4-J 4-J A z z P P G C 4-1 z * P x: CO Ui TO x: D- CO Q_ cm-H CO 0 s-> G O 4-J N G Z C -P O G G P 4-> rH p- r^ p co- rH ■r-i rH CD •H CO. ■r-i G C G 1 i—i co G CD a XI CO 2 P TO E O P G s_^- G P G CO CO c cm O G G C a x: cni— cn o TO c G G CO G rH CO p C -H G o G P P O ■r-i P CO CD G o ■H 4-1 4-J E G CO O 4-J co G -H x: . > CO CO CO P G • G •H rH 4-J G G CO N ■r-i Z G 4_> G cm G TZ -H P P X -H G c o J* C > CO P P- G E CO CD co rH O o CO G G -H p CD O E CO p G ■r-i G ■r-i CO E G G CJ <4- G >s • 4-J CO -p p rH 4-> Z > G p 4-J P X G Z CO CD co CD P TO 0 U- G O G cr -P 4-J P ■r-i r-i CD C G O P P- -r-i G c rH O P- >. P z CO. O Z r- G CO 4-J O rH CO cm a G G G G O CO CD C G G rH > G P TO G G G G CO- o C X CO P co p o -P c P CO CO Ui o P- O z G 4-> -P O ro Z <—i C -r-i p Q. o 4-1 G rH o G c c p O CO G X G z 4-J G G Z 4-J c O G z G z CO. CO- G -H 4-1 TO ■H G G G X P- c g G TO CO ■r-i cm -P P cr iH Z G O CO o CO x: g •H l—i c P co ro ■H CO CO G 4-J P •H CD Q_ z O -P 4-> X p G O-X Ui > rH 4-> x: P G cr G C G G G U- G •H CO Z ZJ G O z .¥ G ■r-i CO- CO O TO ■H X G 0 -h cm G G X o X G ■H x: TO ro c O P 2 CD CO G O 2 p G -P rH 2 cr •H CO. G - E Cn G c p CD •H cm > 1 Q. X O G O r> X o ro -P G J* c CO 14- 4-1 -P cr co co G 4-1 G CO- •H G •H X o i • 2 G cm g CO G G CO G P E CO 4-J 4-J H c c C XJ G rH CO- O Z O X P ZJ rH G G o o ■r-i G X O G O o Ui a E O 14- X P- > G -P a TO lA C CD CO- cm CO 4-> o 2 CO CO CD cm p G o c rH O G CO cr X G c 0 p C CO •H X ■H G TO >^ G cr p •H CL . 4-> G P c c 4-J • cm o 0 ro > G TO TO CO CO CO X 4-J co o z •H CO ro 4-1 G CD P G G G G N r— co G •H O Z G 4-J >s X P 4-J -H G •H X XJ 4-> P- c E c >NrH J G Z P rH cr cr 2 co •H O G l—i G -P rH G O -H G co G ■H C 4-J G G G P rH Q_ C CL E CO. cm TO > P G c 4-J P J* CD O •H G CO X p C co CO G o Z G ■r-i O G G c P P- G CO CD X > XJ G CO O-rH -P X Q. z 4-> P x: o cn a. z o p X 4-1 -P P co a. >N G c x: CO r— c c ■H CD a. TO G G -H cmx ro cm to c G CD XJ G C J* CO G G CO r-i Ui CD G G P > Z G O E G -r-i -P G P r-H CD G G > CD X P 4-J -P -H CO TO g cr x ro u- cm o c •H G -P G -H c ro G 2 ■H c u- G O > c c o o G -H -P G CO X P I— G TO •H G C o G G CO G G CD 4-J G -H •h x: 2 3 -P G >^ CD rH G G 4-> G CO -P E Z •H C X -H O E P CO-CO Q.IA CO U- X o -P o 4-1 G X 4-J tt- o G X 4-1 P- •H lt- I P Z o <4- 4-1 z o X co •H CO 2 G G - G G X G G 4-J •H C X -H S: G C G ro e x: -h 4-> 4-J g cm g c E -H ■H 4-J -P -r-i ro cm 2 c •h cn 4-> C •H -H CD 4-J 2 P O g a. G G >>4J G CO > x: p 4-1 ZJ CO TO ■H 4-J TO i-H C G 0. • X G UJ P CO G G P CO i—i C_) CO G rH -H CO TO G G ■H E TO G P- ZZ O H G CO G G G P G CO -P U Z cr P- -H O E g a G rA P ZJ c o ro g _c 4-J 1—1 CO G Z P G O Z E G TO 4-> C ■H G O. O G 4-1 O G -P G 4-1 >> Z rH cr g •H _¥ E -H -H o (A G P cr o p cr g ro cn o_ c G o -H r-i x u p z 4-> O CO G p o -P 00 C rH G ^ G • P G G P O. CD G NO I--- P- O G CO G P P G P G G G 2 G H CO G r*» ro r^ z on co rH Z c o •H -P CO CO rH G Z rH CL-H O X as G G CO rH rH co co z P- G O Z 4-J P cr -h G G o x: P 4-J G O LA r» g G ~-P G Z E C •H -H -P E r-^ CD G KN > CO c P -H -P X -P C -H •H 2 X -P rH CO G X p o p G XI E Z C G X 4-J G X -P X 4-1 4-J 2 C O G P G cm g p G Z G O X Z -P cr •H C 4-1 -H cr o ^ G G to a c 2: CD X G X X Ui 2 P G -XJ rA E r- g CON E G G G - C CO •H OZ CO X U- • O 4-> g x: CO 4-J a 2 o p G XI E U- G O E 4-J >> C rH G -H G E P CO CD P- Q. G ON C O O O —i -P o ro 4-> c c G O TO 2^ G c G cm p G E G TO c CD W P o 4-> G O TO O 4-J G G G G G CD P Z o X ^ P rH G G CO 4-J •H CO > G -H CO P x: o. x G CD c x: •H TO cr co O G •h x: 4-> CO G D_ G ■-i O o x: •H 2 4-> P G CO TO CO--H o >,X 4-J G •H G P Z o o c x: •H E P O C P- o 4-1 00 G ON ■H rH •H XI 4-1 c G G G G CO cms: C X G CO p x: p- DJU O CL G ON ^ G CN Z . CD o x: 2: 4-> cm g C -H O 4-J CO G X G CO 4-1 CO TO 2 G G P -P CO z G XI CD >s P • CO Q. G -H •h x: TO G G P IZ G XI P E O G E TO •h a co 2: G X •H TO TO G G ZZ -P p c o ro Q. x: g -p p P G G -X X G 4-1 CO O r-i CD G cm^- co o p G 4-> > C O G G G P G G G O- C CO CO P rH G TO c c •H CO x: g -P G rH -P CD -H g x: x: 2 TO G TO cr CO co_ X G G cn ro p z o G c G 2 o c ro p cm o p a. (4- G G P c CO G G XJ ■H TO E G CD m t-i cm g o X P 4-1 CO- ON ON CN TABLE 5 ENABLING FACTORS: ORGANIZATION 1982 ?o With Particular Doctor WHITE 79 BLACK 67 HISPANIC 72 % With Regular Source of Care But No Particular Doctor 11 20 17 § % With No Regular Source of Care 10 14 11 % With Hospital Out-patient Department or ER as Regular Source of Care 20 12 % Waiting Time Over 30 Minutes 15 23 29 SOURCE: Andersen, Giachello and Aday "Updating Access to Health Care Among Hispanics." G P O 4-J G CO >> -P •H C Z E E O C_) G C cr P X > 2 o o 4-> p O ■H 4-> P- r-i G P O G 2 P O P G O >> coGcm-p POQJ4-) G O -p G G C > XX C •H x: o g CO Q. G G H OX G -p ro C G G 2 -H CO G -H G p- x: CO -H -H O X) C. X G -P G p o cm H to x: ro U- O 4-> OH (0 G c c G G C O G G P G G CO c >,-H o •H G G P - Z g----4_> O P c -P rH -h g e x: co c G 4-J O G Q- G G -H •H rH 4-J >n CO -P G O C P G C G E C H H H ffl x: G -H -H G z c ox; o -h c •H 2 C O- XJ JMO (0 XI E O -P G G -H XJ G G -P P E E •H X G p CO G G it- x: cnrH g g Z >,-P C -P G -H QJ >^ rH c x: O -P CO CO Q. P C rH CO co o x: -H X -H ■H O -P P -H G • O. G G r— X -P rH CO -H O C D G -H G Ui X Q_ O G P- 2 CO > -P c •H -H O G TO G c G Z rH G P- -■I-i CO ZJ -H -p o a. -h CO G G -H CO • X G G -P CO CN C G G -P ■H CO G 4-J G C G G G -H G p . G -i-i -H G G c x: G G r-i CO ■H O. X 4-J C CN 4-> x: G ■H P CO r— CO -P CO -P G 4-> W CO G ON G CO J— G G •H P C O cn G G C -H r h cor p >> cr G O. G X C CO C -H G r- X 4-> G x: •H ■H • -H -H -H P - O -H O • c o. G G G 4-> > cn O- G TO G 4-> rH O W C G >, O - CO -H C P -P C >> ^-s O G G -H p. x: XT -H C O-X CD >N G c tor co r>- O -H o 4-1 CL-P CO C X 4-J X G O-CN • a CO P TJ CO-H G -H G rH -H TO G v-x CN (OH C -P >%P- CO > CH<»- • H CO P O G p CO G O -H G P CO. C 00 -H CO E G E X G G >^-H O CD Q. GOP On E G E CO G G QJ to x: X O- G o_ TO -H G i—i O -H XJ O G rH •H P -H O. G G G G W XI P G G G P Q. rH ro > zz o -h P G G E cr p- >^ > ro E a. GOG -P 4-J x co •H G Z •H X P P O E ro p x: p -h G G P- C XJ Q. G G rH G ■H G CL-P O >^ G P G O CO G G ^ PPG CO G G P r-i P G G O 't- X >H C -P P to ar O CO X G CO Z 4-> P • 4-J O -r-i CO CO c CO G J— ^ CD -P XJ E P CD CO cn i ro -h G G E G 00 P C G Z G cr C G G G -h •r-i O P rA u <*- zj >^ CT •H O P -H G P G P 4-> CO • CO- O >> •H ^H CO X O CO -P 4-J G G E CO G 4-J O X r-i E CD G CO CO S G G -P G G >sr^ o g 4-j cr CO 2 Zi rH XJ C 4-J G 03 4-> - tZ G G rH Z CO X O O XJ C CO •H •H Q. G O- G -H C rH G CO E r>- p -h co co rH E O XT G P CO >H CO G E on c cm > >h c •H Q-4-> O (OH ■H CO X o rH O G -H C •I-i XI CD GOO G P- Ui G •H G P ro G CO CO _Q G •* CO O •H 4-J - _c C -P X Q. > > -P r-i rH CO P- Q. G P C 4-J G •H CO P- CO G G •r-i >^ Q. G O O >^ O O O -H rH O P CD CO E P -P C X X G XJ jZ CO O -P >% c > E -H CO >n CO Q. G f_i CO CL'—> r-i G X CD CO CO 4-J -H g - cr CO O *— G Cn G c CO G G -H G p- cm D_ CO "H - O-CN P C XJ CO G C C X G H -H o c g x: r-i *~S O -H X CO CO -P C -H G •H et G 4-J CJNO . 4-J G -P 4-1 r— ■h x cnxj >> Ui z P -H CN O XJ G CO O G g 4-j c c ro x: PC* -} 2 CO O O G C CO • H -H -H rH Q. G -H G O CO E E ~-H X) P P G g x: rH -H XJ 4-J CO OOP O G 4-1 >, P 4-J rH CO G ECG CD -H p x: P CO G P- c ro -P x o -h -h > x: ZJ O P U- r-i CL O -r-i X CO O. E 2 2 ro -P C G CO X G CJ •H p P X G •H CO CO CO 4-> X G > H >> G X G Z -H > •H XJ 4-> •H p G CO P X > CO tt- E X X X G G C X G G -H 4-> g cm G G G O X •H CO G p c x: o oz c A rH G CJ G G C H J* CD CO -P 4-J c -h G —i -r-i CO G C CO CO -H > XI o •H 2 -P G X - ZJ CD Q. H O P XJ G O CO- ■H CD S- X X P- -H G Z O G—I G P X G P G G •H • G G ~ C 4-> G 4-J 4-J G O P -H X >* P j_> p P CO CO G CO ^-n D P- G 4-J O CO O O 4-) O- • CO ■H CT CN X -P CO >%' X -H E JO P O (H ■H G TO H CN G P G XT rA C C CD Q. O G G E G — P CO P Q-CN co zj x X 4-J Q. •H G O Q_ TO • X CD E C G C E -PGP •H rA X CN • J* E CD >* G CO -H P G t*- (4- CO 00 P C G H p G O G -P XJ CO CO o G ON CO H £ z CO G C H CD P P Q. G .r^ r-\ S- 4J >> o rH O rH G Z CO C C >sX tp CO P- -H OP H G CO CO E G X -P C C G O -P co. cm g z P -H -H G ZZ C. 2 -H Z o •n CD x: g GOG X CD X O O 4-J G G -H CJ CO -P -H X -H -H CO c c x: g -h 4-J O- D ■ i-i -H -P •H G G CJ •H CO G G -P > X ■H >^ Q-rH P P > >n >n et r^- p g g G X -P O XJ -P O CO o x: x: cm ON -P G -H P G S CL Z G P- O- P Q- O- c c ih > cm o. z X o. ro Q. CO ■h ro co o X (4- C G co. g it- P- G G -PC X rH G G O CO 4-J cm e p o CD O 4-) ■r-i CO G ^jZ •H G G CO 4-J > G •H 2 c >^ g x: P O P 2 cm G -H G C -H E G -P cm G C 2 G •H -P PGP- CO P < •H G -H > >s XI 4-J C G GOO- X C Q- -p xj g p cm >s e O OI E X G 4-> -H G P D 1 -P Z X) O 4-J G 2 G P i-i E G X CO -P >* G C C C ■H O G CO co p cm-P Q. C CO CO z P c G E -P G G G CO G E P -H O G -P G CO x: G -H X -P P >^ CO G GOG •H G X X 4-> rH G >> G Z XJ C P 4-J X -H X •i-i -P -P XJ P x: x o CO G -H -P P >^-P X - o p ro G G >> •HQ- O -P J* C P r- C0 E O- 4-J G -P P P P XI C o o z O P G CO X CO O O G u- G -H •> OJTJ -P O G 4-1 •H 4-> -H O ^ 2 cm G • G P C a.-P c X O- G G O G P >% p ro co G E CO G G X -P X G 4-> x: p o -P CO G «H P G Q. G O X -P C c G E •H O X C p G Q. C G O G >^ G GO cm c 4-j P U- Q. P {*~ •H NO XI G TO -H -H G CO c O O G Z O X 4-> P rH g x: -P • X G ef P -H G C CO G G CO CO -P 2 CO G G E P TO C 4-J CO CL r-i C pro n c G -H G G C. -r-i O XJ O G 4-J G G -H O X Ui a XJ X O G G X (OHO C C rH rH -H •— E P G E G -H G O -P •H 1— 4-1 -H g g ro co -p CO z -P Z >^-P x: P C CO CO C G Q-4-> -P CO P XJ C C G O -P G G G rH CO E H H H cm e G > ro o_-h ■H Z CL CO O Q- CO. Z o -h •r-i G P X P -P X CO. G G G G CO. P G rH x: zj g o r- co G O -H c c o o o Q. P G 4_> CO P <*- 00 ClZZ O-X •h ro x: x: a o m TABLE 6 Residence COMMUNITY INDICATORS HOUSEHOLDS AND RESIDENCE BY REGION AND TYPE 1982 White Black Spanish-Origin Households 1982 Region o % Northeast % North Centra % South % West 22.1 26.4 31.6 19.9 18.8 20.0 50.6 10.6 18.9 7.5 33.7 39.5 SMSA Central City SMSA Other Non-SMSA Non-Farm Non-SMSA Farm 25.8 40.9 31.3 2.0 57.5 20.6 21.4 .5 50.8 35.0 13.7 .5 SOURCE: Statistical Abstractr of the United States: 1985 Current Population Reports, Series P-60, No. 137, Money Income of Households, Families and Persons in the United States, 1981. (Household information as of March 1982. Adequate payment to providers for care delivered to the medically indigent, that is, the uninsured and underinsured, has become a critical question of national health policy. From the provider perspective, uncompensated care represents the debt incurred after services are delivered but no or inadequate payment is made. It is the total costs of care delivered to: • Medically indigent for which there is no public reimbursement. • Medically indigent for which there is inadequate public reimbursement. • Bad debt incurred when patients do not, but presumably can, pay their bills. Many health care facilities, through receipt of Federal grants, loans or loan guarantees, have accepted an obligation to provide a level of uncompensated care to qualified persons in need of care. These obligations were incurred under Title VI of the Public Health Service Act, also known as the Hospital Survey and Construction Act of 1946 (popularly known as the Hill Burton Program) and Title XVI of the Public Health Service Act, also known as the National Health Planning and Resources Development Act of 1974. The total number of facilities with uncompensated care obligations as of January 1, 1985 was 4,653 of which 2,801 were hospitals. But the volume of care provided through this program is not sufficient to provide adequate health care for the growing number of medically indigent.(24) The problem of uncompensated care affects minority populations disproportionately because minority members are more likely to lack health insurance coverage and thus are more likely to be unable to pay for care. Moreover, urban hospitals and public hospitals bear a disproportionate share of the uncompensated care burden. Community hospitals provided $7.5 billion in uncompensated care in 1982. Of the $7.5 billion, $2.3 billion, 3158 was reported as charity care; the balance, $5.2 billion, resulted from bad debt. These figures from the 1982 AHA Annual Summary are based on hospital charges and not the actual costs of uncompensated care.(25) Analysis of 1981 Hospital Discharge Survey data from the National Center for Health Statistics revealed important information about the kinds of care received by uncompensated care patients: • The cases were most likely to be either maternity or accident-related; • The most prevalent surgical category for this group was obstetrics; • Problems relating to premature birth were common for newborns; and • The patients were likely to have had surgery on the day of admission.(26) 303 Medicare has adopted a prospectively based inpatient hospital payment methodology that is intended to create incentives for hospitals to operate in a more efficient manner. As of December 1984, there are six Medicaid programs that also have similar DRG systems for inpatient hospital reimbursement.(27) Government is not the only cost conscious purchaser of medical care. Health insurance carriers are adopting alternative payment practices and are offering more price-competitive policies in response to purchaser demands. One unfortunate result from this price squeeze has been the increase in the number of economic transfers of patients from private hospitals to public hospitals. Economic transfers - also known as patient dumping - have increased in such areas as Boston, Chicago, Washington, D.C. and San Francisco. A study of economic transfers in the San Francisco area found 63 percent had no medical insurance at the time of transfer, 34 percent had Medicare or Medicaid coverage and 3 percent had private insurance coverage.(28) Exacerbating the problem of economic transfers is the fact that the hospitals receiving the transfers are often in worse financial shape than the originating hospital. These hospitals usually have higher percentages of uncompensated care, higher levels of Medicaid and Medicare patients, and lower levels of privately-insured patients that can be charged higher rates. Uncompensated care charges represent only approximately 5 percent of all community hospital charges. The burden of uncompensated care, however, is not evenly distributed and falls disproportionately on certain types of institutions. Data for the year 1982 show that public hospitals, teaching hospitals, urban hospitals and hospitals located in the South provide disproportionately high levels of uncompensated care. • Public hospitals account for 20 percent of aggregate total charges, and 40 percent of total uncompensated care costs. • Major teaching hospitals (public and private) accounted for 24 percent of aggregate total charges and 36 percent of total uncompensated care costs. • Urban hospitals (public and private) account for 39 percent of aggregate total charges and 49 percent of total uncompensated care costs. • Southern hospitals represent 31 percent of the Nation's aggregate total charges but 48 percent of its uncompensated costs. • Investor owned hospitals account for 8 percent of charges but only 5 percent of total uncompensated care costs.(29) Hospitals that combine these characteristics have especially high levels of uncompensated care. For example: 304 • Major public teaching hospitals account for 5 percent of aggregate total charges but 21 percent of uncompensated care. • Southern large city public hospitals account for 2 percent of aggregate total charges and 12 percent of uncompensated care.(30) In short, public hospitals are bearing a disproportionate share of the uncompensated care burden. The foregoing discussion has focused on uncompensated care provided in hospitals. It is this aspect of uncompensated care for which the most extensive data are available. Less is known about the volume and costs of uncompensated ambulatory care provided in settings other than hospitals. Also, little is known about the extent to which a person's inability to pay causes the deferral of care, and the cost and health status consequences of such deferred care. Several Federal programs, however, address the need to encourage ambulatory care services to the medically underserved. The Indian Health Service is discussed later in this paper. Other programs include: Community Health Centers The Community Health Center program supports about 600 grants to community health centers in medically underserved areas, about 390 in rural areas and 210 in urban areas. While there are approximately twice as many rural grantees as urban grantees, funds are split almost evenly between rural and urban centers. Community health centers provided primary health care services to approximately 5.1 million persons during 1984. Community health centers use a sliding scale fee structure and no one is denied care because of an inability to pay. Fifty-eight percent of all center users had income levels at or below the poverty level and 84 percent were at or below 200 percent of the poverty level.(31) Migrant Health Centers The Migrant Health Program provides comprehensive primary health care services to an important subset of the underserved, migrant and seasonal farm workers and their families. Access to health care is difficult for this group because of its mobility, language and cultural differences and low income. Most states consider migrants temporary residents, rendering them ineligible for Medicaid. Over 120 grantees are funded, covering over 300 rural areas in 35 states and Puerto Rico. They serve over 450,000 migrant and seasonal farm workers, representing 16 percent of the total estimated migrant and seasonal farm worker population in the country.(32) 305 x: CZ P -H >> XJ P o G -P 4-J G X <4- P G -P CO G X O H G X X ON c X O p O x: co cr x 4-J X G CO CO c G C c o CO G CD P- 4-1 cmx ro c • H -P G P G ■r-i ■r-i CO G Z CN -P G G G c x: co G G C -P X > -P P >s CD P >> c G G p ro -P G H G 4-> Ui P G C G 00 X G z O G ro ■H ro cm -4-J > > 4-> CO rH r-i G C G 2 •H Z G G CD X P G 4-> P CO -P CO S. CO D_4-> -P O o X CZ O G H - G G 4-J G X rH XJ r-\ Q.4-1 G CO Z ih ro g U- •H E CO -H ■r-i 4-J O -H CO CO G C G X c G G CO rH G G (0 J)£ > o x: C G ■r-i CD G -P Ui rH -H U) i-i CZ g ro O- c z G P C X G U- XJ G >s-P XJ 4-J CO H P rH CO Ui CD H >s. O. O G -H X co o CO Q. •H c >sx: O ro >> G X G H x: g o cnxi 2 G G >H X -H >n G G G ro G 4-1 p P P PCX G 2 4-J >s 4-J CO P. X -H Q_ G G O -H c G as -P X > X ■H O O rH G -H G G x: G C G o E -P G rH G G o CO G G G p p c G G -P -P CO Q. C r-i X c o. g x: z x p rH X CO G a_ >^ Ct- •H P G G CO z •H CO G CO CO C G-H G G •H -P -P o. CO 00 G > P- E X •H -P X CO XI C G G G X l—i G O G G E C G -P n-H O G Q. rH CD G G >> J^ z •H •H G -H CO CO CO P • g x: G G p O • O P P rH ■H G X G G P G X) X 4-J E C X •H G O A CL-P G -H G •H CN rH Q- Q-rH 4-> c z X P X 4-J P - C O G E c CO -H CO G G P P -H XJ X X -P Q. co c •H o_ 2 O 4-1 rH G C G L_ G co •H Z 4-J Q. •H G G X O 2 -P 4-J ro ^ e x x P G o P CD a o cm rH G G G C P G C_> G o 2 r-i G C 4-J G E G O-x: P G •H p p >> CD r-> -H P >sH o cm G > X G CO rH X G G 4-J X O X 5^4-1 G • G O X) X • X CO x: E CO G CO CO G X CD -l_> >, CO G G •H G CO G 4-J CO E P CO -P c XI o G C G C 4-J CO D c > > CO X CL •h ro co Ui CL P X o •H z iH E G ■H P •h cm CO CD o co co P G E G - X P -H TO Q. CD s P G G -P o ro G G -P Z ro x X P •H G G -H O c •H O -P 2 G ■r-i ON Q. G G CO CJ CO •H X G Z c X rH ■i-i ■r-i G >, 4-J > G G •H r-i G O x: co G rH Q-U_ >% cm > X -P G •s G X o G i—i P i—i E CO p ro co x; G H >N - -POX ■H CO X -H G -H G • X CO G G E •H G •H 4-J O rH 2 ■H CO X o rH -P C O G >- >s. Q-iH G P G X Z G G G 4-J c Q. -P G G CL C CO > O-H C0 CO r— >sLi_ O XJ XI •H CL X c CD G > E G X CD >> X Z ■H G CO P P ■fcr^ G r-H x: 4-> CO > -P ro G G ■i-i ■r-i o rH G P P .v CO c o X X G G Q. C X O P G C G C P ro X G P G O G G G CD P- G P cr g • r-H G G -H - G CO X G cm G O o -P CD Z G o G G C G CO- P 4-1 G rH C G G G P > -P 2 CO G G c P P E G X X o •H p Z X CO CO 4-J Q.X G ■H > CD O -H G G ro p o_ g A -P •H •H G G E G o •H cr 4-1 •H -H CO H 4-1 G O C G > G G • G G O C X ON C 4-1 X P CL O P o X P ro G G cm cox: •H H •i-i O P P -H CO G G 4-J P- •H G G > i-H co r g h rH IA E p G •H P P X G G Z CO G •H G C G H O 4-J O. G P CO. G -P G G ON CO ». IA Z O rH X rH O -P ■H G G rH -P >s >s O > A CD G P rH P -H CN G C V^' G c G a p c CO X CD ro x: xj h G -a- g 2 X G G X CD CO G 4-J - X O . O •H G G G c G G o G rH G HO- 4-J •H C E -P P c G 2 G rH 'i-i G X E X G O >^X) G X X E co en Z G CO C CO P O G rH CO G ro G X CO H -P G -P G P •H CO O P -P c G 4-J CO Q.X -P H CO H G G X CO G P cm g p p- CO Q_ •H G c X O 4-J X cc G ■r-i ro ,-~s G <4- ■r-i p O G H P Q. G > C 4-J QJ CD G ro >mh o. o c o > CD G X 4-1 CO CO- 2 o- rz O O- G Q.4-J G CD G H O H X X G -P -P P o. o P X G G rH N as G j* rA G > •H H > •H G P. 4-1 ■i-i P ro O G rH ■H CD G 2 ■H P CO ■r-i X X CO >s G s^> Ui G 4-J O CO • G > X >^ G >> O G r-i JZ Z o e g 4-j ro 4-> G CO G 4~> X O c i—i 4-1 1— rH CO . Ui G •H X CO H X G rH X CO 4-J P x -h a co p CD -H G G z p o ■r-i iH P •H rH ✓—V G G C p • > H • X 0_ •H H ID H UJ H G G P > O Z 4-J X P X 4-J •H 4-J CO G E CD r-^ a •H z o ro ro p c CZ E g > x: ro G O-O -r-i -P G 4-J O G 4-1 G z G a X CO G > a >s ro ro CO co ro o O G P P OP CO CO i—i u. G P G X G 4-J p G G < P rH 4-J G H G C H Q_<4_ Q. P P X G X O Q-< CO P CO G X CO G G >> G X i—i G CO ■i-i G CO > CZ -rH J* c G P O- o G 2 G - CO- G -* -P c Q_H- G X CO X ro 4-J X X CO Z Ui -P -P P s ON X G CO CD O •H G X 2 CO G G ■H CO O Z H CO P X Q_ O O P CL -H iH XJ • rA G CO P rH P > -P p •■. r— N G Z > p -P CO G H X x: -P X O C > 00 G rH 4-1 c h- X CD a P P >,x o G ■r-i P • CO H X -P G C G C_) ro o i o c CO CO >> G ■H G o -P G P l—i ro G c c G 2 ^ >^ •H cr h X >> CO P E p i o ro > •H G 4-1 •H 4-1 X G Q. ■H P 4_) o ro CJ x: ro o x: g oo p x: CO 2 >— -ui E G G Q. CO O -H .. -H X X •H P c -P G rH G P 4-1 a G r— •H •H 1— H G • G •h G X G X CZ - G X P c I— rH •H co i—i X P ■r-i X CD CD G cr X G G > G > G G G G X -H •H 4-1 G OO-H G Q. r-i CO 4-1 4-1 CD 4-J 1 XJ ■i-i Q. G G -P ■H G P H E G J* >H CO > > rH O- G r^ G G XI 4-J c G G i—i <4- CO Ui ■r-i 4-1 rH G •H G XJ G O 4-1 G Q. O.X o P ro p p > CO G G G j P G X) X ro rH G c G X •H a ro >^ E 4-J CO P G H ro o ro c p G gog -x: rH X X g! o X z G G •H o CD c X > G X X C 1— G c x: i—i G o ro CL CO X O Q-rH CL a.-P 4-J Z| E •H G CD X G X G 2 ■I-i S. I-H X Q. co Mv^ ? •H S CO Q_ TABLE 7 NEED ACCESS FACTORS 1978-1980, 1982 White Black Hispanic % Self-Report Fair or Poor Health* 13.2 23.8 18.7 % Persons with Medical Emergency in Past Year* 14.5 15.8 17.6 o % Families with Member with Serious Illness* 8.8 11.0 11.3 Days of Restricted Activity Per Year (1978-1980)** 18.7 22.3 18.3 Bed Disability Days Per Year (1978-1980)** 6.6 9.4 7.8 SOURCES: *Robert Wood Johnson Foundation Special Report: Updated Report on Access to Health Care for the American People **Trevino, F.M., "Health Indicators for Hispanic, Black and White Americans" per year, while those with family incomes of less than $10,000 averaged 4.6 visits. This difference of approximately one visit between income groups continued across racial groups, with Whites averaging more visits at both income levels.(35) The masking effect of aggregate statistics is evident in the physician visit indicator for Hispanics. While Hispanics averaged 4.4 physician visits per year, Cubans averaged 6.2 visits per year and Puerto Ricans 6.0 visits, well in excess of the White average of 4.8 visits, while Mexicans averaged 3.7 visits per year.(36) The site of physician visit varies between Whites and Blacks. Whites were more likely to see a doctor in his office (69.2 percent) than Blacks (58 percent). Over one-fourth of Black physician visits occurred in hospital clinics or emergency rooms, but only 11.2 percent of White physician visits took place there. Whites are more than twice as likely to consult their doctors by telephone than Blacks. The higher one's income, the more likely one is to have an office visit; the lower one's income the greater the likelihood that a physician visit will take place in a hospital clinic or emergency room. (Table 9) (37) There is a pattern of lower use of preventive health services among minorities. Prenatal care provides an opportunity to identify and treat medical problems and to educate patients about the effects of diet, smoking, alcohol and drug use on the fetus. Close to 80 percent of White women receive prenatal care during the first trimester of pregnancy, while only 62 percent of Black women receive early prenatal care. In the 22 states which report Hispanic birth information, only 60 percent of Hispanic mothers receive first trimester prenatal care.(38) The lower number of minority physician visits for children is reflected in lower vaccination rates. Sixty-seven percent of White children under four had been vaccinated against measles in 1983 versus 57 percent for minority children. Even larger differences in vaccination status were observed for rubella, DPT, polio and mumps.(39) Dental services are used less frequently by minority populations than by Whites. Approximately 56 percent of all Whites four years of age and over visited a dentist in the previous year during the time period 1978-1980 compared to 37 percent for Blacks and 40 percent for Hispanics.(40) During this same period, hospitalization rates for these population groups did not vary substantially among groups. When adjustment is made for age, 10.3 percent of Whites, 11.1 percent of Blacks and 10.2 percent of Hispanics were hospitalized at least once in the previous year. Within the Hispanic population Mexican Americans were hospitalized least (9.6 percent) 308 TABLE 8 NUMBER OF PHYSICIAN VISITS PER YEAR BY ETHNIC/RACIAL BACKGROUND, AND AGE, SEX, AND FAMILY INCOME (NHIS household data: United States, 1978, 1979, 1980) ♦Includes unknown family income. NON-HISPANIC CHARACTERISTIC TOTAL* TOTAL POPULATION 4.7 All Races 4.8 White 4.8 Black 4.6 ALL HISPANICS 4.4 Age in Years Under 17 17-44 45-64 65+ 4.2 4.5 5.2 6.4 4.3 4.5 5.2 6.3 4.5 4.5 5.1 6.3 3.2 4.8 5.9 6.7 3.5 4.2 5.8 8.2 Sex Male Female 4.0 5.4 4.1 5.4 4.1 5.5 3.8 5.2 3.6 5.3 Family Income Under $10,000 $10,000+ 5.4 4.6 5.5 4.6 5.6 4.6 5.1 4.1 5.0 4.8 SOURCE: Health indicators for Hispanic, Black & White Americans Data for the National Health Surve Series 10, No. 148, DHH Publication No. (PHS) 84-1576). TABLE 9 PERCENT DISTRIBUTION OF PHYSICIAN VISITS BY PLACE OF VISIT, AND AGE, SEX, RACE AND FAMILY INCOME (NHIS household data: United States, 1980) CHARACTERISTIC PLACE OF VISIT Physician's a Office Hospital Clinic or Emergency Room Telephone b Consultation Company or Industry Health Unit Home Other & Unknown TOTAL c 67.9 12.9 12.2 0.7 0.6 5.7 Age in Years Under 15 w 15-44 o 45-64 65+ 62.7 66.4 70.7 75.7 12.8 14.3 12.3 10.2 18.1 11.2 10.2 8.9 0.1 1.1 0.9 0.1 0.2 0.5 0.6 1.8 6.1 6.5 5.4 3.3 Sex Male Female 66.2 69.0 15.2 11.4 11.0 13.0 0.6 0.7 0.7 0.6 5.8 5.7 Race White Black 69.2 58.1 11.2 25.6 13.1 5.2 0.6 0.7 0.7 0.2 5.1 10.2 Family Income Under $10,000 $10,000 - $24,999 $25,000 or more 63.3 69.0 70.5 17.4 12.1 9.3 9.9 12.8 14.2 0.4 0.8 0.8 0.7 0.3 0.1 8.3 5.0 4.2 aIncluding Prepaid Group. DDoes not include calls for appointments and other nonmedical purposes, cIncludes unknown family income and races other than white or black. SOURCE: Physician Visit, Volume & Interval Since Last Visit, United States, 1980. Data from the National Health Survey Series 10, No. 144. TABLE 10 AGE-ADJUSTED1 PERCENT WITH 1 OR MORE HOSPITAL EPISODES (NHIS household data: United States, 1978, 1979, 1980) ______Non-Hispanic_____ __________Specified Hispanic___________ Total All All3 Mexican Puerto Cuban CHARACTERISTIC__________Population Races2 White Black Hispanic American Rican______American All persons 10.3 10.4 10.3 11.1 10.2 9.6 11.5 12.2 Sex Male 8.7 8.7 8.8 8.9 7.8 6.9 8.9 11.3 Female 11.9 11.9 11.8 12.8 12.4 12.5 13.9 12.5 Family Income Under $10,000 12.4 12.4 12.4 12.9 12.5 11.3 15.2 13.5 $10,000 or more 9.8 9.8 9.8 9.6 9.2 8.7 8.8 12.6 Education of Family Head Under 9 years 11.2 11.5 11.6 11.3 10.0 9.1 13.4 12.4 9-11 years 11.6 11.7 11.7 11.3 11.6 11.0 13.7 12.9 12 years or more 9.8 9.8 9.8 10.3 9.9 9.9 8.7 11.5 Perceived Health Status Excellent or good 8.5 8.6 8.6 8.5 8.3 7.3 9.0 11.1 Fair or poor 22.3 22.5 23.4 20.0 19.5 19.3 19.9 22.5 ^Age adjusted by the direct method to the age distribution of the total civilian noninstitutionalized population of the United States as of July 1, 1979. ^Includes other races and unknown if Hispanic origin ^Includes unknown specified Hispanic origin. TABLE 11 AGE-ADJUSTED1 AVERAGE NUMBER OF DAYS IN HOSPITAL PER YEAR FOR PERSONS WITH 1 OR MORE HOSPITAL EPISODES (NHIS household data: United States, 1978, 1979, 1980) CHARACTERISTIC Total Population Non-Hispanic All Races2 White Black All3 Hispanic Specified Hispanic Mexican American Puerto Rican Cuban American All persons 8.7 8.7 8.4 11.0 8.8 8.3 10.9 9.0 Sex Male Female 9.7 8.1 9.7 8.1 9.3 7.9 13.3 9.9 10.4 8.3 9.6 7.7 10.1 11.7 13.1 6.6 Family Income Under $10,000 $10,000 or more 10.3 10.3 9.9 11.8 10.1 9.5 11.5 14.1 7.5 8.0 7.8 9.6 7.6 6.9 9.1 6.7 Education of Family Head Under 9 years 9-11 years 12 years or more 10.1 10.0 9.4 11.7 10.5 9.1 13.6 14.2 9.3 9.3 8.8 11.3 9.8 9.7 9.6 *9.1 8.2 8.2 8.1 10.3 7.2 6.7 6.5 6.7 Perceived Health Status Excellent or good Fair or poor 7.0 7.0 6.8 9.0 6.9 6.0 8.5 7.3 13.7 13.7 13.3 14.8 13.3 12.8 14.5 14.0 1Age adjusted by the direct method to the age distribution of the total civilian noninstitutionalized population of the United States as of July 1, 1979. 2Includes other races and unknown if Hispanic origin 3Includes unknown specified Hispanic origin. and Cuban Americans the most (12.2 percent). The age-adjusted average number of days in the hospital for those with one or more hospital stays was lowest for Whites (8.4 days), followed by Hispanics (8.8 days) and Blacks (11.0 days). The number of hospital days experienced is inversely proportional to family income, family education and perceived health status (Table 10, Table 11). (41) If the health status of minority groups were similar and access to health services were equal, one would expect utilization of different types of health care services to also be similar. The Task Force, however, has demonstrated that there are significant differences in life expectancy, morbidity and mortality among major population groups. One would expect, based on health status alone, higher utilization rates for minority population groups than for Whites. Thus, the differences in the use of health services presented in this section are magnified in importance. Utilization measures indicate whether people receive services but do not relate how satisfied patients are with their care. The 1982 Robert Wood Johnson survey, summarized in Table 12, finds Black and Hispanics to be less satisfied with their medical care than Whites for most aspects of recent medical visits. They reported lower levels of satisfaction for travel time, waiting time in the office, time with the doctor, the information they received, the quality of care received and the overall quality of the visit. V. Access to Health Care: American Indians and Alaskan Natives Definitional questions and data availability complicate consideration of access-related indicators for the American Indian-Alaskan Native population. The 1980 Census identifies 1,534,000 Americans as American Indians and Alaskan natives. A culturally and historically diverse group, the Native American population shares some of the demographic characteristics of other minority groups. Census data show it to be a young population, with 10.1 percent of the population less than five years old while only 5.2 percent, less than half the national rate, is aged 65 or more.(42) For those in the civilian labor force, unemployment rates are high, 13.2 percent in 1980, while median family income is low, 84 percent of the national 1979 average. Twenty-seven and one half percent of all Native Americans are below the poverty line. (Tables IA and 2A). American Indian population is concentrated in the 32 Reservation States. A State is considered a Reservation State if the Indian Health Service (IHS) has responsibilities within the State. There are currently 32 Reservation States. Maine, Pennsylvania and New York were added as Reservation States in 1979; Connecticut, Rhode Island and Texas in 1983; and Alabama in 1984. The 32 Reservation States are: 313 TABLE 12 Percent Not Completely Satisfied with Aspects of Recent Medical Visit WHITE, BLACK AND HISPANIC SATISFACTION INDICATORS 1982 White Black Hispanic Travel Time 19 22 29 uj •—• j> Office Waiting Time Time With Doctor 31 20 34 23 37 29 Information Received 17 24 25 Out-of-Pocket Costs 40 43 38 Quality Overall Visit 17 21 24 28 24 28 SOURCE: Andersen, R. M., Giachello, A.L. and Aday, L.A., "Updating Access to Health Care Among Hispanics." Alabama Connecticut Louisiana Alaska Florida Maine Arizona Idaho Michigan California Iowa Minnesota Colorado Kansas Mississippi Montana North Dakota Texas Nebraska Oklahoma Utah Nevada Oregon Washington New Mexico Pennsylvania Wisconsin New York Rhode Island Wyoming North Carolina South Dakota Of the total American Indian population, 1,295,000 reside in these states.(43) Unlike other minority populations, the Native American population is more rural than the general population. Its physical isolation contributes to the types of health care problems which characterize it. The Federal responsibility for providing health services to Indians makes this population unique among minorities and raises unique questions when one considers its access to health care. This Federal role in health care delivery arises from the historical treaty relationships between Congress and individual tribes. The Snyder Act of 1921 forms the legislative basis for IHS services. In 1955 Indian Health programs were transferred from the Department of the Interior to the Public Health Service. The mandate of the Indian Health Service is to provide comprehensive health services to American Indian people and Alaskan Natives living on or near Federal Indian reservations or in traditional Indian country such as Oklahoma and Alaska. The IHS estimated its 1985 service population as 962,000. Thus, not all Indians in Reservation States fall within the IHS service population. Approximately 90 percent of the IHS service population lives in 11 states: Alaska, Arizona, Minnesota, Montana, New Mexico, North Dakota, Oklahoma, South Dakota, Utah, Washington and Wisconsin. In these states, 92 percent of Indians are served by the IHS.(44) The substantial number of self-identified Native Americans not included in this total either live outside the geographic range of its service units or are not members of Federally recognized tribes entitled to its services. There are two principal sources of data on health status and utilization measures for Native Americans. The Health Interview Survey (HIS), the source of already cited information for Black and Hispanic need and utilization indicators, provides information on need-related access 315 indicators and utilization measures. HIS data are drawn from household interviews in about 42,000 households representative of the civilian non-institutionalized population. The small percentage of Native Americans in the general population results in a small sample in the HIS, a sample which despite the limitations inherent to its size represents the Native American population as a whole and not just that portion served by the IHS. Data from the 1978 Health Interview Survey show Native Americans reporting levels of health status and utilization which diverge from the majority patterns.(Table 13) (45) The only source for data on Native American coverage by Medicaid and Medicare is the HIS. In 1978, the HIS reported that 8.3?o of all Native Americans were covered by Medicare and 17.4?o were "probable Medicaid" program participants. The extent to which private insurance, Medicaid and Medicare pay for Native American health care is not fully known. Similarly, the degree to which Native Americans living within IHS service areas seek and receive care from sources other than the IHS and how such care is financed is not now known.(46) Within its service areas the IHS directly operates 47 hospitals, 80 health centers and more than 500 health stations and satellite clinics. Four hospitals and 292 health clinics are operated under the tribal health delivery system which is administered by tribes and tribal groups through contracts with the IHS. Supplemental services not available through IHS' direct or tribal facilities are purchased from appropriate providers under contract.(47) During FY 1984, the IHS directly, through the tribal health program and through contracted purchase of services, was responsible for 4,232,000 outpatient visits and 103,000 hospital admissions.(48) The number of outpatient visits has remained relatively steady since 1981.(49) The number of hospital admissions peaked in 1978 and has varied within a narrow range since then. The average daily hospital patient load in IHS, tribal and contract hospitals continues to decrease in large part because of a drop in the average length of stay which was 9.3 days in 1970 but stood at 4.9 days in 1984. (50) Over the history of the IHS there have been significant improvements in the health status of the American Indian people. The infant mortality rate has dropped to 13.8 per thousand live births compared to the U.S. overall rate of 12.8. Maternal mortality declined to a level at or below that of the U.S. as a whole. Infectious diseases such as tuberculosis have been brought under control. As a result of addressing the acute problems which faced this population, the health problems of the American Indians have changed over the past 30 years. Chronic diseases associated with aging have become an increasingly common cause of death. Accidents and injuries is this population's second leading cause of death. Alcoholism is 316 TABLE 13 NATIVE AMERICAN HEALTH AND UTILIZATION FACTORS INDICATOR White Native American Non-Hispanic Restricted Activity Days Per Year 18.5 24.6 % Persons with Chronic Activity 14.4 18.2 Limitation % Persons Perceived to Be in Fair or 11.3 18.1 Poor Health % Persons with a Physician Visit in 75.9 70.3 ^ the Past Year r—< ?o Persons with 5 or More Physician 21.2 24.0 Visits in Past Year % Persons with one or More Dental Visits 53.4 39.1 in Past Year Number of Dental Visits per Person 1.7 1.2 per Year % Persons with One or More Short-Stay 10.5 12.7 Hospital Episodes SOURCE: National Center for Health Statistics, Health Interview Survey, unpublished data cited in Gerzowski and Adler, "Health Status of Native Americans" an underlying cause for many accidents and this disease, by itself, is a major contributor to death among American Indians. Many of the deaths associated with these causes are preventable. A central focus of the IHS has always been health promotion and disease prevention. Efforts targeted at reducing injuries and health risk factors are underway throughout the program. (51) VI. Asian-American Access to Health Care The 1980 Census identified 3.7 million Asian-American/Pacific Islanders in the American population. Like other minority groups, this population has a young age-structure, with a greater percentage of the population under the age of 5 and a smaller percentage over the age of 65 than the general population. In the aggregate, it differs from other groups in several factors relating to potential access to health care. Of its members aged 25 years or more, 32.9 percent have completed four or more years of college, compared to a national average of 16.2 percent. There is less unemployment among this group. In 1980, of persons in the civilian labor force, only 4.7 percent of asians were unemployed. The 1979 median family income of $26, 456 was 33 percent above the national average.(52) Because Asians make up less than two percent of the nation's population, recent systematic information on their insurance status and use of health services is not readily available. The term Asian American comprises a number of diverse groups, the largest of which include: Japanese, Chinese and Filipinos. Although the Asian population has become more geographically dispersed, it remains highly concentrated in the West, with 56 percent of the total. Asians as a whole lead an urban existence. Ninety-seven percent of the Chinese live in urban areas, followed by Filipinos and Japanese, both 92 percent compared to 71 percent for the White majority.(53) As Table 14 illustrates, there is considerable variation among Asian American subgroups in terms of percentage native born, income and percentage of families below the poverty level.(54) Within the Asian American population is a subgroup which has been recognized as having special problems in seeking and gaining access to health care. Between 1975 and 1982 roughly 1.4 million Indochinese refugees migrated from Cambodia, Laos and Vietnam, with 580,000 ultimately settling in the United States.(55) Since then, the annual number of new Southeast Asian immigrants has dropped and stabilized: in Fiscal Year 1983, there were 61,000 Southeast Asian immigrants, which represented 65 percent of total refugee immigration.(56) These immigrants for the most part came to the United States with little preparation, few resources and no realistic hope for a return to their former homes. The stress of their recent lives is compounded by adjustment to new lives in very different environments. Language, 318 TABLE 14 1980 DATA ON SELECTED DEMOGRAPHIC AND SOCIOECONOMIC CHARACTERISTICS OF ASIAN/PACIFIC ISLANDERS Variable Asian Viet- Hawai- Guama- 1979 Japanese Chinese Fillipi no Korean Indian namese ian nian Samoan Median age, yr, i 33.5 29.6 28.6 26.0 30.1 21.5 24.2 22.6 19.2 % native born 72?o 37% 35% 18% 30% 10% 98% 90% 64% # of children/' 1000 106 82 278 229 236 305 431 408 453 15-24 years old Median year school completed 12.9 13.4 14.1 13.0 16.1 12.4 12.4 12.4 12.3 r—' Mean family size 3.26 3.68 4.15 Mean 1979 family 30,527 26,600 27,194 income $ Median 1979 27,354 22,559 23,687 family income $ 3.86 3.63 4.85 3.96 4.16 5.17 24,670 29,591 15,271 21,495 20,959 16,968 20,459 24,993 12,840 19,196 18,218 14,242 % Families under poverty level 4.2 10.2 6.2 13.1 7.4 35.1 14.1 11.6 27.5 SOURCE: Subcommittee on Cancer in Minorities Report, Volume III, Draft, 1985. conflicting values and a lack of familiarity with Western practices complicate their ease of entry into the health care delivery system. The full extent of our knowledge of the medical and emotional needs of these new immigrants is partially obscured by barriers of language and cultural distance; because they are a new population, our knowledge about their individual characteristics, patterns of utilization and variations among subgroups is fragmented. The refugee population is not evenly distributed across the country. Two-thirds of the population live in 40 American counties and secondary migration to areas of high Southeast Asian concentration after initial resettlement elsewhere is frequent. The population is young with 90 percent of the population under 45 and 80 percent under 35.(57) Because of its age structure, contact with the medical care system often occurs through the need for obstetrical or pediatric care. Funding of medical services for refugees, particularly new arrivals, is through Title XIX State Medicaid programs, for refugees who meet State eligibility requirements, or through Refugee Medical Assistance, for refugees who meet State income requirements. The family composition requirement is waived for the first 18 months a refugee resides in the country, after which time he must qualify for assistance on the same basis as other indigent persons in the State.(58) Reliance on Medicaid for financing of medical services has been criticized because of its link to cash assistance. Fear of losing Medicaid benefits has been seen as an obstacle to employment, where medical insurance may not be available, and to self-sufficiency.(59) The limitation on benefits of State programs has led to difficulties for some immigrants, as in States where prescription drugs are not covered under Medicaid. A Rhode Island multi-year study has found that after initial resettlement, where refugee health screening is required, utilization of health services begins to lessen and decreases steadily over time. Most care sought is ambulatory with community health centers providing 72 percent of outpatient care in Rhode Island.(60) Community health centers in general dominate in the delivery of health services to this population and thus provide the potential both for development of a data base on this population and a network for health education.(61) Lack of familiarity with Western medicine and language difficulties lead to problems in health seeking behaviors and compliance with physician instructions. Many are unfamiliar with the germ theory of disease and lack a surgical tradition. While such persons will seek a physician when they feel ill, they are not familiar with the idea of asymptomatic disease. Compared to other urban poor, there appears to be a delay in seeking care. 320 Language difficulties inhibit care. For example, difficulty in understanding a course of treatment, may cause treatment to stop when a symptom disappears. For those unfamiliar with Western life, accidents are frequent. (62) There is a strong tradition of self-treatment among Southeast Asians. The degree to which such practices as dermabrasive techniques and herbal medicines supplant Western medical care is not well known.(63) Utilization studies in different parts of the country suggest that the Hmong are low utilizers of health services while Vietnamese among Southeast Asian subgroups are high users of services. The latter's prearrival experience with formal health services is cited as an explanation.(64) 321 G . P P- Z G o o .-. 4-J P U- c co CO G x: 2 G 4-J G x: cm cm O •H G G >s4-> ro H X X X o 4-> x: G o -i-i G 4-1 O XJ CZ 4-J X CN G G C G P- QJ 4-1 o •N G X G CD G O G -P 4-J G P 4-1 G X G G P G G G 2 G p x: o c CO 4-> a C 4-> CZ G G 4-1 o a CD •H G o G X 4-J ■I-i P c XI •H CD C P- o -P G ■H G G •H O X cr X >^ G G G O ro G G 4-J 4-J G cn G Z -P E G Q-4-> •H G C H G G E CO 4-1 c r-i co CD 4-) G cm P o c CO •H CO G x: g ■H C G G -P P G H X •H G c G P G -P 4-1 X P 4-1 2 G o C G CO cm CD 1— o C CO P E •H > 4-> c i—i <4- G p Z •H G •H G G X G 4-J >>-P O P • 4-> G CO c X ■H C cn G C P 4-J G ro ro G O G G G Z O E p G O H p P C O CO c c -p CL-P O O G G G •H G CO It- 4-J •H G >> G G G N CO P CO rH CO G P- G E H CD G G c X X o X O H X ■r-i O-X! o 4-1 4-1 G H 4-1 X X G 4-J cm -P c G H X ■r-i P- c •n X Z G •H x: G o •H G C CO >^ 2 4-1 •N > -N> CO X G 4-J c G >^r-^ a -p •N X •H G G P cr CO ro ► H -P 4-1 P G G O ro X H G CD z O H G G 4-1 E XJ P G CO U- "-i G P P G G ro x: o co ro G CO- ■r-i C x: P G E -P •H X •H 4-1 o c • >^ G P CO c p H H G c G XJ P ro ■r-i G cr CO a ro x: o CD X X G X 4-1 H z E 4-J CO X CO G c ro ro o G p G E G ro C H p X E X G 4-J G 4-1 cm 4-) O -P aa CO- CO C G c P C ■ i-i P G CO rH G > o Q. G G 2 p- c P CO- G G CO ■ I-i P E o CO P XJ X 4-J >N G o G G C_) CO 1 CD H P- G G cm 4-) ■r-i G G G rH P <4- C C P c X X G CO G z CO H ■r-i G G G 4-1 4-J G X X o. G X •H > E i—i G CO o 2 P •H E CO G G 4-J G o. cm cm o G X O G G CO CO G c c i—i Q. G X z X X i—i •H H X G ro G 4-J CO -¥ X X G P H o G Z o p G C c CO rH 4-J G o G 2 G G CO ro >% O H CO ■H G 4-1 c G -P 4-1 2 G P G G G G •H x: G CO P G cm C 1 E P 4-1 P G > cm G g cm O O H G G o G G E c u c CO Q. G p p X X o o c ■r-i G CO ef o Q-r— 4-J 2 r-i ■I-i E X Q. G X C •H G •H ^-~. cn X c H G P G G 00 E cr ro r-i G G C CO G G -P Q.NO G G p o i—i Q. P CD G Q. G X H D v_^ » ■ -P G CO o G >, <4- r-i G . G P >> ✓"—•> Z x: c Z O . G G P p 4-J P o CD >^X G x: Z X r- o 4-> z G P G G > , G -P •H G 2 X rH P -P G c 4-J G NO p o G 4-1 cmx C G X G -P 1 > 4-1 P- ro P o CO C ro G X v_* cn G p- P •H G CO C CO CZ I ■r-i O X O 4-> G •s •H CO -P . a .* Ui G G 4-J G < -P G C H c o X G Q. G >^ G CO 4-1 c O 4-> G C G P C G o G . •H 4-> a Q. G cm o •H x: •H c G 4-J G E ■H o 4-> z G •H CO - o p •H X 4-1 ro C z G c 4-J CO 4-J c G CO G •H •H p- •H XI x: P X G >> G G 4-> G E G O G Z •H c G •H x: X x: ■H P 4-1 E G 4-J P -P •H CO a CD G ro G G p X CZ > •H •r-i 2 4-J 4-J 4-J > O •H CO c CO CO •N G i—i X X O i—i P p P P cn -p h •H X G P C P p G X G G >>X CO 4-J et 4-) r-» Z CO G G 4-J G G G «s cr G G •H O cmxj c Q. Q.XJ 4-J G i—i a G 4-1 G Q. G ~ c G G s: G H •H G G E c o z Q. Z •H CO >> G o c G ■H G O G -P G rH o •H p P o CO o X cm G G • H z ax: •H G CN x: G G P ro p C co P •H 4-J P- E Q. o Lt_ p G c >sx: CD G w 4-J G r-i i— ■H rH o co o > G O 4-1 cm-p •H X P .¥ p G i—i O G C G O G p p -P P G P X >N c O P 4-1 CO- G CO H G - ro 2 CD X CO P -P C G z o G G 2 G O •H 4-1 o CZ >^ o G X G H > c G 4-1 c a Q. CO CD a G o G G G CD •H G -p a G CO -P o X G z -P G G G P cr C G X P E E G rH c o •H G E CD .¥ G 4-1 •H G P o r-i •H G 4-J G H C G O G •H •H X X P p Ct- G P c -P G x: co p z x x: H G o CO P G G rH X XJ o O C G O CD O G CO 4-J 4-J G G 4-> G G G P G ro x: C XJ G •i-i ro G CZ c ■H G G E G c ■H -P H a>H CD p P O P X G 4-1 •H o 2Z cm G ■I-i o H G p X X p P- G CO 4-1 ro a G G p •H X •H E •H G XJ cm O G G G C G -P o P- -* G G G -X Q.4-J >^ cm P Q. G H o > 4-J > ro X 4-> _* Q. X •H •H H Q. H o x: XI G G r-i P c G X G o P >^ G ■H •H G O 4-1 X co G X ro O X X CO X z o •H 2 CO -P s G XJ CD 4-J P H 2 CD H G 2 G P H o 4-1 4-> rH G G 4-1 G O C G P- G CO 2 O H rH c G X G CO 4-1 i—i r-i P CZ co Z rH >, c •H 4-> X G G > O x: CO 1 ■H cm C c P CD CO G X Z o l—i H o p G G •H P c 1 c •H o o CO P G P G -P c G G z G G G G G P > 4-J U- X .* X C G G ro 4-1 . •H G O G Z x •H ro C XI C x: 4-J G G XJ ro ■H CD G G C CD X G X -P 4-J -P O G 4-J CO- •H G E o 4-1 CD c cn O G G G z X CO XI o cm G C G G CO X O- G G o CO >, c > ro G C o c or G W G c P G cr G P CO X 4-> G Z 4-> G 4-1 rz co P o •H ro r-i G P. G CO cr z •H ro •N E G a c i—i P- ef P Q P X p G .a •H 4-J x: G CO P O G O G a X G C •i-i z CD O -P G G G a G 4-> CD G x: G rH 4-1 •S X X G G C O P G o G G G > a o 2 E P H \— •H G G G 2 C P p P X L_ X c ^ G ■ O G 4-J G X O O Z O X P rH G P cn z G C G •H CO 4-J O LfN X G G Q. E P G G a Q.4-1 G CD G p G CO tp o G G 4-J > XJ c o •H NO C G cm O >> p C O O • E G > ro U- z O C E o 4-1 4-J v^ CD CO ro C G rH G z •H p a. G E G G cm c G CO G (4- O G CO • p G C G G G CO- i—i ro G rH H G P c C CO r-i CO. o P c G 4-J X H r-H G G •H .