CHAPTER 5 Peptic Ulcer Disease Contents Page Iroduction..-. 2 eee 155 Epidemiological and Clinical Studies... 155 Experimental Studies Gastric Seeretion-.- 2 157 Pancreatic Secretion... 159 summary of Recent Peptic Ulcer Disease Findings. .________ 162 References. — es 163 List of Figures Figure 1.—Gastric ulcer mortality ratios of Japanese (men and “omen combined) by age at initiation of cigarette smoking (1966-70) = eee 156 Figure 2.—Effect. of cigarette smoking on volume of secretin- stimulated pancreatic secretion in humans.......... 160 Figure 3.—Effect of cigarette smoking on secretin-stimulated pancreatic bicarbonate output in humans__-.-- 161 Introduction Previous epidemiological and experimental studies of the relation- ship between cigarette smoking and peptic ulcer disease were reviewed in the 1971 and 1972 reports on the health consequences of smoking (17, 78) and form the basis of the following summary : The results of epidemiological studies indicate that cigarette smok- ing males have an increased prevalence of peptic ulcer disease and a greater mortality from peptic ulcer as compared to nonsmoking males. Among males, the association between cigarette smoking and peptic ulcer disease is stronger for gastric than for duodenal ulcer, but sig- nificant for both. For males, cigarette smoking appears to reduce the effectiveness of standard peptic ulcer treatment and to slow the rate of peptic ulcer healing. The relationship between cigarette smoking and the prevalence of and mortality from peptic ulcer disease is less clear for females than for males. Experimental studies of the effect of cigarette smoking in man, and of the effect of injection and infusion of nicotine in animals, on gastric secretion and motility have produced conflicting results. In dogs, an infusion of nicotine has been found to inhibit. pancreatic and hepatic bicarbonate secretion, thus demonstrating a possible link between cigarette smoking and duodenal ulcer. Recently, additional epidemiological, clinical, autopsy, and experi- mental studies have confirmed the association between cigarette smok- ing and gastric ulcer mortality and have clarified a mechanism through which cigarette smoking might be linked to duodenal ulcer. Epidemiological and Clinical Studies Previous studies of the relationship between peptic ulcer disease and “ivarette smoking have been conducted in predominantly white, West- *'n populations. A large prospective epidemiological study is currently being conducted in Japan. From this study, Hirayama (6) reported Year followup data on 265,118 men and women, aged 40 years and older, representing 91 to 99 percent of the total population in the area of the 29 health districts in which the study was conducted. Both male 155 and female cigarette smokers experienced higher death rates from gastrie leer as compared with nonsmokers. The mortality ratio for cigarette smokers was 181 for males (P<0.001) and 2.15 for females (P<0.05). The mortality ratio for smokers (males and females con. bined) was dose-dependent as measured by age at initiation of smoking (tig. 1). The results of this study, in the context of the genetic and cu). tural differences between Japanese and Western populations, provide a significant confirmation of the association between cigarette smoking and eastrie ulcer mortality. Figure 1.—Gastric ulcer mortality ratios of Japanese (men and women combined) by age at initiation of cigarette smoking (1966-1970). 5.00 ——_, 4.00 + 2 ge 3.00 2.67 £ g iY) GS 2.00 F = 1.59 a “i 1.00 } 2°00 CL “i Z 0 Lo GY VI V4 Nonsmoker >25 <24 <19 Age at initiation of cigarette smoking (years) SOURCE: Hirayama, T. (6). 156 Alp, et al. (7) conducted a retrospective survey of 638 patients, admitted to two Australian teaching hospitals between 1954 and 1963, with chronic gastric ulcer confirmed by roent genographie, endoscopic, or surgical examination. The findings in the patients were compared with information available about the South Australian population obtained at census in 1954 and 1961, and with a control group of 233 subjects matched for age and sex with the ulcer patients. Cigarette use, a family history of peptic ulcer, domestic stress, and aspirin and alcohol intake occurred significantly more frequently among ulcer patients. Alp, et al. (2) found that after surgical treatment, recurrence of the ulcer was significantly more likely to recur among those patients who continued to smoke, drink, and use aspirin (P<0.001). Fingerland, et al. (5) compared the autopsy findings from 765 males with their smoking history. The autopsies were performed without selection during 1965 and 1966 at the University of Hradec Kralové, Czechoslovakia. Peptic ulcer was significantly more frequent among male ex-smokers and male lifelong smokers than among male non- smokers (P<0.02). Among males, a dose-response relationship was found between estimated total cigarette consumption and the presence of peptic ulcer at autopsy. Cooper and Tolins (4) reported results from a retrospective study of the relationship between cigarette smoking and postoperative com- plications among 2,988 males, admitted to 19 Veterans Administration hospitals, for the surgical treatment of duodenal ulcer. Smoking his- tory was obtained for 1,441 of the men, and of these 273 were non- smokers, 1,018 smoked cigarettes only, and 93 smoked cigarettes plus a pipe and/or cigars. The authors found no evidence of an association between either the number of cigarettes smoked per day, or the number of years of cigarette smoking, and postoperative complications, opera- tive mortality, or length of hospital stay. They emphasized that their results must be viewed with considerable caution and listed several potential sources of bias. In addition, they noted, “* * * that these results apply only to the immediate postoperative findings and do not apply to the long-range effects of smoking upon the patient after Surgery for duodenal ulcer disease.” Experimental Studies Gastric Secretion Strupres 1s Humans Morales, et al. (10, 11) studied the effect of cigarette smoking on £astric secretion in a group of 312 patients. The patients included 138 157 with duodenal ulcer, 93 with gastric ulcer, and 81 with other gastro. intestinal disorders, who served as controls. Cigarette smoking was significantly more frequent among the patients with peptic ulcer than among the controls. The chronic effect of smoking on gastric secretion was quite variable, Male smokers among the controls and in the group with duodena] ulcers had a significantly increased baseline acid output as compared with nonsmokers in the same groups (P<0.05). After a subcutaneous injection of histamine, only the group of male smokers with gastric ulcers had a significant increase in acid output over the values obtained for nonsmokers in the same group (P<0.05). Among the smokers in the control group, the relationship between gastric acid output and the number of cigarettes smoked daily was dose dependent. No such rela- tionship was obtained for either of the two groups with peptic ulcers, In these experiments, the acute effect of smoking on gastric secre- tion was slight. In one set of experiments, a group of eight smokers served as its own control. The smoking of two cigarettes prior to collection of gastric juice had no significant effect on acid output as compared to baseline values. After smoking two cigarettes and also receiving a subcutaneous injection of histamine, the patients experi- enced no significant change in gastric acid output as compared to baseline values; 21 male patients, including members from the groups with ulcers and controls, smoked one cigarette 1 hour after an intra- venous infusion of histamine. A transient depression of gastric acid output was noted as compared with the yalues obtained from nine patients who did not smoke. Stepies iw ANTrALs Konturek. et al. (8) studied the effect of intravenous infusion of nicotine on the formation of acute, experimental duodenal ulcers in cats. The authors infused nicotine intravenously in doses comparable to the smoking of four, eight. and 16 cigarettes per hour into cats in whom near maximal gastric acid output had been stimulated with intravenous pentagastrin. The investigators found that nicotine in the two lower doses had no effect upon the gastric acid output stimulated by pentagastrin, but that the highest dose produced a significant de- crease in response, due to a fall in both volume and acid concentration. Nicotine alone failed to alter a negligible basal gastric secretion. In control animals (pentagastrin alone), duodenal ulcers were found in eight of 10 animals. Nicotine at the two lower doses, in combination with pentagastrin, produced ulcers in all 26 animals, At the inter- mediate dose of nicotine, the mean ulcer area was twice that found in 158 the control group. At the highest dose of nicotine. peptic ulcers ap- peared in only two of six animals and the