* Z cr o CO >^XJ p ■H CO G ro O -P r-i G G Ct- X G C G CD CO > 2 •H •H 4-1 ■H -P CD cm-H G P G G H •H CD ■I-i CD c X G o G G >> •H •H o -P X rH X cr X ■r-i P o XI •H G H P G X X o 4-1 P X ro G G O G G G G 4-J cr z X P G p z X 2 > G as ■H c ■I-i Z G G G i—i C cr XJ •H G ro G G O C Q. a. G O P •H G G NO CO -P 4-1 c -P cm X CD P CL co ro G 2 G JK4-1 G c rH E rH G P p g r^ G ro •H ro c O o. e c c G c G /-—v 4-1 CO C •H z G c G P O X ON cm CO i—i c CO O cm G G •H •H CZ o cnrH NO •H rH CO E > •H •H U- G CO 4-J rH c a. z G X 3 cr P c P P- CD ■H cr NO P G p x: o x: G XJ G G •H Q. G G G ■H z P 4-J ■H G ^-^ O P z 14- X4J G 4-J G CO cm p 4-1 O cmx: c 4-J 4-J G P •> Z CD G P X c P G o r-i .* G G l—i P. G C z c Q. cr 4-1 z P p 4-1 G o G G rH CO CO G •H O c i—i x: G E 4-J X ro C H X G ro G O CO X E CZ G G P E G ■H G CD -P CO Z CO 4-J p H O G G X X •* X XJ Q. C 4-J G P •i-i p P G Z G •H •H XJ i—i X z co z cr G C G G CZ O G o G G p G Q. G X G U- C G 2 4-J o 4-J 4-J 4-1 p 4-J CO P CD G ro CC P •H TO C -P >> O C Z C 4-1 •H o CO G z > ro G z G cn •H X G 4-1 C G -P G •i-i o •H •H •H a. to G G x: Z 4_] G C CJ G G cr rH G G Ui P CO G i—i P p 2 G z* rH G G TO G 4-1 p r—j P- H • H X P ro G p X c O X ro ro cm G cm G G G G G -P G G 4-J 4_) z it- E G I— Z X z r— ro O-X CO G a c cm X X C CO cn CD x: G ro 4-J CO CJ ro g >> 4-J E 4-1 4-J G O G ■H i O CO i—i 4-1 G CD rH G •H 4-1 z i—i G 1 r—| f~ G Z G 4-> P G co cr •r-i p P P CL Z c ro 4-1 G X p p 4-1 P G CO ■H c P Z P z G z G o 4-J X G z CJ G cm o ■ H x: l_ o a. P 4-J G G >^ G > G c •H G E < G G •H C X H c o CO H p x: a.-p G ro > O G G CO P ro X 4-J G C ■H G C i—i CD E o p Cn G G CO CD X 4-1 P CD G a. G cm z G G O G •H X x: c 2 CO c o c CN CN m the White population and Spanish heritage persons most likely to use home remedies in the treatment of illness—21 percent as opposed to 12 percent for the White population. However, the degree to which such practices supplant conventional medical care is not understood. For example, although the use of curanderas, lay persons, for health advice is cited as a part of an Hispanic alternate health subculture, few studies report specific information on the number and location of these persons and the type of clientele they serve.(69) Studies have found that the degree of acculturation of Mexican-Americans, including knowledge of English, directly affects medical care utilization.(70) The extent to which being an illegal immigrant influences health seeking behavior has not been thoroughly examined. An array of divergent forces currently are working their influences on the nation's health care system. These include: the increasing physician supply the growth of new financial and organizational arrangements for the nation's physicians the growth of new technologies increased cost-consciousness among the purchases of health care and health insurance. This has led to a sharp rise in HMO membership, the restructuring of many health insurancepackages and increased awareness of the hidden costs of uncompensated care the growing role of proprietary organizations in the provision of health care services and a new business-oriented emphasis throughout the health care system. This has resulted in increased caution in determining the level and amount of indigent care institutions provide an increased emphasis on State-level approaches to solving public health policy problems. Some of these include: State/Local Programs of General Assistance A number of States and/or local governments have established programs to finance health services for poor individuals unable to qualify for Medicaid, under the heading of general assistance medical care. A few States have developed catastrophic health insurance programs. These include Alaska, Maine and Rhode Island. Revenue Pools A few States have initiated revenue pools which are used to finance indigent care. In New York, funds are derived from a surcharge levied on hospital reimbursement amounts paid to insurers. The pool funds are redistributed to individual 323 G 4-J G CO G -P • Z G CO E C X CD G I— G P > x cm G -P O P . P CO it- CO- -flj o ro z >> x c c x •i-i CD O G P > H G O G 4-1 C H P P Lj_ O >, 4-1 Q.H C G H r-i C P CO O G G -r-J-H CO TO G H E G G CO E G 4-J CO G-H G O Q. P x: H CO O 4-> O p- G X G P >^ U CO JT CO C G G Q. CO CO C 2^ G O H 4_> 4-J P- •H C P o ~ o ro -p 4-j x x P g co ro to 4-J Q. G 4-> P g ro c h G -HZ P- >n CO- lt- JT) - G O X P X G O G G C 4-J 4-J D O CO H G P G 4-) i—i G P Z co c ro xj 4-> G H h cm g p Q. X -P G G C C O P Z O x ro <4_ o G G G G 4-J X 4-1 G rH 4-> X • O CD CO 4-> X CO E O -P c co CD 4-J •H H -p C CO 00 C O CO G ■H G G G CD p cm ■H G On H -H G X > G >^ i—i jz P- Q. •H H 4-1 X 2 CO •H X -p CD G O G X G P •H X H •H c -P CO G O G P- J* O > G XJ G G •H ■i-i X c X OZ O CO CO. o H G z XI cm CO- c •H o 4-> O p CO G a z •H CO p > •H c P -P P Q.4-J 4-J G X o o o G •H •H CJ O O- ■H CO p c X p p ■H CD G et 4-J G G Q-4-J o -P r-H •H Q. > 4-1 jz h c G CO G o H o P 4-1 C -X o O c G >^-P CD X p G O P = •H X >^ o P H CJ G Q_ C_J P H CD G c ■r-i •H rH X •H G 4-) > CO ■H G CD 4-> X CD G X G P X CO G > G E CD G X a 2 X O cz h 2 cn O O P X •H ■r-i G 4-1 P. ZJ H o c P X 4-> G G X > rH X •H x: H Q--P o a a G o CO c P -P >* G > z . o o 2: p z z o c -PC - P G o G rH O-X L_ p G ■H O X G G 4-J 4-1 W G G CO ■H E HOGG G a CO rH X > G > G G G ■H Q) P X G >nX CD 4-1 G X P •H G G W XJ CC -H G 1— X ro 4-1 X G CD X G •i-i P •H Z-H C a o •H p X c ■H 4-> Z X -P CT CO z 4-J a o G G G •H > C X) G G G -P • P- o G p > •H 4-1 P z E i—i cn p h G -P X o CO rH CZ X G o •H P- X Q-4-J 4-1 G X 4-1 X XJ G G CO G G Z G G C G G CO p ro E P or X CO P O G G i—i O o o o 4-J >> G X o CD G X •H P X) P -p CL G CO CO > 4-J -P X Ui Ui 2 4-J •H CO Q. Q-H G Q. o r-i •i-i CO ZJ P- CO X c Q. 4-1 z CO G CO G CO G XJ G O Q_ z CO p G G 4-J C G X G P -H G O o G CO G Z X c •H G c x: C G 4-1 G G Q. 4-J G X G z X C E G O E G •H Q. G 4-1 P P >%p- G •HO CO-H O >, O X z G CO O •i-i o 4-1 2: O 4-J G 4-J X 4-1 CO P 4-J X X p- c cm C G P CO CZ •H 4-1 •H CO ro z c p G X Q. Z ■I-i C P- r-i X p X G o •H O > 4-J Q.-P 1 z o o ro G o G P CO X •H CO H 2 4-1 4-1 •H o E H ■H CD G . G cmx: G O P G G 4-J O o X G CO 4-J G H -P X l—i o G G C G CJ 2 CO P G G H H G CO- P Z CD c X X E CO > p 2 co x: p- CO. co C o r> C 4-J G G G G G -P o o G G •H ■r-i G • ■H rH X X —I 2 G -H > La- 4-1 rH G CO G c 4-J E X -P - c G •H H CO E E G ■i-i CO c P G CO G C p CO X G 4-J X 4-> CD CO X i—i G G G P OZ Z G •H C G X •H P p XJ >> cm X 4-j cm o -Q X > G G ro a cm cmx co G ■r-i cm ro O 2 G E 4-1 o 4-> P G o o c -P C X •i-i 4-J P G G Z •H P X ■H c O p p z G o c X co o-z ro CZ 2 Q. G Q r—1 X Q.O. p. G E ■H C cm G •H X c E 2 P G i—i X H o O X O 4-J cm G cm z G cm G o H c a c c O c P- X o X c 4-J 4-J H ro X G G ■H X H O G G G CO z CO G E X J* -P a 4-J G X o G G c Q. G •H CO o G CO X C G G i—i H o Q-x: > E G O O H X G G x: G i—i Q. •H •H -P G a JZ -p 4-1 G G c H o >s-P G •r-i O 4-1 z o C r- > G G X G H r-i G 2 G ro cr cm >^H ■H c G G C Q. 2 H X ro CO Q. E G CZ H G o G X E < H G G P •H H > H • H G G o G ax: P- Q. C O r-i >^ ro G Z X 4-J G O 4-1 G • i—i CO O rj • CO P H i—i •H X P- G c G 4-1 CO C G XJ • p o o G G c G P G G P G CO ■H z a H G 4-1 G p C •H •H 2 G G O G O 4-J X X z G rH co G cm o P X CL-P Z 4-J -P G •H G G o G XI G C XI H G cm-P c G •H CO 4-> CD E CO p O ro c G G z 4-1 E c O z G X -P C G Q. E cm O-H •H P H G c et •H c G G o P G H x: i H Z P G 4-1 c G G o E CO cm G G XJ G CO G G G X > G G G ro P C E r-i Cn XI G Z z P > G CO Q. C P X ■H P CO G z ZJ X G G CO CO G G X CO G 4-1 X CO c CO G XJ G G O 4-> G E G 4-1 G G G P •H 4-J G >> G G G G C cr p G C G CD X 4-J G O G -P 4-1 c CO G r-i Q. H O X o XJ o ■H i—i P G G 4-1 Z H P o G CD Z Q. G 4-J ■H JZ z CO C X G P O H C G C o P Z ■i-i G G 4-J G G o o_ G G G G O P- 4-> H G o p G X G > CO P •H X X G > H c P- CD G H cmx G 2 G CO 4-> G X Ui G H C •H H G rH G CO -P 4-1 G X H X 2 > > 4-1 X E z G CO >s O cn P G E J— Ui H G C ■H 4-1 CZ a. cm cm c 4-J c O 4-1 G •H G G cm o G ■H O o o c c ■r-i X CO P- CO G rH G 4-J Cn H C 2 4-J H Q. ro H G P c X > G a G z X c 4-J G 4-1 x: 4-1 JZ o ro G J* •H P G G G X) r-i ■H C G G 4-J G G G CD 4-J c G P Z G G G •H z 4-1 G i—i G G H r-i H rH G CO Q. o G G CD P o G -P G CL G i—i c cn G 4-> E CO G 4-J G X •H 4-1 X ■H -P P CO- P i—i X C P z 4-J G G 4-> > P -P G X CD CO CO G O 4-J H o P- CZ X ro X CO P O CO G p -P G G -P P XJ O G 4-J 4-J 4-1 X G G P 4-1 cm C CD O ro E G CO c G G G G X G CO H CO X E P- G C P- •H X X co G G -P G 4-J X >^ 4-J cr H c G P o c X -P G z •H CD 4-1 G P CO ro -P G o C G G CO cm *> 4-1 Z X XJ X G X CO X G C H G > G P c G H XJ -P H 2 P G 4-> H i—i G CO -P H O 4-1 CO 4-1 •H CJ CO •H P CO G Z r-i CO G co P cm G G c 4-> P G P 2 4-1 P •r-i G G ro ■r-i P E E G G G G O X 4-1 o -P O cm CO ■H c G O P Q.P- P X E •H Lu -P X CO P- a G 4-1 o CO P- X O X o P -P c X o CO E 4-J G P H P c P- G G i—i P G J* 4-1 X G G G ro G CO 4-1 G CO c cm CO G G G c G Q.H p ■H CL CO c X H G c G G > P CO G X CO G Z XJ 4-J 4-J c o O 4-J JZ H X X O O r— G 4-J CJ o O G H P X c P 4-1 4-J -P cm G G G p P H O c ro 4-1 G G G o G G G o c CD cn co- C ro H 4-1 G E 4-J X cn cn-p c C •■ G 4-1 o XJ o G G ro E C -P c P c •H o >, CO ■H P z co H O P G X O G CO CO ■H w X ■H P G G O G -P r-i CO C X P CO 4-J 4-1 G 4-1 4-J CO P- CO H G G o p G H G P G ■H CO 4-1 o a cm C H G X X H G > > P P CO G 2 4-1 G G G c o rH cr G -P 4-1 O C • co G X G ■r-i P ro Z X •H H G H CO H ro 4-J P G G G -P CJ CD G E O G G P 4-> G o X G G CL G 4-J E ■H G p C >, O CO G G G 2 H XJ E x: P X G P G rH CO ro G G 4-> rH H G P- ■r-i 4-J Z 4-1 XJ •H -P G ■H H z > CO -P G H G rH G i—i G G G G G C H C a. P 4-1 CO P X zx G CO CO CD G X >^x: X X o o o o G co -P CO o co p G G G X -P G r— 4-J 4-1 CJ Q. E a G X CO G s E ro X E x; 4-1 CN CO FOOTNOTES Aday, L., Andersen, R., and Fleming, G., Health Care in the United States: Equitable for Whom?, Sage Publications: New York, 1980. Ibid. Subcommittee on Cancer in Minorities Report, Volume III, Draft, February 1985. Ibid. Aday, L., Fleming, G., and Andersen, R., Access to Medical Care in the United States: Who Has It, Who Doesn't. Pluribus Press: Chicago, 1984. U.S. Bureau of the Census. Economic Characteristics of Households in the United States: Fourth Quarter 1983, Series p. 70-83-4. Trevino, F., and Moss, A., "Health Insurance Coverage and Physician Visits among Hispanic and Non-Hispanic People," in Health United States, 1983. DHHS Pub. No. (PHS) 84-1232. Public Health Service, Washington, United States Government Printing Office, 1983. "Health Insurance Coverage and Physician Visits among Hispanic and Non-Hispanic People." Ibid. Berk, M., and Wilensky, G., "Health Care of the Working Poor," "National Medical Care Expenditure Study, U.S. Department of Health and Human Services, Public Health Service, National Center for Health Services Research. Davis, K., and Rowland, D., "Uninsured and Underserved: Inequities in Health Care in the United States," Millbank Memorial Fund Quarterly/ Health and Society, 61:2. 1983. Access to Medical Care in the United States: Who Has it, Who Doesn't. Subcommittee on Cancer in Minorities Report. Ibid. Rossiter, L., and Wilensky, G., "Out-of-Pocket Expenses for Personal Health Services," National Health Care Expenditure Survey, U.S. Department of Health and Human Services, Public Health Service, National Center for Health Services Research, Hyattsville, Maryland, 1982. 325 16. Andersen, R., Giachello, A., and Aday, L., "Updating Access to Health Care among Hispanics," paper delivered at the American Public Health Association annual meeting, Anaheim, California, 1984. 17. Kasper, J. and Barrish, G., "National Health Care Expenditures Study," Data Preview 12: Usual Sources of Medical Care and Their Characteristics DHHS Pub. No. (PHS) 82-3324. U.S. Department of Health and Human Services. National Center for Health Services Research. 18. Ibid. 19. Montgomery, E., and Paranjpe, A., A Report Card on Health Maintenance Organizations: 1980-1984, conducted for the Henry J. Kaiser Family Foundation by Louis Harris and Associates, Inc., undated. 20. National Center for Health Statistics. Health United States, 1984. DHHS Pub. No. (PHS) 85-1232. Public Health Service. Washington. U.S. Government Printing Office, Dec. 1984. 21. Ibid. 22. American Academy of Pediatrics. Trends in Pediatrician Participation in State Medicaid Programs. Chicago, III: American Academy of Pediatrics, March 1985. 23. Children's Defense Fund, American Children in Poverty. Washington, D.C: Children's Defense Fund, 1984. 24. Tenenbaum, J., "Health Care for the Medically Indigent" unpublished staff paper. National Health Planning Information Center. Office of Health Planning. Bureau of Health Maintenance Organizations and Resources Development, Health Resources and Services Administration, U.S. Public Health Service. 25. Singer, J., Sulvetta, M., Solish, M., and Beatrice, D., Uncompensated Care, Issues and Options, Draft, Waltham, MA: Health Policy Research Consortium, December 1984. 26. Shanks, N., 12 Questions: What Legislators Need to Know about Uncompensated Hospital Care. Washington, D.C: National Conference of State Legislators. Undated. 27. Desonia, R., and King, K., Profiles of State Programs of Assistance to the Medically Indigent: A Report in Progress. Washington, D.C: George Washington University Intergovernmental Health Policy Project, 1984. 28. Ibid 326 Uncompensated Care: Issues and Options. Ibid. Health Resources and Services Administration, Background Book, U.S. Department of Health and Human Services, Public Health Service. Ibid. Ibid. Trevino, F., "Health Indicators for Hispanic, Black and White Americans, Vital and Health Statistics, Series 10, No. 148 DHHS Pub. No. (PHS) 84-1576, Public Health Service, Washington, United States Government Printing Office, September 1984. Ibid. Ibid. Subcommittee on Cancer in Minorities Report. Health United States, 1984. Ibid. p.82 "Health Indicators for Hispanic Black and White Americans." Ibid. U.S. Bureau of the Census, Statistical Abstract of the United States 1985 (10th Edition). Washington, D.C, 1985. Indian Health Service, Chart Series Book, U.S. Department of Health and Human Services. Public Health Service, Health Resources and Services Administration, April 1985. Ibid. Gerzowski, M., and Adler, G., "Health Status of Native Americans," paper presented at the American Public Health Association annual meeting, Montreal, 1982. Ibid. HRSA Background Book IHS Chart Book 327 49. Ibid. 50. Ibid. 51. HRSA Background Book 52. Statistical Abstract of the United States: 1985. 53. Subcommittee on Cancer in Minorities Report 54. Ibid. 55. Kolter, M., Goldstein, M., and Lidsler, C, "Final Report of the Evaluability Assessment of the Centers for Disease Control Refugee Health Program." Macro Systems, September, 1984. 56. Ibid. 57. Moecke, M.A., "Caring for Southeast Asian Refugee Patients in the United States." American Journal of Public Health, 73:4, April 1983. 58. Forbes, S., "Medical Assistance for Refugees: Options for Change." Refugee Policy Group, Washington, D.C, November 1983. 59. Ibid. 60. August, Lynn Kao, "Health Service Utilization Patterns of Southeast Asian Refugees." Rhode Island Medicaid, Refugee Medical Assistance, Cranston, Rhode Island, April 1984. 61. Interview with Dr. Robert Knouss, December 1984. 62. "Caring for Southeast Asian Refugee Patients in the United States," p. 433. 63. Ibid. p. 434-435. 64. "Health Service Utilization Patterns of Southeast Asian Refugees." Also, Strand, P., and Jones, W., "Health Service Utilization by Indochinese Refugees." Medical Care, 21.11. 65. Robert Wood Johnson Foundation. Special Report. Number One/1983. "Updated Report on Access to Healtn Care for the American People." 66. "Health Insurance Coverage and Physician Visits among Hispanic and Non-Hispanic People. 67. Ibid. 328 Access to Medical Care in the United States: Who Has It, Who Doesn't. Health Care in the United States: Equitable for Whom? Chesney, A., et al., "Barriers to Medical Care of Mexican Americans: The Role of Social Class, Acculturation and Social Isolation." Medical Care, 20:9, 1982. Profiles of State Programs of Assistance to the medically Indigent: A Report in Progress 329 Health Education Among Minority Communities Report of the Working Group on Health Education The Task Force gratefully acknowledges Cheryl Damberg, M.P.H., Office o Disease Prevention and Health Promotion, Department of Health and Human Services, for her valuable assistance in preparing this report. 332 Table of Contents I. Introduction II. Background A. Major Health Problems Facing Minorities B. Characteristics of Minority Populations that Impact on the Delivery of Health Education III. Developing a Health Education Strategy for a Minority Population A. Definition of Health Education B. Health Problems Amenable to Health Education Interventions C. Factors to Consider in Developing Strategies IV. Program Illustrations A. Primary Prevention Strategies with Low Income Hispanic Families B. Caide Su Corazon: Weight Reduction for Mexican Americans C. To Your Health - Living with Alcohol D. The California/Baja California Maternity Child Health Care Project E. Healthy Mothers, Healthy Babies Coalition F. Indian Health Service Diabetes Program V. Summary VI. Recommendations A. Information and Education B. Access and U .tilization C. Capacity Building in the Non-Federal Sector D. Financing Issues E. Health Professions Development F. Leadership, Work with Other Sectors G. Research Issues H. Data Issues VII. References 333 Health Education Among Minority Populations I. INTRODUCTION In the last century, dramatic changes have occurred in the leading causes of death and disability for most Americans. As documented in the 1979 Surgeon General's Report, Healthy People, infectious and communicable diseases no longer rank among the ten current leading causes of death.1 Today, the leading killers are heart disease, cancer, stroke, accidents, influenza, diabetes, cirrhosis, suicide, and homicide. Many of the deaths attributable to these diseases are preventable, and thus unnecessary. Disease and disability are not events that are experienced equally by all individuals. For a given individual, the likelihood of developing a health problem depends on a variety of factors, for example, heredity, socioeconomic background, environment, inadequacies in the health care system, and personal behaviors. Furthermore, the probability of experiencing ill health changes, depending upon individual experience with risk factors—behavioral and environmental influences that are capable of causing ill health, with or without previous disposition. Upon examination of the ten leading causes of death, controllable risk factors are identifiable for each. For example, heart disease is related to smoking, elevated serum cholesterol, diabetes, and obesity. In the areas of suicide and homicide, alcohol and stress are two prominent controllable risk factors. Some risk factors increase probabilities for several illnesses, such as cigarette smoking, poor dietary habits, alcohol, and severe emotional stress. It is interesting to note the dominance of lifestyle as a characteristic of each of these cross-cutting risk factors. For U.S. minority populations (Blacks, Hispanics, Native Americans, and Asian Pacific Islanders), as compared to the non-minority population (Whites),2 excess deaths are found in a number of the leading causes of death. The disparity in the incidence and prevalence of certain health conditions for minority populations is a compelling reason to identify ways in which the health status of minorities can be improved. Because behavioral and environmental risk factors are associated with the causes of excess deaths among minorities, more work needs to be done in the area of health education, that is, for those components of the major health problems facing minorities that are amenable to health education efforts. Examples of these include the misuse of alcohol and drugs, use of tobacco, dietary habits, exercise, management of stress, adherence to medical regimens, and appropriate use of preventive services. (Table 1) 334 Table 1 LEADING CAUSES OF DEATH FOR MINORITIES RISK FACTORS Cardiovascular Disease Cancers Homicide, Suicide, and Unintentional Injuries Diabetes Infant Mortality Smoking, high blood pressure, elevated serum cholesterol, obesity, diabetes, lack of exercise. Smoking, alcohol, solar radiation, worksite hazards, environmental contaminants, diet, infectious agents. Alcohol or drug misuse, stress, handgun availability. Obesity. Low birth weight, maternal smoking, nutrition, stress, trimester of first care, age, marital status. Cirrhosis of Liver Alcohol. For the four minority groups identified, as for any group, health education interventions are directed at improving the awareness of individuals and communities about controllable risk factors associated with the causes of excess death and disability. Differences in health status underscore the importance of providing health education to minority populations but consensus has not been reached on how to develop health education programs and strategies, how to affect change, and how to disseminate these strategies. It is somewhat unrealistic to expect consensus, however, due to the constant shifts occurring in these populations as they become more "main stream," and because of the existence of very different segments within these minority groups. The purpose of this paper is to describe how health education can be used to address the health problems of minority groups. The paper starts with a brief overview of the major health problems facing Blacks, Hispanics, Native Americans, and Asian Pacific Islanders and a discussion of minority group characteristics. Next, the paper describes the potential for health education to impact positively on minority health status. Following this 335 section is a discussion of relevant factors to consider when developing a health education program. A sampling is then provided of health education interventions that have been undertaken to illustrate ways in which minority populations can be reached. In the final portion of this paper, recommendations for future health education activities for minority populations are presented. II. BACKGROUND A. Major Health Problems Facing Minorities Based on National Center for Health Statistics (NCHS) death certificate data for 1979 and 1980, areas of excess deaths which contribute to 80% of the disparity in health status between non-minorities (Whites) and minority # populations (Black, Hispanic, Native American, and Asian Pacific) were calculated as the data were available.^ Six major problem areas were identified: cardiovascular disease, cancer, infant mortality, diabetes, violence (homicide, suicide and unintentional injuries), and alcohol and drug misuse. It is worth noting that each of the problem areas for the four minority groups have elements amenable to health promotion and disease prevention interventions. Blacks Using 1979-1980 data, the percent of excess deaths by leading disease category for Black males under 45 years of age was 36.6% for homicide, 14.3% for infant mortality, 9.7% for heart disease, 7.8% for accidents, 4.9% for cirrhosis, and 3.0% for pneumonia, as compared to White males under 45 years of age. For Black males under 70 years of age, compared to their White male counterparts, the percent of excess deaths within each disease category was 18.7% for homicide, 16.1% for cancer, 15.5% for heart disease, 8.1% for cerebrovascular diseases, 6.7% for accidents, 5.9% for infant mortality, 4.0% for cirrhosis, and 3.4% for pneumonia. For Black females in this same time period, the percent of excess deaths for those under 45 years of age by disease category was 21.6% for infant mortality, 13.9% for homicide, 13.1% for heart disease, 5.1% for cirrhosis, 5.0% for cancer, 4.3% for cardiovascular, 4.2% for accidents, and 3.3% for pneumonia, as compared to White females under 45 years of age. The percent of excess deaths for Black females under 75 years of age was 30.5% for heart disease, 10.5% for cardiovascular, 9.5% for cancer, 7.5% for infant mortality, 5.9% for homicide, 5.0% for diabetes, 3.4% for cirrhosis, and 2.6% for nephritis/nephrosis. 336 Hispanics Until recently, national data on Hispanic health status indicators did not exist; therefore, definitive statements could not be made about the problem areas which contribute to the disparity in health status between Hispanics and non-minorities (Anglos). Action was taken to remedy this problem in the last several years, and national data on Hispanics will be forthcoming. While selected State Hispanic data exist, there are significant problems with how the data were gathered. It is impossible to use the State data, with the inconsistencies in the methods of collection, to make generalizations by region or by Hispanic group. It is known that many Hispanics in the United States are living below the poverty line, and that as a group Hispanics have much lower levels of education than do non-minorities (Anglos). Both of these factors ultimately impact on the health status of Hispanics. Some evidence indicates that gall bladder disease, obesity, diabetes, cardiovascular disease, and tuberculosis are among the more significant health problems facing individuals of Hispanic origin. Native Americans The percent of excess deaths for Native American males under 45 years of age, by leading disease category for the period 1979-80, was 49.5% for accidents, 10.7% for cirrhosis, 9.5% for homicide, 8.1% for suicide, 3.3% for heart disease, and 3.0% for pneumonia, as compared to Whites. For Native American males under 70 years of age, the percent of excess deaths within each disease category was 56.7% for accidents, 17.8% for cirrhosis, 10.4% for homicide, 7.3% for suicide, 4.9% for pneumonia, 3.9% for diabetes, and 2.9% for heart disease. For Native American females under 45 years of age, the percent of excess deaths within each disease category was 38.9% for accidents, 20.0% for cirrhosis, 6.8% for homicide, 4.5% for heart disease, 3.8% for diabetes, and 3.6% for pneumonia, as compared to Whites. The percent of excess deaths for Native American females under 70 years of age was 41.0% for accidents, 29.7% for cirrhosis, 11.1% for diabetes, 7.3% for homicide, 5.7% for nephritis/nephrosis, 4.3% for pneumonia, and 3.9% for heart disease. 337 Asian Pacific Islanders Within the major disease categories, the incidence of disease among the Asian population, for males and females, is lower than in the non-minority population. However, the risk for suicide in Chinese females rises considerably after age 45 and increases with advancing age. Areas of greatest concern where incidence is higher for Asian Pacifies, particularly for the most recent immigrants and refugees, are tuburculosis and hepatitis. B. Characteristics of Minority Populations that Impact on the Delivery of Health Education Blacks Demographics The Black population is the largest minority group living in the United States, approximately 26.5 million in number. The majority of Blacks live in urban areas. As is true with the other minority populations discussed in this paper, Blacks are not homogeneous as a group. Blacks exhibit great variation in educational levels, socioeconomic status, and religion. While some similarities exist, differences among Blacks, region by region, group by group, person by person, must also be considered when designing a program. Health Beliefs Although a reliable estimate for the prevalence of folk or traditional health beliefs and health care providers in the Black population does not exist, it is important to be cognizant of the beliefs and practices which exist since they may ultimately impact on a health education intervention. It is believed such traditional beliefs and practices are more prevalent in individuals who have less access to mainstream health care, such as older, lower income, rural dwelling individuals.3 Blacks living in urban settings, as the majority do, tend to use the mainstream system as their first choice of care, but some may still hold traditional beliefs. Few studies to date have examined health beliefs in the Black population. 338 • Sources of Health Information Blacks, in general, tend to have well-organized communities which play an important and active role in health and social matters. In many Black communities, for example, the church serves as a powerful and influential institution around which the community may organize. The church often is the primary institution of social support and social control, and significantly influences the norms guiding the lives of church members.4 Because ministers, deacons, ushers and other church personnel are seen as influential, their potential for serving as health education advocates is great. The importance of the church in Black communities should not be underestimated or ignored in planning a health education strategy. The well-organized and highly credible church organization could be effectively used to screen, educate, and follow-up on individuals in the community who are at increased risk of certain diseases. The extended family, traditionally, has been a primary source of health information for Blacks, in part, due to their limited access in the past to the mainstream health care system. The important role of the Black family as providers of counsel and support continues, and may provide an effective means for conveying health information. Because of the high value placed upon families in Black communities, family members could be used to carry health messages that stress the importance of the individual family member's health for the security and well-being of the entire family. Messages incorporating the theme "do it for your loved ones" have been quite effective in reaching members of the Black community.3 In a study completed in 1982 by Juarez and Associates, Black pregnant women and recent mothers named the physician as the best and most credible source of health information.5 The study also found that it did not make a difference whether the information was given verbally or handed to the women. Nurses were not seen as credible sources of health information by these women because they were more often associated with the negative aspects of clinic care. Women, Infants, and Children 339 (WIC) program nutritionists were also cited as a favored source of health information. In another study, Gombeski et al. collected data on Whites, Blacks, and Mexican Americans to determine the media habits of Houston-area adults.6 According to the results of the community survey, Blacks identified physicians as their primary source of health information, with television as the second major source. Whites, on the other hand, selected newspapers as their primary source and physicians as the second major source. In terms of the most credible source of health information, the physician was overwhelmingly selected by Blacks, Whites, and Mexican Americans. These data indicate that the most available source of information may not necessarily be the most credible source (physician). Because physicians are reported as the most credible source of health information and education, they should be encouraged to counsel their patients about measures to promote health and prevent disease. To reach Blacks effectively as a target group, health education for Black communities should take place in settings relevant to Blacks, such as in the neighborhood, at the worksite, through the media, in the schools, churches, and other community organizations that serve Blacks.4 Because of the important role of the community in Black life, active participation by the community in program development is an essential aspect of a successful intervention. In addition, health education programs in the Black community require an understanding of the relationship between the environment, lifestyle, and the improvement of the quality of Black health.4 Programs may need to address other priorities for Black communities such as education, job training, and child care programs, in addition to addressing the health problem(s). All the major determinants of health must be acknowledged in the design of health education interventions. Programs cannot be problem oriented alone, but must be addressed from a quality of life context. As such, health educators may have to assume advocacy roles to assist Black communities in addressing political and social issues that either impact on health status or are of greater priority than the identified health problem alone. 340 Hispanics The Hispanic population represents a "mosaic" of individuals and groups of individuals.' Because Hispanics derive their cultural make-up from a variety of countries of origin (Mexico, Cuba, Puerto Rico, Central and South America), they cannot be classified as a homogenous population. Furthermore, as a population, Hispanics vary greatly in their level of educational attainment, acculturation, and socioeconomic development. While Hispanics differ among themselves as a group and are in a constant state of change, good generic Hispanic health education programs are possible if carefully developed and tested. Only when a particular ethnic subgroup characteristic affects the behavior in question, such as their media use or access to care, should health education programs target subgroups separately. While the Spanish language is the primary cohesive factor binding Hispanics together, their cultural, economic, and educational diversity dictates that no single health education strategy is likely to suit the hetrogeneous needs of this population.8 • Demographics An important characteristic of the Hispanic population in comparison to non-minorities (Anglos) is its much younger age structure. The Hispanic population has a median age of 23, with one-third of its population under the age of 15. Currently, the Hispanic population is the fastest growing minority group in the United States, comprised of approximately 14.6 million people.3 Of this total, 60% of Hispanics are of Mexican origin. Most Hispanics live in urban areas. As a whole, the level of educational attainment for Hispanics remains far below that of non-minorities. Even though upward mobility in educational attainment exists for subsequent generations of original immigrants, the rate lags far behind those of European immigrants. This disparity in educational attainment contributes to Hispanics having higher rates of unemployment and greater numbers of individuals living below the poverty line than for non-minorities. Among all Hispanics, Mexican Americans and Puerto Ricans tend to fare the worst economically. 341 Hispanic families are typically larger than non-Hispanic families, with Mexican Americans having the largest families among all Hispanics. Due to their generally lower economic status and large family size, Hispanics have high rates of overcrowding. Overall, Hispanic families are experiencing shifts in family structure similar to the general population; for instance, Hispanics have an increasing number of female-headed households and single-member households. • Family Involvement in Health Maintensnce The role of the extended family as an available source of health information for Hispanics is an important issue to consider. Hispanics tend to have strong family support networks. There is a tendency for lower socioeconomic and culturally diverse minorities to be more ethnocentric and isolated from mainstream health care. As a result, they may rely heavily on family members for support and information. Roles in Hispanic families are clearly defined: younger members defer to elders , and men are considered the authority figures in their families. A strong family structure can serve to either encourage or discourage the seeking of health care and the use of health information by individuals.10 Depending on the level of acculturation and socioeconomic status, Hispanic families may serve to reinforce traditional beliefs about health, or they may maintain barriers against participation in mainstream health care.3 For example, one study found that young pregnant Hispanic women were largely influenced by their mothers and grandmothers, who provided them with less accurate health information than their doctor or nurse.5 For many Hispanics, illness is a family affair in that the extended family is involved in making decisions about the course of treatment. Hispanic families traditionally share benefits among members and avoid seeking outside help, since loyalty to the family overrides individual interests.3 Thus, when conveying health education messages, it is essential that the Hispanic family be treated as a unit rather than as individual members in isolation in order to avoid possible barriers and to enhance the 342 educational effort. In addition, lines of authority and decision making should be respected and attended to when designing or implementing health education interventions among Hispanic populations. • Perceptions of Health and Disease The health beliefs of Hispanics represent a mixture of traditional health beliefs and "scientific" or mainstream medical beliefs. According to da Silva, some "Hispanics define health as the ability to work, resulting from good luck or good behavior or from a gift of God. Illness is seen as the presence of symptoms and is often accepted fatalistically."9 it is also worth noting that some Hispanic complaints have no convenient reference point in the lexicon of Anglo medicine, such as empacho (indigestion), caida de la mallera (sunken fontanel), mal ojo (evil eye), and mal aire (bad air). Because Hispanics may have traditional perceptions of health and illness that do not necessarily coincide with contemporary medical beliefs and practices, the dissemination of health information with scientific deriviations may create conflicts for many Hispanics and may not be received or used as intended. Should traditional beliefs be determined to play a role in an Hispanic community, messages should be tailored within existing belief systems whenever possible rather than attempting to completely restructure beliefs and practices. • Personal Characteristics Several individual behavioral factors common among Hispanics can play a key role in both the diffusion of health information and the seeking of health care. Among some groups of Hispanic adult males, "machismo" attitudes are prevalent.3,9 The degree of importance placed on exhibiting a macho attitude will impact on whether or not an Hispanic male will use the health system himself or whether his family will use mainstream health care services. Machismo also is closely associated with one's perceived susceptibility to an illness, and will ultimately affect the degree and ease to which behavior change through health education is possible among Hispanics who possess this personal characteristic. 343 Another important characteristic seen among Hispanic women is modesty.9 For Hispanic women, modesty may be a reaction to or a result of male dominance; in some cases a husband may not want a male to treat his wife for a health problem. It is thus possible that modesty may prevent some Hispanic women from seeking health care, including preventive services and health education. Determining whether such characteristics as these exist within an Hispanic community that has been identifed as the target for an intervention allows for the design of the appropriate health education/information strategy. • Sources of Health Information A study in San Antonio, Texas, which examined the media consumption habits of individuals of Mexican origin, found that no significant difference exists between urban Mexicans and non-minorities (Anglos) in the average number of hours spent each week watching television.6 However, Mexicans were more likely than non-minorities to watch television on Sundays and to watch late evening news broadcasts. Particularly noteworthy is the finding that information presented on national news programs was viewed as more credible by Mexicans than by non-minorities. In the area of radio, Mexicans were more likely to listen to radio than non-minorities. A main reason cited by Mexicans for station preference was the desire to hear Spanish music, since music is an integral part of Hispanic life. With respect to newspapers, Mexicans were more likely than non-minorities to read the sports page and special advertising supplements, but were less likely than non-minorities to read the editorial section. Mexicans in San Antonio were also more likely to subscribe to weekday afternoon and Saturday and Sunday morning editions of a newspaper. The findings from this study indicate that urban Mexicans are sufficiently different from non-minorities to merit a special communication effort to reach them effectively. Individuals and organizations wishing to convey health information to individuals of Mexican background and other Hispanics need to consider carefully 344 Table 2 preferences in the type of media used as well as how that media source is used. In another study, the Baylor College of Medicine, National Heart and Blood Vessel Research and Demonstration Center, completed an in-house community survey of Houston area adults.6 The study examined the sources of health information for Whites, Blacks, and Mexican Americans. As indicated in Table 2, physicians represent a major source of health information for Mexicans, both male and female. MAJOR SOURCE OF HEALTH INFORMATION IDENTIFIED BY ETHNICITY OF RESPONDENT Source of Health Information White N=1604 Mexican Black American TOTAL N=547 N=170 N=2,340 Doctor Newspaper Magazine Television Radio All Other Sources; Don't Know TOTAL 21.8 27.8 12.0 16.0 3.0 19.4 100.0 37.3 9.3 7.0 25.0 3.6 17.8 35.0 13.5 5.0 24.1 3.5 18.9 26.4 22.3 10.3 18.7 3.0 19.3 100.0 100.0 100.0 Table 3 MOST ACCURATE SOURCE OF HEALTH INFORMATION IDENTIFIED BY ETHNICITY OF RESPONDENT* Most Accurate White Source of N=1604 Health Information % Doctor Newspaper Magazine Television Radio All Other Sources; Don't Know 68.1 4.7 9.5 5.4 0.2 12.7 Mexican Black N=547 % American N= 170 % TOTAL N=2340 % 73.5 2.4 4.9 8.6 0.9 64.1 3.5 6.5 5.3 0.0 69.1 4.1 8.2 6.2 0.3 TOTAL 100.00 9.7 100.0 20.6 100.0 12.1 100.0 345 In addition, physicians were viewed as the most accurate source of health information for Mexicans (Table 3). However, many poor Hispanics have little contact with health professionals due to possible cultural and language barriers. As a result, several studies have found that many Hispanics rely on the mass media as a source of health information. • Language For Hispanics, Spanish is still the dominant and preferred language.6 However, bilingualism is growing, with a reported 78% of Hispanics stating they either speak or understand English. Generally, Spanish is the preferred means of communication. Exceptions to this generalization are among those individuals who are younger and more educated, who may have only learned to read English. Individuals who only speak Spanish tend to be concentrated into several groups—recent arrivals, the very young, the very old, and those unable to read any language• Native Americans Native Americans have a number of health problems that warrant health education interventions, especially in the areas of alcohol abuse and fetal alcohol syndrome, unintentional injuries, coronary heart disease, gall bladder cancer, diabetes, and obesity.11 However, in attempting to design an appropriate intervention for Native Americans, an educator should be aware of the tremendous diversity that exists within this small population. • Demographics According to the 1980 census, there are approximately 1.5 million Native Americans residing in the United States. This population consists of American Indians, Eskimos, and Aleutian Islanders. Native Americans represent the smallest minority population living in the U.S. and are concentrated primarily in the Southwestern section of the country.3 Over 50% of this population resides in rural areas. 346 Native Americans are separated into 212 tribes, with the five largest tribes being Navajo, Cherokee, Sioux, Chippewa, and Pueblo. What is important from an educational perspective is that each tribe possesses its own dialect/language, philosophy, customs, and structured tribal government. This diversity further complicates the design of a health education intervention, since no single design will suit the needs of all Native Americans. While common elements can be found among Native Americans, educators and communicators need to address the particular tribal characteristics that will ultimately impact on the development and implementation of an effective intervention. • Cultural Use of Tobacco A substantial problem exists when interventions are targeted at health practices or behaviors that are affected by the cultural milieu. For example, tobacco use is interwoven into Indian cultural and religious practices,12 therefore the prevention of smoking becomes problemmatic. Traditionally, tobacco has been viewed as a sacred substance to be used ceremonially. Smoking education for this population, while not a high priority from a Native American standpoint, needs to focus on improving the awareness of the health hazards associated with the habitual use of cigarettes. Existing organizational structures such as Indian Health Service clinics, Tribal Boarding Schools, and other Indian community organizations should be used to disseminate health information on smoking. Messages must be carefully tailored around existing cultural and religious beliefs. • Sources of Health Information American Indians have three important sources of health information in addition to mainstream health care providers/physicians.3 One of those sources is the extended family, which is a source of health information for other minority groups as well. This is, undoubtedly, due to the perception that family members are more likely to be sympathetic about and understand the health problem than are outside mainstream health care providers. The extended family and clan are integral to most tribes' social 347 structures. Households are frequently composed of the nuclear family plus members from three or four generations. As a result, a health education intervention could be enhanced by the inclusion of more than just the individual or nuclear family in the diffusion effort. Another source of health information is traditional healers/medicine people. In traditional Indian cultures, illness is viewed as a state of disharmony or imbalance. Traditional healing ceremonies conducted by medicine people are used to correct the imbalances causing the disease. Indians tend to use medicine people, in addition to mainstream health care providers, because of existing traditional health beliefs. Educators and communicators need to assess the extent to which this alternative health care system is used by a particular Indian community to determine how and to what degree such beliefs and practices will affect the health education effort. It is always important to remember, however, that changes in health practices will be more acceptable to the Indian community provided they do not challenge the existing culture and beliefs. A third source of health information is the Community Health Representative (CHR), within the Indian Health Service; it serves as a liaison between the tribal community and mainstream health care providers. These representatives are community people who are trained to assist Indian communities in obtaining services, both existing and new; to provide basic health instruction to Indians in their homes and the community; and to coordinate the provision of comprehensive health services to the Indian community. CHRs are important to the success of health education efforts since they serve as a vital source of information to the community. CHRs are highly credible sources of health information and could serve as effective agents of change because they are known members from within the community who have been trained to provide health education services. It should be noted, however, that each tribe has a slightly different CHR program that reflects the particular tribe's health needs and priorities. 348 For Indians, the presence of an individual with the same ethnic background assists in the diffusion of health information.2 This is due to the fact that members from within the Indian community have status and credibility as sources of information. Indians generally prefer practitioners of Native American background, since they possess a cultural understanding and sensitivity toward Native Americans. • Urban versus Rural Environment American Indians are provided with health services and health education through the programs of the Indian Health Service (IHS). Primarily, the IHS serves Indian communities in rural and isolated environments. For those individuals living on reservations, tribal councils, tribal newspapers, alcoholism counselors, and community organizations provide good mechanisms for information dissemination and educational efforts. Within Indian communities there is a great deal of movement between the reservation and urban centers, because of the need to find employment. Indians living in urban areas may have greater access and exposure to television programming and health messages aimed at the general population than do rural Indians; however, they lack the support and diffusion mechanisms provided by Tribal Councils and the Indian Health Service. There also is a lack of Indian-directed radio or television programming for urban dwellers. Several potential avenues exist for reaching urban Indians with health education messages. For example, urban Indians tend to cluster in neighborhoods where they form clubs, community centers, and churches, through which health information could be diffused. • Message Content Because great value is placed on the extended family as well as on ties to former ancestors among Native Americans, messages should be designed to highlight aspects of Indian life and culture that can be used positively to influence health behavior. Messages that stress the importance of "keeping healthy for your children" or "eating and keeping fit as your ancestors did" could be effectively employed.3 349 Asian Pacific Islanders The Asian Pacific population is highly diverse in terms of cultures, customs, and languages, and thus presents many challenges to educators/communicators seeking to disseminate health information. Because no national data exist that break out disease prevalence and incidence by country of origin for Asians, the health educator must assess the needs of each target population on a community level. • Demographics The Asian population in the United States has grown significantly in the last ten years, partially due to an influx of refugees from Cambodia, Laos, and Vietnam. In 1980, the Census Bureau reported 3.5 million Asian Pacifies living in the United States, the majority of whom reside in urban areas. This minority group consists of individuals from Japan, China, Samoa, Korea, Thailand, Burma, Hawaii, Vietnam, Laos, Cambodia, and the Philippines. With such a large number of countries of origin, the Asian Pacific population exhibits tremendous diversity in customs, languages, religions, educational levels, level of acculturation, and socioeconomic status. The range extends from poor and uneducated refugees to more highly educated and financially sound immigrants. The newest wave of Asian immigrants have come primarily from Vietnam, Cambodia, Laos, and Burma. These immigrants are more prone to culture shock than earlier immigrants from Japan and China because the majority have not been exposed to an industrial/western way of life. Generally, the priorities for the newest arrivals are securing housing, jobs, and language skills. Often, health is not a high priority for these individuals who may be experiencing a great deal of frustration in adjusting to American life. • Family Structure For the most newly arrived Asian immigrants and their families, the effects of urbanization, role changes in the family, cultural conflicts, and acculturation create feelings of relative powerlessness and social isolation within 350 American society. As Asian families become more acculturated, with subsequent generations, their family structure becomes more like the majority population.3 Newly arrived families are more likely to live together for both economic and social support reasons, while those who are better acculturated to American life may live alone, may divorce, or may not live with elders. However, family influence upon members remains strong even if the family does not reside together. Males tend to still serve as the heads of the household, and age, experience and seniority are respected. Health education interventions would, therefore, need to consider the role that family influence plays in communicating health information. It might be best to design programs that reach family units as opposed to individuals, and to acknowledge the lines of authority and decision-making within Asian families. Additionally, because the church plays a significant role in most Asian communities, its potential is great for reaching families with health information. Also, messages with a "love of family/children" theme would tend to be appropriate for Asians, given the importance of the family network.3 • Community Boundaries Subgroup communities of Asian Pacifies are frequently self-contained, as in the case of Chinatowns and Koreatowns. Since residents of these communities generally do not travel beyond community boundaries for services and because they often encounter cultural and language barriers, health education/information services need to be brought to the target community. In addition, Asians who require health education/information services in their native language usually find culturally relevant and language-appropriate materials to be virtually nonexistent. To increase the likelihood of success, education/information efforts must involve community outreach to develop suitable programs and materials. By working with existing community organizations, such as ethnic churches, associations of business owners, community centers and the Asian Pacific caucus, programs can be appropriately designed and disseminated at the community level. 351 Health Beliefs Traditional health beliefs and practices may exist for many Asian Pacifies, especially for those who are first generation families, elderly, and live in highly concentrated and segregated Asian communities. It is likely that traditional beliefs and practices are most prevalent among recent Indochinese immigrants and least prevalent among Japanese Americans. One should note that it is extremely difficult to generalize about the different Asian Pacific ethnic cultures; therefore, it is essential for the health educator/communicator to assess the prevalence of folk beliefs and practices, and to then determine how such beliefs impact on the design and delivery of health education interventions• • Sources of Health Information Asian Pacific populations have three basic sources of health information: (1) the mainstream system; (2) the folk/traditional system; and (3) unlicensed medical professionals in the community.3 The decision to seek health care or health information depends on the individual and his/her level of acculturation, the ailement, and the availability and acceptability of health care providers. More often than not, the mainstream system is used. Prior to designing an educational/informational intervention for the Asian population, a careful assessment of recipient needs, beliefs, and sources of health information is required. A paucity of data exists in the area of what sources of health information are used by Asian Pacifies, which precludes one from making definitive statements about the best course to pursue; however, it is possible to speculate about how to develop and place health messages for Asians based on the limited information that is available. For the most recent Asian Pacific immigrants, effective messages may have to be bilingual.13 This may require that mainstream providers of health education and care provide interpreters to ensure adequate communication with members of the target population. Another potential mechanism for disseminating health information is through the print media, such as 352 (1) community and national ethnic newspapers, (2) pamphlets distributed in ethnic food stores and at community centers, and (3) newsletters that are distributed and read weekly at church. With respect to the use of television, stations that provide non-English language programming should be considered as another viable avenue for the diffusion of health information to Asian Pacifies. • Cultural Characteristics The desire for independence and self-sufficiency among Asian Pacifies is strong, and in general there exists a substantial fear of dependency. Because this tends to be true, health promotion and disease prevention messages are likely to be more effective if they convey a need to take care of one's health to avoid a problem, and thus avoid dependency upon someone else.3 Another common characteristic found among Asian Pacifies is their great sense of being in control of or being able to control their own bodies and destinies. Whenever possible, this trait should be capitalized on to convey health messages; that is, the educator should emphasize the role the individual can play in adopting or changing behaviors to control or improve health status. III. DEVELOPING A HEALTH EDUCATION STRATEGY FOR A MINORITY POPULATION A. Definition of Health Education Health education for any population focuses on improving the awareness of individuals or communities about lifestyle/behavioral issues that impact on health status. According to Hochbaum, ...it is true that the more one knows, the more likely that person is to act in a healthy manner.•.knowledge may even motivate one to engage in sound health practices, but knowledge alone is not enough to ensure that the individual will act in a healthy manner or choose positive health behaviors.14 Hochbaum points out that an individual's actions are influenced in many ways, by such factors as competing priorities and incompatible messages. 353 Another definition, provided by Green et al., states that health education consists of "...any combination of learning experiences designed to facilitate voluntary adaptations of behavior conducive to health."15 Furthermore, Green ascertains that while the terms health information and health education are used interchangably, the two processes are distinct entities. For Green, Health education...refers to strategies or learning experiences designed to bring about voluntary adjustment of behavior conducive to health...Information may be one element of a strategy, but most behavioral change requires more than health information.16 Health information, then, is often neither a necessary nor a sufficient component of effective health education interventions. Clearly, individuals who are the target of health education will require some minimum level of information to engage in voluntary behavior change; however, health education ideally assists the individual in making healthful decisions, modifying risk factors, adhering to therapeutic regimens, or resisting pressures to engage in harmful practices. In summary, minority health education interventions seek to facilitate community and individual measures that can foster the development of lifestyles to maintain and enhance the state of health and well-being, as well as to increase public and professional awareness of risk factors that impact on minority health status. Given the breadth of such interventions, it is useful to keep in mind that health education interventions can occur in a multitude of settings—for instance, in schools, worksites, and communities. B. Health Problems Amenable to Health Education Interventions Among the most important variables that ultimately affect the outcomes of health education interventions are the following: • the nature of the target population; • the types of interventions available; • the kinds of health-related outcomes that can be expected from the intervention; • the kinds of health-related outcomes that should be of high priority; and • the outcomes other than health-related that may be expected. 354 All of these issues are critical to the successful planning, adoption, and implementation of health education strategies for minority populations. In order to address the particular needs of the populations, the educator must know the critical problems, the likely delivery mechanisms, and strategies for implementation. He must recognize the types of interventions that are available given a unique set of needs and population characteristics (e.g., interventions may be regulatory or legislative, educational products, organizational structures/change, personnel training, technical assistance, or personnel assistance).17 Given a particular target population, a laundry list of potential problems and proposed outcomes can often be identified, yet to adequately plan interventions priorities must be established. The promise of health education cannot act as a bottomless pit into which are placed all of society's special needs and problems for solution. Only those areas that are directly amenable to the interventions available and the resources that may be devoted to those interventions must be identified. Among the potential impacts of health education efforts, six are appropriate to the populations of interest.18 1. To increase understanding about the philosophy and science of individual and societal health. 2. To increase the competencies of individuals to make decisions about personal behaviors that influence their own health. 3. To increase the skills required by individuals to engage ii\ behaviors that are conducive to health. 4. To encourage the maintenance or adoption of appropriate health-related behaviors. 5. To enhance the skills of individuals to maintain and improve the health of their families. 6. To enhance the skills of individuals to maintain and improve the health of the communities in which they reside. Phillipp and Kolbe have identified the following specific priority areas directly addressable by health education interventions that also are directly related to excess morbidity and mortality among the targeted minority groups: • Smoking: The principal activities here should be directed at reducing the number of persons who start smoking and emphasizing the importance of stopping associated with particular health problems of minorities (e.g., cancer and high blood pressure). 355 • Diet and nutrition: Efforts in this area should concentrate principally on improving consumer choices given a fixed income. • Social support behaviors (stress, coping behaviors): Programs teaching social support behaviors should include coping with stresses associated with suicide and homicide, two particularly prevalent problems among minorities. • Exercise: Programs in this area should be designed to foster behaviors that can be carried through life to ensure fitness and that minimize problems associated with chronic heart disease. • Alcohol and drug misuse: Efforts to minimize or stop illicit drug abuse should also include information about inappropriate or non-use of medications in the treatment of disease. • Maternal and child health issues: Activities here should primarily be directed at enhancing early prenatal care and reducing maternal smoking. • Safety issues: Information and education about safety issues can range from the use of safety belts to avoiding occupational hazards. • Age at first sexual intercourse/unprotected sexual intercourse/number of sexual partners: Activities in this area should include efforts to minimize problems associated with sexually transmitted diseases and teenage pregnancy.19 These priority health behaviors are among the most important to be dealt with among minority populations, and more importantly, they also represent those behaviors most likely to be affected by health education interventions. C. Factors to Consider in Developing Strategies Because a multitude of factors can enhance or impede the effectiveness of a health education strategy (i.e., the creation and dissemination of health information and behavior change messages), the development and implementation of any strategy necessarily entails a thorough planning process. The process forces the planner to consider what contributes to making the health message and its dissemination special for a given population and what may contribute to the success or failure of a health education program. 356 1. Influence of Community Leaders and Groups Most change takes place within the context of a social system.20 Examples of social systems operating in all populations are the family and the community. Social systems can and often do facilitate or impede the diffusion of health information. If the nature of these social systems is not clearly understood by those who implement a health education strategy, the family and community may serve to prevent health information from being diffused as planned. On the other hand, the family and community can serve as tremendous resources—agents of change. Involving the family and community members in the strategy can strengthen the effort by lending credibility and visability to the activity, facilitating acceptance and self-determination, and creating greater awareness of the target community and its culture. Examples of people and groups at the community level who also might participate in a health education program include the following: • Local political leaders who have an obvious incentive to participate in programs designed to help the community. • Church leaders and visibly active church members (e.g., deacons, ushers) can be effective opinion leaders, especially in Black, Asian Pacific, and Hispanic communities. Some Black churches have church nurses who might be especially influential as opinion leaders in the diffusion of health information. In many minority communities, churches serve as social centers as well as religious centers. • Local media and sports personalities have high visibility and credibility in many communities. Although some minority groups have little access to broadcast media, some communities do support minority programming. • School teachers traditionally are respected as influential members of the community, especially in Hispanic communities, and this role may be most prominent when a teacher is one of the few sources of information for non-English-speaking parents. # Alcoholism counselors can be influential leaders for portions of a community. In programs funded by the Indian Health Service, alcohol counselors often have great influence and credibility. • Club presidents can be influential opinion leaders. In many Spanish-speaking communities, hometown clubs consist of people who came or whose families came from particular regions outside the country. Women's clubs also are very 357 important in many communities, as well as fraternities and sororities that often have community service goals. • Local newspaper editors, local chapters of advocacy groups, and professional organizations may also provide effective opinion leaders at the community level.3 While these examples refer to any population at the community level, their relevance to minority populations is evident. One of the major benefits of involving community leaders and groups in the development and implementation of a minority health education program is that their participation helps ensure an accurate understanding of the target population's health beliefs and needs. 2. Community Attributes Community traits that will influence a health education activity might include the relative importance placed on health; the cultural habits of all members or segments of the community; and the level of employment, which will influence the amount of resources available and the degree of self-esteem the community as a whole possesses. In addition, special functions may be assigned to particular members of the community, for example, on the basis of sex. Understanding these components helps ensure the most appropriate point of intervention as well as the nature of the message. 3. Perceived Barriers to Taking a Health Action In planning a health education strategy, the planners need to consider whether the person or group to receive the health information feels susceptible to the condition or illness being addressed and whether they feel the condition is serious.21 Individual characteristics—such as fear, anxiety, a sense of invulnerability, self-determination, and skepticism—also influence behavior.3 In a study conducted by the National Heart, Lung, and Blood Institute, which examined the diffusion of information to culturally diverse populations, five factors were found to represent beliefs concerning the cause or manifestation of disease among some minority individuals. These included fatalism/God's will, naturalism, life balance, supernatural origin, and superstition.3 when attempting to pursuade individuals to take actions to maintain or improve their health it is important to realize that these five factors may create a perceived sense of powerlessness, alienation, and inability to change one's destiny. While not all minority individuals maintain these beliefs, those who do may be less likely to seek out and act on messages that presume individual autonomy and self-control. 358 Another important consideration is whether the recipient of the information believes in the benefits to be derived from the recommended health behavior.22 Belief in the effectiveness of a health intervention will greatly influence the likelihood that an individual will adopt the measure. The likelihood that a person will comply with a recommended health action is in part a function of his or her beliefs about the probable effectiveness of the action in reducing the threat. Perceived benefits are multidimensional and might include an increased chance for recovery, or the prevention and detection of disease prior to symptoms. Other factors such as fear of pain during treatment and complexity and duration of treatment serve as barriers for an individual considering the adoption of a health action. Perceived barriers can best be moderated by underscoring the benefits to be gained by taking the recommended health action. 4. The Environment and Other Barriers to Health The individual and his or her behaviors and perceptions are not the sole source and solution to a potential or actual health problem. Behavioral and nonbehavioral factors contribute to health and disease.1 Nonbehavioral factors that influence an individual's health status include environmental hazards such as those that exist in the home or workplace; biological factors such as genetic make-up, sex, and age; and inadequacies in the health care system. It is essential that other threats to health besides the behavior of the individual are recognized as contributing factors in any population; otherwise, it may appear as though individuals are being blamed for their health problems. Indeed, in some instances, environmental and technological problems may be higher priorities for and more apparent barriers to solving the larger health problem for the individual and the community. 5. Demographic Parameters A variety of demographic characteristics require careful assessment prior to designing a health education strategy. For instance, the average level of education in a particular community, both reading skills and literacy, will dictate how the health message is to be shaped. In the case of preparing health strategies for minority populations, the level of acculturation—the degree and ways in which minorities adopt the beliefs and behaviors of the majority population—will influence the type of message developed. Rural versus urban living creates different issues related to health information and services.3 6. The Nature of the Innovation Examination of the attributes of the health information to be diffused or the health-related behavior change desired is 359 essential to the strategy-building process. New information, or innovations, can be classified according to their (a) relative advantage—is the innovation perceived as better than the existing beliefs or practices? (b) compatibility—is the innovation consistent with existing values, past experiences, and the needs of the receiver? (c) complexity—is the innovation perceived as difficult to do or understand? (d) "trialability"—can the innovation be tried on a limited basis? (e) observability—are the results of the innovation observable to the person trying the innovation and to others?3 Problems may be encountered in any or all of these areas. For example, the relative advantage for a teenager to stop smoking may be low because the health problems associated with the behavior are not readily apparent. Unless the benefits of not smoking are firmly established, the teen will not discard the old behavior. Or, in the area of lowering cholesterol, it is not readily observable to an individual whose diet has changed that serum cholestrol levels are actually lowered. Observable results from a behavior change serve to assist in reinforcing the new behavior. Finally, in the area of treatment for high blood pressure, the initially incurred problems of "trialability" and complexity were overcome by developing a process of stepped care in treatment. Stepped care allows the person to progressively adopt behavior over time and thus integrate the changes into a daily routine. These examples point to the importance of carefully considering the attributes of the innovation as it will impact on the target population prior to developing a health education intervention. 7. Channels of Communication Mass media, group discussions, lectures, role playing, modeling, community organizing, individual counseling—these are some of the mechanisms by which health information within the context of a health education program, can be conveyed to an individual or population of individuals.15 An important element of each of these educational methods is the process of communicat ion. Although mass media channels can be used effectively to create a greater awareness about a health problem, their use to create behavior change is generally less effective.20,23 Among culturally diverse minorities, which tend to have strong social networks, interpersonal channels are a preferred means for conveying a message to elicit behavior change. In each of the four minority populations, physicians are viewed as one of the best interpersonal channels through which to send a health message.3,24 physicians are seen as highly credible, and can play an important role in enhancing motivation and compliance. As such, physicians should be encouraged to engage in patient education practices that serve to improve and/or maintain the health of their patients. In addition, the immediate and 360 extended family are valued purveyors of health information in an informal context. Family members can serve to encourage, support, and reinforce the beliefs and behaviors of individuals who are the target of a health education strategy. 8. Summary of Planning Principles for Minority Populations A number of general guidelines can be offered as planning components that should be considered in the design, implementation, and evaluation of a health education strategy for a minority population. These guidelines recognize the value of health interventions that involve community support and respect differences between populations. Plan a health education message that has scientific consensus on the validity or value of the recommendation. Avoid making assumptions about the health problems, practices, and beliefs of any population—that is, confirm facts through community assessment or research. - Set specific, measurable goals and objectives. Avoid wholesale change—fit new health messages within the context of existing health beliefs and practices. Determine who is providing health care and health information within the community and its origin. - Assess the health needs of the target population and monitor community reactions to intervention efforts. - Involve physicians and other credible agents of change in efforts to convey health messages and to encourage behavior change. - Directly involve the target population in the planning, implementation, and evaluation phases of the health education strategy. In the instance of a minority health education strategy, (a) solicit the participation of bilingual and bicultural individuals of the community and determine preferences for the language to be used to convey the health message; (b) use models who are similar to the target group; and (c) identify the correct cultural context for the message or program. Recognize cultural diversity—for example, tribe to tribe, region to region, rural to urban, levels of acculturation, and generation gaps. Use existing channels of communication to facilitate the dissemination of information. 361 Pretest health messages among members of the target audience for impact, comprehension, personal relevance, believeability, and acceptability—consider the language of the message, the use of illustrations, the reading levels of the target population, the models portrayed in the message, and the environment in which the message is portrayed IV. PROGRAM ILLUSTRATIONS The descriptions of programs that follow are intended to provide the reader with a sampling of health education interventions that have been initiated to address the significant health problems facing the four specified minority populations. By no means are the descriptions of the programs intended to be representative or exemplary of all of the existing interventions of this nature. Rather, they serve to describe a number of approaches that have been employed to address minority health concerns. Although evaluations of the results of these projects are not currently available, this sampling of health education strategies is illustrative of programs that take into consideration those factors cited previously in this document as elements essential to the targeting of a minority population. A. Primary Prevention Strategies with Low Income Hispanic Families The National Coalition of Hispanic Mental Health and Human Services Organizations undertook this primary prevention project in September 1983.25 The purpose of the project was to promote awareness and action for health improvement among Hispanic families through risk reduction, preventive measures, and health education efforts. Several activities were proposed to accomplish the project: (1) the identification of innovative and promising approaches in health promotion and disease prevention aimed at low income Hispanic families in Head Start and other programs; (2) the development of health promotion materials for low income Hispanic families; (3) the development of strategies for Head Start programs to initiate or improve health promotion activities; and (4) the development of a dissemination plan to guide utilization of materials under the direction of the Office of Human Development Services/Administration on Children, Youth and Families/Head Start. This program is unique in that it will use the existing Head Start program mechanism, already serving a large number of Hispanics, as a channel through which to disseminate health information on such topics as stress, exercise, nutrition, smoking, alcohol misuse, and safety/injury prevention. The Head Start program will receive the following materials as a result of this program. 362 • A bibliography of selected consumer pamphlets, primarily in Spanish. Topic areas include prenatal care, mental health, infant care, hypertension, dental health, child health, safety and accident prevention, nutrition, and reproductive health. • A list of curriculums that address alcohol and drug abuse, smoking, nutrition, exercise and fitness, safety and accident prevention, and stress. • A bilingual booklet on health promotion. • A strategy guide on how to incorporate health promotion activities into Head Start programs. B. Caide Su Corazon: Weight Reduction for Mexican Americans Within the State of Texas, individuals of Mexican background are at high risk for cardiovascular disease, in part because obesity is a significant health problem within the Mexican American population. To address this problem, investigators at the Baylor College of Medicine, National Heart and Blood Vessel Research and Demonstration Center in Texas are initiating a demonstration and evaluation project to test the effectiveness of a culturally adopted version of the HELP Your Heart Eating Plan and behavioral weight control program within a Mexican American community.26 The project will assist young Mexican American families in developing a more active approach to family health and adopting dietary and physical activity patterns that promote (1) the reduction of risk for cardiovascular disease, hypertension, and diabetes; (2) the achievement of an ideal weight; and (3) the prevention of obesity and cardiovascular disease in children. This health education intervention will utilize social learning theory and social support as a basis for behavior change. Due to the importance of family social support within the Mexican American family and the familial clustering of obesity and other risk factors, the project will compare a family-oriented intervention to a more traditional individual-oriented treatment program. Families will be followed for one to three years after treatment to evaluate the long-term effectiveness of the intervention. Participants will consist of families in which one or both parents are 20% over ideal body weight, are between the ages of 18 and 45, and have a child in the 3-6-year-old age range. A total of 180 families in two cities will participate in this project. Families will be randomly assigned to (1) a diet-booklet-only group; (2) an individual-oriented intervention; or (3) a family-oriented intervention. 363 The emphasis of the project is to assist families of Mexican descent in developing lasting lifestyle changes in eating and exercise habits, ultimately to impact on cardiovascular health, obestiy, hypertension, diabetes, and fitness. C. To Your Health - Living with Alcohol The Indian Nations/Tribes that participate in the Bureau of Indian Affairs Boarding Schools (Quechan, Cocopah, Apache, Navajo, Mohave, Pima, Hualapai, Supai, Hopi, and others) have intiated an alcohol abuse prevention project to target students 13-18 years of age.27 The purpose of the project is to reduce the incidence of drop-out, injuries, accidents, and arrests related to alcohol abuse and misuse among boarding school students. The project, which consists of developing and implementing a series of educational programs, has as its focus self-responsibility, limitation, and group control in relation to alcohol use. The series of programs will be developed and used within the context of Indian life, relationships, and social structure. This program is especially noteworthy since a key element of its design is the provision of an historical overview of Indian society and use of alcohol among Indians. This will include an examination of learned behavior in Indian society, Indian socialization practices, and the effects of alcohol use on Indian life, both from an individual and societal perspective. The overall program will emphasize the future of Indian life and social function in modern society. D. The California/Baja California Maternity Child Health Care Project Under a three year project grant from the Department of Health and Human Services, Division of Maternal and Child Health, the Health Officers Association of California (HOAC) will address myriad factors that affect the health status of mothers and infants living in the California/Baja California boarder zone.28 Based on State vital statistics, approximately 31% of births in 1982 in California were to women of Spanish origin, 89% of whom were of Mexican ethnicity. A high percentage of these women who are seeking and receiving maternity care services have low incomes. Individuals involved in this project are collaborating closely with the Pan American Health Organization, the U.S.-Mexico Border Health Association, and the California/Baja California Binational Health Council. This project will accomplish the following activities: • convene a series of binational meetings of health professionals; 364 • produce a California/Baja California Health directory on maternal and child health programs; • develop an inventory of MCH health education materials; • design a health education campaign on perinatal care issues; • develop perinatal health education materials, specifically for this population; and • review guidelines for the care of the low income Spanish-speaking pregnant women and medical treatment protocols for those with high risk factors. The California/Baja California Maternity Child Health Care project is unique because it focuses on developing ongoing communication among providers and public health officials from California and Baja California. The purpose of this communication is to foster a binational cross-cultural understanding of current medical nursing and related maternity care services practiced in both countries, including health education services to non-English-speaking individuals. E. Healthy Mothers, Healthy Babies Coalition Healthy Mothers, Healthy Babies is a public education effort carried out through a partnership among government, professional, and voluntary organizations and agencies.30 a coalition of members representing 66 national organizations was formed to: • provide information to promote healthful behavior among pregnant women and women planning pregnancy; • increase their understanding of health risks and the importance of taking personal responsibility for their health and the health of their infants; • motivate women to take action to protect their health, obtain regular prenatal care, and seek other counsel or assistance when needed. Low income and/or minority women are the principal targets of this public education effort. Projects undertaken by the Coalition include the development and dissemination of educational posters and cards to physicians for low income women. These cards and posters were printed in Spanish as well as English. Members of the Coalition also have developed a directory of educational materials, a promotional packet on breast feeding for health professionals, and a television production on the Coalition for local stations to air. 365 Plans for the 1985-86 calendar year include encouraging the establishment of State level chapters of the Coalition; developing a "media materials exchange network," beginning with New York State Health Department materials, which will be made available for use in 10 other States; sponsoring regional workshops to exchange Coalition ideas, projects, and methods; and developing a compendium of program ideas for motivating low income women to seek prenatal care. F. Indian Health Service Diabetes Program In 1979, the Indian Health Service (IHS) funded five diabetes model care projects as part of its Diabetes Program, to develop, document, and disseminate improved ways for preventing premature diabetes-related morbidity and mortality in Indian communities.31 Training is coordinated for model site personnel by the IHS, and active interchange among the sites and other diabetes-related organizations is promoted. Workshops have been held to transmit the latest recommendations for diabetes care and education to IHS providers and to show providers how to implement these recommendations in the context of the IHS. Educational materials and approaches piloted in the model sites have been disseminated widely throughout the IHS. The five model projects are located in Ft. Totten, ND; Albuquerque, NM; Winnebago, NB; Claremore, OK; and Sacaton, AZ. Various types of activities that have been initiated within projects include emphasizing home visiting and teaching; providing community awareness programs and exercise classes; providing extensive follow-up for pregnant diabetics; emphasizing strong community prevention programs and producing culturally acceptable audio-visual teaching materials; using a diabetes test kitchen as a teaching tool; and developing programs in the areas of exercise, pregnancy, and the management of obese adolescent diabetics. V. SUMMARY This paper underscores the important contributions health education and information interventions can make to lessen the disparity in health status between non-minorities and minority populations in the United States. While the paper does not provide a definitive review, it does highlight some important issues to consider when designing and implementing a health education program for culturally and ethnically diverse minority populations. It is hoped that the preceeding discussion of issues will facilitate thoughtful planning and implementation of health education interventions for minorities in the future. 366 . OPPORTUNITIES FOR PROGRESS Information and Education • Several of the major health problems of minorities have components susceptible to intervention through educational efforts. Messages should be developed to target the special needs of the group for the given problems. Important issues to be addressed by such efforts include: • health and prenatal examinations, and positive maternal health habits; • promotion of healthy nutritional habits within the constraints of cultural patterns (e.g., more fiber, less fat, less salt); • deterrence of the use of tobacco products, including chewing tobacco and snuff; • deterrence of drug use, especiallly among youth; • deterrence of alcohol use, especially among youthful drivers and chronic adult alcohol abusers; • promotion of enhanced levels of physical activities for all ages; • enhancing awareness of hazards inherent to certain occupational settings; and • enhancing awareness of where to obtain health services, whether general or specialized. • The Department, in cooperation with the private sector, should assess the availability and appropriateness of health education materials/activities that address the major health problems confronting minorities: cancer, diabetes, violence, alcohol and drug misuse, cardiovascular disease, and infant mortality. Where an information void has been determined to exist, the Department and the private sector should direct a portion of their resources to developing culturally appropriate materials to meet the identified needs. Gaps in information exist especially in the areas of stress, violence, and exercise. • Material developed for use with a target minority population should be screened and tested through the use of focus groups or other assessment techniques with members of the minority population, prior to implementation and dissemination. Such techniques can also be used effectively to identify the particular needs of a population. 367 • Where materials written in languages other than English are needed, they should be developed from the outset of the program, involving professional and lay members of the target community in their development. • Innovative health education interventions should be developed for minorities to be used in churches, worksites, and schools. • Techniques should be explored for using pictures and words to convey health education messages. Such an examination should produce innovative ways of communicating health messages to culturally diverse populations. • A mechanism should be established to facilitate the exchange of health education materials appropriate for the identified target groups. In addition, a compendium of materials and a summary of selected projects should be developed, by disease category, which address the major health problems confronting minorities. • The Department should consider developing public information campaigns that target priority health problems facing minorities; these campaigns should be long lasting in nature and be evaluated for their impact. Such campaigns should be modeled after the successful National High Blood Pressure Education Program. • Messages developed for media use should be based on providing useful tools on which to act, not just admonitions. • Culturally sensitive and problem-specific materials should be developed, where lacking, for use in providing health education services to individuals in DHHS-operated clinical settings. • Providers of health education services in clinical settings should either come from the minority target group and thus possess the required cultural perspective, or be culturally sensitive to the needs of the target group. B. Access and Utilization • Individuals and organizations should work to ensure the integration of culturally sensitive health education programs/activities, delivered by culturally sensitive health care professionals, into health care services that target minority health problems. • Federal, State, and local agencies should use the established communication networks of organizations within 368 minority communities as conduits for the dissemination of information about health promotion, disease prevention, and the use of health services. • Because of the powerful influences of cultural factors developed over a period of many years in people's attitudes toward health behaviors, programs sponsored to motivate individuals to change their behaviors should be prepared to be sustained over time and always be implemented with the cooperation and participation of community support organizations when available. Such organizations can serve to reinforce the central themes of the program/message. Capacity Building in the Non-Federal Sector • The Department should consider funding risk reduction grants to States for programs, including those run by localities and private organizations, that target the health behavior needs of the four minority populations. • Many of the solutions to behaviorally associated health problems of minority groups have their roots in cultural factors which mandate carefully planned and sustained approaches. Health education efforts should embrace a three-tiered approach, including • development of general media-based messages on the target problem; • development of materials to be provided in individual counselling for persons at highest risk; • development of support networks to facilitate and sustain behavior change, e.g., direct involvement of the family, churches, employers, and schools. Financing Issues • HCFA should assess the special health and patient education needs of minorities, which can be provided in clinical settings. Once the needs have been identified, HCFA should implement a series of demonstration programs to identify the best means for reimbursing health education programs provided in clinical settings. • Federal and State governments should examine mechanisms for reimbursing counselling and patient education services provided under Medicare and Medicaid programs. 369 Health Professions Development • A series of training seminars on health education and health promotion techniques/methods for minority populations, directed at practitioners within the National Health Service Corps and those working under HHS grants should be sponsored. The seminars should utilize the existing resources and expertise of the Department of Health and Human Services. • The public and private sectors should examine ways to increase the minority representation in health education, communications, and other health professions. Leadership, Work with Other Sectors • The Department should convene a meeting with leading minority organizations to chart a strategy for diffusing health information among minority groups. The involvement of major voluntary organizations in this activity is recommended. By involving a broad array of agencies and individuals, the Department could mobilize and strengthen private and public efforts to address minority health information needs. Research Issues • Impact and outcome evaluations of minority health education interventions should be sponsored to help plan or modify intervention and to justify the allocation of resources to such projects. • Research studies among the four minority populations should be sponsored to identify more accurately existing health beliefs and practices and to determine what are their sources of health information. • Research needs to be conducted that will elucidate the specific characteristics of minority populations that may impede or facilitate the diffusion of health information. • Researchers should examine different ways of approaching minority populations to affect behavior change. For example: • How do we motivate young Black males to change their behavior in the area of violence? • If there are male and female differentials in behavior patterns, why do they exist and what is the nature of those differences? • What is the nature of minority populations' dietary behaviors? 370 H. Data Issues • The Department should develop a data collection model for application at the local level, so that communities can identify their own needs in order to develop interventions at the local level. • The Department should work with the non-federal sector to develop a data base on health needs, health beliefs and practices, and sources of health information among minority populations. 371 VII. REFERENCES 1. U.S. Department of Health and Human Services, Healthy People: The Surgeon General's Report on Health Promotion and Disease Prevention. DHEW (PHS) Pub. No. 79-55071 Washington, D.C: U.S. Government Printing Office, 1979. 2. Personal Communication, National Center for Health Statistics, October, 1984. 3. U.S. Department of Health and Human Services. Development of Diffusion Strategies Among Culturally Diverse Populations, NIH Pub. No. 84-2697, National Institutes of Health, National Heart, Lung, and Blood Institute: 1984. 4. U.S. Department of Health and Human Services. Health Education and the Black Community. Centers for Disease Control: December 1980. 5. Juarez and Associates, Inc. Healthy Mothers Market Research: How to Reach Black and Mexican~American Women. Office of the Assistant Secretary for Health, Public Health Service, U.S. Department of Health and Human Services, Washington, D.C: September, 1982. 6. U.S. Department of Health and Human Services. Proceedings of the Conference on Communicating with Mexican Americans: Por Su Buena Salud. National Institutes of Health: June 1981. 7. Ramirez, A.G., and Cousins, J.C Hispanic Women's Health Issues: Understanding a Mosaic Population. Paper presented at the American Public Health Association Annual Meeting, November 1983. 8. Ramirez, A.G., Gombeski, W.R., Kantz, J.A., Farge, E.J., Moore, T.J., and Weaver, F.J. Communicating Health Information to Urban Mexican Americans: Sources of Health Information. Health Education Quarterly 9(4): 293-309, Winter 1982. 9. da Silva, CC Awareness of Hispanic Cultural Issues in the Health Care Setting. Children's Health Care 13(1): 4-10, Summer 1984. 10. Warneke, R.B., Intervention in Black Populations. Cancer Among Black Populations. New York, N.Y.: Alan R. Liss, Inc., 1981. 11. U.S. Department of Health and Human Services, Indian Health Service Chart Book Series. Public Health Service, Health Resources and Services Administration, Washington, D.C: June, 1984. 372 12. Personal Communication, U.S. Department of Health and Human Services, Indian Health Service, January 5, 1985. 13. Personal Communication, San Francisco General Hospital, Robert N. Ross Patient Education Resource Center, December 19, 1984. 14. Hochbaum, CM. Health Behavior. Belmont, CA: Wadsworth Publishing Company, Inc., 1970. 