area of ulcer was reduced compared to controls. Shaikh, et al. (74) studied the acute and chronic effects of sub- cutaneously injected nicotine on gastric secretion in rats. Under basal conditions, the volume of gastric secretion was initially depressed, then stimulated, and depressed again as the dose of nicotine was increased. Acid output was decreased over the entire range of nicotine dosage. Pepsin output reflected a similar triphasic response to in- creasing nicotine doses as did gastric secretory volume. In the absence of nicotine, pentagastrin stimulated gastric volume, acid, and pepsin output. The injection of micotine. in increasing doses, administered simultaneously with pentagastrin. resulted in a gradual decrease in response for all parameters. Volume of gastric juice, acid output, and pepsin output were all increased significantly by chronic exposure to nicotine alone. Based on an average smoking dose of nicotine, the dose of nicotine employed in the chronic experiments corresponded to the smoking of three to five cigarettes per day. Thompson, et al. (76) extended the study of rats described above by studying the effects of chronic nicotine injections in vagotomized rats and rats with discrete lesions in the hypothalanius, In sham- operated animals, chronic nicotine injections significantly increased baseline volunie of gastric juice, acid output, and pepsin output. Fol- lowing vagotomy, the nicotine response was completely suppressed. Caudal hypothalamic lesions did not influence the response to nicotine in the presence of intact vagus nerves. Anterior hypothalamic lesions, Tanging from the anterior hypothalamic area to the ventromedial hypothalamus, blocked the nicotine-induced gastric secretory stimula- tion in the presence of intact vagi. The authors concluded that chronic nicotine-induced gastric secretory stimulation is mediated via anterior hypothalamic activation and intact vagus nerves. The importance of local effects remained uncertain. Panereatic Secretion Stupres in Humans Bynum, et al. (3) studied the effect of cigarette smoking upon pan- creatic secretion in 23 healthy young males and females, Five control male nonsmokers were compared with seven male and two female light Smokers (less than one pack of cigarettes per day for less than 3 years) and eight male and one female heavy smokers (more than one pack of 159 cigarettes per day for more than 3 years). Pancreatic secretion Was measured by the double secretin test, using Boots secretin. The exper. ment was divided into two parts for the smokers: A basal collection period and an experimental period during which the subjects smoked seven nonfiltered cigarettes at the rate of four per hour. Light smokers had basal values for pancreatic secretory volume and bicarbonate oyt. put in response to secretin which were not significantly different from controls, After the subjects had smoked, significant depression of both pancreatic volume and_ bicarbonate output was noted (P<.001), Heavy smokers had basal values that were significantly less than in the control subjects (P<0.01). Smoking, however, did not further depress the response to secretin (figs. 2 and 3). Solomon and Jacobsen (15) reviewed some possible mechanisms whereby the increased prevalence and mortality from duodenal ulcer among cigarette smokers might be produced. They concluded that evidence from studies in animals, coupled with the findings of Bynum, et al. (3), supported the hypothesis that the mechanism active in humans involves impaired neutralization of acid secondary to the inhibition of pancreatic bicarbonate secretion. Figure 2.—Effect of cigarette smoking on volume of secretin-stimulated Pancre- atic secretion in humans. 3.0 2.9 2.4 Mean volume 2.5 of 1.97 2.0 pancreatic 2.0 fluid in 1.5? milliliters 1.5 per kilogram body weight 1.9 0.5 Oo i J mw 2 g 2 2 = a a oa a cE 285 85 8 Gj 8&5 6 0 ar “a+ vo -« G > < = ~ s g wd aj xr Ir ‘Significantly different from nonsmoking test within group of light smokers (P <0.001) 3 Significantly different from nonsmoking controls (P <0.01). SOURCE: Bynum, et al. (3). 160 Figure 3.—Effect of cigarette smoking on secretin-stimulated pancreatic bicar- bonate output in humans. 