15. Green, L.W., Kreuter, M.W., Deeds, S.C, Partridge, K.B., Health Education Planning: A Diagnostic Approach, Palo Alto, CA, Mayfield Publishing Company: 1980. 16. Green, L.W., Health Information and Health Education: There's a Big Difference Between Them. Bulletin of the American Society for Information and Science 4(4):15-16, 1978. 17. Kolbe, L., and Iverson, D. Comprehensive School Health Education Programs. Miller, Matarazzo, Weiss, Herd, and Weiss (Eds.), Behavioral Health: A Handbook of Health Enhancement and Disease Prevention, New York, N.Y.: John Wiley & Sons, 1984. 18. Kolbe, L., Predicting the Impact of School-based Health Promotion: Some Variables in the Formula. Presentation to the NIH Health Promotion Subcommittee, National Institutes of Health, Bethesda, Maryland, February, 1985. 19. Phillipp, A., and Kolbe, L. Changes in the Prevalence of Priority Adolescent Health Risk Behaviors. In press. 20. Rogers, E.M. and Shoemaker, F.F., Communication of Innovations: A Cross-Cultural Approach, New York, N.Y.: MacMillan Publishing Co., Inc., 1971. 21. Rosenstock, I.M., Historical Origins of the Health Belief Model, Health Education Monographs 2, 328-35, 1974. 22. Kirscht, J.P., The Health Belief Model and Illness Behavior, Health Education Monographs 387-408, 1974. 373 23. McQuail, D. and Windahl, S. Communication Models for the Study of Mass Communication, New York, N.Y.: Longman, Inc., 1981. 24. U.S. Department of Health and Human Services. Strategies for Promoting Health for Specific Populations. DHHS(PHS) Pub. No. 81-50169, Office of Disease Prevention and Health Promotion: 1981. 25. Personal Communication, National Coalition of Hispanic Mental Health and Human Services Organization, December, 1984. 26. Personal Communication, Department of Medicine, Baylor College of Medicine, January, 1985. 27. Personal Communication, Indian Health Service, January, 1985. 28. Personal Communication, Health Officers Association of California, January, 1985. 29. Personal Communication, U.S. Department of Health and Human Services, Public Health Service, Region V, January, 1985. 30. Personal Communication, U.S. Department of Health and Human Services, Public Health Service, Office of the Assistant Secretary for Health, January, 1985. 31. Personal Communication, U.S. Department of Health and Human Services, Indian Health Service, February, 1985. 374 Minority and Other Health Professionals Serving Minority Communities Report of the Working Group on Health Professionals MINORITY AND OTHER HEALTH PROFESSIONALS SERVING MINORITY COMMUNITIES Report of the Working Group on Health Professionals Contents Acknowledgements ............. 378 Introduction .............. 379 Part I. Summary and Conclusions.........382 Part II. Selected Variables Influencing Health Professionals: Numbers, Types, and Distribution ....... 394 A. Minority Populations and Communities and their Health Professional Resources 1. Black 2. Hispanic 3. Asian/Pacific Islander 4. American Indian B. Minority Health Professionals 1. Distribution 2. Development 3. Practice Part III. Why do the Differences Exist and How do They Contribute to the Health Status Disparities? . . . 484 Appendices I. Counties with 20 percent of Population in Any One Minority Group...........495 II. Minority Health Professions School Graduates .... 509 III. The Treatment Practices of Black Physicians: Summary . . 541 IV. Letter to External Community ........ 545 377 ACKNOWLEDGEMENTS This report represents a coordinated effort under the guidance of the Bureau of Health Professions, Health Resources and Services Administration, Public Health Service. We, the working group on Health Professionals, gratefully acknowledge the following individuals for their contributions: William A. Darity, Ph.D., Professor of Public Health and Dean of the School of Health Sciences, University of Massachusetts at Amherst for his research, editorial and technical assistance, Everett R. Rhoades, M.D., Director, Indian Health Service, for his counsel and technical support, Howard V. Stambler, Director, Office of Data Analysis and Management, and his staff including, Roger B. Cole, Chief, Information Systems Branch, Leonard A. Drabek, Chief, Technical Analysis and Coordination Branch, and Ernell Spratley, General Statistician, for the development of information on the U.S. and minority populations and the U.S. health professionals, largely through their Area Resource File, Vivian Chen, Program Analyst, Division of Medicine, Remy Aronoff, Program Analyst, Analysis and Evaluation Branch, Division of Disadvantaged Assistance and Blake C. Crawford, Writer-Editor, Division of Associated and Dental Health Professions, all of the Bureau of Health Professions for compiling data, graphs and tables, and developing narratives to present this information on minority health professionals, and communities. Additional appreciation is extended to Marcella Murphy, Medical Services Assistant, Commissioned Personnel Operations Division, Public Health Service, Leonora Surosky, Secretary, Kathy Owens, Secretary, Stacey L. Williams, Clerk-Typist and Deborah A. Hunter, Clerk-Typist, Division of Disadvantaged Assistance, Donna Breslyn, Clerk-Typist, Division of Medicine and Pamela Dobson, Clerk-Typist, Division of Associated and Dental Health Professions for their clerical and secretarial support in preparing portions of this document. Special appreciation to Susan Eddins, Secretary, Office of the Director, Bureau of Health Professions, for her assistance in all support phases throughout the development of this activity and the preparation of the final report. William A. Robinson, M.D., M.P.H. Clay E. Simpson, Jr., Ph.D. Frank A. Hamilton, M.D., M.P.H. 378 INTRODUCTION BACKGROUND During the process of developing this broad study to examine the persistent differences in health status between the nonminority and minority U.S. populations, it became apparent that several cross-cutting factors were probably contributing to these differences. Some of these factors were the relative socio-economic status of the population groups; their dietary habits and nutrition status; the availability and accessibility of health facilities and personnel; occupational and environmental conditions; and others. At the direction of the Task Force Coordinating Committee, working groups and subcommittees were established to attempt to analyze the influences of these individual cross-cutting factors. This report was developed by the working group on Health Professionals, whose charge was to examine the importance of the availability of health professionals as a factor influencing the disparities in health status between nonminority and minority communities. Keeping in mind that other working groups would be addressing such issues as access to health resources, and the financing of health care as other important factors, this report was to address the following questions: • What are the variables concerning health professionals that create, enhance or foster differences between the nonminority and minority populations? • How do these differences subsequently contribute to the evidenced disparity in health status? • Why do these differences exist? In further consideration, the working group was also asked to respond to these questions: • "What are the patterns of (a) minority health professional development, distribution, and practice, (b) health professionals serving minority communities, and (c) minority health professionals serving minority communities?" • "What do health professionals expect of their clientele, and what does the client expect of the professional? Can expectations be meshed and if so how?" In determining the availability of health professional resources, the working group chose to attempt to analyze the following variables: (a) the numbers and types of health professionals appropriate to address the major health problems under study, and (b) the geographic distribution of these individuals in comparison to the location of minority population groups. 379 In an effort to make the analysis of "numbers, types and distribution" most meaningful for the diseases and conditions under study (e.g. cardiovascular disease, cancer, etc.), an initial attempt was made to develop a list of health professionals involved in providing health care for a sample patient encountering the health system with a single, reasonably common sample condition. Using a non-fatal traffic accident (violence) as an example, the list might include, among others: • Emergency Medical Technicians • Emergency Room Specialists and Nurses • Medical Records Staff • Clinical Laboratory and Blood Banking Personnel • Radiologists and Radiographers • Anesthesiology Specialists • Operating Room Nurses, Technicians, et al. • General and Specialty Surgeons • Intensive Care Nurses and Staff Nurses • Medical Social Workers • Physical Medicine and Rehabilitation Specialists • Occupational and/or Physical Therapists By extrapolating to include each of the health problems under study, in a variety of settings, it seemed apparent that virtually all of the different types of professionals would eventually have to be considered if a comprehensive study were to be undertaken. Such a study is planned. The multiplicity of personnel engaged in providing health care, while critical to the overall well-being of the patient also makes it particularly difficult to demonstrate that the role of any one (or few) is the keystone to improved health status. If the task had been to review the availability of health professionals in general, it would be comprehensive and require review of large amounts of data. Fortunately, systems providing such data on a national, regional, and often state-wide basis have already accomplished basic data collection and analysis functions. [One of the most comprehensive is the Bureau of Health Professions' Biennial Report to Congress on the Status of Health Professions Personnel.] The assigned task, however, was to analyze the availability of health professionals to the minority communities, thereby significantly complicating the review process. The first obstacle came in attempting to define a minority community in such a way that the subject minority group (Black, Hispanic, Asian/Pacific Islander or American Indian) could be analyzed within specific geographic boundaries for which there were also data on health professional resources. The system identified to facilitate the definition and analysis processes was the Bureau of Health Professions' Area Resource File (ARF). 380 The second obstacle to the task was encountered in trying to collect data on minority health professionals, especially data that could be analyzed in the context of the previously defined minority communities. Although statistics on the development of minority health professionals are reasonably available, information on their distribution and practice is poorly accessible, if at all available at the "community" level. The third, and perhaps most dominant constraint on this report was the time alloted to accomplish the work. As a result, rather than this effort comprising a definitive study, it constitutes the first phase or beginning of what will, hopefully, evolve into a more detailed and complete examination of the health professional resources available within minority communities. THE REPORT Because of the interrelationship of the two sets of questions posed to the working group, this report has been structured to address them concurrently. The first section of the report presents the conclusions drawn, while summarizing the highlights of the report. Needs for further intervention and/or monitoring are also described, within the context of "Recommendations." The second, and most quantitative portion of the report, presents descriptive information on the four minority populations (and "communities"), and the health professionals who serve them (both nonminority and minority). It discusses who they are and where they are, emphasizing differences between the minority and nonminority populations and their resources. The data included in this report were derived from numerous sources. Although there is some consistency for cross referencing information, most data sources generated their own terms, or groupings of minorities, or other indices. The third portion of the report attempts to focus on the more subjective parts of the task, including trying to answer the "how" and "why" questions, and discussing health professional/provider - client expectations. 381 PART I. - SUMMARY AND CONCLUSIONS This report represents the completion of the first phase of a study to examine the availability of health professionals to minority communities. The effort began with the examination of the minority groups themselves, first at the national, then state, and finally county level to attempt to define boundaries for a "minority community." The progression of the review from the larger to the smaller geographic units was designed to identify the largest unit which might reasonably be called a "community," for which health professional resource information was available or could be quickly generated. Because of the differences in the sizes and the patterns of distribution of the minority groups, no single standard for comparing all of them at the state/county level could be agreed upon. As a result, the working group arbitrarily selected to examine data for U.S. counties having a population of greater than 20 percent for any of the minority groups. [A lower density of five or ten percent was used for some Asian analyses.] The review focused on analyzing the attributes of those counties taken collectively within a state in comparison to the rest of the counties within that state. The cluster of counties was to serve as a proxy for a "community." The analysis of state/county level data was directed first towards the demographics of the respective minority groups, then towards the issue of health professional resource availability. Although the socio-economic and related characteristics of the target (key) counties were generally consistent with the national picture of these minorities (e.g. higher infant mortality and poverty rates), these features could not be ascribed specifically to the minority groups in question. Further, there was no mechanism for examining differential characteristics of minority sub-groups, particularly among Hispanics and Asian/Pacific Islanders. These same problems carried over into the review of resource availability. The methodology did not allow for full examination of variances due to urban versus rural environment; the specific influences of the presence of other minority populations; features of the general population; or several other potentially significant factors. Much of the discussion of the availability of health professionals focused on physicians. While inappropriate as a sole measure of resource availability, physicians were felt to be appropriate as a partial barometer, especially since the methodology was still evolving, and trends for other professions, where available, followed similar paths. Generally the data suggest: • Many high density minority counties have substantially lower numbers and percentages of health professionals than their low-minority counterparts. • The Indian Health Service plays a critical role in providing needed resources in many American Indian communities. 382 • Where high density minority counties have numbers and percentages of health professionals which equal or exceed the other counties, it is generally found to correlate with the location in those counties of significant health professions training resources. The numbers of health professional faculty usually overstate the availability of practitioners for patient care. • More specific data and analyses are required to answer specific questions on specific communities and sub-groups, however some more accurate generalizations can probably be achieved through re-structuring of existing county-level data analyses. Data on minority health professionals are quite mixed in availability, specificity and reliability. Student data are generally much better than that on practicing professionals, but both need to be more closely analyzed for relevance to the minority sub-groups and communities. With some notable exceptions among sub-groups of Hispanics and Asian/Pacific Islanders, minorities are substantially underrepresented among students and practitioners of virtually all major health and allied health professions disciplines. As a consequence they are poorly available as a major resource to assist their respective communities. The same patterns of paucity/scarcity persist among minority health researchers and science faculty in health educational and training institutions. This may be especially crucial and have negative consequences for rate and degree of progress being made in improving the health status of these minorities and their communities. Although definitive evidence is not available, existing studies of interpersonal and sociologic behavior suggest that the availability of health professionals who are from the same cultural background as their patients, may provide for a greater awareness of and sensitivity to the health-influencing factors which determine a positive health outcome. More study is required to evaluate the importance of this area to the health status of individual minority sub-groups. Some minority-specific summary comments are provided below. Blacks • The Black population, although perhaps not as diverse as some other minority populations, does encompass several sub-populations. These vary not only by socio-economic status, urban versus rural residence and educational attainment, as customary documentation suggests. They also vary by factors such as the homeland of their parents/grandparents (the U.S., Africa, the Caribbean), religion, family customs, etc. • In the 23 states and the District of Columbia, which provided the focus for this report, the heavily Black counties have larger poverty-level populations and larger number of Aid to Families with Dependent Children (AFDC) recipients per 100,000 population. The percentage of AFDC recipients was generally higher in urban Black counties. 383 • Black Americans have higher infant mortality than the overall population. In virtually all states, both in the key counties and elsewhere, Blacks had higher infant mortality rates than nonminorities. The single exception was for those Blacks living outside the key counties within the State of California, where Blacks had a marginally better infant mortality rate than nonminorities. • General hospital utilization figures show that inpatient days per 1000 population were higher in the key counties of 22 of 23 states. • More than half of the Black population of the target area lived in urban counties, which had professional to population figures slightly or substantially below national and national urban rates (with lower rates for counties with higher Black concentrations). Rural counties with Black populations of 20 percent or more had professional-to-population ratios (for all professionals) well below the figures for all rural counties and far below national averages. • Black-physician-to-Black-population ratios were significantly lower than the overall physician-to-population ratios for the key counties in all 23 states, although 19 of these had higher overall physician-to-population ratios than the remaining counties. Only in the District of Columbia and the key California counties were there as many as 100 Black physicians per 100,000 Black population. All other states were far behind. The picture appears similar for other health professions. Aggregate national data on dentists, Registered Nurses (R.N.), optometrists, and pharmacists also show consistently low Black-professional-to-Black-population ratios, well below professional-to-population ratios for the nonminority population. • The proportions of Black health professionals serving Black populations are not likely to change appreciably in the near future. In virtually none of the states considered does the percentage of Black graduates of medical, dental, and pharmacy schools (the three disciplines examined) approach the percentage of Blacks in the population. Thus, even if the numbers of Black graduates continue to rise, it is not likely that they will significantly alter Black professional-to-population ratios in the near future. • The relative paucity of Black and several other minority health professionals has been documented, with implications not only for their availability to provide care to their communities, but to participate as faculty in schools which train those professionals; and to engage as researchers and scientists in studying the problems which affect their communities. • All of the above suggest, as it did for other minority groups, the need for more specific, targeted attention to the individual Black communities to not only address the availability of health professions resources, but also the general adequacy of health services. 384 1980 Black Population in the U.S., by State 00 Ln All counties with less than 20% Black population At least one county with 20% or more Black population Hispanics The Hispanic population is generally concentrated in only nine out of fifty states. A critical factor which becomes problematic when reviewing the data on the "Hispanic communities" is that they are actually a group of communities brought together for analysis because of a common ethnicity and language base. There are important demographic, life-style and other differences among these communities, varying with origins in Mexico, Puerto Rico, Cuba, Europe and Central and South America. Each sub-population eventually must be examined as independently as possible to develop a true picture of availability and accessibility of health professionals. Further, because "Hispanic" denotes ethnicity rather than racial identification, collection of data relevant to these populations has lagged significantly behind some other groups. Cultural and language variations have either been difficult to define or considered not of sufficient importance for inclusion in many data collection efforts. Another factor found in analyzing data for the Hispanic community and health professionals may be the presence of significant numbers of individuals who are not citizens. In the field of medicine, for example, of the 3,655 Hispanic physicians in residency training in September 1983, more than 1,700 were graduates of foreign medical schools. The implications of the presence and contributions of these individuals, particularly those whose non-citizen status introduces other variables, should be the subject of further study. The availability of health professionals to Hispanic communities seems to present a mixed picture. The widespread variation in residence patterns ranges from dense urban populations in some states (California, Florida, Illinois and New York are some examples) to more sparsely populated rural areas of others (New Mexico, Texas and Colorado are other examples). Further, there are mixed patterns even within these states, which complicate the picture even more. Some areas seem to have large numbers of health professionals available, and in a few counties even large numbers of Hispanic health professionals. However it has not been shown that these health providers are geographically spread so as to be made available to all those Hispanics in those counties. More detailed review of specific county level data, followed by analysis of selected sub-county level data is necessary to provide clarity to the circumstance of the individual sub-populations of Hispanics. Mexican-American and Puerto Rican Hispanics appear to continue to be under-represented in the education and practice of various health professions but in many professions the picture is clouded by the possible over-representation of other Hispanics. More definitive data analyses are required. Other national, state and education data need to reflect the existence of the sub-populations which comprise the Hispanic minority, and the degree to which they have available resources to provide improved health and to pursue health professions careers. 386 1980 Hispanic Population in the U.S., by State 00 All counties with less than 20% Hispanic population At least one county with 20% or more Hispanic population Asian/Pacific Islanders • Demographically, Asian/Pacific Islanders tend to live in dense populations in a few states (Hawaii, California and New York), but in much smaller concentrations in several other states. In fact these three states are the only ones containing counties in which Asian/Pacific Islanders constitute more than 5 percent of the population. As a consequence, analyses of data at the county level regarding health status, numbers and types of health personnel, etc. were made ineffectual by the small population figures. • Hawaii is by far the largest single Asian/Pacific Islander community in the U.S. However, because of its size and its political definition as a State, it was deemed inappropriate to include that State for bases of comparison with other Asian/Pacific Islander communities. • At least three critical factors complicate the examination of data on Asian/Pacific Islander populations and communities. One is degree of heterogeneity which the term "Asian" masks. The residents of these communities have origins in a large number of countries and cultures from throughout the continent of Asia, and include many island nations of the Pacific. Many are well-known and live in larger communities (e.g. Chinese, Asian Indians, Philippinos, etc.) while others are lesser-known or more recent immigrants living in smaller communities (e.g. Laotians, Samoans). They encompass a broad spectrum by socio-economic status and other demographic variables. Most U.S. data collection efforts do not provide a mechanism for analyzing these sub-populations. • A second critical factor that is frequently encountered among the Asian/Pacific Islander communities is the geographic proximity to other minority communities such that the latter, and usually larger, minority group dominates the limited focus on "minority concerns." County level statistics are generally not discrete enough to provide meaningful information on Asian/Pacific Islander populations. Sub-county level data will have to be examined to obtain a realistic picture. This is particularly so when attempting to look for differences between Asian sub-groups. It is also important that eventually Asian/Pacific Islander sub-groups be compared to other Asian/Pacific Islander communities and not solely to the nonminority or large "other minority" communities. • A third factor which must be considered in the analysis of data on Asian/Pacific Islander health professionals and their communities, is the presence of significant numbers of individuals who are not citizens. In the field of medicine, for example, of the 5,632 Asian/Pacific Islander physicians in residency training in September 1983, more than 4,000 were graduates of foreign medical schools. The implications of the presence and contributions of those individuals, whose non-citizen status introduces other variables, should be the subject of further study. • In broader analyses, Asian/Pacific Islanders as a group appear to be disproportionately over-represented both in the education/training and practice of the health professions. Only to a limited degree have data been available to demonstrate which sub-populations contribute to that over-representation by profession. 388 • The data which have been analyzed are not sufficient to provide definitive statements on the availability of health professionals to Asian/Pacific Islander communities, especially to sub-populations of the "Asian/Pacific Islander" minority group. 389 1980 Asian Population in the State of California, by County Less than 5% 5% and over .__J------T 390 1980 Asian Population in the State of New York, by County LO Less than 5% 5% and over American Indians • American Indians represent the smallest of the four minority groups with total numbers, by 1980 Census count, of approximately 1.4 million. This represents a 70 percent increase over the 1970 Census, which undoubtedly is due to increased numbers of people declaring this identification, rather than the result of changes in vital events, i.e. births and deaths. • Counties having greater than 20% Indian population are spread over 10 States. In part this is a reflection of the location of reservations; otherwise it reflects strong tendency to live as community in single counties rather than dispersing. • Generally, in addition to being poorer, Indians suffer from increased health problems, such as diabetes and cirrhosis. They also have generally lived in communities where health professionals and other health resources were deficient. Fortunately the Indian Health Service has compensated for some of these deficiencies, in selected communities, but real problems remain. • Despite relatively small numbers there is a great deal of heterogeneity in the American Indian population, primarily due to the existence of more than 300 tribes. Customs and cultures vary widely. The nature and formality of the relationships between these tribes and communities, and local, State and Federal governments also show wide variation. To address the concerns of the individual Indian communities will require information on the specific make-up of the residents of that community and inevitably, sub-county and tribal data will have to be obtained and analyzed. This further suggests it might be appropriate to compare certain Indian communities to other Indian communities, and not solely to the nonminority or other larger minority populations. • Another major factor which must be considered in any review of Indian communities is the varying presence and role of the Indian Health Service. More time and specific data are required to complete the review of the role of this important resource. 392 1980 American Indian Population in the U.S., by State LO LO All counties with less than 20% American Indian population At least one county with 20% or more American Indian population PART II. - SELECTED VARIABLES INFLUENCING HEALTH PROFESSIONALS: NUMBERS, TYPES, AND DISTRIBUTION PART II. A. - Minority Populations and Communities The principal reason for taking a look at minority communities in the context of this working group's charge, was to define the boundaries for comparisons of the relative availability of health professional resources between minority and nonminority communities. Other reasons are that the demographic profile of the population groups and the environments in which they live should be key factors considered when analyzing the need for and use of various health professionals. Some of the differences between and among the minority populations, are being discussed in other working group reports. Nonetheless, some re-statement of the differences, and the demographics, is presented to facilitate the discussion of this topic. To define "community" for each of the four minority populations the first activity was to identify an appropriate sub-population of the total U.S. population of each of those groups. Listings of the States were generated, based on 1980 Census data, through the Bureau of Health Professions1 Area Resource File (ARF). A review of the data derived confirmed that the four groups were broadly spread throughout the 50 States, each with its own pattern of dense and sparse concentrations, and each different from the nonminority population. Table 1 displays the Distribution of the U.S. Population by State and by Race/Ethnic Category, while Table 2 displays the Distribution of the State Populations by Race/Ethnic Category. Tables 3A-D display Rankings of Twelve States Having Largest Population of Individual Minority Group, by Number and Percent. Because the statistics showed that only the State of Hawaii (and the District of Columbia) among the States had a significant enough minority population to have that state be possibly deemed a "minority community," a further analysis was conducted of county-level population data. Through further use of the ARF System, a listing was generated of all U.S. counties having a population of at least twenty (20) percent in any one of the four minority groups. The 20 percent level was selected as having enough of a critical mass to meet a general concensus definition of "community," with the exception of Asian/Pacific Islanders, as noted below. See Appendix II. A. Table 4 displays a listing of the 50 States showing the number of counties within each, with a population of at least 20 percent in one of the four minority groups. 394 H2: > — io—ooooo — oooo< KiKiM\i^oifi^oMfl^ff.oNe>r>ojff"- jo-oineMioMoajj- — ct- >rcM — -*CT->*}r>.in.««>inr<.rM-e.O(Mm xOCO'-ooi)>--io-(M-(\ioo»-o*o«)-in^-e »o ct- .3- -o — — m ct- r-* oo o — -o-j--o-eo — r~.<\i* — o«)-»'ifi*r^o^tooiN^io«ioioj-MoiOff'>-iOfg1r o»o ^r — 10 oo -o r^CM —^cm -o <*r in»o — «o *» — — ■*■ •* -o — «" •» — oCT-<\i,oCT-«><\jr^c\i«Ooooin«or-jmr^f-.pgr».oOt\i,,-aO*3 — MONMff'j-- ooN^rtKio»t3^in- — o— ^oiOMn- cm oo — oinr-j—m,r»o(\i<\ir^>j- — *j-c\i-r-oCT-o--o.oao<\ir~.mCT-NONOiri-- — r^-inNN'TOoo** <«o«or-* — inrtrs.CMCMCMr»*i CM- CM — l*>CM -^ CSJ CT- r»CM—— — CM •*> ifl -O to CM r»» 10 in — i«inK)»NeioiMK)9>eo4io^in4-e'e''iANN — O IOCM in—*»»» — — IOCM fs.— CM «- K> — — ►ho] — s2! O-»-OrOOrv(^ff.MO^)ff.(M|)Kl-^.O^00^OrOa'i/lC0«OO'Off'O^rvOO-^ff-Kl-JTN^|--INi^|ir-J o«j-oou"i — ooo — ooo — oo — ooooo(M(sio<=<\ioooo>o<\j-r— — — — o o o cm o f> — oo^omo •0«»T«»Or^CMinCM—*^C*^K»CM — NO-Ninifl«9»e4>-O<04'inoONeiO4'4'O«IM^4'NinNIO«9B>*JN«<« (-a) roomff'ino-(Oi7'-orsMM-ff.j->floro-orNO'-po-r<.o.OlOFOOOO(Mn^^lOO>.Off>no-»Ts^O>rs-lOKl('jrNO'0'0>irNJ--NMno«0-00(OKlf>.N * .«-.*-cm — — .»>' ro J-OvJ-OOK10K)000000 CT-rooOCT-mom — r~- o >j- in r- — r^. ct- oo ^ o- \ro<\ior^ — mN*» in— » ro cm n«. — o -o -o ms — 10 — ■or-or>»**CMK>CM*io*o — ro — ror%«»»CM»OCT>oooo>»T'^«o*rorOCT»- oo r- .o — -o -oo — cm (\iom*«ho« yo-ooor"">>j-r^cM*-CM-oCT-ooior^ — fs.CMf~.— **«>inCT-CM-««»T>.inr^r^CM>j>*c* — aiAinnn oo ct» — i>. r«. — o — &> *t*t*o — N«-*«Minio UI (M CT- m 00 — .OrO*CM* NO o o -o oo #on k»oo—cm *ro«ar ..»* — cm cm o>cm — oom o> *k> o cm o o» r^r^ o — — CM— — — — O— — C* — — OOvJ-CMOO-OTCM — — — — O— N*NO(MOOOOK)OMMOm- — mo— OCMinoOCMCMOCMO ^l^^o«u1MON^J^lrlMlrlu^^^.J^a-oo^^J«cou^(^l^l-OlO•-lOlsJ■o«)lOrO(Oooo^o^3^J•^^|l^lo•3^oo^ — uoMioinrv6>ooinooNoine7>in'tfo» — «o oo oNO>NO>iONoors*]irio^o4«3iooNK)«>«^No«4,o>NeiriM«oinKi<-oi04'4'HN«>oo>eK)o>«e4i lAtO'09'>>"M^>00 4'Omo<£rvlO^IOOrs^MOION«04>^4) rop»- *» «c cmct» — oor^oo^roo^o^p^oo — — o«-ino^Ki«Nooo>rK)iooeoo''>o^r>4'4' ooorocMtocMinoo*- — ct- ro ct- cm o oo — ro a» — — roooos*>'*»n».orr^«»>- ct-o-j»-«»>o>o*-oct- — M>«ororoincMr»«a* — * — «»• ■*• CT- CS1CSI O o — o — — o — — oosrcjooincM— — — — o— CM* — — NooseioeM\ief — mo— o cm w>o ocm — ono into — mincM**»«o»o«or in — iA«««i»Neemr*.mo«o-ooinrooio^-CM-or-»oomin*oooo- n^mso^ms oootn — roo^r«.io»oCMo«r>ro — cMor»r>»o.»or«.r«.i*>ooo>CMCT»- cm ct» r»« ooo^>o-o>o«^-o«i«Oot-.N>ofj'-oD-ior^--o^io--«ir>ifi inio — •»«»>r*c*f>»*«o^»'>*i*»«a9o»oK>oin«»»«tfr^CMino-«3<*ic»»oo«»»CM«o — cm r*. in >o ro r» — o— c* — — «n»m« •«-H5-a5^«00»l04'40'f(MO--^-lO.(\l-«3JOfJ^Oin«5inO'fJlO^^MO-(M\(^-^iog'0< TV CM It CM IO — in cm cm i»> ♦ — «orm t> » — r-in oiocm — «-«- — in* — ' ©ox •c « a o ui a. x>o oo ,__<_> < Of •< * CKXKUQ DCZZ K«(rtNlOZZ(nHKK< N<2332atKH^uzo33ZX 3««UUOOeU.OZHHHMXMJCCSIIEItZZZZZZZZOOOt.gtlAinKh3»3323 hOh«1(9 OX> tn zzz>zz xomuJxxujiA o->»i oh KKoo»o».o» on «0 • » — — -onr.ro— — — •> • oo m * m» — <«> o» * r» ro ro *to«M«or» — a}«o)Mi«- * o« mm •*• * r» k» * «n * r» r» » o»*m oo f* m »»» ,.> * •• m — »mcMo»omr»o> — .#*«>* ro cm <••*«••_-*- -r» ^ --CM- -0BOO*o-CMO- rorop^fMoo>omoO''<'OCMoooo"--CT-*fM-*CT.^>r^mM»>omr~.CMioo»tocM- r»mo*»ro*e>moror^oro«of'»OMj — CMoo*r^cMoo — ooro mj or- o — o>o*cMmr* — «>ioin*»incM*p««a»io- ro«o — — to — cm— CMoow>o*oto»- mCT--oooomo»or«. * o — — mo>r>» too — «o*-o>>ocMr>.o>m»o — r»-cM* — ©to*m*r>.OM»roor»Msrooororo*cMoroCT« — iocm* OlO *— **CM — m O r» ro ro <0 CM M> CM CT» r»CM— M»CM >o * m»o— to -*> •-•- **M» — — ©> — * * — CM >Ninso>o>^»iA»iA«o>oi pmCT>ro*mooioro*M>mmr»- — o CM oom— •• — oo— o — oooooo — ooooooo — O— •— OOOO — OOOOOO— O — CM • OB o OO.o*CMCMrOr*M>rvrO*CT>CT'— romr^O-rOroroCT-m — •oCT-CMrvOM>*roooCMOCM*»moo>rr--or-.CMoo*oCT-f^ — M»IOr^in— CMtO— OinCM — m*rOCMCMP»*— *CM*M>CT»CT>OMJOOCMr*mCT>M»«Om — — fv — mCMCM*MJOO** *M»rotncMOCM- -otomo-or-sCMrotomoo — oc*M»M»CMf»c*o — w»roror««j» — cm — r«.c*in*»roc*>CM*CMM>inio><*oo iocm in —*r* — —rocM r»» — -OCMO-o»mCT-f> — CM — CM — IO — — — CMr» — tOiOCM— *0>CMIOf»mr» — — — CM — O — ^— Ol oo-omooooooooo- 00090000000000*0 — 00000— 100m — 0900-000 — 00 — 00 — ■o*r-or-.cMincM — *m»o»m»iocm — cmo— CMinmoooo* — o>o*moor»oro**c>>CM**CMmcMroooCT>tor»>oOM) ro o in ct- mo — too* — onnw* t**«ooiOM»f^w» — oo- r^o**oooinoo-omincMMsroo*r-io — o >o CT«*M»^*r>.C»>*— o ms ro m ct-m — o — iOM»mr«.oo- OlO CM oto IO M»in — IOCM CM— —— — ~ *lO —IO — — 0*M» —— CIO — * in CM —mom — NO*lflN- rs. CT> 00 MS CT- CM — oJC»Mn» mt0M)r0*O — to* — CMO*t0— C*CMtO- — torn- *to*mcMM>9>*mr^rorooO — ^>ncMmtor>>to— CMr«CM- toro*ocMor^ — mr>*- m — M>M>mo*o>oOM>< rOCMr-v— OMJMJMJ — IO — *M»r»**CMlOCM*lO*0— IO— ior>»o»CMioe*otoo*r*to*ioioo»- ooc m> — M»- CM— — — »orof.rocM«oCT-ocMr-.>ocM*in — — rv — CMO>oomrom-o*to*mr^oorommoOM3*CT>moocMioo^*CT>CT> — M>*in ooO*r^CMinooOoOCT- (M*M>m0-toto 0000CT-OCT-Or0CT-M>O'-o oo < m < a. ui lu (j cmx V-X OOXr-i~ii-i < Li/Ihx<< k< xx tr> < < lu E -J — OQul OX>» m ooz ^O-OXi-i r-NOMiftO »X~»E>-XX ZOIAUZZUM tjE*?1 hBMN-«)ZZv1inh>1< r-hO3333«(0(MjUIZO33zy-il-l>-i**-lEEEEEEEEXXZZZZZXOOO 396 Table 3A Black Population •53% of Blacks live in the South •States with the largest Black population as percentag e of total 1. [District of Columbia] 70.2% 7. Maryland 22.7% 2. Mississippi 35.2% 8. North Carolina 22.4% 3. South Carolina 30.4% 9. Virginia 18.9% 4. Louisiana 29.5% 10. Arkansas 16.3% 5. Georgia 26.8% 11. Delaware 16.2% 6. Alabama 25.6% 12. Tennessee 15.8% •States with the largest number of Blacks 1. New York 2.41 mil 7. North Carolina 1.32 mil 2. California 1.82 mil 8. Louisiana 1.24 mil 3. Texas 1.70 mil 9. Michigan 1.20 mil 4. Illinois 1.67 mil 10. Ohio 1.08 mil 5. Georgia 1.46 mil 11. Pennsylvania 1.05 mil 6. Florida 1.34 mil 12. Virginia 1.01 mil Table 3B Hispanic Population •More than 60% of the Hispanic popul ation lived in three States, California, Texas and New York. •States with the largest Hispanic population as percentage of total 1. New Mexico 36.6% 7. Florida 8.8% 2. Texas 21.0% 8. Hawaii 7.4% 3. California 19.2% 9. Nevada 6.8% 4. Arizona 16.3% 10. New Jersey 6.7% 5. Colorado 11.8% 11. Illinois 5.6% 6. New York 9.5% 12. Wyoming 5.1% •States with the largest number of Hispanics 1. California 4.54 mil 7. New Mexico 0.48 mil 2. Texas 2.98 mil 8. Arizona 0.44 mil 3. New York 1.66 mil 9. Colorado 0.34 mil 4. Florida 0.86 mil 10. Michigan 0.16 mil 5. Illinois 0.63 mil 11. Pennsylvania 0.15 mil 6. New Jersey 0.49 mil 12. Massachusetts 0.14 mil 397 Table 3C Asian/Pacific Islander Population •60.0% of Asian/Pacii :ic Islander population lives in the Pacific Division of the West •States with the largest Asian/Pacific Islander popula tion as a percentage of total 1. Hawaii 61.2% 7. Maryland 1.60% 2. California 5.6% 8. Oregon 1.60% 3. Washington 2.7% 9. Illinois 1.50% 4. Alaska 2.1% 10. New Jersey 1.50% 5. Nevada 2.0% 11. Utah 1.40% 6. New York 1.9% 12. Virginia 1.30% •States with the largest number of As ian/Pacific Islanders 1. California 1.31 mil 7. New Jersey 0.11 mil 2. Hawaii 0.60 mil 8. Virginia 0.07 mil 3. New York 0.33 mil 9. Pennsylvania 0.07 mil 4. Illinois 0.17 mil 10. Maryland 0.07 mil 5. Texas 0.13 mil 11. Michigan 0.06 mil 6. Washington 0.11 mil 12. Florida 0.06 mil Table 3D American Indian Population •50.7% of American Indians live in the West •States with the largest American Ind ian population as pe rcentage of total 1. Alaska 16.0% 7. North Dakota 3.1% 2. New Mexico 8.2% 8. Nevada 1.8% 3. South Dakota 7.0% 9. Wyoming 1.8% 4. Arizona 5.7% 10. Washington 1.5% 5. Oklahoma 5.7% 11. Utah 1.4% 6. Montana 4.8% 12. Oregon 1.2% •States with the largest number of American Indians 1. California 0.23 mil 7. Washington 0.06 mil 2. Oklahoma 0.17 mil 8. Texas 0.05 mil 3. Arizona 0.15 mil 9. Michigan 0.04 mil 4. New Mexico 0.11 mil 10. New York 0.04 mil 5. North Carolina 0.07 mil 11. Montana 0.04 mil 6. Alaska 0.06 mil 12. Minnesota 0.04 mil 398 Table 4 Nimber of U.S. Counties With Over 20 Percent Minority Population, By State TOTAL NO. OF COUNTIES WITH >20 PERCENT AMERICAN ASIAN/PACIFIC COUNTIES BLACK HISPANIC INDIAN ISLANDER ALABAMA 67 37 ALASKA 23 ARIZONA IA 8 3 ARKANSAS 75 27 CALIFORNIA 58 11 1 COLORADO 63 12 CONNECTICUT 8 DELAWARE 3 DISTRICT OF COLUMBIA 1 1 FLORIDA 67 13 1 GEORGIA 159 108 HAWAII 4 4 IDAHO 44 ILLINOIS 102 4 INDIANA 92 2 IOWA 100 KANSAS 105 1 KENTUCKY 120 1 LOUISIANA 64 46 MAINE 16 MARYLAND 24 10 MASSACHUSETTS 14 MICHIGAN 83 1 MINNESOTA 87 MISSISSIPPI 82 65 MISSOURI 115 2 MONTANA 57 5 NEBRASKA 93 1 1 NEVADA 16 1 NEW HAMPSHIRE 10 NEW JERSEY 21 1 1 NEW MEXICO 32 28 3 NEW YORK 62 3 2 NORTH CAROLINA 100 56 2 NORTH DAKOTA 53 3 OHIO 88 1 OKLAHOMA 77 OREGON 36 PENNSYLVANIA 67 1 RHODE ISLAND 5 SOUTH CAROLINA 46 40 SOUTH DAKOTA 67 TENNESSEE 95 8 TEXAS 254 27 70 UTAH 29 1 VERMONT 14 VIRGINIA 98 47 WASHINGTON 39 1 WEST VIRGINIA 55 WISCONSIN 72 1 WYOMING 23 399 Thirty-seven states have at least one county with a single minority's population of 20 percent or greater. In eight states, over half the counties have a single minority's population of 20 percent. The number of states* having at least one county containing a "20 percent or more" minority group are: Black (23), Hispanic (10), American Indian (9) and Asian/Pacific Islander (2). Note: Some states have more than one type of minority county with greater than 20 percent of that minority population. Further, because of the relatively few counties with a 20 percent or greater Asian/Pacific Islander population, a lower density of 10 percent or even 5 percent was used in some analyses. Minority County Population Distribution Of the 3,077 total U.S. counties, 1,265 (41 percent) are urban, 1,157 (37.6 percent) are rural, and 655 (21.3 percent) are metropolitan, where "urban" denotes that not more than 25% of the population lives outside a city, and "rural" denotes that not more than 25% of the population lives within a city, and "metropolitan" encompasses all others. The counties with large minority populations are distributed among metropolitan, urban and rural counties somewhat differently than are the counties without large minority concentrations. The smallest differences occur in heavily Black counties. Counties with heavy concentrations of Blacks (20 percent or more) are predominantly urban in nature, more so than are the counties with smaller concentrations. Over 45 percent of the 503 counties with heavy Black concentrations were urban as compared with 40.2 percent of the counties without such concentrations. Heavily Black counties were somewhat less likely to be in metropolitan or rural areas than their counterparts. Among Hispanics, differences were more widespread. Counties with a heavy population of Hispanics were less than half as likely to be in metropolitan areas as low-Hispanic counties and significantly more often were found in rural areas. Almost half (49.9 percent) of the Hispanic counties were rural counties, while only 38.4 percent of the non-Hispanic counties were rural. American Indians were distributed in much the same way (by county) as were Hispanics. Nearly 47 percent of the counties with high density of Indians were rural, as compared with 37.5 percent of low density Indian population counties that were rural. American Indian counties were far less prevalent as either metropolitan or urban areas. By way of contrast, Asian/Pacific Islander population centers were heavily concentrated in urban areas. The distribution of the "high minority population counties" within these sub-groups is provided in Table 5. * Includes the District of Columbia. 400 Table 5 Distribution of High Minority ( 20.0%) Counties by Minority Group and Location Total Met # ro % Urban # % Rural # % |U.S. Counties 3077 655 21.1 1265 41.1 1157 37.6 |Black |A11 Other 503 2574 99 556 19.7 21.6 229 1036 45.5 40.2 175 982 34.7 38.2 |Hispanic |A11 Other 154 2817 15 640 9.7 22.7 63 1202 40.9 42.7 76 1081 49.4 38.4 |American Indian |A11 Other 32 3045 5 650 15.6 21.3 12 1253 37.5 41.1 15 1142 46.9 37.5 IAsian/Pacific | Islander | (Over 10%) I All Other 6 3071 0 655 0 21.2 4 1261 66.7 41.1 2 1155 33.3 37.6 The variations in these statistics between the four minority groups reinforces their individuality and the need to review separately their communities and their health professional resources. Consequently the following discussion reviews in turn, information derived from the Area Resource File pertinent to the "high minority population" counties for Blacks, Hispanics, Asian/Pacific Islanders and American Indians. For each minority group, the discussion presents demographic characteristics followed by data and discussion of the degree to which health professionals are available to that minority group. In developing a descriptive picture of the minority communities, data compiled from the ARF which may affect health outcome were analyzed. The data selected for review by this group were the following: 1. Personal income 2. Infant Mortality Rate (IMR) 3. Aid to Families with Dependent Children 4. Level of poverty 5. "Health Manpower Shortage Area" and "Medically Underserved Area" 6. General hospital utilization 7. Physician manpower and distribution 8. Distribution of other health professionals This section of this report is divided according to the four ethnic/racial groups; Blacks, Hispanics, American Indians and Asian/Pacific Islanders. It utilized the aforementioned selection factors to evaluate the gross health and economic status of minority communities and the relative availability of health professional resources. 401 The second part of this section examines data on minority health professionals — their development, distribution and practice. NOTE TO THE READER It is critical at this point to remember that the process for fully delineating the various minority communities could not be completed within the time constraints of this study. As a proxy measure, the working group has chosen instead to analyze data available, primarily from the Area Resource File, on those counties or groups of counties within various states, if they had a 20 percent population of a minority group. [The exception regarding Asian/Pacific Islanders was noted above.] There would be many flaws in attempting to directly compare events or circumstances in these analyses to, for example, a specific Chinese-American or Puerto Rican community in a given city or county. Nonetheless, an effort was made to proceed in reviewing the county-level statistics to see if any consistent patterns or trends might be identified among the several study variables by state, regional or national perspectives. It was also felt that this examination of county-level data represented a necessary preliminary step (just as the prior review of national and state data had been), in presenting what can evolve into a more definitive analysis based on review of other inter-county and/or sub-county level data. PART II. A. 1. - Black Communities and Health Professional Resources This section of this report will focus on the Black community by documenting various differences between densely populated counties (i.e. greater than 20% of the population being Black) and the rest of a particular state. For this presentation, 503 counties, out of a total of 3077 U.S. counties, from 23 states and the District of Columbia were included as the target or "key counties." The population of these counties and the District of Columbia consisted of 15,797,000 Black Americans or 60 percent of the U.S. Black population. Among States containing counties with a Black population greater than 20 percent, there was a broad range of variance from one county in one State to 108 counties in another State. The following States illustrate the broad range of Black residents: one county each in Kansas (41,280); Michigan (829,990); New Jersey (320,827); Ohio (340,046); and Pennsylvania (638,064), to such large representation as Arkansas with 28 counties (326,419); Alabama, 37 counties (820,050); South Carolina, 46 counties (841,632); North Carolina, 40 counties (1,104,506); Louisiana, 46 counties (1,064,228); Mississippi, 65 counties (769,860); and Georgia, 108 counties (1,335,170). The states having the greater numbers of key counties are primarily "southern" states accounting for 487 of the 503 key counties. However the key counties in the remaining states still account for 6.5 million of the 15.8 million Blacks in all these counties. See Table 6 for the full listing of these 23 states. 402 M CJ CO rH PQ 3 o •H 4-1 CO rH 3 CL o P-. OONvOvOOO^fCNOuOOOMOOOOr^vOvOvO^CNO^O ifliCinNfOMJiNOOCONHCfivOHNNO^vOrOOCOsr O^vocoooiHCMOcNiocMt^cricor^oor^iooovooo^o OS#t0\OS0s0SOSOsOS0S0SOS0SOSOs0\ov0\0S0SOSOS0S0' Ov0l^l^CX3in>a-CMrHt^>*r^c^C^CNO0N!tHvO^CMvDHCMOOJM OOCOO-d-CM ONCONCMcn^HCOvOOO^CMrs o Ok t^ as m CO CM 3 •H CO CO cu •H •H ^J 4J B 3 3 3 rH O O u u X) 4-1 01 O 4-1 vO a 4J cu o cu rH •H rH cu U X cn 4-1 CO CO H 3 ■H •H Q 3 13 o 3 •H CO 4-1 CO CO rH cu 3 4-1 a CO o 4-1 PM C/3 ^ U CO CO c cu o •H •H 4-1 4-J 3 CO 3 rH O 3 u a o >> P-i cu ^ mcOsOtNHmMr\O^Hooino'ONNa\Noo(N^roo ^HNOvOsO^H-stiAsOcgiO(N--tNaia>(NNiO«tCO- ■H r>> 4-1 OJ c M 3 o a COvOOOOOOCMOOCMr^ sjinnmio>jooiNvO COOrHvOOvOmCMrH OS OX OS OS OS OS OS rH 00 rH CO "O rH CN OOr^OOCOOOrHOO-^-COOOrHOOOl^. oocT>iniNst 3 cu o M cu p* o II m -J- CO rH rH CM rH r^ m vo sO MO r^c)>cMCMinO'-ju.vOin>si-oo sOsOOOAOCMsDCslOOCOHcNvOOOOvO^ONiOCft rH rH rH rH rH rH CM co O rv oo cm CO CM CO 00 . O 3 O •H CO P- at r-l CO M cu Ui c CO 5*. 