12 11.5 10 8.4 8.1 Mean hourly 8 7.1? output of ia ald 6 5.5? milliequiva- 4 lents per hour 2 0 4 se fs Bs Fo BS 5 & 5 wt £ 5 835 $&§ 3835 #6 S E Ew E> E> 6 Oo z = 3 S z Pm CD ? as °s —_ aad r zr ' Significantly different from nonsmoking test within group of light smokers (P <0.001). Significantly different from nonsmoking controis (P <0.01). SOURCE: Bynum, et al. (3). STuDIEs In ANIMALS Konturek, et al. (7) extended his research on the mechanism of nicotine-induced inhibition of pancreatic secretion in the dog, using the design previously employed (9). Infused secretin alone led to a sustained increase in pancreatic bicarbonate output. Intravenous nico- tine, at all four doses of infused secretin, produced a significant in- hibition of pancreatic volume and bicarbonate output (P<0.05). Infused nicotine appeared to inhibit competitively the effect of secre- tin on pancreatic secretion of fluid and bicarbonate. Topical (intraduo- denal) nicotine failed to affect significantly the response to infused secretin, Stimulation of endogenous secretin by an acid infusion into the duodenum produced the expected pancreatic secretory response. Nicotine either applied to the duodenal mucosa or injected intra- venously significantly inhibited the pancreatic secretory response to endogenous secretin. Nicotine had no significant effect on total pancrea- tie protein output. Nicotine did not alter the cholecystokinin-induced stimulation of pancreatic secretion. The authors concluded that nico- tine may inhibit pancreatic secretion of fluid and bicarbonate both 161 by a direct effect on pancreatic secretory mechanisms, acting as a com- petitive inhibitor of secretin, and by a secondary effect on the duodena] mucosa. depressing the endogenous release of secretin by acid. Robert (/2) studied the potentiation of active duodenal ulcers by nicotine administration in the rat. Subcutaneous infusion of pentagas. trin and carbachol resulted in the dose-dependent formation of duo. denal ulcers within 24 hours. Nicotine alone produced no ulcers, Increasing doses of subcutaneously infused nicotine, in combination with the other two agents, resulted ina steadily increasing dose-related incidence and severity of the duodenal ulcers. Robert noted that Konturek, et al. (9) found that nicotine inhibited pancreatic and biliary bicarbonate secretion in dogs, and that Thompson, et al. (16) found that acute doses of nicotine in rats either depressed or did not alter gastric secretion, He concluded that the most probable mechanism by which nicotine potentiated acute duodenal ulcer formation in the rat was via a suppression of pancreatic secretion. Robert, et al. (13) further tested this hypothesis by infusing acid via the esophagus of rats in doses found to cause duodenal ulcers in one-third of the experimental animals. One group of rats also received a subcutaneous infusion of nicotine. Another received nicotine, but only water was infused via the esophagus; 31 percent of the animals receiving acid but no nicotine had duodenal ulcers; 93 percent of the nicotine-acid group had duodenal ulcers, while none of the nicotine. water group had ulcers. The ulcers in the nicotine-acid group were more numerous, extensive, and deeper than those in the animals which received actd alone. Summary of Recent Peptic Uleer Disease Findings Tn addition to the findings relating cigarette smoking to peptic ulcer disease, summarized in previous reports on the health consequences of smoking (7, 28) and cited in the introduction to this chapter, recent stiulies have contributed further to our understanding of the USssochit or: L. The finding of a significant dose-related excess mortality from gastric ulcers among both male and female Japanese cigarette smokers, in a large prospective study, and in the context of the genetic and cultural differences between the Japanese and pre- viously investigated Western populations, confirms and extends the association between cigarette smoking and gastric ulcer inortality. 162 bo (3) 4 ~~ on (6) (8) (9) (10) (12) . Data from experiments in several different animal species sug- gest that nicotine potentiates acute duodgnal ulcer formation by means of inhibition of pancreatic bicarbonate output. . Cigarette smoking has been demonstrated to inhibit pancreatic bicarbonate secretion in healthy young men and women. Peptic Ulcer Disease References ALP, M. H., Hisxop, I. G., Grant, A. K. Gastrie ulcer in South Australia, 1954-63. 1. Epidemiological factors, Medical Journal of Australia 2(24) : 1128-1132, Dec. 12, 1970. ALP, M. H., Hisxop, I. G., Grant, A. K, Gastric ulcer in South Australia, 1954-638. 