3 TJ 3 •H •H CO ^ CO > CO CU CO CO CO u CO CO •H cO ^ CO 3 cO CO r-l M r-l U rH CJ CO •H B CO o T3 •H O C CO O •H cO 60 CO 3 cu o J^ CO 3 CO 3 MH •H 00 c CO co 3 CO rH -H •H O •"3 >* rfi CO X cu CO •H X CO •H . r-l M •H •H CO 4-1 •H S^ ,3 CO CO 4-1 o 3 4-1 3 cO 00 CO r±& rH CJ O O rH X) 3 3 3 M o CO CO & £ u •H 3 3 3 X M rH u CO rH CU rH 3 CO 0) O ££ $ •H CU CU o rfi cu 0 Demographics of the Black Counties/Communities In each state, a review was made of several ARF indices, to help define differences between the target counties and the counties which comprise the rest of the state. If a differential need for health resources were to be established between the high and low density Black communities, there should be some related factors which could be examined to lend supportive evidence. One available measure for comparison of general hospital utilization in "Inpatient Days." In analyzing the inpatient days per 1,000 population among the key counties, there was a significant difference between the densely Black communities and the rest of the state. Table 7 clearly shows this gap between the key counties and the rest of the states. It suggests that Blacks were more likely to spend more time in the inpatient hospital setting at a rate of up to two to five times that of their counterparts in the rest of the state. Only Florida, and questionably Texas, do not support this trend. This may mean that the Blacks have more serious illnesses or more complications related to their illnesses. Table 7 Inpatient Days - Per 1,000 Population (1982) State Key Counties Rest of S Alabama 1,691 1,248 Arkansas 1,836 1,171 California 1,181 848 District of Columbia 2,828 __ Florida 1,220 1,486 Georgia 1,542 864 Illinois 1,686 1,161 Indiana 1,908 1,106 Louisiana 1,447 807 Maryland 1,889 635 Michigan 1,659 1,082 Missouri 5,578 1,162 Mississippi 1,608 1,123 New Jersey 2,141 1,154 New York 2,045 1,305 North Carolina 1,325 1,097 Pennsylvania 2,002 1,317 Ohio 2,150 1,313 South Carolina 1,213 1,047 Tennessee 2,181 1,320 Texas 1,364 1,320 Virginia 1,572 977 404 From available national data, individuals of all racial and ethnic groups from the lowest socioeconomic levels have the highest mortality rates. However, the level of poverty among Blacks is most striking. For example, in New York State, the rate in the key counties was 23.7 percent and in the rest of the State 9.0; in Missouri 21.8 percent and in the rest of the State 10.8; in Pennsylvania 19.9 in the key counties and 9.0 for the rest of the State; in Illinois 13.6 in the key counties and 8.1 for the rest of the State; in Indiana 10.8 in the key counties and 9.0 for the rest of the State; and in Ohio 11.2 in the key counties and 9.9 for the rest of the State. It appears that the differentials in poverty rates were much closer in the Midwestern and upper Central States than in the Eastern and Southern States. See Table 8 for these data. Table 8 Percent Below Poverty Level (1979) Differential Percentage Key Rest of Point Percent State Counties State Differential Difference Alabama 20.3 16.1 4.2 26.1 Arkansas 21.0 16.7 4.3 25.7 California 13.3 10.2 3.1 30.4 D.C. 17.3 — — — Florida 17.9 13.1 4.8 36.6 Georgia 18.8 11.2 7.6 67.9 Illinois 13.6 8.1 5.5 67.9 Indiana 10.8 9.0 1.8 20.0 Kansas 13.7 9.6 4.1 42.7 Kentucky 19.3 17.2 2.1 12.2 Louisiana 20.6 13.4 7.2 53.7 Maryland 14.4 6.1 8.3 136.1 Michigan 14.0 8.9 5.1 57.3 Missouri 21.8 10.8 11.0 101.9 Mississippi 26.1 16.3 9.8 60.1 New Jersey 17.5 8.3 9.2 110.8 New York 23,7 9.0 14.7 163.3 North Carolina 15.7 12.3 3.4 27.6 Ohio 11.2 9.9 1.3 13.1 Pennsylvania 19.9 8.6 11.3 131.4 South Carolina 17.6 11.5 6.1 53.0 Tennessee 17.3 15.7 1.6 10.2 Texas 16.5 14.5 2.0 13.8 Virginia 14.0 9.3 4.7 50.5 405 Another reflection of the level of poverty in a community is the rate of Aid to Families with Dependent Children (AFDC) per 100,000 population. Table 9 shows that the rates of AFDC were two to three times higher among the key counties as compared to the rest of the state. Differences in the rates of AFDC between the key counties varied according to their geographic designation of rural versus urban. In reviewing the 23 states, and the District of Columbia, which contained the key counties, the distribution of families receiving AFDC is concentrated in non-rural areas. There were only four states (Florida, Mississippi, Texas, and South Carolina) which had a greater proportion of rural population in the key counties for Blacks than in the rest of the state. Table 9 Aid to Families with Dependent Children Rate Per 100,000 Population and Percentage Rural (1980) Key Count ies Rest of State States AFDC/100,000 % Rural AFDC/100,000 % Rural Ratio Alabama 6030.9 31.1 2646.6 52.9 2.3:1 Arkansas 5734.0 35.5 2036.0 59.3 2.8:1 California 6835.2 1.0 5355.4 12.7 1.3:1 Dist. of Columbia 15,530 0.0 — __ __ Florida 4785.0 20.8 2317.0 15.1 2.1:1 Georgia 5096.8 31.0 1786.6 51.1 2.9:1 Illinois 9234.0 1.1 2564.4 31.3 3.6:1 Indiana 5302.5 2.0 2030.4 46.1 2.6:1 Kansas 8315.2 0.9 2371.1 35.9 3.5:1 Kentucky 5788.2 30.5 4528.7 49.5 1.3:1 Louisiana 5994.8 30.5 2622.3 33.3 2.3:1 Maryland 9,385.9 13.6 1922.4 24.0 4.9:1 Michigan 12,379.5 1.6 5471.0 38.6 2.3:1 Missouri 14,343.0 2.7 2880.6 35.0 5.0:1 Mississippi 8,231.2 55.5 3119.1 45.2 2.6:1 New Jersey 13,631.2 0.0 5309.1 12.4 2.6:1 New York 10,778.2 0.0 4561.6 21.2 2.4:1 North Carolina 4294.9 44.3 1761.8 64.9 2.4:1 Ohio 6973.2 0.3 4317.9 30.9 1.6:1 Pennsylvania 15,870.6 0.0 3526.6 35.8 4.5:1 South Carolina 5679.0 47.7 2198.4 40.0 2.6:1 Tennessee 5666.2 11.1 2454.8 52.8 2.3:1 Texas 2869.9 42.6 2100.0 18.7 1.4:1 Virginia 4561.8 28.3 1779.6 38.8 2.6:1 406 In all target 23 States and the District of Columbia, the infant mortality rate was higher among Blacks than nonminorities. The range was 23.7 per 1,000 for Blacks in the key county versus 9.7 in the rest of the State in New Jersey for a differential deficit* ratio of 1.44 or 144 percent or in Ohio 28.5 for Blacks in the key county and 12.5 for nonminorities for a differential deficit ratio of 1.28 or 128 percent to 19.8 for Blacks in the key county in Kansas versus 16.3 for the rest of the State for differential deficit ratio of .176 or 17.6 percent. The range of infant mortality rates in Kentucky was 28.6 for Blacks in the key county to 18.9 for nonminorities. In the rest of the State the rate for Blacks was 21.3 and for nonminorities 11.9. The lowest rates for Blacks were in New York (19.6), North Carolina (19.7), and California (19.7) for the key counties, while for nonminorities it was (14.5), (12.0); and (16.7) respectively. Table 10 shows that infant mortality rates were demonstrably higher for Blacks in comparison to the nonminority population whether in urban or rural high density or lower density settings and in no case in any of the areas was it lower than the nonminority population. To this point, each of the demographic features examined was consistent with a picture of lower health status. The working group felt that at least so far, the groupings of counties as a proxy for "communities," had not obscured some major factors known to be relevant to health status. The effort moved on to the examination of health professions resources in these same counties. The percent that is necessary for the Black population to gain parity if the nonminority rate remained stable. 407 3 If 13* 4> O 00 C^N)ln^N00wiaiui4>O0!)OU)VD4>sO0ClU(J>Or-'MH MMMtototototo^toiotototototototo^rotol-'toio co co uj \o h o o o ^ m m !o ^ a b c^ 4> w b oo 'm b b o^ p p p 6 p p K g g P p 6 P 6 K vo g p g p p 6 g p sO KJ sD CO sO Im H a U H H C> ON Oi O In Ui M CO 00 CO 'eft U) UiC>VDNvflC^C^C»C»Nja)4>HN4>N(^4>U)HwLn OsVOC»WMOO4>HljJN3P<^OVOC0UrOC0b00VDV0 OO^WKJOOsC^WNJlO^PsDCOvDUONOHavOOH OOVOOO^^Mt^H^vOaOlCO^COOltOOOMMvjbtJ* pppppKp£voio£g!i>gooo^Plo££ PC»VOHOO^O^slHs>)COs7iWVDlos^4>OON)COMLnNl Pfcpo^ro^GPIo'PGPgullo'vogKgS, g££ • • ....... ........... ... ^O^MMOv/iOO"«JtOUJ*«J0>0>*^l00UJ00Ui Ov UJ O b,fc!H!,h'NKNKfo^NNHN3rON:N)MN3N5 t_i |_i kj ^OOMyiWOHOHvONPcA^HHLoONlvL) I O Vt5 H OONlnH4>0>CJiUil>)r0 4>l-' Ui W l) H H '^O C^ O M VO U) P p C g P P p vo g P g g g g p vo g P vo P , £ & P M H N3 H l^ O 4> 00 4> O CO ^ N3 VO VO H W to In VO b>VOO> Availability of Health Professionals In examining the availability of health professional resources, it was apparent that more information was available regarding physicians than any other health providers. For the sake of brevity, much of the discussion which follows will focus on that data, with reference to other health disciplines as time and space permit. The working group did not feel this to be inappropriate since the methodology was still evolving. A more equitable presentation by discipline is anticipated to be forthcoming. To review the availability of physicians to minority communities, a reference point was needed. Since the county data was based on the 1980 Census, it was felt appropriate to use the national physicians per 100,000 population ratio for 1981 as an index. That figure was 186. In 12 of the 23 states the ratio of physicians per 100,000 population in the key counties was lower than the national figure. Generally these were the same states previously described as "Southern," containing the largest number but also least densely populated of key counties. Where the ratio exceeded the national figure, a cursory review of the states suggested a correlation to the location of medical schools. This could significantly overstate the numbers of physicians assumed to be available directly for patient care. A review of physicians per 100,000 population indicated that in all but four of the 23 States (District of Columbia excluded), the number of physicians per 100,000 population were greater in the key counties than in the rest of the State. These exceptional States were Florida, Kentucky, Michigan and Texas. In the other States the number of physicians per 100,000 population in the key counties ranged from 103 in Texas and 108 in Georgia to 402 in New York and 635 in Missouri. When comparing the number of Black physicians per 100,000 Black population (excluding California, where insufficient records are available, and Kentucky where the rate was too low to compile), the range was 7.8 per 100,000 population in Mississippi to 16.2 in South Carolina to 69.3 in Illinois, 70.5 in New York and 105 in the District of Columbia. In none of the States except the District of Columbia did the ratio of Black physicians per 100,000 Black population exceed 40 percent of the national ratio for all physicians to population. See Table 11. There were no generally consistent patterns observed in the change in physician population comparing the key counties to the rest of the State, across various states, between 1970 and 1980. The increases or decreases were not consistent with the population growth nor its decrease. For example in Alabama, the physician population grew by 69 percent in the key counties and 66.3 percent in the rest of the State, which indicates similar growth. In Florida, physician population increased by 81.5 percent in the key counties and by 109 percent in the rest of the State. 409 ro 3 3 co h-1 < Cu 3 CU 25 ro <-i ro H CO ro CO OQ co OJ H- 3 T3 H« 0) O •-< P4^ ro oj 3 3 rt CO 3 OJ O CO OJ OQ 3 ro o M h> ro co co o o HHNJHN)WH4>N)a\HHWHH MOCM000W^O^4>H4>L0MI-' UJUJOOOv/i4>UJtOUJV/iUJOOUJ"»-JUJ H O hrj O I—' a> r—• »-»« M O O CO H> rl rl rt 3 oo h- n O H« P- H« H-m si n CO rt O Hi n o 3 3 cr H« 03 LOrONHHWHHrOHH tOrO4>O<-n0Nt-'tO00Lni-n OUJUJOOOtOVOUJONUJVO ■P- r—i O M H1 r-1 r-1 r-* t-1 IO r-1 (-* l-'tO NJ H I-1 M Ui4>^OOLOUlUOC^OMONHsOHWsON3slCO | slCONlM4>UlOMOOCOWVOCOvON3(JNvOONOluJ I VO VO N 00 4> Ui IO tOMt-'tOtOVOtO 4> 4> H LOMOMv>COOvOlOHvOONM4>VOZ WC^HCOHWN3MOCOOVOCOlfl> V0 4> 4> H 00 CO 4> ON (3> 4> O O to f-1 ^«J to I ON M I Ln vo U3UJV/ir-itOONtO^JON4> On CJ\ H WN30NWN304>N) tOtO PHN30COOONOMONMCOOv02UCOVOW4>UiHH I O^J ..............>........I • • H4>0NN30>HvlOl4>4>COH00O VOOM>)vON)OHW 4>tO ro o o 3 3 rt H« ro co M OJ rrj o ro K H rrj M O O ID O 3 " M O 03 O rt O H» O 3 rtf 3d >-< CO n H- 03 3 CO P- ^J CO H- o H- 03 3 CO I-1 03 o H- 3 rt hri P- 3d ro ^ CO IO H- U3 O H- C/3 03 rt P 03 CO rt ro hd CO ro H e-> r-> o o H> o CO •4 rt o M o H« o n rt rrj o o V i-h 3 M n 03 o rt m H> 3 O 3 3 cr H- 03 H 03 CT In Georgia, the physician population grew by 61.4 percent in the key counties and by 102.5 percent in the rest of the State. In Illinois, from, the physician population in the key counties grew by 34.7 percent compared to a growth of 66.7 percent in the rest of the State. In Louisiana, the population in the key counties grew by 47.5 percent and by 117 percent for the rest of the State. In Maryland, the growth of the physician population in the key counties increased 31.6 percent, and increased by 136 percent in the rest of the State. Other Health Professionals per 100,000 population: Examination of the distribution of health professionals in key counties (greater than 20% of population was Black) in all the twenty-three states and the District of Columbia showed a higher rate of health professional per 100,000 population than the rest of the state. The exception to this finding was the State of Kentucky. See Table 12. These figures for pharmacists, registered nurses, dentists and optometrists follow the trends of the physician data and are also felt to be influenced by the locations of their respective educational institutions. The raw data necessary to investigate this and related questions has been compiled and will be analyzed in the near future. 411 Table 12 Health Professionals Per 100,000 Population STATE M.D.s PHARMACISTS R.N.s D.D.S.s OPTCMETRISTS Rest Rest Rest Rest Rest Over- Key of Over- Key of Over- Key of Over- Key of |Over- Key of all Counties State all Counties State all Counties State all Counties State 1 all Counties State 1. ALABAMA 125 159 75 58 63 50 293 345 215 30 34 24 1 6 6 6 2. ARKANSAS 121 153 94 58 64 54 268 319 223 29 31 28 1 10 10 10 3. CALHOvNIA 229 286 198 30 34 29 410 443 392 56 57 55 1 H 12 11 4. DIST of COIIMBIA 544 - - 74 - - 896 - - 75 - - I 7 - - 5. FLORIDA 179 150 183 39 38 39 486 401 497 41 35 42 1 8 6 8 6. GEORGIA 141 174 75 42 44 36 360 433 204 36 40 26 I 5 6 5 7. TTITNOIS 183 243 126 41 45 38 517 530 503 48 54 43 1 11 12 10 8. INDIANA 128 223 99 52 77 45 428 559 386 39 49 35 1 H 9 11 9. KANSAS 155 299 144 45 50 45 510 669 497 41 41 41 1 10 5 11 10. KENTUCKY 132 114 132 47 37 47 337 409 335 37 30 37 8 8 8 11. LOUISIANA | | 155 178 99 37 I i 41 29 292 325 210 34 35 32 I 5 6 5 1 12. MARYLAND | 1 1 266 | 1 1 333 | 1 I 218 | 1 1 1 1 44 | 1 1 61 32 475 504 455 50 53 48 1 6 7 5 Table 12 (continued) Health Professionals Per 100,000 Population STATE M.D.s PHARMACISTS R.N.s D.D.S.s OPTOMETRISTS Rest Rest Rest Rest Rest Over- Key of Over- Key of Over- Key of Over- Key of Over- Key of all Counties State all Counties State all Counties State all Counties State all Counties State 13. MICHIGAN 158 169 154 57 63 55 452 493 438 49 46 50 10 8 11 14. MISSISSIPPI 106 117 76 57 60 50 273 280 252 28 29 25 6 6 7 15. MISSOURI 162 685 106 59 91 55 447 890 390 44 136 33 9 11 9 16. NEW JERSEY 189 298 175 44 52 43 480. 574 469 60 70 59 6 6 6 17. NEW YORK 267 405 214 41 44 39 566 573 563 67 74 65 8 9 7 18. NORTH CAROLINA 151 163 130 46 46 46 433 474 363 33 33 33 6 7 9 19. OHIO 162 296 140 37 38 37 495 598 478 45 64 42 8 5 9 20. PENNSYLVANIA 186 341 161 48 72 45 615 594 619 50 56 49 10 12 10 21. SOUTH CAROLINA 132 138 115 57 56 61 351 360 321 31 30 33 7 6 9 22. TENNESSEE 156 272 102 52 64 47 344 494 272 42 56 35 9 11 8 23. TEXAS 151 105 155 43 44 43 310 260 314 37 36 37 7 6 7 24. VIRGINIA 171 181 163 47 50 44 424 458 395 42 44 40 7 7 6 PART II. A. 2. - Hispanic Communities and Health Professional Resources Inferences about health facilities and manpower resources available to Hispanic populations must be made cautiously. The basic area profiles of the Bureau of Health Professions' Area Resource File (ARF) provided basic data on health facilities, manpower resources, population, and income of Hispanic and non-Hispanic areas. These were supplemented by data on the size of the Hispanic population and the aggregate number of Hispanic physicians available in the target counties. However, in the absence of Hispanic-specific data on incomes, rates of utilization, numbers of other health professionals, and the like, one cannot necessarily conclude that Hispanics have the same access to resources or use these resources to the same degree as the general population of the target areas. This is particularly true since Hispanics are a minority, although a sizable one, in all but one of the areas examined. Further, there was no information in this data base on the extent to which these Hispanic populations and health professionals were of Mexican-American, Puerto Rican, Cuban, or other Hispanic origin. Therefore, no judgements have been made about the degree to which Hispanic communities have access to health professionals with similar ethnic and cultural backgrounds or whether this is even a significant consideration. Following the process outlined earlier, the effort was continued in comparing the key counties and the rest of the state for the high density Hispanic counties. Because there are only nine states in the U.S. having counties with a greater than 20 percent Hispanic population, some discussion of each of those states is provided below. Florida, New York and New Jersey, contain Hispanic counties which are overwhelmingly urban; California, has urban, rural and mixed areas; the southwestern and western states of Arizona, Colorado, New Mexico, Texas, and Washington contain primarily rural counties. Few common factors cut across each of these communities; perhaps the most striking is that except for Florida (Dade county), the Hispanic communities have much lower Hispanic-physician-to-Hispanic-population ratios than the overall physician-to-population ratio for the county or group of counties in which they are located. Some Hispanic communities, both urban and rural are located in areas that are relatively richer (in terms of per capita income, or health-professional-to-population ratios, or number of health professions schools) than the non-Hispanic areas of their states; other Hispanic communities are located in areas that are relatively poorer by these measures. See Table 13. 414 Table 13 Selected Characteristics of Hispanic* and Non-Hispanic Communities Hispanic Primary M.D.'s Care Pop. Percent M.D.'s Per M.D.'s Dentists R.N.'s | Percent Percent Density Below Per 100,000 Per Per Per j Pop. Hispanic Nonwhite (Persons/ Income Poverty 100,000 Hispanic 100,000 100,000 100,0001 State & Number (000s) Pop. Pop. Sq. Mile) Per Capita Level Pop. Pop. Pop. Pop. Pop. | of Counties (1980) (1980) (1980) (1980) (1980) (1979) (1979) (1980) (1979) (1979) (1977) | Florida Hispanic (1) 1,626 36 22.8 796 $10,582 15.2 301 409.5 112 52 541 | NorHHispanic (66) 8,121 14.7 156 $ 9,135 13.3 148 58 39 474 | New York Hispanic (2) 2,597 28 45.4 40,582 $11,414 23.7 555 89.3 174 90 711 | I NorHHispanic (60) 14,961 15.7 313 $10,098 11.2 209 85 63 540 | |New Jersey I Hispanic (1) 557 26 22.6 11,850 $ 9,832 16.7 158 84.0 71 50 395 | I Non-Hispanic (20) 6,808 15.9 911 $11,088 8.7 184 73 61 488 | I California I Hispanic (11) 9,849 27 29.6 250 $10,938 13.1 217 39.1 85 50 380 | I Non-Hispanic (47) 13,819 28.3 118 $10,988 10.0 218 83 60 432 | |Arizona I Hispanic (8) 890 25 18.4 21 $ 8,726 14.3 181 53.0 72 34 589 | Non-Hispanic (6) 1,828 16.1 26 $ 9,238 12.8 168 68 42 592 | |New Mexico I Hispanic (28) 1,105 41 20.8 10 $ 7,886 16.7 149 44.9 58 36 383 | I Non-Hispanic (4) 198 41.8 16 $ 7,996 22.7 92 43 26 322 | |Texas | Hispanic (90) 3,884 51 19.3 32 $ 7,605 21.5 119 63.1 48 29 256 | | Non-Hispanic (164) 10,345 21.0 73 $10,349 12.1 155 60 40 331 | |Colorado Hispanic (12) 225 35 15.7 11 $ 7,946 16.0 129 25.1 50 44 521 | Non-Hispanic (51) 2,665 9.9 32 $10,269 9.5 196 77 55 597 | Washington Hispanic (1) 13 22 13.5 7 $10,030 12.2 66 0 43 36 365 | Non-ttLspanic (38) 4,119 8.2 64 $10,237 9.8 172 66 60 550 | *Hispanic community is considered to be a county or group of counties with Hispanic populations greater than 20 percent. Florida Florida has one county with a Hispanic population greater than 20 percent — Dade County, which is approximately 36 percent Hispanic. Dade County is significantly different in several ways from the other overwhelmingly urban areas with 20 percent Hispanic concentrations (New York, 2 counties and New Jersey, 1 county) as well as significantly different from rural and mixed Hispanic communities elsewhere in the U.S. In 1980, its population was 1.626 million and 1.1 percent rural, compared to 8.121 million (18.7 percent rural) for the rest of Florida. With 796 persons per square mile it is much denser than the rest of Florida considered in aggregate (156 per square mile) but far less dense than Hispanic target areas in New York (40,582 per square mile) and New Jersey (11,850 per square mile). The minority population was 22.8 percent for Dade County and 14.7 percent for the rest of the state. The 1980 income per capita was $10,582 for Dade County, compared to $9,135 for the rest of the state, but the percent change over the preceeding 10 years was comparable (296.6 and 305.1). Percentages of the population receiving AFDC were comparable for Dade and the rest of the state (2.5 for Dade, 2.6 for the rest of Florida) but considerably smaller than Hispanic target areas in New York (10.8) and New Jersey (10.7). The percentage below the poverty level was 15.2 percent for Dade County and 13.3 percent for the rest of Florida, slightly below the Hispanic target area in New Jersey (16.7) and considerably below the Hispanic target area in New York (23.7). Ambulatory visits per 1,000 population were comparable for Dade County and the rest of the state (1,042 vs. 1,064), although the percent change from 1970-82 was far higher for the rest of the state (96 percent for other counties, 40 percent for Dade). Inpatient days per 1,000 population were higher in Dade County than the rest of Florida (1,674 to 1,415), although the percent change from 1970-82 was higher in the rest of the state (60 to 39). Dade County and the rest of the state showed striking divergences in the numbers of active, non-Federal physicians available, and to a lesser degree in other health professionals. Dade County had 301 M.D.'s per 100,000 population, compared to 148 per 100,000 for other counties. Between 1940 and 1980, growth rates for M.D.'s were 1,158.1 percent for Dade County and 795.2 percent for the rest of the state, although for the last 10 years, growth was higher outside Dade County (118.4 to 80.4 percent). Dade also had more active, non-Federal dentists per 100,000 population (52 to 39). Growth in real numbers was slightly faster outside Dade County (11.4 percent to 9.4). Employed R.N.;s per 100,000 were 541 in Dade, 474 elsewhere. Dade also had a higher percentage of pharmacists per 100,000 in 1980 (44 to 38) and optometrists per 100,000 in 1981 (9 to 7). Health professions schools were located largely outside the target county. Dade had 1 medical school (175 graduates), and 5 R.N. schools (423 graduates). The rest of Florida had 2 medical schools (212 graduates), 1 dental school (59 graduates), 29 R.N. schools (1,964 graduates), 1 pharmacy school (190 graduates), and 1 veterinary school (39 graduates). 416 Dade County has a substantially higher Hispanic-physician-to-Hispanic- population ratio than the general physician-to-population ratio for the county as a whole. Although 36 percent Hispanic, the county had 2,375 Hispanic physicians, giving it 409.5 Hispanic physicians per 100,000 Hispanic population. The county overall had 301 physicians per 100,000 population. Dade has an extraordinarily higher Hispanic-physician-to-Hispanic-population ratio than any other Hispanic target area in the U.S., including New York (89.3 and New Jersey (89.0). New York New York State has two counties with a Hispanic population greater than 20 percent—Bronx and New York, which together have a roughly 28 percent Hispanic population. In 1980 these two counties had a combined population of 2.597 million (0 percent rural), which is about 1/6 of the rest of the state (14.961 million, 18 percent rural). The percentage of minorities is also nearly 3 times greater in the Hispanic target counties (45.4 to 15.7). Income per capita was higher in the target counties ($11,414 vs. $10,098). Even so, the target counties had a significantly higher percentage of AFDC recipients (10.8 to 5.5 percent) and of persons below the poverty level (23.7 to 11.2 percent). Ambulatory visits per 1,000 population were more than twice as high in the target counties (2,728 to 1,244) although the percent change from 1970-82 was higher elsewhere in the state (71 to 18 percent). Inpatient days per 1,000 population were nearly twice as high in the target counties than elsewhere (2,449 to 1,345) although this figure declined 12 percent from 1970-82 in the target counties while rising 15 percent in the rest of the state. The target counties had a far higher proportion of M.D.'s, both in aggregate and in all specialties except general practice. These counties had 555 physicians per 100,000 population (compared to 209 elsewhere); and in primary care specialties, they had 100 in internal medicine (compared to 37), 24 in ob/gyn (compared to 13), and 29 in pediatrics (compared to 15). Only for general practice were the numbers per 100,000 comparable (21 for the target areas, 20 elsewhere). Although the target counties had far higher concentrations of physicians, growth rates compared to the rest of the state were much lower. In real numbers, physicians declined 12.1 percent from 1940-80, up 2.2 percent in the last 10 years in the target counties. Elsewhere, physicians increased 148.2 percent in the last 10 years. The target counties also maintained higher numbers of other health professionals per 100,000 for dentists (90 to 63), R.N.'s (711 to 540), pharmacists (53 to 39) and optometrists (11 to 7). Both communities showed declines in the real number of pharmacists and optometrists in recent years. A slight decline in the number of R.N.'s (although growth in numbers per 100,000) occurred in the target counties between 1972 and 1977. The target counties had considerable concentrations of health professions schools: 6 medical (926 graduates), 2 dental (233 graduates), 14 R.N. (1,346 graduates), 1 optometry (39 graduates), and 1 podiatry (101 graduates). The rest of New York had 6 medical (791 graduates), 1 osteopathic (231 students, no graduates reported that year), 2 dental (109 graduates), 80 R.N. (5,894 graduates), 4 pharmacy (624 graduates), and 1 veterinary (76 graduates). 417 The target counties had 652 Hispanic physicians and a Hispanic population of 730,385. The Hispanic-physician-to-Hispanic-population ratio of 89.3 per 100,000 is the highest outside Dade County, but far below the overall physician figure for these two counties (555 per 100,000). New Jersey New Jersey has one county with a Hispanic population greater than 20 percent—Hudson County, which is roughly 26 percent Hispanic. In 1980, its population of 557,000 (0 percent rural) was roughly 1/12 the size of the rest of New Jersey (6.808 million, 11.9 percent rural) and far denser (11,850 vs. 911 persons per square mile). It had a larger percentage of minorities than the rest of the state (22.6 to 15.9), lower per capita income ($9,832 to $11,088), a larger percentage of AFDC recipients (10.7 to 5.9), and a greater percentage of persons below the poverty level (16.7 to 8.7). The number of ambulatory visits per 1,000 population were similar (1,237 in Hudson County vs. 1,211 elsewhere), although the growth from 1970-82 was much higher in Hudson County (181 percent vs. 79 percent). Inpatient days per 1,000 and 1970-82 growth rates were slightly higher in Hudson County (1,359 and 23 percent) than in other New Jersey counties (1,260 and 18 percent). Except for primary care where there is approximate parity, Hudson County lags behind the rest of New Jersey in M.D.'s and M.D. specialists per 100,000 population. The target county had 158 M.D.'s per 100,000 (compared to 184 elsewhere) and in primary care specialties. In Hudson County between 1940 and 1980, the number of physicians grew 5.1 percent and the number per 100,000 population grew 23.1 percent, with larger increases in the past 10 years (15.2 and 26.1 percent respectively). Growth was more significant elsewhere in the state: the number of physicians increased 178.6 percent from 1940-80 and 45.2 percent in the past 10 years while physicians per 100,000 population grew 43.6 percent overall and 39.9 percent in the past 10 years. The rest of the state had greater numbers per 100,000 of other health professionals, in some cases by margins of 10 to 20 percent. Both areas have seen declines in real numbers of pharmacists and optometrists in recent years, comparable increases in real numbers of R.N.'s, and increases in numbers of dentists. Health professions training occurs almost exclusively outside the county. Hudson had only 3 R.N. schools with 63 graduates, while other New Jersey counties had 2 medical schools (229 graduates), 1 osteopathic (117 students), 2 dental (156 graduates), 36 R.N. (2,065 graduates), and 1 pharmacy (104 graduates). Hudson County had 122 Hispanic physicians and a Hispanic population of 145,249. Its Hispanic-physician-to-Hispanic-population ratio (84.3 per 100,000 was just slightly behind that of the target counties in New York. 418 California California has 11 counties with Hispanic populations greater than 20 percent, and which together have roughly a 27 percent Hispanic population. Demographically somewhat different than the other 47 California counties, these 11 are very similar to the rest of the state in availability of health manpower resources. In 1980, the 11 target counties had a population of 9.849 million, 6.2 percent rural, compared to a population of 13.819 million (1.4 times larger) for the rest of the state, which is 10.5 percent rural. The target counties are more densely populated with 250 persons per square mile, compared with 118 per square mile elsewhere. The two areas were comparable in income per capita ($10,938 for the 11 counties vs. $10,988), in percent change of per capita income over the past 10 years (299.3 vs. 306.4), and in percentage of minority population (29.6 vs. 28.3 percent). However, the target counties had a larger percentage of AFDC recipients (6.8 to 5.2) and percent below the poverty level (13.1 to 10.0). The target counties were slightly lower in ambulatory visits per capita (1,173 to 1,278) Lut growth in visits per capita was slightly higher from 1970-82 (26 to 19 percent). These 11 counties had both a larger number of inpatient days per 1,000 (1,022 to 921) and percent growth from 1970-82 (9 to 5). Physician-to-population ratios were virtually identical for the two regions. Comparing the target counties to the rest of California, M.D.'s per 100,000 population were 217 to 218. Growth in real numbers of M.D.'s has been substantially higher outside the target counties: 348.7 percent between 1940-80 with 208.3 percent in the last 10 years for the target counties, 540.7 percent with 295.9 percent in the last 10 years for the rest of the state. However, when adjusted for population, these growth rates are much closer, 60.6 percent for the target counties and 56.8 percent for the others in the last 40 years, 31.4 percent and 29.5 percent in the last 10 years. The rest of California had a higher ratio of dentists per 100,000 population (60 to 50) and employed R.N.'s (432 to 380 with indistinguishable growth rates in recent years. Professional-to- population ratios were virtually identical for osteopathic physicians, pharmacists and optometrists. Both areas have considerable health professions training resources. The Hispanic target counties had 2 medical schools (314 graduates), 1 osteopathic (162 graduates, 2 dental (236 graduates), 35 R.N. schools (2,194 graduates), and 1 pharmacy school (128 graduates). Other California counties had 6 medical schools (680 graduates), 3 dental schools (305 graduates), 51 R.N. (2,902 graduates), 2 optometry, 2 pharmacy, 1 podiatry, and 1 veterinary school. The Hispanic target counties had 1,055 Hispanic physicians and a Hispanic population of 2,697,924. Thus, while the 11 counties had one of the highest overall physician-to-population ratios for Hispanic areas in the U.S. (217 per 100,000 population), they had one of the lowest ratios of Hispanic physicians to Hispanic population (39.1 per 100,000). 419 Texas Texas has 90 counties with Hispanic populations greater than 20 percent. In fact, 51 percent of the total population of these counties is Hispanic—the highest of any of the target communities. This is the poorest Hispanic target area in per capita income, and except for the small group in Washington state, is the poorest in terms of available health resources. In 1980, the target area had a population of 3.884 million compared to 10.345 million (2.6 times greater) for Texas' other 164 counties. Classed as 19.6 percent rural, the target area is only 1.2 times smaller in land area than the rest of Texas and had an overall population density of 32 persons per square mile, compared to 73 per square mile elsewhere. The minority populations were similar—19.3 percent for the target area and 20.3 percent elsewhere. The similarities largely end here. Income per capita was $7,605 compared to $10,349 elsewhere, although growth rates were comparable from 1970-82 (336.0 vs. 344.4 percent). The AFDC population was 3.3 percent compared to 1.7 percent elsewhere, but the percent below the poverty level was 21.5 percent compared with 12.1 percent elsewhere. Only New York (23.7 percent) had a poverty level population within 4 percentage points of this figure. Much medical care in the target area appears to come on an ambulatory basis. Ambulatory care visits per 1,000 population were 1,492. While lower than in Arizona or New Mexico, this was substantially higher than the 952 reported for other Texas counties. The percent growth in ambulatory visits from 1970-82 was 146 percent, nearly double the 74 percent for other counties. Inpatient days per 1,000 were 1,214, compared to 1,341 elsewhere, with 12-year growth rates in these figures comparable for the two areas (30 and 32 percent). In 1979, the 90 Hispanic counties had 119 M.D.'s per 100,000 population, compared with 155 per 100,000 for other Texas counties. This rate was below that of the Hispanic target area in Colorado (129) and well below the target areas in New Mexico (149) and Arizona (181). While M.D. general practitioners per 100,000 were nearly equal in the 90 counties and elsewhere in Texas (23 vs. 24), the target area was well below the rest of Texas in all other specialities. M.D. growth rates in real terms from 1940-80 and 1970-80 were lower for the target area than for the rest of Texas, although M.D.'s per 100,000 grew slightly faster in the target area between 1970 and 1980. The Hispanic counties had lower health-professional-to-population figures than the rest of Texas for osteopathic physicians (4 per 100,000 vs. 8 per 100,000), pharmacists (38 vs. 45), R.N.'s (256 vs. 331), dentists (29 vs. 40) and veterinarians (16 vs. 23). Relatively little health professions training is conducted in the target area, which has 1 medical school (137 graduates), 1 dental school (129 graduates), and 14 R.N. schools (795 graduates). Other Texas counties had 6 medical schools (801 graduates), 1 osteopathic (328 graduates), 2 dental (252 graduates), 42 R.N. (2,758 graduates), and 1 optometry, 3 pharmacy, and 1 veterinary schools. 420 Although the Hispanic counties had only 119 physicians per 100,000 population overall, they did have 1,238 Hispanic physicians for a Hispanic population of 1,963,334. The Hispanic-physician-to-Hispanic-population ratio of 63.1 per 100,000 exceeded that of Arizona, Colorado, New Mexico, and California. Colorado Colorado has 12 counties with Hispanic populations greater than 20 percent; overall, these counties are about 35 percent Hispanic. In terms of income differences between the target area and the rest of the State and in terms of limited availability of health professional resources and training opportunities, this area is more like the Texas target area than those of Arizona or New Mexico. In 1980, the 12 target counties had a combined population of 225,000, 32 percent rural, with a population density of 11 per square mile, compared to 2.665 million elsewhere, (18.3 percent rural) with a population density of 32 per square mile. The target area had a 15.7 percent minority population (compared to 9.9 percent elsewhere), a substantially lower income per capita ($7,946 vs. $10,269), a larger percentage of AFDC recipients (6.1 vs. 2.3), and larger percentage of persons below the poverty level (16.0 vs. 9.5). Ambulatory visits per 1,000 population were lower in the 12 counties (1,147 vs. 2,279) with a lower percent change from 1970-82 (132 vs. 239). Inpatient days per 1,000 population were higher (1,684 vs. 1,123), but the percent changes from 1970-82 (16 percent) were identical for the two areas. Except for general practitioners, the Hispanic target area had consistently lower physician-to-population ratios. Overall, the 12 counties had 129 M.D.'s per 100,000 population (compared to 196 for the rest of Colorado). Physician supply per 100,000 has grown somewhat more rapidly in the target areas than in the rest of the state. Between 1940 and 1980, the active supply of M.D.'s grew 48.5 percent (compared to 255.5 percent elsewhere) with a 39 percent increase from 1970-80 (compared to 57.3 percent elsewhere). Yet the supply per 100,000 population in the target area was up 138 percent over 40 years (compared to 21.9 percent) and 34.2 percent over the past 10 years (compared to 17.3 percent elsewhere). The target area had more pharmacists per 100,000 population (56 to 47), but fewer osteopathic physicians (3 to 10), dentists (44 to 55), R.N.'s (521 to 597), optometrists (7 to 9), and veterinarians (23 to 35). Both regions have had a decline in pharmacists in recent years. Very little health professions training is conducted in the 12 target counties: only two R.N. schools with 64 graduates. Elsewhere in Colorado, there are 1 medical (133 graduates), 1 dental (21 graduates), 10 R.N. (708 graduates, 1 pharmacy (62 graduates), and 1 veterinary school (121 graduates). The target counties had 20 Hispanic physicians for a Hispanic population of 79,603. Thus, in addition to having one of the lowest physician-to-population ratios among the targeted states, this area had the lowest Hispanic-physician- to-Hispanic population ratios, 25.1 per 100,000. 421 Arizona Arizona has eight counties of its total 14, with Hispanic populations greater than 20 percent. Together, these counties are about 25 percent Hispanic. Both target counties and other areas have undergone substantial population growth in the past 10 years. In 1980, the 8 target counties had a total population of 890,000, 22.4 percent rural, with a population density of 21 persons per square mile. Minority populations were 18.4 percent for the target counties and slightly lower (16.1 percent) for the rest of the state. The target area had a lower per capita income than other Arizona counties ($8,126 to $9,238), and a slightly higher percentage of AFDC recipients (2.0 to 1.7) and persons below the poverty level (14.3 to 12.8). Ambulatory visits per 1,000 population were higher in the target counties in 1982 (2,026 to 1,694) although the rate of growth between 1970 and 1982 was lower (66 vs. 166 percent). Inpatient days per 1,000 were virtually identical for the two areas (1,082 vs. 1,080), although the rate of growth from 1970-82 was lower in the 8 counties (12 vs. 30 percent). For most specialties, the group of counties with concentrations of Hispanics had approximately equivalent or slightly better numbers of M.D.'s per 100,000 than the other 6 Arizona counties. Overall, the target counties had 181 M.D.'s per 100,000, compared to 168 elsewhere in the state. The growth in M.D.'s per 100,000 has been large in both areas of the state. Between 1940 and 1980, the number of M.D.'s grew 712.6 percent in the target counties and 1,112.4 percent outside, with both areas seeing a 92 percent growth in M.D.'s per 100,000. However, between 1970 and 1980, M.D.'s per 100,000 grew more than twice as fast in the target area (68.6 to 30.2 percent). Both areas had comparable numbers of other professionals per 100,000 population, including dentists (34 in the target counties vs. 42 elsewhere), employed R.N.'s (589 vs. 592), pharmacists (59 vs. 54), optometrists (7 vs. 8), and osteopathic physicians (15 vs. 17). In recent years, both areas have shown slight declines in dentists, and considerable growth in employed R.N.'s. The state has few health professions schools, most of which are located in the target counties, which had 1 medical school (86 graduates), 2 R.N. schools (296 graduates), and 1 pharmacy school (46 graduates). Elsewhere in the state were 10 R.N. schools with 483 graduates. The target counties in Arizona had 119 Hispanic physicians and a total Hispanic population of 224,383. The Hispanic-physician-to-Hispanic-population was 53.0, somewhat higher than neighboring New Mexico, but well below the overall physician-to-population ratios in Arizona. New Mexico In New Mexico, 28 of 32 counties have Hispanic populations greater than 20 percent. With a combined Hispanic population of about 41 percent, these counties have the second greatest concentration of Hispanics among the areas examined—slightly greater than Dade county but below the Texas target area. 422 Both regions of New Mexico had substantial population growth between 1970 and 1980. With a population of 1.105 million in 1980 (26.3 percent rural) and a population density of 10 persons per square mile, the target area had about 5-1/2 times more people than the other 4 counties, which were 36.6 percent rural with a population density of 16 per square mile. Both areas were comparable in income per capita ($7,886 in the target areas in 1980 vs, $7,996 elsewhere) and percent change in per capita income over the previous 10 years (321.1 vs. 337.0). The target area and the rest of the state had low percentages of the population receiving AFDC (3.9 vs. 4.9), although both areas, especially the non-target counties, had large percentages below the poverty level (16.7 and 22.7). In 1982, ambulatory visits per 1,000 population were 1,758 for the target counties and 4,251 for the other 4 counties. (The latter is an anomaly; nothing approaching this figure was reported for any other area.) The percent change in ambulatory visits between 1970 and 1982 was 174 percent for both areas. Inpatient days per 1,000 population were higher in the target counties (1,016 to 832) and in the target counties, grew 20 percent between 1970 and 1982, while declining 17 percent elsewhere. For most specialities, the number of M.D.'s per 100,000 population was greater in the target counties than in the rest of the state, including 149 M.D.'s of all types (vs. 92). Both areas had high growth rates of M.D.'s from 1940-80 (453 percent in the target area and 494.3 percent outside), with higher growth of physicians per 100,000 population in the target area (136.6 vs. 77.3 percent). Outside the target area, al of this growth was in the latest 10 years (77.9 percent); during the same period, the percent change in the 28 counties was 44.7. The target area also had higher professional-to-population ratios for osteopathic physicians (8 per 100,000 vs. 2 per 100,000), pharmacists (48 vs. 38), employed R.N.'s (383 to 322), and dentists (36 to 25) but a lower ratio of optometrists (8 to 12). Growth in the number of R.N.'s per 100,000 population was more than twice as high between 1972 and 1977 in the target area. The state has few health professions schools, most of which are located in the target area, including 1 medical school (75 graduates), 8 R.N. schools (415 graduates), and 1 pharmacy school (52 graduates). Elsewhere in the state are 3 R.N. schools with 23 graduates. The 28 target counties had 202 Hispanic physicians and a Hispanic population of 449,749, for a Hispanic-physician-tp-Hispanic-population ratio of 44.9 per 100,000. Washington One of Washington's 39 counties has a Hispanic concentration greater than 20 percent—Adams county, which is about 22 percent Hispanic. Adams is a rural county in western Washington: in 1980, it had 13,000 persons, 66.1 percent rural, with only 7 persons per square mile. The rest of the state had a population of 4.119 million, is 26.3 percent rural, and had 64 persons per square mile. 423 In 1980, Adams's income per capita was similar to the rest of the state ($10,030 vs. $10,237) and had grown somewhat more in the previous 10 years (372.7 vs. 304.8). The percentage of the population receiving AFDC was slightly higher than the rest of the state (4.3 to 3.7) as was the percentage below the poverty level (12.2 vs. 9.8). Ambulatory visits per 1,000 population were lower in Adams County (913 to 1,119 but had grown more (147 vs. 60 percent) from 1970 to 1982. Inpatient days per 1,000 population were much lower (497 to 840) in a state that had among the lowest totals reported anywhere. Further, both Adams and the rest of the state saw declines in inpatient days per 1,000 (3 and 2 percent respectively) between 1970 and 1982. Similarly, all health-professional-to-population ratios, except for M.D.'s in general practice and ob/gyn, D.O.'s (the county had 2 osteopathic physicians), and pharmacists, were below those of the rest of the state. The county had 66 physicians per 100,000 population (compared to 172 elsewhere). The county has no health professions schools. The state has 1 medical school (173 graduates), 1 dental school (97 graduates), 20 R.N. schools (1,141 graduates), 2 pharmacy schools (129 graduates), and 1 veterinary school (79 graduates). The county has no Hispanic physicians and a Hispanic population of 2,950. PART II. A. 3. - Asian/Pacific Islanders' Communities and Health Professional Resources The caveats noted earlier regarding the difficulty in reviewing data at the county level for relevance to specific minority communities, need their greatest emphasis here. The Asian/Pacific Islanders as a minority group are widely spread through the U.S., with a reasonably high density in only a few states. Because they rarely appear even in double-digit percentages in county-level population figures, it is difficult if at all possible to attribute events seen in those counties to that minority group. Further, since Asian/Pacific Islanders are among the most heterogeneous of all U.S. minority groups, the numerous sub-groups provide even more complication in attempting to define a "community." For purposes of this discussion, Asian/Pacific Islanders are defined as being from Japanese, Chinese, Filipino, Korean, Asian Indian (e.g. Pakistan, India), Vietnamese (e.g. Laotian, Cambodian), or Hawaiian and other Pacific Islander (e.g. Samoan) ethnic origin. Precise figures on the numbers of each of these groups are difficult to obtain, thereby making it almost impossible to make meaningful conclusions for any of them individually. Nonetheless, the methodology applied elsewhere was also applied to the statistics for "highly dense" counties for Asian/Pacific Islanders. The process was modified in deference to the small number of counties meeting the "greater-than-0-percent" threshold, by lowering that level to "greater-than-five-percent." [This represented a two-edged sword in that while providing some increase in the number of counties available for analysis, it reduced the already moderately low density of population even further.] 424 According to the 1980 Census, Hawaii, California and New York contain the counties with the largest concentrations (five percent or more) of Asian/Pacific Islanders. Hawaii has, by far, the largest single Asian/Pacific Islander community in the U.S., however, because of its unique size and its political definition as a State, it was deemed to be inappropriate for inclusion for basis of comparison with other Asian/Pacific Islander "communities." The problems inherent in attempting to analyze the contribution or impact of five percent of a county's population were acknowledged to be great. This was so much the case that the working group almost immediately began to look for alternative approaches to defining an Asian/Pacific Islander community. One thought was to simply recognize the reality of the dispersal of this minority population across the entire nation, and ignore the limits imposed by a five percent county population standard. For example, by arbitrarily lowering the data threshold to reveal counties with 0.5 percent or more Asian/Pacific Islanders. The population, as noted in Table 14, is dispersed across 51 states in 541 counties. Although the 0.5 percent in a given county may seem small, the actual numbers that it yields provide a larger population base for more specific characteristics to be drawn. Table 14 U.S. Counties Sorted by Percent of Asian/Pacific Islander Population Percent Asian/Pacific Islander Number of Percent of Total Population Counties 293 Counties* 0.5 to 0.99 54.2 1 169 31.2 2 39 7.2 3 22 4.1 4 3 0.6 5 2 0.4 6 2 0.4 7 4 0.7 8 1 0.2 10 1 0.2 21 1 0.2 60 1 0.2 61 1 0.2 62 1 0.2 67 1 0.2 * Fifty States and the District of Columbia contain 0.5 percent or more Asian/Pacific Islanders in a given county. A total of 541 counties were observed. 425 Another approach would be clustering counties that are adjacent to one another in a given geographic area to increase the numbers available for study. For example, percentages of Asian/Pacific Islanders in the Washington, D. C, metropolitan area may appear small when looking at individual county data. They only represent one percent of the population in the District of Columbia, two percent in Howard and Prince Georges counties (both in Maryland), three percent in Montgomery (Maryland), Alexandria City and Fairfax (both in Virginia) and five percent in Arlington County (Virginia). Each of these counties are geographically adjacent and together represent 86,068 Asian/Pacific Islanders, not an insignificant figure when combined. Another example of the many complexities that arise when examining the data on Asian/Pacific Islanders, can be illustrated through the use of figures on health professionals. The percent distribution of Asian/Pacific Islanders health professionals in 1980 are provided in Table 15. Table 15 Percent Distribution of Asian/Pacific Islander Health Professionals According to Ethnicity, 1980 _________________Physicians_____Dentists_____Pharmacists Total Asian/Pacific Islanders 100 100 100 Japanese 5 39 18 Chinese 16 29 41 Filipino 25 7 8 Korean 10 8 10 Asian Indian 37 13 18 Vietnamese 2 1 3 Hawaiian and Other 5 2 3 Pacific Islanders Thirty-seven percent of all Asian/Pacific Islander physicians were Asian Indian, while Filipino physicians make up 25 percent and Chinese 16 percent. The large number of Asian Indian and Filipino physicians reflects the influence and impact of including those who have been foreign trained. The ethnic composition of dentists and pharmacists are more similar where health professionals from Chinese and Japanese heritage better reflect the composition of Asian/Pacific Islanders population in this country as a whole. The variation among ethnic groups and where they settle, the differences between new immigrants and older immigrants, and the impact of being an American citizen for many generations, coupled with socio-economic factors, are just a few examples of how Asian/Pacific Islanders select their communities and how health service needs vary. 426 As a result of all these considerations, it was decided that for consistency with the other minority presentations, analyses should be developed for the States of California and New York. Those narratives follow shortly. It was further decided that additional analyses need be undertaken utilizing a variety of data manipulations including use of standard metropolitan statistical areas (SMSA's); sub-county and inter-county data comparisons across, rather than within states; comparisons of one sub-group versus another; etc. Several of those latter analyses are planned for future study. New York Two New York counties (Kings and New York) have populations in excess of one million persons each, and are urban. The Asian/Pacific Islander population in these two is 5.3 percent, compared to 1.1 percent for the rest of the state. The Black population for the two counties is 20.1% compared to 12.2 percent for the rest of the state (60 counties). In 1980, the personal income per capita for the two counties was $12,114 and the personal income per household was $28,396. These county figures were significantly higher than those for the rest of the state ($9,868 and $20,527, respectively). Contrary to the wealthier population base in these two counties, the population receiving Aid to Families with Dependent Children per 100,000 total population was 10,777 compared to 5,220 for the rest of the state. The percentage of the population below the poverty level in 1979 was 15.6 compared to only 12.4 for the rest of the state. These combined findings suggest significant polarization of resources. The Infant Mortality Rate for "other" (than nonminority or Black) is 6.4 compared to 5.8 for the rest of the state. The mortality per 1,000 population is 11.2 vs. 9.5 elsewhere in the state. Portions of both counties have been designated as Medically Underserved Areas and Health Manpower Shortage Areas for primary care. Other shortages exist in some areas for dental, vision care and psychiatric manpower. Fourteen of twenty urban National Health Service Corps sites within the state in 1980 are located in the two counties. General Hospital Utilization has differed significantly between these two counties and the rest of the state since 1970. From 1970 to 1982 Two Counties Rest of State % Change in ambulatory visits +5 +75 % Change in patient days -18 +18 This suggests marked reduction in use of general hospitals, even while the physician population grew 7.4 percent from 1960 to 1970, and another 7.7 percent from 1970 to 1980 while the total population fell by 5.8 percent. The increase in physician population from 1960-1980 does not recover the decrease of 26.7 percent noted between 1950 and 1960. 427 The increase in physician manpower in the two counties was relatively insignificant when compared to the increases seen in the rest of the state during the same time periods. In other urban counties, the increases were 52 percent for the years 1960 to 1970 and 19.6 percent for 1970 to 1980, while the overall population fell 3.2 percent. This was in addition to the 72 percent increase seen between 1950 and 1960. In 1950, nearly two-thirds of all the state's physicians (18,975) were in these two counties. By 1980, even with the aforementioned increases of the past two decades, there was an actual decline in the number of physicians (16,083) to account for only one-third of the state's total. Indeed, in 1940 there were more physicians in these two counties (16,889) than in 1980. Some measure of the distribution of these changes among physicians can be seen by reviewing the numbers of specialists between 1975 and 1979. For example, while the number of hospital-based general practitioners in the rest of the state increased by 26.1 percent, in these two counties there was a 14.1 percent decrease observed. Obstetricians/Gynecologists showed a similar, but less severe pattern. The numbers of Internists and Pediatricians in office-based practice increased modestly (14 percent and 4.7 percent, respectively). In the rest of the state, however, there was minimal change in these counties (2.9 percent and 1.0 percent, respectively). When reviewing all active non-Federal physicians, the total number of physicians per 100,000 population in the two counties (471) far exceeds that for the rest of the state (212). However, when reviewing perhaps the most critical area of specialty for primary care, i.e. Family Medicine/General Practice, there is no appreciable difference (22 vs. 20). Among other health professionals, the number per 100,000 population fluctuated as follows: Professions Two Counties Rest of State Pharmacists (1974-1980) -23.8% -13.5% Employed RN's (1972-77) + 1.1% +25.0% Optometrists (1972-1981) -12.2% - 8.4% Dentists +10.1% + 9.5% California Three-fifths of the state's population can be found in these 11 of the 47 counties. All 11 counties are urban communities. Seven of the 11 counties have a population in excess of one million persons while another three counties have between 250,000 and one million persons. The eleventh county has in excess of 100,000 persons. The total Asian/Pacific Islander population for the state is 2.9 percent compared to 7.2 percent within the 11 counties. The personal income per capita is somewhat higher in these 11 counties than for the rest of the state ($11,376 vs. $10,297). The Aid to Families with Dependent Children per 100,000 total population is higher for the 11 counties (6,353 vs. 5,063). There is a slightly higher percentage of the population below the poverty level in the 11 counties (11.8 percent vs. 10.5 percent). 428 The Infant Mortality Rate for "other" (than nonminority or Black) is the same for these counties as it is for the rest of the state, i.e. 6.0. The overall mortality per 1,000 population is 8.0 in these counties vs. 7.7 for the rest of the State. Of the eleven counties, a portion of ten of the eleven counties have been designated as both Medically Underserved Areas and primary care Health Manpower Shortage Areas. Portions of the eight of the eleven counties have dental manpower shortages; two have vision care and two have psychiatric manpower shortages; one entire county and part of another have podiatric manpower shortages. For the rest of the State, five counties and portions of the 19 of 47 counties have been designated Medically Underserved Areas, while two entire counties and parts of another 39 counties have been designated primary care health manpower shortage areas. Entire counties, totaling 21, have podiatric manpower shortages; parts of eleven counties have dental shortages; one entire and portions of three counties have psychiatric manpower shortages. Thirteen of the twenty urban National Health Service Corps sites within the State in 1980 were located within the eleven counties, as were five of the 29 rural sites. Physicians were disproportionately concentrated in the eleven counties, totaling 261 per 100,000 population, compared to 186 per 100,000 for the rest of the State. The distribution of nurses (both RNs and LPNs), based on numbers per 100,000 population and per 100 hospital beds seemed more even. The distribution of other health professionals more closely approximated the distribution of the general population between the 11 counties and the rest of the State. 11 Counties Rest of the State Total Population Dentists Podiatrists Optometrists Pharmacists IA, ,715,000 9,408 9,408 1,871 8,581 8, ,953,000 5,852 5,852 1,250 4,771 Health Professions training institutions are also disproportionately concentrated in the 11 counties, including 5 of the 8 medical schools, 4 of the 5 dental schools, 54 of 86 nursing schools, the only podiatry school and three pharmacy schools and 1 of 2 optometry schools. 429 Table 16 Selected Characteristics of Asian/Pacific Islanders and Non-Asian/Pacific Islander Cannunities State County Population (000's) (1980) % Asian/ Pacific Islander Population Population Density (Persons/ sq.mile) (1980) AFDC Pop/IOOK Population Income Per Capita (1980) % Below Poverty (1979) Infant* Mortality Per 1,000 Births (1980) M.D. Pop/IOOK Population (1980) .£> New York Key County (2) Rest of State (60) 3,320 14,238 5.3 1.1 25341 299 10777.1 5220.5 12,114 9,868 15.6 12.4 6.4 5.8 484 216 U) O Cali fornia Key County (11) Rest of State (47) 14,715 8,953 7.2 2.9 818 65 6353.4 5063.2 11,376 10,297 11.9 10.5 6.0 6.0 258 181 *Figures represent 't)ther" which includes Asian/Pacific Islanders, minority, American Indians. PART II. A. 4. - American Indians/Alaskan Natives' Communities and Health Professional Resources American Indians/Alaskan Natives represent the smallest of the four minority groups with total numbers, by 1980 Census count, of approximately 1.4 million. This represents a 70 percent increase over the 1970 Census, which is due to increased numbers of people declaring this identification, rather than the result of changes in vital events, i.e. births and deaths. Difference % Difference Between Between 1970 1980 1970 & 1980 1970 & 1980 American Indian 792,730 1,364,033 +571,303 +72.1 Eskimo, Aleutian Islander 34,525 56,367 + 21,842 +63.3 Total 827,255 1,420,400 +593,145 +71.1 This minority population is geographically spread throughout the continental United States and consists of more than 600 tribes and other sub-groups. Alaska is the State with the highest percentage of American Indian inhabitants, i.e. 16 percent of the total. California, however, has by far the largest number with nearly a quarter of a million inhabitants. Despite the relatively small numbers, the heterogeneity of the American Indian population is quite striking. Customs and cultures vary widely, as do the nature and formality of the relationships between these tribes and communities, and local, state and Federal governments. Efforts to address the concerns of individual Indian communities require information on the specific make-up of the residents of that community. Inevitably sub-county and tribal data have to be obtained and analyzed. Counties having greater than 20 percent American Indian population are spread over 10 states. In part this is a reflection of the location of reservations; otherwise it reflects a strong tendency to live as a community in individual counties rather than dispersing. The State of Alaska was excluded from this analysis for essentially two reasons. First, it has no "county" structure and by itself has only a 16 percent Alaskan Native population. Second, as a state it has a different resource base than the counties being used for comparative analyses. The State of California, although comprising the largest single state population of American Indians, was also not a part of this analysis since none of its counties met the 20 percent threshold level. The working group agreed that future analyses would have to utilize a methodology which assured some examination of the Indian populations of those two states. However, it was also agreed to apply the same standards at this time as applied to county analyses for the other minority groups. 431 In all states with sizable American Indian populations, those counties with 20 percent and over Indians have substantially lower physician to population ratios and greater percentages of persons living below the poverty level than all other counties in the states. The counties with these Indian populations also show lower numbers of hospital inpatient days and slightly higher infant mortality. (See Table 17). Table 17 Comparison Among States of Key Indicators: Counties with 20 Percent or More American Indians and All Other Counties Active M.D. Infant % Population Number Physicians/ Mortality below poverty Inpatient 1000K pop. Rate level days/1000 pop. State MC AOC MC AOC MC AOC MC AOC Arizona 69 189 13.0 12.3 28.5 12.1 533 1,123 Montana 59 137 19.1 11.8 20.4 11.4 1,918 1,482 Nebraska 97 149 31.9 11.3 23.9 10.4 2,143 1,674 New Mexico 92 164 10.9 11.7 27.1 11.6 1,610 1,932 North Carolina 72 154 18.4 14.4 24.8 14.3 1,067 1,244 North Dakota 67 139 11.8 12.1 27.1 11.6 1,610 1,932 Oklahoma 42 132 10.2 12.8 22.8 13.0 590 1,273 South Dakota 26 121 19.4 10.0 39.4 14.7 448 1,915 Utah 49 170 8.5 10.4 32.9 10.3 489 815 Wisconsin 0 161 8.9 10.3 NA 8.5 NA 1,399 Average 57 152 15.2 11.7 27.3 12.3 1,053 1,389 MC = Minority county(ies) AOC = All Other Counties When 1970-1980 changes in health indicators are examined, there are three mortality areas in which counties with 20 percent and over American Indian population are worsening: (a) deaths from cancer are increasing in the counties with high proportions of Indians at a faster rate than in all other counties; (b) deaths from cirrhosis of the liver increased n average of 45.7 percent in Indian-population counties but dropped in all other counties but dropped in all other counties by an average of 15.1 percent; (c) there was an increase in almost all Indian populated counties of deaths in the 15-24 age group while at the same time, deaths in this age group were falling in all other counties. (See Table 18). These findings, from Bureau of Health Professions ARF file data, are consistent with Indian Health Services records of trends in Indian Health. 432 The total U.S. Indian infant mortality rate has been falling and by 1979 was 14.1 compared to 13.1 for the entire U.S. Alcoholism-related mortality among Indians is six times greater than for the U.S. population. And, common causes of deaths of young adults—accidents, suicide and homicide—are significantly higher in the Indian population than in the total U.S. population. Brief narratives for the 10 states included in the analysis follow. They relate the socio-economic disadvantage of the inhabitants of these high density American Indian counties, as well as the relative scarcity of health professions non-Federal resources. This is followed by a brief discussion of the Indian Health Service. Table 18 Compairson Among States of Selected Mortality Patterns, 1970-1980: Counties with 20 Percent or More American Indians and All or Other Counties Mortali tv Rate Percent Change 1970 - 1980 Cirrhos MC is of Liver AOC Cancer 15-24 Age MC i Group | 1 State ML AOC AOC I Arizona +88.6 -22.5 +26.6 +21.5 -38.1 -16.6 | I Montana +103.9 -12.9 +19.0 +7.8 +18.5 +0.7 | I Nebraska +92.0 -29.4 +2.4 +9.8 +100.0 -11.2 | New Mexico -27.2 -26.4 +28.9 +24.7 +20.2 -12.5 | I North Dakota +98.0 +8.8 +23.0 +8.3 +32.7 +9.3 | I Oklahoma -23.6 -21.8 +26.2 +3.4 +15.3 -1.1 1 I South Dakota +20.9 -3.2 +3.2 +9.4 -7.8 -17.6 | I Utah +55.1 -30.4 +93.8 -3.3 +62.3 -22.5 | I Wisconsin NA NA NA NA NA NA | +45.7 -15.1 +27.1 +12.2 +22.6 -9.7 | MC = Minority County(ies) AOC = All Other Counties New Mexico 32 total counties Counties with 20 percent or more American Indians: McKinley 65.7 percent San Juan 33.0 percent Sandoval 27.2 percent Rural Population. The percentage of rural population in minority counties was twice that in all other counties in New Mexico, 48.4 to 24.7. 433 Income. Personal income by household was almost equal for both minority counties and all other counties, $23,361 and $23,372 respectively. However, in minority counties the population percentage below the poverty level was 26.0 vs. 16.3 in all other counties. Health Professions Schools. There are 13 health professions schools in New Mexico. Of these, only two nursing schools are in minority counties and, only 23 of 415 nursing graduates are from the two minority counties' schools. Health Practitioners. Active M.D. physicians per 100,000 population are 164 in all other counties and 92 in minority counties. There are also disproportionately small numbers of dentists, veterinarians, optometrists, and pharmacists, and there are no podiatrists in minority counties. The RN to population ratios are relatively close. From 1970 to 1980, the minority counties have shown a greater increase in M.D. physicians than in all other counties. The number of minority counties M.D.'s increased by 133.3 percent compared to 45.2 percent in all other counties. None of the minority counties had a decreased physician/population ratio while 6 of 29 all other counties had a lower ratio in 1980 than in 1970. The minority counties increase held for all M.D. specialities, for example, pediatricians increased from 3 to 8 and surgical specialists increased from 15 to 26 between 1975 and 1979. Hospital Utilization. In minority counties ambulatory visits per 1,000 population were twice those in all other counties but inpatient days were only two-thirds of what they were in all other counties. The change in both these ratios between 1970 and 1982 widened the difference in both types of utilization between minority counties and all other counties in the State. Infant Mortality. Infant deaths dropped in both minority counties and all other counties between 1970 and 1980 but are slightly higher in all other counties, 11.7 to 10.9 per 1,000 live births. Utah 29 total counties County with 20 percent or more American Indians: San Juan 45.9 percent Total Population. The one minority county in Utah contains only 12,000 persons of whom 5,500 are American Indians. Income. 32.9 percent of the minority county population lives below the poverty level compared to 10.3 percent of all other tween 1970 and 1980 but are slightly higher in all other counties, 11.7 to 10.9 per 1,000 live births. Key County. Personal income by household is about $6,000 less. Thirteen percent of the minority county population receives AFDC while only two percent of all the other counties population are recipients. 434 Health Practitioners. Although the minority county has only a 12,000 population one would expect more health practitioners than there are: six general practitioners and no surgeon, osteopath, dentist, veterinarian, podiatrist, optometrist or pharmacist. All other counties appear to have a normal complement of those practitioners. There are nurses in the minority county but they too are well below the ratio for the rest of the State. Hospital Utilization. Minority county inpatient days per 1,000 population are low and dropped almost 50 percent between 1970 and 1982. In all other counties inpatient days increased. Infant Mortality. Between 1970 and 1980 American Indian infant mortality fell from 31.6 to 6.1. This 1980 rate is below the total for all other counties and the rate for nonminorities and Blacks. Other Mortality. There were increases between 1970 and 1980 in minority county death rates from cirrhosis of the liver and cancer which did not occur in all other counties. There were large increases in the death rates for several age groups but this may be attributable to the small size of the population. Wisconsin 71 total counties County with 20 percent or more American Indians: Menominee 89.4 percent Rural Population. The counties minority population is 100.0 percent rural in a State which is 35.8 percent rural. Total Population. About 60 percent of the 3,000 people in the minority county are younger than 25. In all other counties only 42 percent of the population is under 25. Income. The AFDC level in the minority county is extremely high: 31 percent of the population are receiving assistance compared to 4 percent in all other counties. However, those below the poverty level in the minority county dropped by 38 percent between 1969 and 1979 vs. a 6 percent decrease in all the other counties. Health Practitioners. In a minority county of 3,000 people the only health practitioner is one general practitioner. There are no others, not even a nurse. Hospital Utilization. Data are not available. Infant Mortality of American Indian in the minority county has decreased by half, to 9.3. Total infant mortality for the rest of the State is 10.3. Other Mortality. Data are not available. 435 Arizona 14 total counties Counties with 20 percent or more American Indians: Apache 74.9 percent Navajo 47.6 percent Coconino 27.9 percent Rural Population. Although the State population is largely urban, 57 percent of the counties minority population is rural. Income. There is a rather wide disparity between minority counties and all other counties poverty population percentages. 28.5 percent of the population in minority counties and 12.1 percent in all other counties are poor. This population increased in both areas though by only 19 percent in minority counties and 35 percent in the rest of Arizona. Health Professions Schools. In the minority counties there are two nursing schools graduating 36 nurses in 1979. There are also a number of allied health schools. Health Practitioners. Arizona's non-twenty percent or more American Indian population counties have the highest M.D. ratio, (189), of all States considered in this analysis but the minority counties M.D. ratio is still only 69. There are no podiatrists but sizable numbers of dentists, optometrists and pharmacists. The ratio of nurses approximates that of minority counties in other States which have nursing schools located in them. In the late 1970s, there were increases in all health professionals except pharmacists. Hospital Utilization. In the minority counties, the ratio of ambulatory visits to inpatient days is six to one. In all other counties in Arizona the ratio is closer to one and one-half to one. Infant Mortality. In both minority counties and all other counties infant mortality is about equal but it appears that American Indian infant mortality in the minority counties has been reduced more than that for any other group. Other Mortality. Cirrhosis of the liver mortality dropped by 23 percent in all other counties but increased by 89 percent in minority counties. North Dakota 53 total counties Counties with 20 percent or more American Indians: Sioux 64.7 percent Rolette 57.6 percent Benson 28.7 percent Rural Population. The entire counties' minority population is rural. One-half of all the other counties population is rural. 436 Income. Almost 14 percent of minority counties population receives AFDC vs. one and one-half percent of all other counties population. The minority counties poverty population is 27 percent compared to 11.6 percent in all other counties. In both areas the poverty population diminished by about the same amount between 1970 and 1980. Health Practitioners. The M.D. physician population ratio in all other counties is twice that in the three minority counties, 139 to 67. Minority counties M.D.'s did increase by 8, 125 percent, between 1970 and 1980. The only M.D.'s identified in the minority counties are GP's; there are no specialists in a population of 24,000. Hospital Utilization. There is an extremely high ambulatory visit ratio in the minority counties—seven and one-half per year for each resident. This is so even though the inpatient day ratio is nearly as high as in all other counties. Infant Mortality. For both the minority counties and all other counties the rate is about the same, 11.8 and 12.1. Other Mortality. In the minority counties death from cirrhosis of the liver rose by 98 percent between 1970 and 1980 and there wee marked increases in death rates for certain age groups, including the 15-24 ages. Nebraska 93 total counties County with 20 percent or more American Indians: Thurston 34.0 percent Rural Population. The minority county rural population is 100 percent rural; in all other counties the rural population is 36.8 percent. Income. Nine percent of the minority county population receives AFDC assistance while only two percent of all the other counties population receives AFDC. The personal income by household in the minority county is only about $2,500 lower than that for all other counties. The poverty level difference, however, is between 23.9 percent (minority county) and 10.4 percent (all other counties). Health Practitioners. With a total population of 7,000 the minority county has no dentists, veterinarians, podiatrists, or optometrists. The active M.D. physician population ratio is 97 per 100,000, highest among the minority county considered in this analysis. Hospital Utilization. Both in ambulatory visits and inpatient days the minority county population is well above the rest of the State, e.g., the ambulatory visit ratio is three and one-half times that of all other counties. 437 Infant Mortality. Even though hospital utilization is very high in the minority county, its infant mortality rate is 31.9 compared to 11.3 in all other counties. Other Mortality. Cirrhosis of the live was 92 percent higher in 1980 than in 1970 in the minority county. It decreased by 30 percent in all other counties. The minority county 15-24 and 35-44 age groups showed marked increases in 1980, 100 percent and 167 percent, respectively. Oklahoma 75 total counties Counties with 20 percent or more American Indians: Adair 33.4 percent Cherokee 26.0 percent Delaware 20.7 percent Rural Population. 82.2 percent of the minority counties population is rural vs. 31.5 percent of the population in all other counties. Total Population. The minority counties and all othe counties populations are similar in that both are predominantly nonminority. The minority counties population is 26.2 percent American Indian; all the other counties it is 5.2 percent American Indian. Income. Per capita income and personal household ncome in the minority counties are substantially below that in all other counties. In fact, household income in all other counties is over $10,000 more than in the minority counties. Between 1969 and 1979, the poverty level populations of both minority counties and all other counties were diminished by about 15 percent but there is still a larger percentage of the minority counties population defined as living in poverty—22.8 to 13.0 percent. Health Professions Schools. As in most other American Indian minority counties, there are no health professions schools. Health Practitioners. The number of minority counties M.D.'s had remained relatively stable between 1940 and 1970 but between 1970 and 1980 increased by 12, or 75 percent. There were also increases in minority counties pharmacists, optometrists, dentists, and veterinarians. Hospital Utilization. The inpatient days to ambulatory visit ratio is about one to one in all other counties but one to three in minority counties. Infant Mortality. Infant mortality is slightly lower in the minority counties than in the rest of Oklahoma. Total Mortality. Deaths of young adults and deaths by motor vehicle accidents were significantly increased between 1970 and 1980 in minority counties but not in all other counties. 438 Montana 57 total counties Counties with 20 percent or more American Indians: Big Horn 46.2 percent Glacier 45.9 percent Roosevelt 36.9 percent Blaine 31.8 percent Rosebud 24.3 percent Rural Population. A large portion of the population in all other counties is rural (45.3 percent) but an even larger part of the minority counties population is rural (74.5 percent). Income. Per capita income in both minority counties and all other counties are about the same and have increased from approximately the same base in 1970. However, the percentage of population in poverty is almost twice as high in the minority counties (20.4) as in all other counties (11.4). Health Professions Schools. There are none in the five minority counties and there are no junior colleges or colleges. Montana has 13 colleges, junior colleges, and universities, four RN nursing schools, one pharmacy school and numbers of allied health schools. Health Practitioners. Considering the 49,000 population of the minority counties, there appear to be adequate numbers of dentists, optometrists, pharmacists, and nurses. The M.D. physician population ratio, however, is 49 compared to 137 in all other counties. Hospital Utilization. The pattern in ambulatory visits and inpatient days is similar to that between most other American Indian minority counties and all other counties: the minority counties population is much higher in ambulatory visits and lower in inpatient days. Infant Mortality. Infant mortality among the minority counties American Indian population remains much higher than it is for the total of all other counties population—18.4 to 11.8, even though it was reduced by one-half between 1970 and 1980. Other Mortality. Deaths attributable to cirrhosis of the liver increased by more than 100 percent while decreasing in all othe counties. The death rate for the 15-24 age group also increased significantly in the minority counties; it did not in all other counties. 439 South Dakota 67 total counties Counties with 20 percent or more American Indians: Shannon Todd Buffalo Ziebach Dewey 93.4 percent 77.6 percent 70.8 percent 58.1 percent 57.9 percent Corson Jackson Mellette Bennett Lyman 47.3 percent 43.4 percent 38.8 percent 38.6 percent 23.5 percent Rural Population. Almost the entire minority counties population is rural (93.2 percent) while 50.7 percent of all the other counties population lives in rural areas. Total Population. Two-thirds of the minority counties population is American Indian but only 2.6 percent of the population in othe counties in South Dakota is American Indian. Income. The poverty population in minority counties is almost 40 percent compared to 15 percent in all other counties. The AFDC ratio in minority counties is six times higher than in all other counties—12,814 to 2,292. The poverty population increased in minority counties between 1970 and 1980 while it was dropping in all other counties. Health Professions Schools. There are no health professions schools in the minority counties and 43 in the State, including 25 dental assistant schools. Health Practitioners. There are no podiatrists or optometrists in the minority counties even though the minority counties have a population of 46,000. There are three dentists and three pharmacists but this is substantially below their representation in the rest of the State. Even the nurse ratio to population is only a third in the minority counties of what it is in all other counties. The M.D. physician to population ratio (26) in the minority counties is the second lowest among the 8 States being analyzed. Hospital Utilization. Ambulatory visits per 1,000 population are 4,157 compared to 448 inpatient days per 1,000 population in the minority counties. Comparable numbers in all othe counties are 1,059 and 1,915. Infant Mortality. Infant mortality is twice as high in the minority counties as in all other counties. Other Mortality. Deaths from cirrhosis of the liver significantly increased from 1970 to 1980 and deaths in the 5-14 age group also were much higher in 1980 than in 1970 in minority counties but not in all other counties in South Dakota. 440 North Carolina 100 total counties Counties with 20 percent or more American Indians: Robeson 34.9 percent Swain 24.2 percent Rural Population. Three-quarters of the minority counties population is rural compared to slightly over one-half all other counties population. Total Population. 34.0 percent of theminority counties population is American Indian but only 0.5 percent of the population in all other counties in North Carolina is American Indian. Almost the entire American Indian population of North Carolina is in two minority counties. Income. Personal income by household is appreciably higher in all other counties, $22,455 to $17,588 although percentage increase since 1970 has been higher in the minority counties. Ten percent more of the minority counties population than of all the othe counties population is still below the poverty level. Health Professions Schools. There are no health professions graduates in the minority counties. There are also no LPN, dental assistant, or dental hygiene schools in the minority counties but 58 in the State. Health Practitioners. The active M.D. ratio is twice as high in all other counties as in minority counties, 154 per 100,000 population to 72. Both minority counties are considered medically underserved and 71 of 98 all other counties are MUA's. There are 27 National Health Service Corps sites in North Carolina; one in minority counties. Nurses, like M.D.'s are twice as well represented in all other counties as in minority counties. In all other counties there has been a gain in all medical specialities but in the minority counties there has been a drop in specialists in internal medicine, obstetrics/gynecology, and surgery. Hospital Utilization. Ambulatory visits are about the same for both minority counties and all other counties but inpatient days are roughly 20 percent higher in all other counties. Infant Mortality. The infant mortality rate in minority counties is high, 18.4, but has decreased from 25.6 in 1970. The rate in all other counties was almost as high in 1970 but fell further by 1980, to 14.4. Other Mortality. Mortality for all age groups fell markedly from 1970 to 1980 in minority counties except the 15-24 age group which rose slightly. 441 Indian Health Service Another major factor which must be considered in any review of Indian communities is the varying presence and role of the Indian Health Service. More time and specific data are required to complete the review of the role of this important resource. The following tables present available information on the location and distribution of Indian Health Service resources. Table 19 Physician (M.D.) Per 100,000 Population Ratio and Indian Health Service Hospitals and Health Centers, in Counties with 20 percent or more American Indian Population Non-Federal Physician Ratio/Counties IHS Fac ilities 20% or more Less than 20% Hospitals American Indian American Indian and Med. Health State Population 69 per 100,000 Population 189 Centers 4 Centers Arizona 2 I Montana 59 per 100,000 137 3 2 Nebraska 97 per 100,000 149 1 1 New Mexico 92 per 100,000 164 3 1 I North Carolina 72 per 100,000 154 0 0 I North Dakota 67 per 100,000 139 2 0 Oklahoma 42 per 100,000 132 1 1 I South Dakota 26 per 100,000 121 3 2 1 Utah 49 per 100,000 170 0 0 Wisconsin 0 per 100,000 161 0 0 In some states with counties having 20 percent or higher American Indian population, IHS health services in the counties help to compensate for low non-Federal physician ratios. In others like North Carolina, Utah, and Wisconsin there are no IHS facilities in those counties with high American Indian populations. 442 Table 20 Number of Counties Having Over 10 Percent American Indians With Indian Health Service Facilities Counties With Facilities Total Counties Hospitals Health Centers Either Hospital or Health Center 70 No. % 22 31.4 No. % 26 37.1 No. % 44 62.9 Although only about a third of the counties have hospitals and a third have health centers, there is little duplication of the two. Almost two-thirds of the counties have one or the other. Hospitals are much more likely to be found in the counties with over 20 percent American Indian population while health centers are more common in those counties with under 20 percent American Indian populations have neither an IHS hospital nor a health center. There are substantial IHS resources available but there is not uniform availability to the entire U.S. American Indian population. If Oklahoma is excluded from the count because it has no reservations exactly two-thirds of the counties have reservations within or on their borders. Of 19 counties with over 35 percent American Indian population, only one is not a reservation county. Only eight of 37 counties with 10-20 percent American Indian populations are counties with reservations. 443 Table 21 Counties with Over 10 Percent American Indian Population American (%) American Indian Indian Population Population County/Stat e OK 1980 1980 McCurtain 3,626 10.0 Okanogan WA 3,111 10.2 Mineral NV 645 10.4 Atoka OK 1,345 10.6 Neshoba MS 2,515 10.6 Pushmataha OK 1,253 10.6 Sawyer WI 1,385 10.8 Johnston OK 1,171 11.3 Rio Arriba NM 3,379 11.5 Fremont WY 4,501 11.5 Coal OK 719 11.9 Graham AZ 2,730 11.9 Mountrail ND 917 11.9 Osage OK 4,727 12.0 Mcintosh OK 1,891 12.2 Craig OK 1,834 12.2 Hoke NC 2,557 12.5 Beltrami MN 3,920 12.7 Hill MT 2,293 12.7 McKenzie ND 911 12.8 Ottawa OK 4,203 12.8 Seminole OK 3,719 13.5 Hughes OK 1,944 13.6 Mayes OK 4,389 13.6 Gila AZ 5,154 13.9 Latimer OK 1,385 14.1 Valencia NM 8,636 14.1 Sequoyah OK 4,462 14.6 Okfuskee OK 1,627 14.6 Alpine CA 169 15.4 Alaska 64,357 16.0 Lake MT 3,162 16.6 Ferry WA 978 16.8 Jefferson OK 2,030 17.5 Charles Mix SD 1,709 17.7 Caddo OK 5,525 17.9 Mahnoman MN 1,008 18.4 Roberts SD 2,110 19.3 Delaware OK 4,960 20.7 Lyman SD 907 23.5 Swain NC 2,493 24.2 Rosebud MT 2,406 24.3 Cherokee OK 7,987 26.0 Indian Health Service Facilities Hospital Health Center Reservati In or Touch County X X X (6) X X X X (2) X X X (4) X X X X X X X X 444 Table 21 (continued) American (%) American Indian Indian Indian Health Service Population Population Faci! Lities County/Stat e NM 1980 1980 27.2 Hospital Healt :h Center Sandoval 9,471 Coconino AZ 20,949 27.9 X Benson ND 2,277 28.7 X Blaine MT 2,223 31.8 X San Juan NM 26,866 33.0 X Adair OK 6,210 33.4 Thurston NE 2,443 34.0 X X Robeson NC 35,507 34.9 Roosevelt MT 3,865 36.9 X Bennett SD 1,174 38.6 Mellette SD 872 38.8 Jackson SD 1,491 43.3 San Juan UT 5,622 45.9 Glacier MT 4,882 45.9 X Big Horn MT 5,126 46.2 X Corson SD 2,459 47.3 X Navaj o AZ 32,215 47.6 X (2) X (2) Rolette ND 7,020 57.6 X Dewey SD 3,107 57.9 Ziebach SD 1,342 58.1 X Sioux ND 2,341 64.7 X McKinley NM 37,115 65.7 X (2) X Buffalo SD 1,270 70.8 X Apache AZ 39,042 74.9 X (2) Todd SD 5,688 77.6 X Menominee WI 3,014 89.4 Shannon SD 10,575 93.4 X In NOTES: Oklahoma does not have reservations. Alaska is not divided by counties therefore the entire state i included in the table. 445 Table 22 Recent-experienced Civilian Labor force 1/ In Health Occupations and Health Occupations ClOsters in the United States by Racial/Ethnic Category: April 1, 1980 All Races Total minority 2/ Black (not Hispanic) Hispanic Asian/ White Native Pacific Other (not American Islander Minority Hispanic) 26,840 163,220 6,070 4,301,270 560 1.510 50 96.100 50 1,230 60 17.220 510 41,920 890 357,840 190 3,830 70 115,880 60 430 30 32,920 40 560 10 23,310 0 100 0 7,340 30 580 20 20.330 3.860 42,950 1.270 1,114,310 250 6,580 170 130,440 330 2,480 60 47,510 210 970 30 39.840 50 430 20 16.160 110 930 40 37,800 80 300 30 30,100 240 620 60 35,650 210 630 30 24,810 0 360 40 8,080 70 350 10 17,520 860 13,500 240 191,210 40 550 20 44,130 180 440 30 12.420 380 1,060 100 82,100 2,790 5,900 440 333.130 B90 3,970 240 120.630 830 2,440 80 139,750 1,920 4,830 250 215,270 11,690 19,700 1.7*0 901,210 190 910 20 40,660 Total, Health Occupations..... 5.442,160 Managers, Medicine «. Health... 110,880 Medical Scientists............ 20.070 Physicians.................... 433,260 Den11 s ts...................... 125,290 Veterinarians................. 34,360 Optometrists.................. 24,610 Podiatrists................... 7,780 Health diagnosing practitioners n.e.c......... 21,540 Registered nurses 3/.,........ 1,205,300 Pharmac Ists................... 145,640 Dietitians and dietetic , techn lc lans................. 67,270 Inhalation therapists......... 48,740 Occupational therapists....... 17,760 Physical therapists........... 43.000 Speech therapists............. 41,300 Therapists n.e.c.............. 43,070 Physicians' assistants 4/..... 30.440 Medical science teachers...... 9,010 Health specialties teachers... 19,550 Clinical laboratory technologists and techn lc lans................ 243,980 Dental hyglenlsts............. 46,190 Health record technologists and technicians... <......... 15,150 Radiologic technicians........ 96,310 Licensed practical nurses..... 435,100 Health technologists and technicians n.e.c........... 152.540 Dental assistants............. 158.120 Health aides, except nursing.. 292,050 Nursing aides, orderlies and attendants.............. 1,370.120 Optical goods workers......... 46,870 Denta I lab and medical appliance technicians....... 40,750 1,140,900 14,800 2,050 75.410 9,410 1.430 1,290 440 1,210 170,990 15,210 19,750 8.910 1,600 5,290 3.240 7,410 5,630 930 2,030 52,770 2.060 2,730 14.240 102,040 31,910 18,300 76,780 476,920 6,220 9,000 73,503 9,480 880 13,240 3.130 520 250 280 260 95,370 4,720 14,400 5,110 770 2.930 2,040 5,100 3,010 260 1,230 28,000 700 1,480 7.900 77,050 20,170 6,640 52,300 372,210 1,960 2,770 209,740 3,100 630 18.850 2,190 390 430 60 320 27,540 3,490 2,400 2,590 330 1,280 710 1,390 1.750 200 370 10,100 750 600 4.000 15.060 6.640 8.390 17.480 71,500 3,140 3,760 220 2,280 50 39,670 1/ The "recent experienced" civilian labor force Is defined as civilian persons employed 1n 1980 or unemployed having civilian work experience between 1975-1980. 2/ Includes all race/ethnicIty categories other than white. Due to Independent rounding figures may not add to totals. 3/ According lo data from the 1900 National Sample Survey of Registered Nurses the total number of R.N.'s employed In nursing in 1900 was 1,272,851. Of these 1,151,221 were white (non-hlspanlc), 54,585 Black (non-hlspanlc) 30,470 Aslan/PaclMc Islander, 3,045 American Indian/Alaskan Native, 17,938 Hispanic and 15,592 of unknown race/ethnicity. 4/ According to data published In the Third Report to the President and Congress on the Status of Health Professions Personnel there were 11,000 PAs In the U.S. of whom 6\B00 were esFlmated to be active. 446 Table 23 Percent Distribution of the Recent-experienced Civilian Labor Force 1/ In Health Occupations and Health Occupations Clusters In the United States by Racial/Ethnic Category: April 1, 1980 Black All Total (not Races minority 2/ Hispanic) Hispanic Asian/ White Native Pacific Other (not American Islander Minority Hispanic) Total, Health Occupations..... 100.0 Managers, Medicine fc Health... 100.0 Medical Scientists............ 100.0 Physicians.................... 100.0 Dentists...................... 100.0 Voter inar lans................. 100.0 Optometrists.................. 100.0 Pod latr ists................... 100.0 Health diagnosing practitioners n.e.c......... 100.0 Registered nurses 3/.......... 100.0 Pharmac Is ts.......T........... 100.0 Dietitians and dietetic techn lc lans................. 100.0 Inhalation therapists......... 100.0 Occupational therapists....... 100.0 Physical therapists........... 100.0 Speech therapists............. 100.0 Therapists n.e.c.............. 100.0 Physicians* assistants 4/..... 100.0 Medical science teachers...... 100.0 Health specialties teachers... 100.0 Clinical laboratory technologists and techn lc ians................ 100.0 Dental hyglenlsts............. 100.0 Health record technologists and technicians............. 100.0 Radiologic technicians........ 100.0 Licensed practical nurses..... 100.0 Health technologists and technicians n.e.c........... 100.0 Dental assistants............. 100.0 Health aides, except nursing.. 100.0 Nursing aides, orderlies and attendants.............. 100.0 Optical goods workers......... 100.0 Dental lab and medical appliance technicians....... 100.0 21.0 13.5 3.9 0.5 3.0 0.1 79.0 13.3 8.5 2.9 0.6 1.4 0.0 86.7 14.2 4.4 3.1 0.2 6.1 0.3 85.8 17.4 3.1 4.4 0.1 9.7 0.2 82.6 7.5 2.5 1.7 0.2 3.1 0.1 92.5 4.2 1.5 1.1 0.2 1.3 0.1 95.8 5.2 1.0 1.7 0.2 2.3 0.0 94.7 5.7 3.6 0.8 - 1.3 - 94.3 5.6 1.2 1.5 0.1 2.7 0.1 94.4 13.3 7.4 2.1 0.3 3.3 0.1 86.7 10.4 3.2 2.4 0.2 4.5 0.1 B9.6 29.4 21.4 3.7 0.5 3.7 0.1 70.6 18.3 10.5 5.3 0.4 2.0 0.1 81.7 9.0 4.3 1.9 0.3 2.4 0.1 91.0 123 6.8 3.0 0.3 2.2 0.1 87.7 7.8 4.9 1.7 0.2 0.9 0.1 92.3 17.2 11.8 3.2 0.6 1.4 0.1 82.8 18.5 9.9 5.7 0.7 2.1 0.1 81.5 10.3 2.9 3.1 0 4.0 0.4 89.7 10.4 6.3 1.9 0.4 1.8 0.1 89.6 21.6 11.5 4.1 0.4 5.5 0.1 78.4 4.5 1.5 1.6 0.1 1.2 0.0 95.5 18.0 9.8 4.0 1.2 2.9 0.2 82.0 14.8 8.2 4.2 0.4 1.9 0.1 85.2 23.4 17.9 3.5 0.6 1.4 0.1 76.6 20.9 13.2 4.4 0.6 2.6 0.2 79.1 11.6 4.2 5.3 0.5 1.5 0.1 88.4 26.3 17.9 6.0 0.7 1.7 0.1 73.7 34.6 27.0 5.2 0.8 1.4 0.1 65.4 13.3 4.2 6.7 0.4 1.9 0.0 86.8 18.6 5.7 7.7 0.5 4.7 0.1 81.4 1/ The civil recent experienced" civilian labor force Is defined-as civilian persons employed In 1980 or unemployed having ian work experience between 1975-1900. 2/ Includes all race/ethnlclty categories other than white. Due to Independent rounding figures may not add to totals. 3/ According to data from the 1900 National Sample Survey of Registered Nurses the distribution of employed R.N.'s by racial/ethnic categories was: White (non-hlspanlc) 90.4 percent, Black (non-hlspanlc) 4.3 percent, Asian/Pacific Islander, 2.4 percent, American Indian/Alaskan Native, 0.2 percent. Hispanic 1.4 percent, and 1.2 percent of unknown raclal/ethntc background. 4/ According to data published In the Third Report_to the President and Congress on the Status of Health Professions ~ Personnel there were 11,000 PAs In TKe ITS" orwhom~B,B00 were esflmalTd IcTEe active. 447 PART II. B. - Minority Health Professionals One factor which the working group has viewed to be crucial in assuring the availability (and enhancing accessibility) of health professionals to minority populations and communities, is the availability of minority health professionals who reflect the racial, ethnic, cultural and other characteristics of those communities. Later in this paper a discussion is presented to support this view. At this point, however, an effort was made to briefly present some statistics reflecting the degree of the participation of these minority groups in the health professions. Since the actual presence of these health providers in minority communities was the desired measure of availability, data were obtained to document the degree of this presence. Part II. B. 1 presents this discussion of Distribution of Minority Health Professionals, while Parts II. B. 2 and II. B. 3. respectively, discuss the Development of these professionals and their Practice Patterns. PART II. B. 1. - Distribution of Minority Health Professionals A review of data from the 1980 Census shows, from a national perspective, varying degrees of participation among the four minority groups, among the several health professions and occupations listed. Tables 22 and 23 provide the detail of these census data, by number of persons and percent distribution, respectively. Among mental health professionals, there are only small percentages of minorities. Asian-Americans are represented in psychiatry and almost six percent of social workers are Black, but otherwise mental health practitioners are almost all nonminority. See Table 24. Both at the M.A. and Ph.D. levels there are minimal numbers and percentages of minority clinical psychologists. See Table 25. The overwhelming majority of staff in these units are nonminority. There are slightly more Native Americans and Hispanics in the alcoholism only units than in the combined units. See Table 26. With few exceptions, the census data show underrepresentation of the minority groups when compared to their percentages in the total population. This is especially so for the more prominent professions, i.e. among registered nurses, physicians, dentists, etc. The degrees of underrepresentation vary widely, but the generally common thread is that the minority numbers are disproportionately low. 448 Table 24 Percent Practicing Professionals by Race/Ethnicity Asian/ Non- American Pacific Minority Black Indian Islander Hispanic Unknown | | Psychologists 95.6 1.4 0.2 1.1 0.7 0.9 | I Psychiatrists 69.4 1.5 0.4 5.6 2.5 20.7 I | Social Workers 88.5 5.8 NA 1.6 1.8 2.2 | I Nurses 91.5 3.7 0.3 2.0 1.2 1.3 I NOTE: Primary data sources are membership surveys of the American Psychological Association (1982), American Psychiatric Association (1977), National Association of Social Workers (1982), a HRSA survey of registered nurses (1980), and an NLN survey of 1980-81 graduations of minority students from basic baccalaureate nursing programs. Table 25 Number of U.S. Clinical Psychologists Asian/ Non- American Pacific I Level Total Minority Black Hispanic Indian Islander Unknown I Ph.D. 16,519 15,885 159 164 182 6 123 I Master's 2,486 2,360 38 19 29 8 33 Percent U.S. Clinical Psychologists Asian/ Non- American Pacific I Level Total Minority Black Hispanic Indian Islander Unknown Ph.D. 100.0 96.2 1.0 1.0 1.1 0.4 0.7 | Master's 100.0 94.9 1.5 0.8 1.2 0.3 1.3 449 Table 26 Racial and Ethnic Characteristics of Staff in Units Providing Alcoholism Treatment Only and in Units Providing Combined Alcoholism and Drug Abuse Treatment Alcoholism Only No. % Combined Total | I Race/Ethnicity No. % No. % | I American Indian 718 2.5 252 1.5 970 2.2 I Asian/Pacific Isl 308 1.1 193 1.2 501 1.1 | 1 Black 3,719 13.3 2,116 13.3 5,835 13.3 | I Hispanic 1,414 5.0 593 3.7 2,007 4.5 | Nonminority 21,608 77.8 12,721 80.1 34,329 78.6 | j Total 27,767 100.0 15,875 100.0 43,642 100.0 | The exception to the aforementioned trend appears to be the Asian/Pacific Islander minority group which seems frequently "over represented." One concern in making such a statement, however, is that closer review of selected census files shows significant variation between the subgroups of both Asian/Pacific Islander and Hispanic health professions. Table 27 displays this data: Table 27 Percent Distribution of Hispanic and Asian/Pacific Islander Professionals by Subpopulation, 1980 Physicians Dentists Pharmacists Hispanic 100 100 100 | -Mexican 19 25 41 | | -Puerto Rican 8 4 11 1 | -Cuban 26 28 23 | I -Other Spanish 48 40 25 | | Asian/Pacific I slander 100 100 100 | | -Japanese 5 39 18 | | -Chinese 16 29 41 | | -Filipino 25 8 7 | | -Korean 10 8 10 | | -Asian Indian 37 13 18 | -Vietnamese 2 1 3 I | -Hawaiian and o ther 5 2 3 | Pacific Islander Thus apparently while some subgroups appear quite overrepresented, others show patterns similar to other underrepresented minorities. 450 Again, more detailed data are required to complete the analysis and answer specific questions for specific communities of subpopulations. Nonetheless, the process was continued to review and display the data which were available. The four graphs which follow present national statistics for minority physicians (Graph 1), dentists (Graph 2), pharmacists (Graph 3) and registered nurses (Graph 4) per 100,000 minority population for all four minority groups. [Note with caution - the scale of magnitude (Y-axis) changes with each graph.] 451 U.S. Manpower/Population Ratios Physician/100,000 Ethnic/Racial Population In 1975 (1970), there were 731.2 Asian/Pacific Islander physicians per 100,000 Asian/Pacific Islander population. By 1980 this was 1,197.4. For Black Americans this figure per 100,000 Black population was 26.6 in 1970 and 50.7 in 1980. For American Indians the figure was 22.9 in 1970 per 100,000 American Indian population and 36.0 in 1980. For Hispanic Americans the figure was 113.9 per 100,000 Hispanic population in 1970 and 129.1 per Hispanic population in 1980. For nonminorities (which includes 96.26 percent nonminority Americans) the figure in 1970 was 146.4 per 100,000 nonminority population and in 1980 it was 197.8. Graph 1 Minority Physicians per 100,000 Minority Population in the U.S., 1975 and 1980 1197 oo - 731 xxx - .*•■ ./ . '•• '"■• '. 198 .•■ ■■■' / SS'' HA 129 ±40 "*'l ''t "*' ii^f y' .••' y nn UU - / .•■ . ' ■■ *•• *■•. •■' ••■■' / //' ■•■ / K "•■■•"""'■■ .'' 51 ..•••■' .'y / 36 ' X '■'■. • / . ••■'' / .-■'' 27 '■•■-. '•., "'•■ /./"'/ \ V '■•• *■•. '•. - 23 '"V/".. i ■-•!''•••''' ..■•* .<•■ .<■ ■>' .-■' ; '••.''••.'' S* y' '■•■•! '••• '■ • / s '••■ ''••. '"'• ■■> '•-. '■ ■ .-■' .•■' "■■> '••. "' t' .' "'•• '•• s ' / / . 10 _ ■>'// .■■' y t X. X v "'■"• *■•«. "'* W-X ■•• \ '■•-. /' ..' .' ' / X X-' \X ■•.. \ •-., / s .■•' •' / /"' . '--X.X x$ / S ,'1 . xx''' ■••.. \ •> **-. V X > .■•' .. < •■' i • X x / / t . '••• X' / / y / .,.•• ,< •» "'•■ '• 1 ■-. '■■•• *■ ■' y / ■■' .' t ' y / \. •-, '< ' <■' .'■' ''■•• X *■ ' •*''. '■■■. X ''- / ' . '■■/','"'• Asian Black Hisp. 1975 Amer. Indian 1980 Nonmin. 452 U.S. Manpower/Population Ratios Dentist/100,000 Ethnic/Racial Population In 1970 there were 89.4 Asian/Pacific Islander dentists per 100,000 Asian/Pacific Islander population. In 1980 the number was 109.4 Asian/Pacific Islander dentists per 100,000 Asian/Pacific Islander population. For Black Americans in 1970, there were 10.5 Black dentists per 100,000 Black American population. In 1980 this figure was 12.0 per 100,000 Black population. In 1970 there were 8.3 American Indian dentists per American Indian population. In 1970 there were 11.3 Hispanic dentists per 100,000 Hispanic population. This figure was 15.0 in 1980. For nonminority dentists the figure was 50.0 per nonminority dentists per 100,000 nonminority population in 1970. In 1980 this figure was 64.1 per 100,000 nonminority dentists. Graph 2 Minority Dentists per 100,000 Minority Population in the U.S., 1975 and 1980 120 110- 100- 90- 80- 70- 60- 50- 40- 0 109 89 // . ..■' ,' / / .-■' ,•■' ..-' / '-■','"'■■■■'■'■ 616 0.6- >.\'"\ 0.5-i / / Y y y / .■•■'* / Y ,«•■' ,• , •' y y .' .-' . / .< ••' /' Y * ' /,/ . '■•. "•- '• 425 '■■•. \ "^ '•• x \ \ '■. i. ■••, ,( *•. v * • *■. '•« ■•. ■«. 0.4- 365 «•■ / / •■" ./ . •■ . 0.3- 289 '•-. '■■■•''', '■• 'N X vXV 241 272 '■• x ■■.. ..XX, 0.2-0.1" 0- 176 *» Y .••' •• ' Y .'' / / / XX Asian Black Hisp. Amer. Indian 1975 1331 1980 Nonarin. Labels on bars are actual numbers, not numbers in thousands. 455 The trends in minority enrollment in health professions schools are similar to those described above for practitioners. It is unlikely that in the near future the proportions of practicing minority health professionals will increase at a higher rate than they have since 1975. Just as the national trends showed increases in professionals for each of these disciplines among the nonminority population, so too were there increases in each minority group. However since the increases were "across the board," the gap between nonminority and minority was not appreciably changed. As a second phase in reviewing the availability of minority health professionals, an effort was made to relate information based on analyses at lower than national level. A particularly timely study by Spratley, et al. presents this information in a creditable fashion. This study examines the "Location of Minority and Nonminority Health Professions," by geographic region and division; metropolitan vs. non-metropolitan area; county population size; situation in primary care manpower shortage areas; and poverty status of county residents. These analyses were accomplished for physicians, dentists and pharmacists. To better appreciate the distribution of minority health professionals, Table 28 was developed to show the relative availability of minority physicians to those minority populations previously described in selected counties. The patterns seen with these figures however suggest no clear trends across the States reviewed. Even through use of the Area Resource File, it has not been possible to obtain detailed information on the availability of these minority professionals to examine resources at the community level. Data on selected professions, notably medicine, will be most readily accessible and will be examined as its availability will permit. For example, preliminary reports on a study recently completed by the Rand Corporation, surveying 1,000 physicians who graduated in 1975 shows that the minority physicians were much more likely (11 percent versus 6 percent) to locate their practices in health manpower shortage areas, than nonminority physicians. 456 Table 28 Number of Physicians, Population and Physician per 100,000 Population Ratios in Counties with a Significant Minority Population 1 "1 1 1 Counties with 20% + Black Pop. | | Counties with 20% + Hispanic Pop. 1 1 1 Black | HispanicI 1 Phys. | Phys. | 1 Per | Per | 1 No. of 100,0001 No. of 100,000 | 1 No. of Black Black Black | 1 No. of Hispanic Hispanic Hispanic I I State Counties Phys. Pop. Pop. I ICounties Phys. Pop. I Pop. I 1 ~l | Alabama 37 219 804,738 27.2 | _ __ __ __ | Arizona — — — — | 8 I 119 224,383 | 53.0 | Arkansas 27 65 318,388 20.4 | 1 — — I — I — I |California — — — — | 1 H 1,055 2,697,924 | 39.1 I IColorado — — — — | 1 12 20 | 79,603 | 25.1 | |Dist. of Colum. 1 467 444,808 105.0 | 1 — — ! — I — 1 IFlorida 12 64 264,381 24.2 | 1 1 2,375 | 580,025 | 409.5 | IGeorgia 106 442 1,303,220 33.9 | 1 — — — | — 1 IHawaii — — — — I I — — I — | — I |Illinois 4 980 1,414,734 69.3 | 1 — 1 — — I — 1 IIndiana 2 80 279,680 28.6 | 1 — 1 — | — I — 1 I Kansas 1 14 41,274 33.9 | 1 — 1 — — I — 1 I Kentucky 1 0 16,695 0.0 | 1 — — | — I — I | Louisiana 45 195 1,023,876 19.0 | 1 — — I — I — 1 I Maryland 9 443 728,137 60.8 | 1 — — I — I — | Michigan 1 479 823,871 58.1 | 1 — — — — (Mississippi 64 60 766,449 7.8 | 1 — — — — I Missouri 2 98 211,263 46.4 | 1 — — — — | Montana — — — — I | — — — — I I Nebraska — — — — | I — — — — |New Jersey 1 210 311,630 67.4 | | 1 122 145,249 | 84.0 | |New Mexico — — — — I | 28 202 449,749 44.9 | iNew York 3 937 1,328,927 70.5 | 1 2 652 730,385 | 89.3 | | North Carolina 56 292 1,092,088 26.7 | I — — — — 1 j North Dakota — — — — 1 I — — — — I Ohio 1 203 338,254 60.0 | | — — — — 1 |Oklahoma — — — — I 1 — — — — I | Pennsylvania 1 181 633,597 28.6 | I — — — — |South Carolina 39 128 790,288 16.2 | I — — — — | South Dakota — — — — I I — — — — |Tennessee 8 261 496,114 52.6 | | — — — — 1Texas 27 76 241,778 31.4 j I 90 1,238 1,963,334 63.1 lUtah — — — — I — — — — |Virginia 49 233 748,580 31.1 | I — — — — | Washington — — — — | 1 0 2,950 0.0 | Wisconsin 1 — — ^— —~• 1 —— "'~ ~ j i 1 |Los Angeles 1 1,123 925,832* 121.3 | 1 1 1 901 12,065,503 1 43.6 |San Francisco 1 1 119 84,334* 141.1 | 1 1 110 I 84,194* 1 130.7 *The minority population group comprises less than 20% of the total population in these counties. 457 Table 28 (continued) Number of Physicians, Population and Physician per 100,000 Population Ratios in Counties with a Significant Minority Population Counties with 20% + Indian Pop. | | Counties with 5.0% + Asian Pop. | Indian | Asian Phys. | Phys. | Per j Per j No. of 100,0001 No. of 100,000 | No. of Indian Indian Indian | I No. of Asian Asian Asian | 1 State Counties Phys. Pop. Pop. I ICounties Phys. Pop. Pop. j I Alabama _ _ IArizona 3 8 92,159 8.7 | | — — — — | I Arkansas — — — — 1 — — — — | ICalifornia — — — — 1 1 9 3,025 853,215 354.5 | 1Colorado — — — — | | — — — — | |Dist. of Colum. — — — — | | — — — — | IFlorida — — — — | | — — — — | IGeorgia — — — — I I — — — — | IHawaii — — — — | 1 4 642 449,289 142.9 | I Illinois — — — — | | — — — — | IIndiana — — — — | 1 — — — — | IKansas — — — — | | — — — — | IKentucky — — — — I I — — — — I | Louisiana — — — — | | — — — — I I Maryland — — — — I I — — — — 1 | Michigan — — — — I I — — — — I I Mississippi — — — — I 1 — — — — I iMLssouri — — — — | | — — — — | I Montana 5 0 18,502 0.0 | | — — — — | | Nebraska 1 0 2,436 0.0 | I — — — — | |New Jersey — — — — 1 | — — — — | |New Mexico 3 18 73,452 24.5 | I — — — — | I New York — — — — | 1 2 2,102 168,061 1250.7 | | North Carolina 2 5 38,000 13.2 | I — — — — | I North Dakota I Ohio I Oklahoma 3 0 11,638 0.0 | — — — — 3 8 19,157 41.8 | — — — — |Pennsylvania — — — — I I — — — — | |South Carolina — — — — I 1 — — — — | | South Dakota 10 0 28,885 0.0 | I — — — — | | Tennessee — — — — 1 — — — — | Texas lUtah 1 0 5,622 0.0 | | — — — — | | Virginia — — — — 1 I — — — — I | Washington — — — — 1 I — — — — IWisconsin 1 0 3,014 0.0 | ^— — — — |Los Angeles 1 18 52,809* 34.1 | I 1 2,095 418,966 500.0 | |San Francisco 1 0 3,374* 0.0 | | 1 283 144,619 195.7 j *The minority population group comprises less than 20% of the total population in these counties. 458 Summary. The total minority representation in health occupations in 1980 was 21.0 percent compared to a U.S. minority population of 20.1 percent. However, the minority percentages are heavily inflated by the large numbers of minority nursing aides, orderlies, and attendants (477,000), registered nurses (171,000), and licensed practical nurses (102,000). The percentages of minorities in professional health occupations, excepting nursing, are far below the minority percentages in the U.S. population. Likewise, in mental health and mental health-related occupations such as psychology, social work, and alcoholism and drug treatment, the professional staffs are overwhelmingly nonminority even when the clientele is substantially minority. The graphs on pages 82 to 85 show that there has been some recent improvement in the ratios to population of minority physicians, dentists, pharmacists, and R.N.s but this has only been in the context of a trend for all segments of the population. Relative to the nonminority population, the minority health professional ratios per 100,000 minority population have remained about the same. In Table 28, the gap between minority populations and the nonminority population is clearly illustrated. The total U.S. physician-population ratio is about 213 per 100,000. No physician population ratio for State aggregated counties having 20 percent or more Black or American Indian populations even approach 213. Most are below 100. There is one state (Florida) where the 20 percent Hispanic counties' ratio exceeds the national ratio; all others are markedly lower. PART II. B. 2. - Development of Minority Health Professionals Because of the historic and documented continuing underrepresentation of minorities in the health professions, and the question of the current and future availability of professionals to serve minority communities, part of this study was focussed on the development of minority health professionals. As a step beyond looking at whether health professionals were available to minority communities, the subcommittee sought to briefly examine whether the education of health professionals seemed available to those same minority communities. If it has proven difficult to attract practitioners to many of these communities, perhaps success could be enhanced if the educational and training programs and facilities were both available and accessible to those minority communities. The purpose of this analysis was to document the degree to which the numbers of minority students in health professions training suggests progress in narrowing the gap of underrepresentation. It was also to examine the immediate availability of health professions training programs to the counties with significant minority populations, and the degree of success of those minority students in graduating from public versus private institutions. With the understanding that government (i.e. public schools) has a greater responsibility than the private sector for assuring equity and addressing the health and educational needs of all citizens, a review of public and private accomplishments was undertaken. 459 For the first phase of this review, an examination of U.S. trends in first-year minority enrollment for the various health disciplines is provided in graphic and tabular form. The most significant increase in Black first-year medical students occurred in the late 1960s and early 1970s when the percentage of Black first-year students rose from 2.7 percent to 7.5 percent. The percentage then fell to under 7 percent and remained at that level throughout the rest of the 1970s and into the 1980s. The increase in Black undergraduates similarly, peaked, fell, and then stabilized at a somewhat lower level. See Figure 1. Asian/Pacific Islander first-year enrollment of medical students was only slightly above their representation in the U.S. population until 1975-76. It then increased precipitously and is still increasing. It is between three and four times higher than the percentage of Asian/Pacific Islanders in the U.S. population, and much higher than the percentage of Asian/Pacific Islander undergraduates. See Figure 2. The Hispanic trend lines resemble those for Blacks although at lower levels. The differences are that Hispanic first-year enrollment continued to increase for a longer time before leveling off and that the U.S. Hispanic population is increasing more rapidly than the Black population. See Figure 3. The American Indian first-year medical school enrollment percentage exceeded the Indian percentage of the U.S. population in 1973-74. However within two years, the percentage declined to what would have been a more expected level based on previous first-year enrollments. The aberrant 1973-74 figure may have been due to a successful program in that single year to increase American Indian students. See Figure 4. For each of the last five years for which there are data, the Indian percentage of first-year medical school enrollment has been 0.4. See Table 29. 460 Figure 1 Blacks as a Percent of the United States Population, of Undergraduate Students, and of First-Year Enrollees in Schools of Medicine: 1968-69 Through 1983-84 Percent 12 i— 10 ...;— '—Percent of Population — Percent of Undergraduate Students — Percent of First-Year Enrollment Ol______L I I______I______L J______L J______L 1970-71 75-76 Academic Year '80'81 461 Figure 2 Hispanics as a Percent of the United States Population, of Undergraduate Students and of First-Year Enrollees in Schools of Medicine: 1968-69 Through 1983-84 Percent 7 »** i^ .%• »» %*' L%* !»» ..•"' ...*'' Percent of Population Percent of Undergraduate Students Percent of First-Year Enrollment J______L J______I J______L 1970-71 75-76 Academic Year '80'81 462 Figure 3 Asians as a Percent of the United States Population, of Undergraduate Students, and of First-Year Enrollees in Schools of Medicine: 1968-69 Through 1983-84 Percent 6 r— Percent of First-Year Enrollment Percent of Undergraduate Students Percent of Population- • • • • • • • • • • • • ..-•'i • •• • • .♦. J_______L J______L 1970-71 75-76 Academic Year '80'81 463 Figure 4 American Indians as a Percent of the United States Population, of Undergraduate Students, and of First-Year Enrollees in Schools of Medicine: 1968-69 Through 1983-84 Percent 0.8 r— 0.6 0.4 0.2 - ....•••" • -• ••••••••••* • -• ..................5 ••••••• • • • '---Percent of Undergraduate Students Percent of Population -Percent of First-Year Enrollment J__________I 1970-71 75-76 Academic Year '80'81 464 Table 29 First-Year Enrollment in Schools of Medicine in the United States, By Racial/Ethnic Category: Academic Years 1968-69 Through 1983-84 TOTAL FIRST Racial/ethnic category NON- MINORITY YEAR MINORITY UNDER- FIRST ENROLL- FIRST REPRESENTED YEAR Academic MENT YEAR MINORITIES Mainland ENROLL- year ENROLL- Mexican Puerto Other American Other MENT V MENT J/ Black American Rican Hispanic Indian Asian minority Number of students 1968-69 9,863 413 292 266 20 3 3/ 3 121 3/ 9.450 1969-70 10,422 641 501 ..440 44 10 1/ 7 140 "7/ 9.781 1970-71 11,348 998 808 697 73 27 "7/ 11 190 "7/ 10,350 1971-72 12,361 1.280 1,063 882 118 40 "37 23 217 "7/ 11,081 1972-73 13,677 1,437 1,172 957 137 44 "37 34 231 3T 12,240 1973-74 14,159 1,631 1,301 1,027 174 56 "37 44 259 71 12,528 1974-75 14,763 1,839 1,473 1,106 227 69 "37 71 275 91 12,924 1975-76 15,295 1,787 1,432 1,036 224 71 TT 60 282 73 13,508 1976-77 15,613 1,891 1,462 1,040 245 72 62 43 348 81 13,722 1977-78 16,136 2,002 1,607 1,085 246 68 157 51 395 3/ 14,134 1978-79 16,501 2,046 1,594 1,061 260 75 151 47 452 "3/ 14,455 1979-80 16,930 2,237 1,735 1,108 290 86 188 63 502 "1/ 14,693 1980-81 17,186 2,344 1,772 1,128 258 95 224 67 572 "37 14,842 1981-82 17,268 2,683 1,918 1,196 300 105 247 70 765 "T/ 14,585 1982-83 17,245 */ 2,840 1,904 1,145 305 114 278 62 936 1/ 14,405 1983-84 17,146 _£/ 2.889 1,906 1,173 301 109 248 75 983 ~3/ 14,257 Percent 1968-69 100.0 4.2 3.0 2.7 0.2 * 3/ * 1.2 3/ 95.8 1969-70 100.0 6.2 4.8 4.2 0.4 0.1 "37 0.1 1.3 "7/ 93.8 1970-71 100.0 8.8 7.1 6.1 0.6 0.2 1/ 0.1 1.7 1/ 91.2 1971-72 100.0 10.4 8.6 7.1 1.0 0.3 "J/ 0.2 1.8 "7/ 89.6 1972-73 100.0 10.5 8.6 7.0 1.0 0.3 "37 0.2 1.7 0T7 89.5 1973-74 100.0 11.5 9.2 7.3 1.2 0.4 -1/ 0.3 1.8 0.5 88.5 1974-75 100.0 12.5 10.0 7.5 1.5 0.5 "37 0.5 1.9 0.6 87.5 1975-76 100.0 11.7 9.4 6.8 1.5 0.5 03 0.4 1.8 0.5 88.3 1976-77 100.0 12.1 9.4 6.7 1.6 0.5 0.4 0.3 2.2 0.5 87.9 1977-78 100.0 12.4 10.0 6.7 1.5 0.4 1.0 0.3 2.4 3/ 87.6 1978-79 100.0 12.4 9.7 6.4 1.6 0.5 0.9 0.3 2.7 "37 87.6 1979-80 100.0 13.2 10.2 6.5 1.7 0.5 1.1 0.4 3.0 1/ 86.8 1980-81 100.0 13.6 10.3 6.6 1.5 0.6 1.3 0.4 3.3 "7/ 86.4 1981-82 100.0 15.5 11.1 6.9 1.7 0.6 1.4 0.4 4.4 "J/ 84.5 1982-83 100.0 */ 16.5 11.0 6.6 1.8 0.7 1.6 0.4 5.4 ~J/ 83.5 1983-84 100.0 _5' 16.8 11.1 6.8 1.8 0.6 1.4 0.4 5.7 1/ 83.2 * Less than 0.05 percent. 1/ Residents of the Comnonwealth of Puerto Rico are not considered to be members of any minority group and are "" Included In this table only 1n the TOTAL FIRST-YEAR ENROLLMENT and the Mainland Puerto Rican data columns. _2/ Includes all minority radal/ethnlc categories except Asian and Other minority. V The categories "Other Hispanic" and "Other ■Inorlty" were not 1n use 1n these years. 4/ Excludes 9 students for whoa racial ethnic Information was not available. 5/ Excludes 4 students for whoa racial ethnic Information was not available. 465 Prior to 1972-73, the number of minority first-year LPN enrollment was greater than the number of minority first-year RN enrollment. Since then, the RN enrollment number have continued to rise while the LPN enrollment number fell for several years and then began to increase slightly at the end of the 1970s. See Figure 5. The percentage of minority first-year LPN enrollment has always remained above that for RN students. However, in the last several years the gap has been closing. See Table 30. 466 Number and Percent Minority First-Year Enrollment in RN and LPN Programs in the United States in the United States and Possessions: Selected Years 1971-72 through 1980-81 Number 16,000 14,000 — 12,000 — u,uuu • 8,000 — 6,000 — 4,000 — 2,000 — 0 I L Minorities LPN Programs Minorities RN Programs J_____L J____I 1973-74 77-78 Academic Year Percent 20 15 10 i— Minorities LPN Programs ••-o 1—Minorities RN Programs J_____L 1973-74 77-78 Academic Year J_____I rxi H- 00 c n ro Table 30 FIRST-YEAR ENROLLMENT IN REGISTERED NURSE AND PRACTICAL NURSE TRAINING PROGRAMS IN THE UNITED STATES AND POSSESSIONS, BY TYPE OF PROGRAM AND RACIAL/ETHNIC CATEGORY: SELECTED ACADEMIC YEARS 1962-63 THROUGH 1980-81 1/ TOTAL Racial/ethnic cateqory First-year FIRST-YEAR m' Inorlty Academic ENROLLMENT enrollment Blai Hispanic Other minority year Percent Percent Percent Percent of total of total of total of total first-year first-year first-year first-year Number enrollment Number enrollment Number enrollment Number enrollment All registered nurse programs » 1962-63 48,259 * * 1,456 3.0 * * • • 1965-66 59,049 * * 1,891 3.2 * * * * 1968-69 60,598 * * 3,735 6.2 * * * * 1971-72 91,896 9,889 10.8 7,088 7.7 1,866 2.0 935 1.0 1974-75 89,706 11,322 12.6 8,159 9.1 2,080 2.3 1,083 1.2 1977-78 101,438 11,212 11.2 7,313 7.2 2,520 2.5 1,379 1.4 1980-81 102,540 13,799 13.5 8,537 8.3 3,515 3.4 1,747 1.7 RN baccalaureate degree programs 1962-63 8,867 * * 433 4.9 * * * * 1965-66 11,590 * * 554 4.8 * * * * 1968-69 14,111 * * 842 6.0 • * * * 1971-72 26.758 3,509 13.1 2,407 9.0 667 2.5 435 1.6 1974-75 29,479 4,911 16.7 3,650 12.4 807 2.7 454 1.5 1977-78 35,442 4,366 12.3 2,905 8.2 970 2.7 491 1.4 1980-81 32,548 5,390 16.6 2,797 RN 8.6 associate dec iree 1,813 programs 5.6 780 2.4 1962-63 3,317 * * 173 5.2 * * * * 1965-66 8,555 * * 558 6.5 * * * * 1968-69 17,808 * * 1,871 10.5 * * * * 1971-72 35,863 4,958 13.8 3,550 9.9 1,034 2.9 374 1.0 1974-75 38,581 5,096 13.2 3,495 9.1 1,069 2.8 532 1.4 1977-78 46,755 5,515 11.8 3,580 7.6 1,318 2.8 617 1.3 1980-81 53,127 6,993 13.2 4,668 8.8 1,477 2.8 848 1.6 RN diploma programs 1962-63 36,075 * * 850 2.4 * * * * 1965-66 38,904 * * 779 2.0 • * * * 1968-69 28,679 * * 1,022 3.6 • • * * 1971-72 29,275 1,422 4.9 1,131 3.9 165 0.6 126 0.4 1974-75 21,646 1,315 6.1 1,014 4.7 204 0.9 97 0.4 1977-78 19,241 1,331 6.9 828 4.3 232 1.2 271 1.4 1980-81 16,865 1,416 8.4 1,072 6.4 225 1.3 119 0.7 AT 1 practical nurse programs 1962-63 27,085 * * 4,455 16.4 * * * * 1965-66 36,768 * * 6,669 18.1 * * * * 1968-69 44.917 * * 7,804 17.4 • * * * 1971-72 57,567 11,183 19.4 8,545 14.8 1,965 3.4 673 1.2 1974-75 46,530 8,313 17.9 5,795 12.5 1,927 4.1 591 1.3 1977-78 53,002 8,279 15.6 5,883 11.1 1,655 3.1 741 1.4 1980-81 51 ,335 8,993 17.5 6,252 12.2 2,010 3.9 731 1.4 * Data for minorities other than black were not collected until 1971-72. 1/ Data for academic years 1962-63 through 1968-69 are based on those first-year students 1n schools responding to ~" question on minority enrollment; data for 1971-72 through 1980-81 are based on those students in schools responding to question on minority enrollment, male enrollment, or both. 468 In the second phase of the review, a comparison was made of the minority graduates (by racial/ethnic group) from health professions schools in counties having greater than 20 percent of that minority versus schools elsewhere in the State. Medical, dental and pharmacy schools were analyzed. A series of tables have been prepared displaying this information and are provided in Appendix II. B. 1. A discussion of what those tables display is included in this part. In addition, there are four tables comparing percentages of public and private medical, dental, and pharmacy school graduates for the four racial/ethnic groups considered in this report provided in Appendix II. B. 1. A comparison of minority graduates from health professions schools in counties having greater than 20 percent of that minority versus schools elsewhere in the same States. Black graduates. Although 51 U.S. medical schools are in counties with over 20 percent Black populations, in no county except where Howard University and Meharry Medical College are located, do Black medical school graduates exceed 14 percent. The average percentage of Black graduates from the 51 schools is 7.4. When Howard and Meharry are excluded, the average is 5.0 percent. From the 35 medical schools not in counties with 20 percent and over Black populations, the percentage of Black graduates is 3.8. The number of Black graduates in the 20 percent and over Black population counties is 510 (76.8 percent) compared to 154 (23.2 percent) in the other counties. Of the 51 medical schools, 18 (35.3 percent) are public. Of the 35 schools, 30 (85.7 percent are public. Private schools are more likely to be located in areas with Black communities and to have, overall, higher percentages of Black graduates than public medical schools in the States being considered. See Table 31. Table 31 Black Medical Graduates | Location of Number of Schools | School Public Private | Counties 20%+ Black | Population | Counties under 20% | Black Population 18 33 30 5 Number of Number of Total Minority Black Graduates Graduates Graduates 6,898 4,071 923 549 510 154 469 A similar pattern exists for dental schools. Both Howard and Meharry have greater than 50 percent Black graduates, but in only one other county (that of the Medical College of Georgia) are there more than seven percent Black dental graduates. The average percentage of Black graduates from the 26 schools in counties with 20 percent and over Black population is 5.4. If Howard and Meharry are not counted, the percentage is 2.5. It is 2.0 percent in the 14 dental schools in all other counties in the same States. In the 20 percent and over Black population counties, there are 145 (82.9 percent) Black graduates but only 30 (17.1 percent) in all other counties. This difference is largely a reflection of the graduate totals for Howard and Meharry. Twelve (46.1 percent) of 26 schools in the 20 percent and over counties and 11 (78.6 percent) of the 14 other schools are public. When Howard and Meharry are eliminated, the percentage differences between public and private dental school graduates are also eliminated. See Table 32. Table 32 Black Dental Graduates Number of Number of Number of | Location of Schools Total Graduates Minority Graduates Black | I School Public Private Graduates | Counties 20%+ Black I Population 12 14 2,678 324 145 I Counties under 20% Black Population 11 3 1,559 147 30 | 470 For 22 pharmacy schools in the 20 percent and over Black counties, the average percentage of Black graduates is 6.7. In all other counties the average of 17 schools is 2.9 percent. Three predominantly Black schools (Florida A & M University, Howard University and Xavier University) are in the 20 percent and over counties, and a forth predominantly Black school (Texas Southern University) is in an under 20 percent Black county. The number of Black pharmacy graduates in the 20 percent and over counties is 146 (71.9 percent) and 57 (28.1 percent) in the other counties in the same States. Of the 22 schools in over 20 percent counties, 13 (59.1 percent) are public. In all other counties, 13 of 17 (76.5 percent) are public. Differences in percentages of graduates between public and private pharmacy schools are attributable to relatively large number of graduates from the public predominantly Black institutions. See Table 33. Table 33 Black Pharmacy Graduates Number of Number of Number of Location of Schools Total Minority Black School Public Private Graduates Graduates Graduates Counties 20%+ Black Population 13 9 2,189 350 146 Counties under 20% Black Population 13 4 1,956 201 57 471 Hispanic graduates. In counties with 20 percent and over Hispanic populations, there are six medical schools graduating 7.1 percent Hispanics. The figure is 7.5 percent for 16 schools outside those counties. The reason is that several schools in California not in Hispanic counties but close to them enroll fairly sizable numbers of Hispanic students. The number of Hispanic medical students graduating from schools not in areas defined by this report as Hispanic communities, is 143 (71.1 percent) versus 58 (28.9 percent) in 20 percent and over Hispanic counties. The medical school catchment area for Hispanic students extends beyond their immediate communities, particularly in California. The proportions of public and private schools in both the over and under 20 percent Hispanic counties are about the same - 65-75 percent public. The public medical schools in California and New Mexico have the highest percentages of Hispanic graduates. The percentages of private medical schools, except for the University of Miami, are much lower. See Table 34. Table 34 Hispanic Medical Graduates Number of Number of Number of Location of Schools____ Total Minority Hispanic School Public Private Graduates Graduates Graduates Counties 20%+ Hispanic Population 4 2 816 134 58 Counties under 20% Hispanic Population 13 3 1,911 298 143 472 There are only three dental schools in 20 percent and over Hispanic counties. Their average percentage of Hispanic graduates is 10.3 which is four points higher than for the eight dental schools in other counties in the same States. There were only 88 Hispanic dental graduates in the year being analyzed. Fifty (56.8 percent) were from the eight schools not in Hispanic communities and 38 (43.2 percent) were from the three schools in 20 percent and over Hispanic counties. Four public dental schools had percentages of Hispanic graduates between ten and sixteen percent. No private dental school had more than 4.5 percent Hispanic graduates. See Table 35. Table 35 Hispanic Dental Graduates Number of Number of Number of Location of Schools Total Minority Hispanic School Public Private Graduates Graduates Graduates Counties 20%+ Hispanic Population 2 1 368 74 38 Counties under 20% Hispanic Population 5 3 774 117 50 Four Hispanic graduates of pharmacy schools, the pattern is almost identical to that for dental graduates. Hispanics are 9.1 percent of graduates in 20 percent and over Hispanic counties and 7.5 percent in all other counties. There are four schools in the Hispanic counties and eight outside them. The number of graduates in Hispanic counties is 26 (31.0 percent) and 58 (69.0 percent) in other counties. Public pharmacy schools in Texas and New Mexico have the highest percentages of graduates and private schools are lower. See Table 36. Table 36 Hispanic Pharmacy Graduates Number of Number of Number of Location of Schools Total Minority Hispanic School Public Private Graduates Graduates Graduates Counties 20%+ Hispanic Population 3 1 287 107 26 Counties under 20% Hispanic Population 7 1 769 198 58 473 American Indian graduates. There are so few American Indian graduates of medical, dental and pharmacy schools that the comparison by type of county has no meaning. There are no medical, dental or pharmacy schools in counties with 20 percent or more American Indian population. See Table 37. Table 37 Number of | Number of Number of American Location of Schools Total Minority Indian School Public Private Graduates Graduates American Indian Medical Graduates Graduates I Counties 20%+ I American Indian Population 0 0 0 0 o 1 I Counties under 20% I American Indian Population 9 4 1,296 126 American Indian Dental Graduates 15 I I Counties 20%+ I American Indian Population 0 0 0 0 o 1 I Counties under 20% I American Indian Population 2 2 296 9 American Indian Pharmacy Graduates 1 I Counties 20%+ I American Indian Population 0 0 0 0 0 I Counties under 20% I American Indian Population 2 2 699 41 5 | Note: For Blacks, Hispanics, and American Indians, "minority communities" are defined as counties with 20 percent or greater minority population. For Asian/Pacific Islanders, the definition is counties with 2.0 percent Asian/Pacific Islander population and over. 474 Asian/Pacific Islander graduates. The comparison here uses 2 percent of the population as a benchmark because there are so few U.S. counties with sizable numbers of this minority group. It also only involves counties in the States of Hawaii, New York and California. There are more Asian/Pacific Islander graduates from the schools in counties with 2 percent and higher Asian/Pacific Islander populations. However, for all three disciplines, the location of only one or two schools with high Asian/Pacific Islander enrollment in the counties accounts for the difference. See Table 38. Table 38 Number of I Asian/ I | Number of Number of Pacific | Location of Schools Total Minority Islander | School Public Private Graduates Graduates Graduates | Asian/Pacific Islander Medical Graduates | Counties 2.0%+ | Asian Population 3 6 1,327 250 127 | | Counties under 2.0% j Asian Population 9 1 1,276 164 59 j Asian/Pacific Islander Dental Graduates | Counties 2.0%+ I Asian Population 2 3 611 84 28 f I Counties under 2.0% | Asian Population 2 2 346 51 34 | j Asian/Pacific Islander Pharmacy Schools | Counties 2.0%+ Asian | Population 1 2 381 138 104 | | Counties under 2.0% | Asian Population 3 1 584 80 54 | Summary: Overall, these comparisons shows that minority health professions graduates are more likely to be graduated from schools located in minority communities (counties), but even in these counties the percentages of minority graduates are still much lower than the minority representation in the total county population. Asian/Pacific Islander graduates do not conform to this pattern because they are so overrepresented among health professions graduates relative to their proportion of the total population and because there are so few large communities of Asian/Pacific Islanders. Subpopulations of Asian/Pacific Islanders are differently represented in the health professions and the total overrepresentation may disguise some underrepresentation of Asian/Pacific Islander subgroups within specific health professions. 475 For both Blacks and Asian/Pacific Islanders, there are higher percentages of graduates from private medical, dental, and pharmacy schools than from public schools. The only public health professions schools graduate percentages which exceed those for private schools are Hispanic dental graduates and American Indian pharmacy graduates. Among individual States, there are examples of higher percentages of public school graduates but this would be expected given the small number of schools in most States. There are private, predominantly Black health professions schools which influence the Black public-private ratio but there are no private, predominantly Hispanic, Asian/Pacific Islander or American Indian health professions schools. The data indicate that public health professions schools in the U.S. are generally not recruiting and graduating minority students as well as private schools. PART II. B. 3. - Practice Patterns of Minority Health Professionals Information on the practice patterns of minority health professionals was sought as an aid in defining the specific role and contribution of these individuals in delivering health care to minority communities. Although information was sought on a variety of health disciplines, the only detailed information obtained addressed the field of medicine, particularly Black physicians. Efforts will continue to identify such information with respect to other professions and other minority groups. Nonetheless, the analysis of the available data is presented below. There was only limited information available for review relevant to physician location intentions. Most notable was a study conducted by the Bureau of Health Manpower a decade ago, looking at the "Influence of Preceptorship and Other Factors on the Education and Career Choices of Physicians." In examining the preferences for practice location of medical students and residents who were engaged in preceptorship programs, the findings included the following: • The probability of preferring an inner-city practice location was highest for those who attended high school in an inner-city community, had average or less than average financial support from family or savings,... (and) were minority. • The probability of preferring a rural or small town practice location was highest for residents who attended high school in a rural or small town community, had less than average financial support from family or savings,... (and) were nonminority. • The probability of preferring another urban/suburban location was highest for residents who attended high school in an urban or suburban community, had above average financial support from family or savings,... (and) were nonminority. 476 The predisposition of these future doctors to prefer a probable practice location similar to that of the community in which they were raised was not surprising; nor was the probable influence of their racial/ethnic and family socioeconomic background. This does suggest evidence to support the potentially positive outcomes of training more minority health professionals, and persons from rural areas and small towns, to help meet the needs of underserved inner city and rural areas. Upon completion of medical school, the initial indicator of the future specialty practice plans of a physician is the postgraduate or residency training that is sought. The length of this additional training varies with the nature of the specialty selected, but generally requires three to five or more years. Table 39 shows the Racial/Ethnic Background of Residents on Duty as of September 1, 1981, 1982 and 1983. Table 39 Ethnic Background of Residents on Duty September 1, 1981, 1982, and 1983 I Ethnic Background 1981 1982 1983 I Black (non-Hispanic) 3,472 3,307 3,379 American Indian or Alaskan Native 197 152 111 I Mexican-American 714 697 743 I Puerto Rican 1,223 1,227 1,343 I Other Hispanic 1,414 1,396 1,587 | Asian/Pacific Islander 6,468 5,762 5,632 I White (non-Hispanic) 53,196 55,417 58,576 I Total 66,684 67,958 71,371 In a more detailed analysis (Graettinger and Swanson) covering 1977 through 1983, the National Resident Matching Program and the Association of American Medical Colleges collaborated to provide the data on numbers and percentage distribution of the residents, by specialty, by gender and by racial/ethnic group, as shown in Tables 40 and 41. Separate figures are not available for Asian/Pacific Islanders or American Indians, the former because they were not deemed underrepresented and the latter because their scarce numbers (fewer than 50 per year) would not lend to meaningful separate analysis. 477 Table 40 Specialty Distribution by Gender and Ethnic Group 1977 1978 1979 1980 1981 1982 1983 4> CO MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WCMEN MEN WOMEN MEN WOMEN FAMILY NONMINORITY 1064 191 1415 303 1488 349 1461 345 1425 368 1485 419 1416 415 PRACTICE BLACK 52 20 68 29 56 29 49 24 44 24 45 41 42 43 MEX-AMER 26 8 32 8 43 10 53 6 46 12 44 11 36 13 MAINPR 4 1 5 2 7 3 5 2 7 1 8 2 5 1 INTERNAL NONMINORITY 2837 598 3521 813 3493 943 3544 992 3391 1030 3678 1096 3577 1218 MEDICINE BLACK 165 72 174 78 157 92 172 95 165 108 180 124 177 110 MEX-AMER 24 2 27 9 31 6 37 7 32 8 43 16 54 10 MAINPR 13 1 13 4 14 4 22 8 10 10 15 12 14 10 PEDIATRICS NONMINORITY 632 348 701 499 731 526 712 522 680 565 694 569 651 613 BLACK 33 52 35 60 43 67 36 66 27 74 32 64 33 69 MEX-AMER 12 5 12 7 16 7 9 7 15 11 10 14 14 9 MAINPR 4 2 1 3 3 6 6 7 7 2 2 7 7 7 OBSTETRICS NONMINORITY 349 146 481 206 485 274 501 279 493 300 513 339 467 344 BLACK 64 34 57 30 44 45 58 44 45 50 46 34 30 44 MEX-AMER 10 2 8 3 8 4 9 5 13 5 10 6 12 6 MAINPR 2 0 3 3 1 4 3 5 8 4 5 2 3 3 PSYCHIATRY NONMINORITY 300 109 358 160 340 173 312 148 352 200 425 213 354 208 BLACK 15 10 20 10 24 15 16 16 14 11 10 18 13 14 MEX-AMER 1 2 10 2 6 1 3 1 6 3 6 1 7 4 MAINPR 2 2 0 0 2 1 2 1 1 0 1 3 2 1 GENERAL NONMLNORITY 1102 133 1403 172 1597 208 1650 189 1752 234 1867 213 1729 259 SURGERY BLACK 79 16 78 16 76 9 73 12 62 18 75 18 89 28 MEX-AMER 10 1 20 5 24 1 17 2 25 3 24 3 19 2 MAINPR 2 0 7 2 7 2 12 3 14 1 11 2 7 0 Table 40 (continued) Specialty Distribution by Gender and Ethnic Group 1977 1978 1979 1980 1981 1982 1983 MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN SURGICAL NONMINORITY 180 U 287 12 324 22 341 23 359 22 476 29 481 26 SPECIALTY BLACK 9 3 8 3 10 0 4 3 9 0 7 5 9 1 MEX-AMER 2 0 2 0 6 0 4 0 1 1 3 0 4 0 MAINPR 0 0 1 0 1 0 3 0 2 0 1 1 1 0 PATHOLOGY NONMINORITY 205 55 215 74 227 93 219 73 203 110 213 102 233 120 BLACK 3 4 6 2 2 4 4 3 0 2 2 6 0 3 MEX-AMER 1 0 3 0 4 2 2 3 0 1 4 1 5 2 MAINPR 0 0 0 0 2 0 0 1 0 0 1 1 0 0 HOSPITAL NONMINORITY 322 61 438 91 487 109 541 136 572 140 614 148 722 159 SPECIALTY BLACK 10 8 12 11 16 9 17 8 28 7 16 19 15 14 MEX-AMER 5 0 5 2 4 3 7 2 7 1 10 2 8 3 MAINPR 1 0 1 0 3 0 3 1 3 0 0 2 2 2 OTHER NONMINORITY 94 22 122 31 100 24 94 31 99 41 104 47 132 55 BLACK 4 1 3 1 6 2 7 2 7 1 3 0 2 3 MEX-AMER 2 0 0 0 0 0 0 1 0 1 1 0 0 0 MAINPR 0 0 0 0 1 0 0 0 1 1 0 0 0 1 TRANSITIONAL NONMINORITY 662 126 812 185 814 194 896 178 912 232 921 250 767 202 BLACK 44 17 39 19 43 16 32 29 41 27 42 24 34 26 MEX-AMER 16 4 14 1 13 0 15 5 8 3 14 2 9 4 MAINPR 1 1 1 0 1 2 1 0 2 1 3 0 8 0 Table 41 Percentage Distribution of Men and Women by Ethnic Group 1977 1978 1579 1980 1981 1982 1983 MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN FAMILY NONMINORITY 13.73 10.61 14.51 11.90 14.75 11.97 14.22 11.83 13.92 11.35 13.51 12.23 13.45 11.47 PRACTICE BLACK 10.88 8.44 13.60 11.20 11.74 10.07 10.47 7.95 9.95 7.45 9.83 11.61 9.46 12.11 MEX-AMER 23.85 33.33 24.06 21.62 27.74 29.41 33.97 15.38 30.07 24.49 26.04 19.64 21.43 24.53 MAINPR 13.79 14.29 15.63 14.29 16.67 13.64 8.77 7.14 12.73 5.00 17.02 6.25 10.20 4.00 INTERNAL NONMINORITY 36.62 33.22 36.10 31.93 34.63 32.35 34.50 34.02 33.12 31.77 33.47 32.00 33.97 33.66 MEDICINE BLACK 34.52 30.38 34.80 30.12 32.91 31.94 36.75 31.46 37.33 33.54 39.30 35.13 39.86 30.99 MEX-AMER 22.02 8.33 20.30 24.32 20.00 17.65 23.72 17.95 20.92 16.33 25.44 28.57 32.14 18.87 MAINPR 44.83 14.29 40.63 28.57 33.33 18.18 38.60 28.57 18.18 50.00 31.91 37.50 28.57 40.00 PEDIATRICS NONMINORITY 8.16 19.33 7.19 19.60 7.25 18.04 6.93 17.90 6.64 17.43 6.31 16.61 6.18 16.94 BLACK 6.90 21.94 7.00 23.17 9.01 23.26 7.69 21.85 6.11 22.98 6.99 18.13 7.43 19.44 MEX-AMER 11.01 20.83 9.02 18.92 10.32 20.59 5.77 17.95 9.80 22.45 5.92 25.00 8.33 16.98 MAINPR 13.79 28.57 3.13 21.43 7.14 27.27 10.53 25.00 12.73 10.00 4.26 21.88 14.29 28.00 OBSTETRICS NONMINORITY 4.50 8.11 4.93 8.09 4.81 9.40 4.88 9.57 4.82 9.25 4.67 9.90 4.44 9.51 BLACK 13.39 14.35 11.40 11.58 9.22 15.63 12.39 14.57 10.18 15.53 10.04 9.63 6.76 12.39 MEX-AMER 9.17 8.33 6.02 8.11 5.16 11.76 5.77 12.82 8.50 10.20 5.92 10.71 7.14 11.32 MAINPR 6.90 0.00 9.38 21.43 2.38 18.18 5.26 17.86 14.55 20.00 10.64 6.25 6.12 12.00 PSYCHIATRY NONMINORITY 3.87 6.06 3.67 6.28 3.37 5.93 3.04 5.08 3.44 6.17 3.87 6.22 3.36 5.75 BLACK 3.14 4.22 4.00 3.86 5.03 5.21 3.42 5.30 3.17 3.42 2.18 5.10 2.93 3.94 MEX-AMER 0.92 8.33 7.52 5.41 3.87 2.94 1.92 2.56 3.92 6.12 3.55 1.79 4.17 7.55 MAINPR 6.90 28.57 0.00 0.00 4.76 4.55 3.51 3.57 1.82 0.00 2.13 9.38 4.08 4.00 GENERAL NONMINORITY 14.22 7.39 14.39 6.76 15.83 7.14 16.06 6.48 17.11 7.22 16.99 6.22 16.42 7.16 SURGERY BLACK 16.53 6.75 15.60 6.18 15.93 3.13 15.60 3.97 14.03 5.59 16.38 5.10 20.05 7.89 MEX-AMER 9.17 4.17 15.04 13.51 15.48 2.94 10.90 5.13 16.34 6.12 14.20 5.36 11.31 3.77 MAINPR 6.90 0.00 21.88 14.29 16.67 9.09 21.05 10.71 25.45 5.00 23.40 6.25 14.29 0.00 Table 41 (continued) Percentage Distribution of Men and Women by Ethnic Group 1977 1978 1979 1980 1981 1982 1983 MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN MEN WOMEN SURGICAL NONMINORITY 2.32 0.61 2.94 0.47 3.21 0.75 3.32 0.79 3.51 0.68 4.33 0.85 4.57 0.72 SPECIALTY BLACK 1.88 1.27 1.60 1.16 2.10 0.00 0.85 0.99 2.04 0.00 1.53 1.42 2.03 0.28 MEK-AMER 1.83 0.00 1.50 0.00 3.87 0.00 2.56 0.00 0.65 2.04 1.78 0.00 2.38 0.00 MAINPR 0.00 0.00 3.13 0.00 2.38 0.00 5.26 0.00 3.64 0.00 2.13 3.13 2.04 0.00 PATHOLOGY NONMINORITY 2.65 3.06 2.20 2.91 2.25 3.19 2.13 2.50 1.98 3.39 1.94 2.98 2.21 3.32 BLACK 0.63 1.69 1.20 0.77 0.42 1.39 0.85 0.99 0.00 0.62 0.44 1.70 0.00 0.85 MEX-tAMER 0.92 0.00 2.26 0.00 2.58 5.88 1.28 7.69 0.00 2.04 2.37 1.79 2.98 3.77 MAINPR 0.00 0.00 0.00 0.00 4.76 0.00 0.00 3.57 0.00 0.00 2.13 3.13 0.00 0.00 HOSPITAL NONMINORITY 4.16 3.39 4.49 3.58 4.83 3.75 5.27 4.66 5.59 4.31 5.59 2.07 6.86 4.39 SPECIALTY BLACK 2.09 3.37 2.40 4.24 3.35 3.13 3.63 2.65 6.33 2.17 3.49 5.39 3.38 3.95 4> 00 MEX-AMER 4.58 0.00 3.76 5.40 2.58 8.82 4.48 5.12 4.57 2.04 5.91 3.57 4.77 5.66 h-■* MAINPR 3.45 0.00 3.13 0.00 7.14 0.00 5.26 3.57 5.45 0.00 0.00 6.26 4.08 8.00 OTHER NONMINORITY 1.21 1.22 1.25 1.22 0.99 0.82 0.92 1.06 0.97 1.26 0.95 1.37 1.25 1.52 BLACK 0.84 0.42 0.60 0.39 1.26 0.69 1.50 0.66 1.58 0.31 0.66 0.00 0.45 0.85 MEX-AMER 1.83 0.00 0.00 0.00 0.00 0.00 0.00 2.56 0.00 2.04 0.59 0.00 0.00 0.00 MAINPR 0.00 0.00 0.00 0.00 2.38 0.00 0.00 0.00 1.82 5.00 0.00 0.00 0.00 4.00 TRANSITIONAL NOroiQCRITY 8.55 7.00 8.33 7.27 8.07 6.66 8.72 6.10 8.91 7.16 8.38 7.30 7.28 5.58 BLACK 9.21 7.17 7.80 7.34 9.01 5.56 6.84 9.60 9.28 8.39 9.17 6.80 7.66 7.32 MEX-AMER 14.68 16.67 10.53 2.70 8.39 0.00 9.62 12.82 5.23 6.12 8.28 3.57 5.36 7.55 MAINPR 3.45 14.29 3.13 0.00 2.38 9.09 1.75 0.00 3.64 5.00 6.38 0.00 16.33 0.00 In the authors' words: "The number of 'Blacks' has increased by 12 percentage points but their representation in the graduating classes decreased by 1.6 percentage points because of 1.7 percent fewer Black men and a 0.1 percent more women. Those who have reported themselves to be of 'Mexican American Chicano' background have increased by two-thirds and represented about 1.5 percent of graduates in 1983 because of increases in the percentages of both men and women. Students from the Commonwealth of Puerto Rico who live in the continental United States, 'Mainland Puerto Ricans,' have doubled during the period but make up only half a percent of graduates. From 1977 to 1983 the percentages of the women steadily increased in each ethnic group; in 1983 women constituted 26 percent of nonminority graduates, 44 percent of Blacks, 24 percent of Chicano, and 34 percent of Mainland-PR. "The patterns of the percentages of men and women of the other ethnic background that have entered various specialties are similar to those of the nonminority with only a few exceptions. A greater proportion of Black women than men have entered Family Practice in the past two years. In Internal Medicine the percentage of Black men remains greater than women. In hospital-based Support Specialties (Anesthesiology, Emergency Medicine and Diagnostic and Therapeutic Radiology) the percentage of women from the other ethnic groups has often been greater than men. The significance of differences in percentages cannot be assessed in the two Hispanic groups because of their small numbers." Patterns of specialty distribution between minority and nonminority physicians are similar but there are some differences. For example, both Black men and women have been entering obstetrics in higher percentages than nonminorities. This is changing because in the last several years the number of Black men becoming Obstetricians has declined. In pediatrics, the ratios between nonminority men and women and between Black men and women are about the same but more Black men and women are entering the field therefore their percentages are higher. In 1981, both Black men and women began to have higher percentages of entrants into internal medicine. For Black women, that trend reversed again in 1983 when a higher percentage of nonminority women entered the specialty. Family practice percentages had always been lower for Black men and women until 1983 when the Black women percentage exceeded that of nonminority women for the first time. This was also true in general surgery; until 1983, there had never been a higher percentage of Black women than nonminority women. Between 1978 and 1981 a greater percentage of Black men than nonminority men were choosing psychiatry. Since then, this has not been so. In hospital-based support specialties (anesthesiology, emergency medicine, diagnostic and therapeutic radiology, and pathology), the percentage of Blacks has been lower than nonminorities. This discussion of physicians in residency training has served as the basis for comparison between the "soon-to-be" and "currently professionally active" physicians. Table 42 compares the profile of primary specialty for Black physicians (from National Medical Association tabulations) to the profile of all physicians (from American Medical Association tabulations). The specialty distribution for the Black physicians is generally similar to that of all physicians, as was the case with residents, except the somewhat heavier concentration in primary care specialities, including obstetrics and gynecology. 482 Table 42 Primary Specialty Distribution of Black and Total Professionally Active Physicians Black ] ftiysicians Total Phy sicians Number Percent Number Percent Total 7,625 100.0 430,745* 100.0 Primary Care 3,372 44.2 165,383 38.4 General and Family Practice 1,264 16.6 60,594 14.1 Internal Medicine 1,489 19.5 75,211 17.5 Pediatrics 619 8.1 29,578 6.9 Primary Care with Ob-Gyn 4,273 56.0 192,583 44.7 Other Medical Specialties 295 3.9 27,242 6.3 Allergy 8 0.1 1,527 0.4 Cardiovascular Diseases 97 1.3 10,378 2.4 Dermatology 93 1.3 10,378 2.4 Dermatology 93 1.2 5,825 1.4 Gastroenterology 46 0.6 4,464 1.0 Pediatric Allergy 3 0.0 398 0.1 Pediatric Cardiology 8 0.1 704 0.2 Pulmonary Diseases 40 0.5 3,946 0.9 Surgical Specialties 2,184 28.6 113,704 26.4 Colon and Rectal Surgery 6 0.1 754 0.2 General Surgery 658 8.6 34,651 8.0 Neurological Surgery 36 0.5 3,498 0.8 Obstetrics and Gynecology 901 11.8 27,200 6.3 Ophthalmology 189 2.5 13,281 3.1 Orthopedic Surgery 158 2.1 14,572 3.4 C)torhinolaryngology 81 1.1 6,529 1.5 Plastic Surgery 15 0.2 3,245 0.8 Thoracic Surgery 22 0.3 2,085 0.5 Urology 118 1.5 7,889 1.8 Other Specialties 1,774 23.3 124,416 28.9 Aerospace Medicine 15 0.2 629 0.1 Anesthesiology 157 2.1 16,845 3.9 Child Psychiatry 40 0.5 3,295 0.8 Neurology 62 0.8 6,226 1.4 Occupational Medicine 36 0.5 2,623 0.6 Psychiatry 393 5.2 28,524 6.6 Public Health 83 0.8 2,924 0.7 Physical Medicine and Rehabilitation 24 0.3 2,355 0.5 Pathology 99 1.3 13,956 3.2 Radiology 225 3.0 21,283 4.9 Other & Unspecified 640 8.7 25,756 6.0 Pediatric Neurologists 3 560 * Excludes Not Classified, Inactive and Address Unknown 483 Perhaps most germane to this entire subject of practice patterns is the discussion of the treatment practices of minority health professionals. It is probably the truest measure of availability and accessibility for the subject minority communities. Unfortunately it is also an area in which little reliable data seems available. One significant study was accomplished a decade ago through oversampling of Black physicians in concert with the National Center for Health Statistics' conduct of the National Ambulatory Medical Care Survey. An Executive Summary of the study is provided in Appendix II. B. 2. Some of the general findings of the study were: The treatment practices of Black physicians do not differ markedly from the treatment practices of non-Black physicians insofar as NAMCS data reveal treatment practices. However they do differ in these respects: 1) Black physicians appear to ask patients to return at a specific time more often than non-Black physicians. 2) They apparently make more use of three categories: • blood pressure check, • drug prescribed, and • injection. 3) Black physicians are far more likely to see Black patients. 4) A Black physician is much more likely to see patients in a metropolitan setting. In order to provide similarly reliable data for the practices of other minority health professionals, additional data collection and analysis efforts need be undertaken. Without it there is no mechanism for appraising the contributions of any providers who serve minority or other communities. PART III. - WHY DO THE DIFFERENCES EXIST AND HOW DO THEY CONTRIBUTE TO THE HEALTH STATUS DISPARITIES Why do the differences in availability of health professional resources exist between minority and nonminority communities? How do they contribute to the health status disparities? These questions, as posed to the working group, included inferences that (1) there were differences in the availability of health professionals to the respective populations; (2) that those differences could be measured on a common basis across the four minority groups and the nonminority population; (3) that some rationale could be discerned to explain those differences; and (4) that a causal relationship could be found between greater or lesser numbers and types of these professionals and disparities in the health status between the nonminority and minority populations. 484 In reviewing the data analyzed to this point, it is not possible to approach even inference number one with any degree of confidence. The examination of data at the national, state, and finally the county level has generally failed to provide information which is specific to address the requirement for a "community" focus. The process undertaken to reach this point has facilitated the development of several listings of U.S. counties, and the States, which clearly provide for more timely focus on the identification of minority problem study areas. Further, the Area Resource File has provided a useful tool for the collation of data from the 1980 Census, and several statistical files containing information on health professionals, health facilities, et al. Another positive aspect of this activity has been the interest generated within non-Federal circles, especially among minority and health professional communities. As part of the planning stage of developing this report, the working group prepared and mailed a letter to scores of minority organizations having health or health-related interests. Note Appendix III. A. These entities were encouraged to share available information relevant to the working group's charge. Federal investment in this effort is apparently stimulating these and other external sources to also direct resources to these questions. Consequently, even following the submission of this document/report, the Bureau of Health Professions and other interested parties will continue to move to answer these questions, through appropriate data development and analysis. This process might begin with restructuring the county level data across, rather than within States, matching for variables other than size; or pursuing the development and analysis of more detailed data at the subcounty level; or both. Because of this commitment to continue to address the aforementioned issues, the working group has been willing to move beyond some of those issues, which are appropriate for primarily objective statistical analysis and comment on some of the more subjective issues. These can be, and are, quite varied. PROVIDER/PATIENT EXPECTATIONS One area which the working group was specifically asked to review was that of the expectations of physicians and other providers compared to the expectations of their patients/clients. This was deemed to be of special significance since the success of many, if not most, health care interactions depends upon the mutual positive participation of both provider and patient. An example might be a patient with visual or vision-related problems who approaches an optometrist or ophthalmologist for care. The patient might expect that the provider will: 1) make a thorough and appropriate examination; 2) make a correct diagnosis of the problem; and 3) prescribe the appropriate treatment and follow-up. On the other hand, the provider might expect that the patient will: 1) answer all questions in a complete and truthful manner; 2) adhere to the prescribed treatment regimen; and 3) return for follow-up as requested. 485 Statements as to lack of data minority-to-minority. Some of the findings of available studies on this topic are discussed below: Nine studies have been carried out with regard to patient/physician relationships. However, there is little to suggest that there have been studies made concerning what physicians expect of or from patients other than compliance. For example, patients' compliance with medication directions is apparently what most physicians expect from patients (Haynes, 1976). Kirscht and Rosenstock (1977) discussed patient adherence to anti-hypertension procedures. Hershey, et al. (1980) discussed self-reported medications taking in a random selective study in a large urban hospital to determine compliance. In other studies not directly related to compliance, Korsch, et al. (1969) and Hulka, et al. (1975) examined doctor and patient relationships. In other areas indirectly related to expectations, Evans, et al. (1984) in a large NIMH study examined 29 psychotherapists at an urban psychiatric out-patient clinic who participated in a special orientation program designed to: (1) increase the therapist's knowledge about low-income and minority patients; (2) increase the therapist's sensitivity to minority patients' requests and problems; and, (3) increase the effectiveness of the therapeutic encounter. Data indicated that therapists improved significantly in knowledge, sensitivity, and effectiveness as a result of the therapists' special orientation program. In another study, Bradshaw (1978) concluded that trainee psychiatrists working with Blacks and with other minority patients can be integrated into existing and relatively traditional residency programs. White (1977) points out that since most of the deliverers of health care are nonminority, they must become sensitive to the traditions, values, and attitudes of the ethnic groups of color. It was noted that schools of nursing were beginning to include cultural differences in nursing curricula, because the majority of nurses who practice are not aware and are not sensitive to the needs of minority patients. The literature suggests that studies of health professionals' expectations of their patients are in an exploratory stage. Also, in an exploratory stage is what patients' expectations are of the health professionals. Many studies on patients' satisfactions and assessments of physicians' competence could be interpreted as an approach. Hayes-Bautista (1969) in a mental health study described Chicano patients in an urban Chicago setting who were making decisions concerning physicians' competency based on their own criteria. It was pointed out that problems to overcome in the health professional/patient expectations realm include: (1) language; (2) diagnostic differences; (3) acceptability; (4) treatment expectation; (5) therapist's sensitivity; (6) lack of knowledge by the therapist/patient; (7) differing views of problematic behavior to its management between psychiatric professionals and minority groups; (8) transference; (9) counter-transference; and, (10) transcultural therapy. 486 The meshing of health professional/patient expectations has possibility. Acosta, et al. (1980) describes a 3-year study to assess the effects of patient orientation, therapist orientation and the interaction of two upon the psychotherapy process and outcome for Mexican American, Black, and Anglo American working class and low-income patients. Based on this summarization, it appears that in the areas of psychology and psychiatry more exploration in health professional/patient expectations than in the general areas of health and medical care is needed. It, therefore, appears that the need for refined studies related to health professional exploration of patients, particularly that of physician and patient expectation beyond compliance, is critical. Such studies would be most useful if race of patient and doctor or other health professionals can be analyzed as independent variables. For example, race and/or ethnic cross identification could be used as control variables so that a model could include minority/nonminority; minority/minority; and nonminority/minority. Such an approach would allow for exploration of various relationships using for example Black physicians, nonminority physicians, and Black and nonminority patients. This could also be done with Hispanics, American Indians and Asian/Pacific Islanders using the nonminority population as the major comparison group. This would also be of value when examining relationships between the various minority groups. For example it is perceived that the majority of foreign nationals, especially among physicians, who are health providers in the U.S., provide services to a large degree to minority populations other than the minority they represent. The implications of such studies could more closely link health needs with the expectations of both providers and consumers. This is extremely important since there is a possibility that health needs and expectations could be met more effectively by variety of health professionals. Based on such studies, it may be possible to conclude that other professionals such as a nurse or a nurse practitioner could better meet the needs and expectations of patients. In summary, it can be said that there have been limited studies with regard to health professionals/patient expectations. Most of these studies involve psychotherapy with little evidence of primary care, promotive health care, and preventive health care. Therefore, in order to meet the ultimate objective implicit in attempts to mesh the expectations of health professionals and patients, more study and detailed analysis is needed. It is expected that such findings will contribute to the provision of adequate quality care by competent, knowledgeable, and sensitive health professionals. HEALTH PROFESSIONS EDUCATION Another area of discussion which rose during the process of reviewing the health professions education statistics, was possible reasons for the underrepresentation of some minority groups and subpopulations in many health disciplines. Even with a significant Federal investment to address aspects of this problem over the last decade, the underrepresentation noted among the health providers persists in the institutions which develop those professionals. 487 For years, many institutions providing these training programs debated the issue of whether these minority students were "qualified" to gain entry to these programs, rather than whether these students had the potential to develop the necessary competencies to provide health care through that discipline. The fact that some institutions more readily accepted the latter as the more valid issue is demonstrated in the earlier tables displaying minority graduates. The institutions having significant numbers of the underrepresented minority graduates have several common attributes. Among them are multi-racial/ethnic admissions committees; some sensitivity to the populations which these minorities represent; and the presence of a flexible pedagogic approach which can accommodate differences between individuals as well as cultures. Perhaps most important, however, is the distribution of these minorities among the faculty of the schools. Since faculty serve as both mentors and role models, it is known that they influence attitudes, practice habits, and even choice of practice by their students. It is less appreciated, but just as important, that they generally constitute the bodies which determine whether individual students will be given the opportunity to pursue such a career. Statistics support the contention that the minority populations showing underrepresented minority students are also underrepresented in faculty positions in medical and dental schools, and presumably others as well. This problem affecting minority health professions training, cannot be resolved through Federal assistance alone. More collaborative efforts must be undertaken with the health professional organizations and training institutions, as well as other public and private sector interest if some reasonable resolution is to be achieved. Other critical factors include: 1) the disproportionate number of low-income students among these minority groups, in a period of escalating education costs, 2) competition for careers from non-health sectors and 3) significant problems encountered by some students during earlier stages of their basic education. They also must be addressed through such collaborative efforts. Without attention to these and related issues it is unlikely that there will be in the foreseeable future any noticeable improvement in the rates in which underrepresented minorities participate in health professional careers, as a vital resource for the minority communities which they represent. HEALTH SERVICES RESEARCH AND BIOMEDICAL RESEARCH Another area suitable for at least brief discussion is the degree to which health services research and biomedical research focus on minority populations and communities, and the diseases which primarily harm them. Related to this is the degree to which minority health professionals and scientists participate in these two research arenas. 488 The underrepresentation of Blacks and some other minorities in the biomedical sciences and other science fields has been a major factor influencing the quantity and quality of research available to address minority health problems. The paucity of numbers has been well documented by the American Association for the Advancement of Science, and others. The reasons for this underrepresentation have been cited as many, and some efforts have been initiated to address some of them. More needs to be done, both within the Federal government and the rest of the public sector, and in collaboration with the private sector. The recent report of a study "The Career Achievements of National Institute of Health Predoctoral Trainees and Fellows," conducted by the Institute of Medicine of the National Academy of Sciences, has confirmed that: 1) Graduate students who receive NIH predoctoral support are more likely to obtain doctorates than other biomedical science graduate students, and 2) they were considerably more likely to have subsequently received NIH post-doctoral fellowships or traineeships, 3) they were more likely to become involved (at later stages in their careers in NIH-sponsored activities, 4) they applied in larger numbers and were significantly more successful in receiving research grants from both NIH and the National Science Foundation, and 5) they have authored more articles and have moved more frequently to faculty appointments at the leading 100 research universities. The participation of Blacks and other underrepresented minorities in these and other Department research activities should be monitored, and more targeted efforts made to increase participation at all levels. The exaggerated incidence and prevalence figures for many of the diseases and conditions studied in this global report seem to provide ample documentation that the events and occurrences affecting the health status of minority persons and their communities are not merely a microcosm of the nonminority world. Therefore to approach these problems from a rational viewpoint requires a perspective different than that applied to the nonminority world. Health Services Research and Biomedical Research are two arenas in which the perspective of the investigator can and does influence not only what hypotheses are tested, but how the "objective" results are interpreted. Even on reviewing reliable statistical data affecting the health status of minority individuals and populations, it is important that the perspectives of minority researchers representing those populations be given paramount consideration. 489 There are numerous factors which influence a person's health status which are related to that person's life-style, environment (home and community), dietary practices, cultural practices and customs, and others. The degree to which knowledge and understanding of these variables affects minority individuals and communities is similarly related to the degree to which there will be success in confronting minority health problems. These concerns are particularly appropriate when you ask the research questions "how" and "why," but also have relevance for objective questions such as "what," as addressed in the collection of health statistics. The numbers and percentages of minorities engaged in the research and data collection of health issues is decidedly low among several minority populations. The perspectives of potential minority researchers, that could be shared, remain unknown, perhaps contributing to the lack of success in addressing these health disparities. Another consideration in this same arena, is the degree to which resources are targeted on health problems with high adverse effects on minorities. This argument does not relate necessarily to a primarily minority condition such as Sickle Cell Disease, but more to the more common afflictions which take a heavier toll on the overall population. Is "alcoholism" in American Indians the same disease as "alcoholism" in the nonminority person or community? This is not a facetious question, but one which has implications for both social and biomedical research. This issue has been explored in the Chemical Dependency Subcommittee Report, also in this document. The same question might be asked about several other diseases and minority groups. The health professional and other resources which are directed towards these problems are inevitably directed towards the further investigation of the disease or condition from the nonminority perspective. This is done even in the face of data which demonstrate that the problems are more severe and costly to the minority populations. It has further been demonstrated that selective oversampling of target minority groups can provide immensely helpful data on those groups, which otherwise would be unavailable or too costly to obtain. Inhibitions and/or restrictions in the use of this technique can only continue to delay the investigations necessary to advance the status of the health of minorities. It becomes obvious that further attention needs to be drawn to these and related issues. 490 REFERENCES Acosta, F. X., Evans, L. A., Yamamoto, J., & Wilcox, S. A. (1980, November). Helping minority and low income psychotherapy patients, "Tell It Like It Is." Journal of Biocommunications, 7^(3), 13-19. American Council on Education, Office of Minority Concerns. (1984). Minorities in Higher Education, Third Annual Status Report. One Dupont Circle, Washington, D.C. Anderson, T. A. (1980, March). New Americans...new dental patients. Dental Management, 20(3), 28-36. Association of American Medical Colleges. (1985). Unpublished Data. Washington, D.C. Association of American Medical Colleges. (1984). Physicians for the Twenty-First Century, The GPEP Report. One Dupont Circle, N.W., Washington, D.C. 20036 Association of American Medical Colleges, Office of Minority Affairs. (1985, March). Minority Students in Medical Education: Facts and Figures II. Washington, D.C. Association of American Medical Colleges. (1983, October). Medical Education - Institutions, Characteristics and Programs. A Background Paper. Washington, D.C. Association of American Medical Colleges. (1984). Minority Student Opportunities in United States Medical Schools 1984-85. 8th Edition. Washington, D.C. Bradshaw, W. H., Jr. (1978, December). Training psychiatrists for working with Blacks in basic residency programs. American Journal of Psychiatry, 135(12), 1520-1524. Child Neurology Society. (1985). Unpublished data, per telephone conversation, Dr. Lawrence Lockman, Society Official. Coggeshall, P. E., Brown, P. W. (1984, November). The Career Achievements of NIH Predoctoral Trainees and Fellows. Institute of Medicine, National Academy Press, Washington, D.C. Equitable Life Assurance Society of the United States. (1983, August). The Equitable Healthcare Survey, Options for Controlling Costs. New York, New York. Evans, L. A., Acosta, F. X., Yamamoto, J., & Skibleck, W. M. (1984, January). Orienting psychotherapists to better serve low income and minority patients. Journal of Clinical Psychology, 40(1), 90-96. 491 REFERENCES (continued) Flaskerud, J. H. (1980, January). A total for comparing the perceptions of problematic behavior by psychiatric professionals and minority groups. Nursing Research. 29(1), 4-8. Graettinger, J. S. & Swanson, A. (1984). Specialty Distribution of U. S. Medical School Graduates for the PGY-1 Year - 1977-1983. National Resident Matching Program and Association of American Medical Colleges. Hanft, R. S., Fishman, L. E., & Evans, W. J. (1983, June). Blacks and the Health Professions in the 80's: A National Crisis and A Time for Action. Prepared for the Association of Minority Health Professions Schools. Hayes-Bautista, D. E. (1978). Chicano Patients and Medical Practitioners —Sociology Knowledge Paradigm of Lay-Professional Interaction. Social Science and Medicine. Part A, 12(2A), 83-90. Haynes, R. B. (1976). A critical review of the 'determinants' of patient compliance with therapeutic regimens. In D. L. Sackett, & R. B. Haynes (Eds.), Compliance with Therapeutic Regimens. Baltimore and Chicago: The Johns Hopkins University Press. Hershey, J. C, Morton, B. G., Davis, J. B., & Reckgatt, M.J. (1980). Patient Compliance with Anti-Hypertensive Medication. American Journal of Public Health. 70, 1081-1089. Hulka, B., Kupper L., & Cassel, J. (1975). Doctor-Patient Communications and Outcomes Among Diabetic Patients. Journal of Community Health. 1, 15-27. Karsch, F. V., & Morris, M. J. (1979). Gaps in Doctor-Patient Communications. New England Journal of Medicine. 280, 535-540. Kirscht, J. P., & Rosenstock, I. M. (1977). Patient adherence to anti- hypertensive medical regimens. Journal of Community Health. 3, 115-124. Levy, D. R. (1985, April). White Doctors and Black Patients: Influence of Race on the Doctor-Patient Relationship. Pediatrics. 75(4). Mechanic, D. Public Perceptions of Medicine. (1985, January). New England Journal of Medicine. 312(3), 181-183. Poma, P.A. Accent on Communication. (1985, January). Journal of the American Medical Association. 253(4), 506. Rack, P. H. (1977). Some practical problems in providing a psychiatric service for immigrants. Mental Health Sociology. 4_(3-4), 144-151. Report on Medical Education in the United States 1983-1984, 84th Annual. (1984, 28 September). The Journal of the American Medical Association. 252(12). " ' 492 REFERENCES (continued) Scientific Manpower Commission. (1984). Scientific Engineering and Technical Manpower Comments. Washington, D.C. Sherman, S. N., Tonesk, X., Ph.D., & Erdmann, J. B., Ph.D. (1982, June). Datagram: 1981- 1982 Enrollment in U.S. Medical Schools. Journal of Medical Education. 57, 495-498. (and prior Datagrams). Standeven, M. W. (1977, October). Social sensitivity in health care. Nursing Outlook. 25>(10), 640-643. Tate, B. I., & Carnegie, M. E. (1965, February). Negro Admissions, Enrollments and Graduations - 1963. Nursing Outlook. 13, 61-63. Todd, J. G. (November/December 1982). The Indian Health Service - A Comprehensive Health Care Delivery Program. Journal of Environmental Health. 45(3), 112-117. U.S. Department of Health & Human Services, Bureau of Health Professions, Office of Data Analysis and Management. (1984, May). The Area Resource File (ARF) System, Information for Health Resources Planning and Research. (DHHS Publication No. HRS-P-OD-84-6). U.S. Department of Health & Human Services, Bureau of Health Professions, Division of Disadvantaged Assistance. (1984, September). Minorities & Women in the Health Fields 1984 Edition. (DHHS Publication No. HRS-DV 84-5). U.S. Department of Health & Human Services, Bureau of Health Professions, Office of Data Analysis and Management. (1984, July). An In-Depth Examination of the 1980 Decennial Census Employment Data for Health Occupations - Comprehensive Report. (ODAM Report No. 16-84). U.S. Department of Health & Human Services, Bureau of Health Professions, Office of Data Analysis and Management. (1984, May). Report to the President and Congress on the Status of Health Personnel in the United States, Volumes 1 & 2. (DHHS Publication No. HRS-P-OD 84-4). U.S. Department of Health & Human Services, Health Resources Administration Office of Health Resources Opportunity. (1979, April). The Treatment Practices of Black Physicians. (Publication No. (HRA) 80-628). U.S. Department of Health, Education & Welfare, Bureau of Health Manpower. (1978, March). Influence of Preceptorship and Other Factors on The Education and Career Choices of Physicians, Executive Summary. Hyattsville, Maryland. (DHEW Publication No. (HRA) 78-74). U.S. Department of Health & Human Services, Bureau of Health Professions, Office of Data Analysis and Management. (1985). Location Patterns of Minority and Other Health Professionals. Rockville, Maryland 20857. (ODAM Report No. 4-85). 493 REFERENCES (continued) U.S. Department of Health & Human Services, Bureau of Health Professions, Division of Disadvantaged Assistance. (1985). Minorities & Women in Undergraduate Education: Geographic Distributions for Academic Year 1982-83. Rockville, Maryland 20857. U.S. Bureau of the Census, Population Division. (1985). Unpublished data. Washington, D.C. U.S. Department of Health and Human Services, Health Resources and Services Administration, Indian Health Service. (1984, October). Indian Health Service Hospitals and Health Centers. Rockville, Maryland 20857. U.S. Department of Health and Human Services, Alcohol, Drug Abuse and Mental Health Administration, Office of Special Populations. (1985). Unpublished Data. Rockville, Maryland 20857. Vaughn, J. C. (1982, October). Educational Preparation for Nursing - 1981. Nursing and Health Care. H8) . Vaughn, J. C, & Johnson, W. L. (1979, September). Educational Preparation for Nursing - 1978. Nursing Outlook. 13, 608-614. (Also prior issues for 1966 and 1969). Waitzkin, H. W. (1984, 2 November). Doctor-Patient Communication. Journal of the American Medical Association. 252(17). White, E. H. (1977, March). Giving health care to minority patients. Nursing Clinics of North America. 12(1), 27-40. 494 APPENDIX I Counties with 20 percent of Population In Any One Minority Group Key to Appendix I Minority Group: Counties with 20 percent of Population in Any One CONAME STNAME PCTBKPOP PCTHSPOP PCTIDPOP PCTASPOP POP 80 County Name State Name Percent Black Population Percent Hispanic Population Percent Indian Population Percent Asian Population Population 1980 495 Appendix I Counties with 20 percent of Population in any One Minority Group rOrOOOOOOOO^^^^^O(NJO>^rOinOrO^^O^^^^OC\|rOO^rOO^^^SOr^ oooooooooooooooooooooooooooooooooooooo»-ooooo'^oo'--oooooooa CSOOOOOOOOOOOOC\IOOOOOOOOOOOOOOOOOOOOO\3-ON3-Or^rO'-C\lr*-C\IONOrOOOOOOOOO r-. cm *- *- n3" (S|sONTNT(\llOCM^^ ^V'T-T...O'-'-'-OO'-'-O'-N'-OOO'-'-T-'-O',«-OOOO'-'-OO'-(\|f0v00>'-'0SN'"(>^sX^vOr^^^r^^r^CNJs0^vOOCO^^rONrO<\|rOinOin^OOOrOvTrOOOOCM^OOlO t\j»a-ror^coinroc\j^sTsoNTCMc>coc\Jr>.CNirOv3-in>y^ C\J>3-C\lM3^rO^srNTrOCMincSICMr^M3^C\I^C\lr^00U^rOvJ-rOvONTrOC^ CMCMCMUO<\IK>nTCM 00 C/> c*uj xt/>iucaiJJ =>0 OHJCQI-^0:3XX-iOO_------- 496 Appendix I (continued) oo D. o 0_ rO m >» »-cm M3CMm CM CM iOCMCMr«.\T SO VO>0 IO r- o 0_ z M 0_ o CL- IO < r- U 0. O^OOlOOSOON\t^^OWvo^^^ONOO^»*vtOC>0'-00«aiIO(\JCOvtlO\tininNN'-lOON»"l')NON\TsOO»-IO OOOOOOOC3OC3OOOOOOOOOI0'-CMr0v0,r-CMr>«.»-CMCMI0'-OOO'-OOOOOOO»-OO»-OOOOC3OOOO CM 0_ o 0_ Q o 0. CMrON^IOrONrO^rOr^^t^OUOinNT^CMOCMvyf^rOr^tOOCNI^lOrO^r^ 0OOC3C3OOOOOOOC0OOC3C3O0»-«-»-'-O'-»-»-'-O'-»-OO'-OOOOOOC3C3'-OOOOOOOOC3C3OOO CL. o 0_ IO x r- O CL v0^rO>^^0>^\TvOir»C\|NOOvOsTO^rONO(MCOs0^v43^^r^W\rcONTOrONr^lO o^oo^o^oc>o^^^o^^^ooo>io^vor^r>-^iAiO(Mo>^CM^ir^^r^NT^iOrOrOr^^CM^ CM m CM CM CM CM CM CM \T *- CM CM IO IO CM SO h- sT CM vd" IO IO IO \T IO 0_ o 0- « r- O Ou SOf^*OO^vOO«C^^^OCO^Or^CM^N©sTOvOCMr^l0^i^^in ^r^JNC^O^vt^fl^fN^OO^ONW^OOO^^^IlO^T^JI01fl^OON*-C\IOOOOOOOOO'-OOOO^vtOO^^ONvt(^JO»- IO sT CM CM nT \T HO IO CM CM CM \T IO NO lf» CM \T CM IO -~ ~- I**. CM »- CM so CM IO CM \T CM s3" CM CM < IH LU CO o <£<<£<<<<<<<<< O (/)(/)(/)(/10(/)trt»1W01CO(/)(/)(/l(/)(/5(OWWZZZZZZZZZZZZOOOOOOOOOOOOLL CO COCO 00 00 COCO COCO COCO CO 0000 CO 00 COCO 00 000000000000 ►H OL CO ►h CO UJ o u CO OH co 1-o UI rH QZUJ r-i 0l O _l r-O < < Q K Z UJ -HZ K- <0 z ZUJI-Q Z O IH UJ CO r- z CO r-Q-lH IF <- O—lUJrHOOX—IC0U.ZCCZQC CO < Q£ UJ UJ CQ U-CfT 3 O Z> -JrHLL. <0—lO«rHJ •HOUJCLOi:ZWCCtZZZ-lZQ_iZOLLOQ|- • o < UJ Ct: LU LU< HJ H M I HHOLUSISI-ZOCCZIU rno-J—I—I—i! ZZZZO0.Q.(/)DZU.Hiis/_)j;Zi:v)«1p-><<«OOX_i l—l OQ.ai0h«QQOOm-iZQ.(/) 497 86V rncz XJIH mm Z/TJ co >oomm> 2ZIZ111/I >mz^-n~o /DWWHrnm OZ70/TJ zcoco o rHC/) xxxooomrnrnmmoooooooooooooo >»^7Jrc<2r>000fnfnX1^0000rrr »73zm>->6d»coo oh>eoi—ntH>> H^nmozHzzmrQro^x^zm^oiiz^s HOz--njTH cmo 7"? c/>om z m—J m r-cz o> Hmi m ^ r-73 Cd/O 30 —I H OH -< oon XXX >>> HH7) HXr- >>H X30 O Z O n X m m ooo > J>3> Z3r- oox r-mo mzc 73 Z CO CO CD CO CZCZCJC -473r--< -l/*;r-J> cornoz o CO 09 CO CO CO /TJI— rHrn> omcozr- ontd o 7i7< XZ C/)f— rH IH m r-z -< r- co>>cz—i >-mz j> ^75-DrH^ mHror ^JZMZO C/)Z 71 oo oooooooooooooooooooooooooooooooooooooooooooooooooooooo-n-n mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmr-r— oooooooooooooooooooooooooooooooooooooooooooooooooooooooo OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO^tH HHHHHHWHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHDO >3>>3>>>3>>3>3>>>>3>3>>3>>3>>>>3>>3>3>3>3>>3>3>>>3>>X>>>>>3>>>3>>3>]>>3>> >>>>>> NCwr0C^CMC*.a>v>n.a>0^l\)l\}CM^ljiC*rOLnCNCwW.l>.C^.P>^ C^(^J^X^ON^^O-Ot\)^00r\)^Ch-^-O(^NWl^N4Ci0v0v0^OUUlv0N(^N00v0NI\)^O^^-^N Woc^oirOorM-^O^Ov^rMO^l\)ON-^j>oi^^-*JrO^C>oLn-^^orNO 0OOO-.-j^00-«-i^O^0-.--0-.-.^-0-iJ0-0000000>J-«0----»---»OO0O-'O-'OlN0 ^Osui^rM.^Oi.a>fMC>.ui.*^soo.t>^OJOv/i^cr>.p.oosoc^ OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO u-»ooroooo-.ocMJ.woojN-«-.-.o-»-'Ww^Mro-ooo--.oo-.voro-.NN-'i\)o-.i\)-.jN-.oro-.oN^ro 0O00OO000000O00000000000000000000000-.0000000000000000000 (^rs)^OLMOO-^^^^NONN(^Ln^^^W(^UO^^^^ONWN-WOvO^O-Ln^-(^NOXv004N--l\)0-.(^ON^ Ln — -cn ^ln _*. _. _» -~j _* _» _. _. ui oo r\j _» _» o -. -* oo ro -» -*^Ul|\)0N000000-vJ00^J00Uls0-a.s0XNs000O'" CfNsOOsrs)Ln<^00XNL/1vOCNL7lXs r\ , .. nn *r> IONJ>OLnOOOOOC7s ^OX^(Jls0O-.CyMCM00UlUlJ> -^-fcOioo inj -» -. o -. -. oo ro _* (j4-~triro -viroo -. -»-. c* -»-»-»m _. cm _. _ _. ^^vO^*OWO«WOO^WLnvOMOWUl^ChW4>^NMMWlJlC^*OLnOLnOOON^WChtnO^ .4>C*lOOOsO.CN^-vl^-xJCWvOO.i>.a>ONrM.i>C^C»ONUi.C*^r\}WLn ONvo^c»or\)>ouiLn-*rMUirooc>>oooN>o^vo^ijisowrMX*-»-vi-. ox^oo^HjnONOiooocD^ONONCM C>^Ul^.a>C«L/lC0C000^v/l.^UlO.^00O0N.DOC/«C«rM0N0*00^ psnujjuoo) i xipuaddy Appendix I (continued) O NOsOsTCMsO 00 CO »- CMr^sT o_ mr>.iricMn O r^sosTr»-co 0_ »0 so o Q_ z M z: 0. o 0. CO < I- o 0_ ON^Os*^IOCMOOO^CM^^CM^O^^IOO^C\JrOOC\J^Na-CNJrO^OlO^OO^CMCMrOCMCM^CM «-0000000000000«-00000000^-OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOC\IO so so Ol o Ol Q u Ol MOONM'-NONOO'-'-OIO'-'-O'-'-OO'-NOOOlM'-IOOr-T.T.ooOlMMO'-OOONIOOO^OOOO'-^IO oooooooooooooooooooooooooooooooooooooooooooooooooooooooo Ol o Q. CO o Ol ^osTsrior>-ininwsTinsTr».^c\jowsor^^^ro^c>Lnc\ioo lONN'-'-O'-OO'-OOO'-N'-'-O'-OO'-'-'-'-'-NO'-'-O'-OOO'-'-'rOOOONO'-O'-'-OO'-OOOONN Ol o Ol i*. OQ h- o CL sosyr^^CMCM^iooinocoo>^^sTi^wcM^inr^osysTsrso^CMCMso^oioc^ sOCMlO^sOlOsOsO^COWr^eD^sTlTI^CSsOsT^sOsOr^sOlOCMr^N^st'sTsTONO^OCOsOinr^ IOsrCMrOrOsriOsTvTCNJsTrOrOsTIOC\JIOinC\lrOsrinU^rOIOsrrOC\lsOsTsOsOC^ UI z: CO <<<<<<<<£<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<< OOOOOOOOOOOOOOOOOCDOOOOOOOCDOOOOOOOOOOOOOOOOOOOOOOOOOOOOrHrH oro^cMo^cec^c^c^ctrc^cecdt^crrc^Q^c^ 00000000 0000 000000 0000000000000000000000000000 000 0 00003 3 LUUJLULlJUJUJLULUUJWUJLUUJlUllJLULULUlJJLUlU ooooooooooooooooooooooooooooooooooooooooooooooooooooooxx GC >- UJ UIX UI Of Ol IHCO X —I UJ UI OC XQ UIO >-Z COLi-O J— —1 21 UI O IH ZQ.Z Z—iZr- UI h- _J UJU-I— ZUJQiUIUJOZOZX S«iS<—lO>-UJOrHOi ?■ CfTO Q3ZZ02UJXOOZQ< < UJOOZQrHOrHrH_iOCr:»—OOh-UIOUJ--iQiOU-OUJOQirHCO—lQiQi 3<<<32Z r- O O oosot: oo _i ZUIUJXtOZ HD HI- _lOUIrHXrH_l XC0UJO^btif-73 r--t mo oz M >?0 zo >c o m >z COM XO I—IOOCOC0C0C0C0C0/O 3>-H —I —| —I —I —I —I l-l LOO J>H2 C0T3 J> 2>73 X-< o 3> >o zx oz 73 X m CO 3> "0 t_X >m 3r- mm coz > • o X or- x> >z 73a r— m co r—l ^. K-l \ I I 11 \ / ■ n ■—• \ t *^ r •* r* r~ ~— *— ■ i ■ w \ # r—• 730—ICZX3>XXCOO-rm"nncdH^o >oo» 0»h> OC0 —I r— zhi-^h khz oz om m m zhxo >rn cm x mooz tjco m mi— m to ">r-73 m z o > c O m oooco >3>3>ih Hr-om >OOZ i>o< O00 M CM 1— r-rn r- >cz m H 73i>737<- hzhiti cooo o -HOZ r- m—I 3> >H m zm 73 -0O>37* CZOr—>> r-om?^XM3> CO > M 7* Z M O rn 73 r-r-r-i-i-r-r-r-r-r-r-r-r-r-r-r-i-r-r-r-r-r-i-r-r-r-r-r-i-r-r-r-r-r-i-r-r-r- i-r-r-r-r-r-r-r-?«;7^MHMMMMXX oooooooooooooooooooooooooooooooooooooooooooooomJ>zzr—r-i—r—J>3> cccccccccccccccccccccccccecccccc HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH(/)HHHHHH>> c/>cowc/>cocococococy>wv>coc/>cy>wc/>c/>coc/>c/>wc/>w MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMIHMMMIHMMMMMI-lOCOZZOOOOMM >3>>>>>>>3>>3>>>>>>>J>>>>>>3>>>3>3>>>J»3>>>3>>3>3>>>>>J>J>>>?^ >3>mihmm zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz-< cocococo 3>>>a>>>>>>J>>>3>>>>>>>3>3>>>>a>>>>3>3>J>3>>>>>3>3>>3>>3>>>3>> rOL/iXNCMC^rv)i^c^r\3cwc^c^*>LnrocMCMiN3^rMi\)s7iCMj>vji oo^jo -» osoxNooo»MOooo^^uiuiciNos-.^\OLnONooooNo-a.-^^r\3J»w-.-»^LnCNvn |\jsooso-»-a.^^c*wo-a.rs>uiLn.a>.i>sooN-»rMO>.rMroo^ oo-oo-o^ro- ^^-.oroo-»-»^roon-^o--.NONHo-.-A-- _._._*.o_»_r\)-»roi\)-*CM^Nooo_.o^oo--Joo Ul00N0C0LnO-vJCMCM0NCMO«.a>r0.i>r000.pss0J>00-.^r0CMLn.C*Ln» ooooooooooooooooooo-»ooooooooooooooooooooooooo-»oooooooooo 0-»OIS)IN)OOIS)Ln-»-»-.OOCMO-*^-.rM-»-»INJ-»l\)ISJ-»0-*-»00-.OrMISJ OS Os OOOOOOOOOOOOOOOOOOOOO — OOOOOOOOOOOOOOOOOOOOOOOO — OOOOOrOO-A-sJ o — -» — -*o — -*OocM-»-a.-»-*ro-»oLnCMLnCMCMrooo.c»ON-»r\)-*o-».C»-»Oo-»oo-» |\).J>-»-.orOo-»oo^4.t>os-.ro-»CM,^J Ui _» ^ui _» CM —IN) _->JUir>) rO _»_»_». .t> .e* ro oo os .t> oo cm ro cm ro -» cm ro ro cm Ln cm cm -» cm Ln -» cm os ro cm -» -* os ro ro -» -»os vji-» .*mnj Ln ro o ^j os ro os ui_»-sicm ^rOsOCW4>-*OOO.P*0.^-»-»sO-vlrOOLn.C»ONs0^s04>l71sOO^.\JC/44>CMsO-*<^LnrO-^ r\j^oosrsj-*LnONforo-*so.p»«ro-*x*rooorosnoowso^ocM-»^-- Lnoowwoosro>oLn-*rorooroui»CMOox^x^^-*ONo^oN-»roLnLnJ>J>-»^soroooow CMOsON-»"*J-viLnoocMX^oox^Ln^Jso^JCMroLnvo-* uiCMC/4LnCM"vi^soro-*CMLnro-»^-»Ln^CMC»^CM.^ooooouiCMsn-.oLnJ>-^ro (panujrjuoo) i xipuaddy 105 "OOZZ >7*om ZHXS OMCZH l—OS COO >eomz mm x > 333 oo> ZZ/Q HJOCO OOX om> 3 r- m r- 73 ■< 33r- >>o /OOZ MMZ ocoo zom zoo Hmm z-nm Or oo r-73 zm I— i— i— r— 7\ m>>a>rn >ZCT1Z 7*-73X-)3>-T3 mm m -< m Z/OITI/0 OOH m>-H r-j-n m om z-n mm corn 73 co o z o 3> < M CO m>co TICOCO -n"o> mrnjQ 7373a. co rn o z z > XXX COM 3r-Z -030 xmco POCO m -< CO oom 737373 mm > zmz >Z/"s omr- 3> M mo oo 73< 73>~i mz COO nnononnnncfltji>>>zzzHW"D^on oor-i—r-xx>>om-H3O3>OM3>O/0mox "03>>3>>OMXJi—r-Z—)M3>- Hi-<73Hnn»iHH>H2znowmz-in^ >o T^eo —l7SOO>r-c3>z> com—icom mm > 73~Z.CAr--Z.73 "HO m COCO z > mo m 2: 73 IO m o 73 O m co rn 73 o w >3> r—r— >>>>>>>>> COCO COW CO C/JCOCOC/) COCO CO Cy>COCOC/)COCOCOC/>C0CO COCO C/)C/)C/)C/)C0COC/)C^ c/>c/>coc/)c/>c/>cococy>coooooooc>coc/>cocy}oococ/}Ooc/>^ C0COC/>C>O0COCOO0COOOCOCOC0C/>C/)C/>COCOC/>C0C0COC/>COC0W coco cocococococococococococococo coco cocococococococo co cocococococococococococococococococo>zzzzzzzzzz HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHZODOODDdDOD ^mT0^_0"O"O"0"O"O"O"OTJ"0"O"0"0"O"0"O"0"O"0'O"O"O"0"0"O'O"0"O"O"0"O"O"0"0"O"O"0'O"O"O"O mm^mT0^^^m^^^^^'O"O"O"O"o_0"O"0"o"o"O"O"O"O'O"O"O"O"O"o"OTJ"O"o"O"O"O",o"O"o"0 ^CMOsro*»roLnMLnCMWLnMCMCMCMroLnroLnooj>uic^-»jj>4Ni\)CMrocMX»ON^ C5o^*s^osocMsOLn^os00j»o-*ON^CNWisjNOLnLn-*Ln-*o^^^oo^o^^oooNUiLnrO-^so-^ vocMONroso^rososoroo-»j>sosocM^CM-*osc3roosos-» — oo-.ro^oN4>c3ooocn-»ocMCM-»vo-*osi^vnCMWCMOscMro"*Jro-»oo _^OOOOC3C3O<3O-»-»OOO-»OOOO-'O-»^OO-^OO-'C3OC3CDOCD-»Cir0O-»OOC>CD-kOOCDOr\)-'-»-*O-. C^OOOOOOCMLnOOLj1*>JOO^CMONLnCJirO^OJ>sOX^CMONOO^sOOOCMCM-.-.-sJsOOOCOON^OvOOLnU1^ OOOrOOOOOOOOOOOMOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OO^f>OO^^ONO^OCNO^^fi^OOOOOOuO^O^O^^OOOO->OOOOO->^lNN-'ONUlUIO00^lH OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOrOOOOOO -0)0-OON-W^NJNNOOaNlO^->OOOl\)WOJOOO^-.OtMW-»^000-WO-iJjSlLnUirOOl(M^MMN^ ro ro cm os •>« ro cm -» -* -» cm ro ro j> cn cm jmji -» -» **J cm -* os -» _» -.rouiro os _. ro cm ro -*• -». -* -»j> -» -» cm cm cm os ro -* os _* cm -^ cm oo OOONC^vOCMOssOLn-»^0-»^OOM^-*0-.CMs0^rOCMrOO-.>OOOONl^ONOs^OsrOC»"sJsOLnLnW -» oro^40CMj>ro"*JOscM-* Lno^JLnroo-.sooo-» roLnsososooooroosOLnsoosorosooo-viosso-* oocmoooooivio\-»ooson-vjon-vi os_»_»X"-ososoo-»o"*Jroosso-*oocM^-.*>ooos-» CM^Jsoj>roo-.roo-. oox,>-visoLn-«JosoNcjiosoNCMsoooj>ooo-visoroLncM-»J ^wisj^oN^ONOocMj>x»Ln-»oooLnoO)rooN-»LnCM-»ooocM^woo^cMC»rsjijisooN (panuT^uoo) i xipuaddy Appendix I (continued) OOrO|vsTO»nsT*OOCMOr.mcMfllflNC\|UNOMO»-OsT«0 — -.— -.^ — .^ocro\3-o^ocovjosor*-oooNCMOinsoia->so CNIsTCaOONOsiOsTCNJsOlOCMsOr^sTr^rOsTin sOsT^soiOOiOsOsTsa-OONi/NsT'-CNIinoOCsl cm m »-«- m sj-»-cm »-10 co cm r>- «- O^N^OOOOstsTN^NsO^OOON^NlflNNO^OOsTC>OONIOOin(NlN^OUNC\I^WIOOvtOOlONN^N1^vO OCOOOOOOOOOOOOOOOOOOOOOOOOOO'-CM'-OOOOOOOOOOOOOOO'-OOOOOOOO'- CNJC\IO^OO(NJIOC30C\JOOC\I^OOsoC3^^^CMOOC>ONrOC3CNJ^s0^r^^^eOONsOsro^r^r^^ OOOOOOOOOOOOOOOOOOOOOOsO'-in>O^^OON*"0'-OOOOOOOOOOIftOtfO»-'-NiOONOO \T IO sT IO CM IO so *" CMIO ITlsOlT|^^IOincOOcOrOCM^iOOsOCMC^OOI^r^CMCM^insocO^^^COr^sO>T ^00^(\IOOO^O^»-0'-»-0'-0000»-(S|OOOOOONvOvOCOOrH.ONO'-(\lsTN»-*OC>IO>*»-IOvT'-rs.'-T-10IAvO CMiOCMrOsTrOir»rOlO0OsrsrCMC\|IO^CCCMrOr>.CMCM'^COinCMsT SvTOOSNN>OOrONjOOvTN^^WNtf^^^OOO^Nr\Jin^O^^O»O^WOOr1^(\lNNOeOO'-vOCOfNllOsTNN '-l0^olOlflO'-^JN^co'0'-^ln»o^oc^cO'■^)lnoooooo^N^lo(Moo•-'-oooovto»-oovT'-o»-oooooo CM sT in IO so rO CM CM so u^ IO r>-sT rO IO IO so IO IO IO CM sT IO »- ClQ.0_0l OlO-ClQ. c)_OuOLQ.a.o.cvcLa.o.Q.CLa.Q.Q.Q. CLOLQ_CJ_Q.OLQ.Q.Q.Cl-C^OLQ.CLa.Q. CO CO CO CO CO CO CO CO CO CO CO CO COCO COCO COCOCOCOCOCOCOCOCOCOCOCOCOCOCOCOQiQi CO COCO CO CO CO COCO CO coco CO coco co cor>z> HHHHHHHHHHHHHHHHOO cocococococococococococooocococococo CO CO COCO CO CO CO CO CO CO CO COCOCOCO CO CO CO << zz << hl- zz lOO zzzzzzzzzzzzz:zzzzzzzzzzz zz oo zz >- UJ ->zz i-oi Z«2222 OLU LULULUUJ zzzzzz >-oo IUJOO COMM QSXX ILL! UIUJ oooo uooo oooo oooa ooooo ooooo ooooo ooooo oooooo oooooo xxxx LU UJ LULU zzzz 2222 LULULUUJ zzzz xxxx UIUJ UIUJ 2222 LULULUUJ zzzz xxxxx LU LULU LULU zzzzz 22222 LULU LULU LU zzzzz xxxxx LULULUUJ UI zzzzz 22222 LU LULU LULU zzzzz xxxxxx LULU LULU LULU EEZZEE 222222 Ii I Ii I Ii I Ii I Ii I Ii I < >- Eh QiUJI-r- C£Sj£rHO UImOO o-a-cyco >-z UIO s^COX QiO-r- <<3< COCOh-r-p-2 z o -i o z o coz zo UlMUlMl— Qixzs^co orco>-_iz << => oo BQOCO OOM -jmz: - >- Ol CO « 03 O IZ OC O- O LULU :zm MO LUQO >OM LUCQCO COUJQi OC0Z5 OOX OCQir- xo< UlCOZ COQQi C03LU uixea UI < 0. u.pao-zQQ< < _J «r- UJ __ >" HJJZ3HJ Z UI CtJLU<>3i-i z: < CtT Qi O<»~'<0i>- COQ r-QiO h<-i zq<lO oososrco*-CMoor». sT^^sTsosTsoCM 00>OLT-IOONN CMOsT O00 sTNO <-CM«- O Z Ol O 0. to < r- O 0. ^rOCNICMin^lOO^OCMCMO^^inCM^I^OOOCO^rO^OsTOsrsTrO^rO^OCMCN^r^^CNI^r^ 0000'~C\linOOOOOOOOOOOO»-0000000000 00000000000000'»-00000000000 Ol o a. Q o 0_ U^^^T-cNjCMCNJrOOsTO^lO^CNJCM^CNJCNIsoCMCM^iOCMOCMOvTCMCOI^inoC^ ^oosToooooo^oooooocMo^ooooooooocMooCMoooooooooooo«-ooooo'-jaro'-so 0_ o CL CO X h- o Ol .-.-CMC>WsOtn^sTv3-inc>OCMsOinsoO^«sOsOOsOvr^CMr>.OCMr^ oxV".-'i^rOr^rO^O'r-'o^^ooo^o^rOoooo^o^ooo^ so sT IO sT NO »- CM 0_ o Ol s^ pq h- o Ci. o^orv.o>srcoinr^CMr^^CMinc>in^CMCMeocMsrr^sr^soc3s^c>^N oQooT-'wT-\ftT.'»MiflNio^oososT^ovjoNio»n\fsio«a,niin^ oooo-CN-so-ONOOrocsj^so^oor^osT *r- 22222 UI UI UI LU LU zzzzz oo orroi QiQiOO oo >->-xx l-l- zzqcq: LULUOO oo CHOC P3 Qi o co>-z o o Z2CO MLUZ s^Z< Oi ouj: o oa _J" -JXUICO Q-QiOQi oooo QOLULL. !*iCO ZUI <£r- Qi< u_o CO CD oi>- LU OQi _l Q Q CQUI _l QiXr-Qi ZZ HLUOO-O QiZUlO >ZLL.U-UIU. CO MM_JOUI ZUI-JMZI-UJUIUJ OMitir-QiX h-J JOQOlOI-ILUv/ -OUJUKUJOO-i-JEEZEZZZOQ.Q.Q.(L(LQ.aapf QiZ UIO >H- oa. ZZ Ol <<3000iMZ XX0M3UI30 z < O X zzo ooz ZICOM o< CO_l 0_r- zo 3 CO CO c 73 o CO CO CO C"0 3> 33>r- H30C mHO 733>3> z CD c 73 O TOOZ3 o>zxj XZCOi- r-Omco >m?3o ZC0/O/O oc-om 700m MO 030 0 7K r-r-7K >>m fZ.-Z.73 50OC0 m>x ZC0 3> COHZ m 73 3> t/)7JZmOOHn?Dr7JI"J>mmJ>r-/gj>«ir-^wni mi*--s TJ73Tjm7J?3/Dnorrr7JWW/Oi^czcj)rnmmi->?3H> rn-7* o zoHOmmco ooa?3?j(/»zm7or-o> r-m7orn o OOmr-r-H HZOm H -<-ir- r~ rrmm 3> oz oom o zm o m m"o Z 73 z m z x cotocoz r-iomM or-zr- crncoco XHOO HZZ m ZZZE<-jM 3>3>3>3>>- zi^mn^H rim m mm z ZZ r- O r- O o z 3> 3 cocococooococococococococococo 000000000000000 ccccccccccccccc CO CO CO CO CO CO CO 0000000 ccccccc CO CO CO CO oooo cccc xxxxxxxxxxxxxxxxxxxxxxxxxx OOOOOOOOOOOOOOO >>>>3>>>>3>>3>3>>3>3> 737373737373737373737373737373 OOOOOOOOOOOOOOO OOOOOOO >3>>3>3>3>> 73 73 73 73 73 73 73 OOOOOOO OOOO 3>>>3> 73737373 OOOO CO CO CO CO OOOO CCCC XXXX oooo 3>>>3> 73737373 OOOO COCO OO CC COCOCOCOCOCOCOCO OOOOOOOO cccccccc xxxxxxxxxx 00 3>3> 73 73 OO OOOOOOOO >>>>>3>3>3> 73 73 73 73 73 73 73 73 OOOOOOOO TJO zr- Z3> cox - 3> OO r— r~ >3> XX 00 33 >3> HHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH, zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz 3>J>>J>3>3>D>>3>3>>>3>3>3>3>3>>3>3>3>3>>3>>3>>>3>3>>3>3>>3>>3>>>> ozzzzzzzzzzz xooooooooooo Y-i 73 73 73 73 73 73 73 73 73 73 73 o-i-i-i-i-^-i-i-i-<-*-* XXXXXXXXXXX OOOOOOOOOOO >3>J>3>>>>3>3>3>3> 7s 7*. 7\ 73 73 73 73 73 73 73 73 OOOOOOOOOOO HHHP-P-P-P-P-P-P-P- >3>3>»HMMMMMMM zzzzzzzz 3>>>>3>>3>a> CO H z roosro JNrocMC ui os on ro ro cm Ln ro un ro .t> cw ro *« *« x^ Ln cm cm cm in ro cm xn in on ro cm cm ro cm cm xs un ro j> cm |SjrONOJ>OlJlOOOs — Os(NOO-»sOJN-»ONrO|SJsOJ^-vlOOvOLn-»OLn-vl|SJOOJjNJ^XvrO^ CMCMLn|NJC/4CMLnOONCMO^rOorONNO-.OsoOOIjnJ>«CMNO-»J>.l>ONOscj^ "0 O H CO 7* "0 o "0 oj^^^OO-»-»-»OO-»-»OOO-»-.-»O-^O-»-»-»-»O-*-»-»O-*-»-.CMOOr0OOCMOOO-»OOOO-»OOOOOO ^l*r0Ln^^0s^^000000XN->J^00^r0-*0S^00O00CMON0CM00X»0sXNC3U1OOLn^0s00000N0^ -0 O H X CO "0 o "0 rorocM OMjnro ro -»OOOOOOOO-.OOOOOOOOOOOOOOO-»OOOOOOOOOOOOOOOO0sCMOJ>-*J00OOOCMOOO.i> ^l^l^jvjLoroLLj^o-*-*-'-*-*^"^ ■v O o "O o •v OOOOOOOOOO 0000 0000000000000 oooooorooooooo—■ OOOOOOOOO OOOOOO roo-kONXNOsoj>-.rorooo-*-*roroxNoXN-*-*oc^c»xs-*oM TJ O H > CO -0 o no ■o O ro ro ro _ Os ro • Cm -* TJ Os cm O O l-*4sOrO "TJ I X>> -vioo 00 I 00O1CM o (ponujzjuoo) j xipuaddy Appendix I (continued) ^insosor^src>rowco^anrowinso^o^roin»oo^iocoior^NTr>-cnos^ sT^ONsorOsosTCMCMO^O|^^insT^iOoOCMinsoroOOr>-CM^rOr».CMr«.'Oi^ Or^^^sTCOCMIOrOlOCOrOOOrOinin^ONrOrOO^OcOrOin^rOsOinr^CMN^ io^ininroiocM^r^cMrN.inioosTsrr^CMooroinwsoooinr^oocM^c>c3N<^c^ *- r-. CM CM CM CM r>-r>-ro NO •«-CM CMoor*. »-cm»-o CM*- ro »-cm •-»-«-*- r»*~ ro *-«-«- sT r* O CM *~ Mt *~ OOIOOO^OIOO»-U10rOO^NinrOU^rO^OC>(O^NO»-0(ONOO^N^IOino^ON^^^^>f^OOIOOlf|K)'-C\l OOOOOOOOOOO oo oooooo «j- OOOOOOOO *- OOOOOOOOOOO*- OOOOOOO »-oorooooo sOCOrOO>srincOsTso^CM^OOsro^^CMr^l0^lOlOsTlOOsTCMCMCM^sTO^sr^inCMOO^rOCM OOOr^r^lO^CO^r^WOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOCM ior>.srinsTCMiooNr^in sorooNTin^^^ror^r^^r^incosooNsooo»coc>ONsoCM^o^»c>^^coino>ONr^insoos^ ooooooo'-ooo»"'-'-oooo^3•'■^^^ln^o'•M^ooN^o^'-o^o^J•^o^tf<3■^l0^o^oeol'l'-^'-(^^^0'-«ootf CMsTrOsTsT sTOO'-rOrOr^ IO IO CM CM CM sT NO so IO sT CM r^ 00 CM sT so IOCM so IO CM O O O O^^O^OON^inKO^CANNMWvtCOsTOOsTOONsTNininOrOtfOOOOW\tO^C>(>sTC\lsTNeONvO*O^IOO'- OOOOOOOO^OCNOsO'~OON'-sO,^'~0«-CMr>-CNJOO^soCNJOinCMrOsO^OOOinOlO^OOOs^OIOr^sTin'T-OCMsO cm m 10 m 10 cm sr cm cm cm cm*- cmcm *- cm »-cm << MM oo S^Sj/ << << MM oo << <<<<<< ooo <<< xxxxxxx 33 oo coco 33 oo CO CI 333 ooo CO CO CO ooo SjSNj/S^UJUl <<< LULU qqqcoco coco XXXUIUI MMMZZ 333ZZ O O O LULU CO CO COMM LULU LULU UIUJ LU LULU LULU LU CO CO CO CO CO CO CO CO CO CO CO CO LUUJUJUJUJUJC0COC0COCOC0COCOCOCOCOCOCOCOCOCOCOC0C0COCOC0COCOCOCO ZZZZZZ<<<<<<<<<<£<<<<<<<<<<<<<<<<<<<<<<<<<<<< zzzzzzxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx IjJLULULUIUIULULULUIULULULULULUIJJLUIULULULULULULULULULULULUUJL^ MMMMr-h-H-MMI-H-MMMMMI-HI-H-l-l-MMMMMMI-l-HHI-r-MMMMMMMMMM MO Z M-JZ O UI-C0Z ZLLC0UJsJi< ZU.0i2OZ UI30UIO- oaoaoQ-)_j UI h-Z MO -JZQ _|>-Qi>- M<<<< NOLL.XX LU -J < OZ QiO UICO Qm 3Q <£< -•Z >-z pao Z < Qi Z OCOCO UI o C020>- MCOCOLU30 QiUIOUl Qi LU CO S»i LU 2 O Qi MOM M CO ZM CO O UI>-QiZCOMM Q xuis^eauio MM_ioiai o z O UJ CO co -m -lOQi —1< —i o UIU.U.U. UI z o I- co CO LU< UIOZN illHH Qi oiai<< LLILOO 505 Appendix I (continued) sTrOrOOcOCMinsyONOONCMoOr^COCOCO'OrOONWONsysrsO'-in^ONOsTWWsrsOCOC^ OO^MoOOO^sOC^C\llOONsTCMsTrOsOO*ONsTr^insO*-C\IOCT*OCOsTsOWCMO^ rO'-00inf^inC\|inCMCMCM»-r>.soOS'-syininCMrOsOinCMso sor»-sorOsOCOrO*-rOOsorOCMCMOsO»-*-Nr»-cOCMsorOr<-NXOCOOro •-m so ro so r^ CM O ro NO CM NO CM *-O mso NO CM ro 00 m O On oO O O sT r>- »~ NO CM On sT m r<0 CO "- sT m sT CM so m O sT »- 00 *~ 00 CM ro 00 »-»-sT ro in sT oo cm cm ro in ro ■a- ro »- cm »- *- rorocM «- io *- so *- «- *- »- *- sT »-m CMCM sT ro m \X 00 CM CM ro CM CM CM o»-»-osysycMCM*-*-»-r>-o*-CMo»-ooinr>-*-ocM»-o»-ocMooro»-oo*-*-ro»-ino^oo'-rorooo»-oo»-CM»-ro O0000000000000^00000^0000000000---000000000000000000000000 OOOOCMsrrOCM*-0>lOsTin»-C\IOC\|.-0^rorOOCMsTOCMOinCMsTsTv3-sTCMOs^ OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOCMOOOOOOOOOOOOOOOOOOOOOOO oosroosysrsosoinro»-i^ocMCM'-incMCMinin'-rof^CMocMONU^r^cosoo>rosTso*-ooo^ »inoooNiniO'-o»-r».ro»-oor>.Nroiv.iocNir^CMororocMinf^sTrv.osToorocoin'-r^ CMN0CM CMro roooCM CMmsr on so sT oo m m ro r*. rorororo ro CM on sT CM CM sT ro sT r^ ro CM soro sT CM 0^vfi«v0^lOONOrOONONOvOlOOinO(\JONOONOeON^N^^rO^COT-OvO*O^O^ONOO>-^CO'00'-vT(vJ^ O^^N^sT^»00«ffNvfOOCOOONOIOtfvOON100N10sttf^\rOOOunT.o\f\fr-OOrOOOPJOOt\l'-OV'-'-'-'-'- •j- CM IO ro CM CM CMN0 «- CM CM CM CM NO CM CM CM CM CM C0COC0COC0COCOCOCOC0COCOCOCOC0c0COCOCOCOCOCOCOCOCOCOC0COCOCOCOCOCOC0COCOCOCOClC^ <<<<<<<<<<Q—i CO—I C0<-Q <<-z LUO -IM SjiCO 03 OO X> M< QLUQ QiOL oo COO QiO U.Z LUrH LUCO 2LU z ZQi M< ->N^ o >->-oi QLULU uizpa zzui UJH_I S^S^Sj/ UI Luzy _lO< -JMO <z o>-o oz coo ; hh QQi <<< zzz . K>ZLUaOLU«QUJLLCQcOZZZxiljH 0»-'-sr'-00'-C>CMO'-OOONT»-CMC^ oooooooooooooooooooooooomoooooooocooooo*-oooooooooooCMooooo sr Ol o Ol CO X M o Ol c>»-sTinroo»-oNOCMosTroinooinooroinoooo'-o'-inc*!^ooNToosoo^o>ONrooNfN. vO^OrslOsT'-OOOlfllOOSlflNIO'-'-OvOinS'OOvlO'-OOO'-pJO'-OOO'-OOO'-OO'-OOO'-O'-OO'-OOO On CM sT CM sT IO CM CM m so NO CM ON CM 00 NO CM CM r-. OO 0_ o 0_ pa M o Ol orOOr^OsTOCMrOinsoOCM»-0>»-'-sosoOsTso^CM'-sorOinCMsrNTrOOsosOsO'-inosTrOsorOOOOsC^ ooosroiosyoc\JO'-r^Nrc\ic\ic\iosoo'-cM'-oo^socooror^CMC3iooooorOsrc3o>so'-o^c^ CM CMsT CM »- ro IO CM CM in sT CM sT r>-ro CM sT in >J" CM ro in ro NO ro CM sT CM sT CM ro ro ro ro \T CM <<<<<<<£<<<<<<<<<<<<<<<<<<<<<<<< LU zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz Z CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO CO HrHMMMMMIHIHIHMMMMMMMMMMMMIHMMMMIHMMM < <<<<<<£<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> CO 2 LU z o o o Oi-iO QiOiM QiM >-Om Qi z QiZM UJOQi UI QiM UlUIQiQi ZQ>OUI O-l Miti M<-lO-l D_XM< 33>>2 O M o Qi Z UI M -icax -JQiCO - oo M< paoi-i CQ<_J LUXM 222 Qi ZUI O-J co *: JZ < Z<<3 3M-J-D S«i<< Z o<<< >-fsJNCO Qi o >- 0- M UI 2 x s: m q _i lu 0>ji< OUI ZUI O-l z s^ MOX-JUJ MQi-©-Om 2M Qi-JQi MOO—lUI U.OZLUMM O Qi OOX QipaUl 3ZCO <03coy — ^-^-^-^xa. CM t-.-*- r*. «- CMO'-'-ioooro»-ininoroo»-inroo OOONOOOOOO'-OOOOOOO — ororoooo»-'-oro»-»-ooc\IONN3- oooooooooooooooooo 00 r^r^oor*.oor>.Nj"NOCMr».invoCM'-»-ONCMro OOO'-'-'-OOOONO'-LT-oNO CM r».ovoNoeoNoooN»rinNOON*-inoo^"o r«.roNONOovroovoinr^N»r»-ocM'-oooo va-CMCMCMNj-rONONOCMNOCMNlinsOsONI oz <<<<<<<<<<<<- z< 32 xo MM QiM OO ZZ QUI QiO Q i -uiz:coz XZZXMQiCOMZiO 2rHrH030iCOCO