2. Symptomatology and response to treatment. Medical Journal of Australia 1(7) ; 372-374, Feb. 13, 1971. Bynum, T. E.,, Sotomon, T. E., JoHNson, L. R., Jacopson, E, D. Inhibition of pancreatic secretion in muan by cigarette smoking. Gut 13(5): 361-365, May 1972, Cooper, P., ToLrns, 8, H. Relationship between smoking history and compli- cations immediately following surgery for duodenal ulcer. Mount Sinai Journal of Medicine 39(3) > 287-292, May-June 1972. ) FINGERLAND, A., Husak, T., BENbDLova, J, Contribution to the investigation of the effect of cigarette smoking. Sbornik Vedeckych Praci Lekarske Fakulty Karlovy University v Hradci Kralové 14(2) : 221-234, 1971, Hirayama, T. Smoking in relation to the death rates of 265,118 men and women in Japan. A report of 5 years of followup. Presented at the Amer- ican Cancer Society’s 14th Science Writers’ Seminar, Clearwater Beach, Fla., Mar, 27, 1972, 15 pp. Konturek, 8. J., Date, J., Jaconson, EB. D., Jounson, L. R. Mechanisms of nicotine-induced inhibition of pancreatic secretion of bicarbonate in the dog. Gastroenterology 62(3) : 425-429, 1972. Kontrurek, S. J., RavEck1, T., THor, P., DEMBINSKI, A., JAcoBson, E. D. Effects of nicotine on gastric secretion and ulcer formation in cats. Pro- ceedings of the Society for Experimental Biology and Medicine 138 (2): 674-671, November 1971. Konturek, 8. J., SoLomon, T. E., McCreteut, W. G., JoHNSoN, L. R,, Jacosson, E. D. Etiects of nicotine on gastrointestinal secretions. Gastro- enterology 60(6) : 1098-1105, June 1971. Mopars, A., Sitva, S., ALCALDE, J., WaISSBLUTH, J., Ramos, J., Bey, H,, Sanz, R. Cigarrillo y secrecion gustrica. I. Analisis de la secreci6n gdstrica en pacientes digestivos fumadores y no fumadores. (Cigarettes and gastric secretion, I. Analysis of gastric seeretion in smoking and non- smoking ulcer patients.) Revista Medica de Chile 99(4}) > 271-274, April 1971. Moraes, A,, SILVA, S., Osorio, G., ALCALDE, J., WalsssLutu, J. Cigarrillo y secrecion gastrica. II. Efecto del cigarrillo sobre la secrecién gastrica. (Cigarettes and gastric secretion, IL. Effect of cigarettes on gastric secre- tion.) Revista Medica de Chile 99 (4) : 275-279, April 1971. 495-025 O-—73 12 163 (72) Rogert, A. Potentiation, by nicotine, of duodenal ulcers in the rat. Pro. ceedings of the Society for Experimental Biology and Medicine 139(1): 319-322, January 1972. (13) Rosert, A., Stowr, D. F., NeEzaMis, J. E. Possible relationship between smoking and peptic ulcer. Nature 238 (5320) : 497-498, Oct. 15, 1971. (14) Suarixu, M. 1. Tuompson, J. H., Aures, D. Acute and chronic effects of nicotine on rat gastric secretion. Proceedings of the Western Pharma- cology Society 13: 178-184, 1970. (15) SoLomon, T. E., Jacopson, E. D. Cigarette smoking and duodenal-uleer dis. ease. New England Journal of Medicine 286 (22) : 1212-1213, June 1, 1972, (16) Tiiompson, J. H., Grorce, R., ANGULO, M. Some effects of nicotine on gastric secretion in rats. Proceedings of the Western Pharmacology Society 14: 173-177, 1971. (77) U.S. Pusric Hearty Service, The Health Consequences of Smoking. A Re port of the Surgeon General : 1971. U.S. Department of Health, Education, and Welfare. Washington, DHEW Publication No. (HSM) 71-7513, 1971, 458 pp. (178) U.S. Pustic HEaLrH SERVICE, The Health Consequences of Smoking. A Report of the Surgeon General : 1972. U.S, Department of Health, Eduea- tion, and Welfare. Washington, DHEW Publication No. (HSM) 72-6516, 1972, 158 pp. 164 CHAPTER 6 Pipes and Cigars Contents Introduction.....--------- The Prevalence of Pipe, Cigar, and Cigarette Usage__.._..___ The Definition and Processing of Cigars, Cigarettes, and Pipe Tobaccos..-_--.-- 2-2 Chemical Analysis of Cigar Smoke Mortality Overall Mortality_.-....... 22222 Mortality and Dose-Response Relationships Amount Smoked Inhalation._.________-_ ee Specific Causes of Mortality Caneer__.___-. 2. Cancer of the Lip Oral Cancer_____-.___ eee Cancer of the Larynx.____._-___._._____________.. Cancer of the Esophagus Lung Caneer_.__.__- 2-2 Tumorigenic Activity Experimental Studies Chronic Obstructive Pulmonary Disease (COPD) Gastrointestinal Disorders Little Cigars Conclusions References List of Figures Figure 1.—Inhalation among pipe smokers by age___________ Figure 2.—Inhalation among cigar smokers by age—Ham- mond... -- 222-28 Figure 3.—Depth of inhalation among cigarette smokers by age—Hammond_.____-___._-_-2.--.-__ Figure 4.—Percent distribution of 130 brands of cigarettes and 25 brands of little cigars by tar content Page 171 173 175 177 179 180 183 189 189 190 191 193 197 203 210 210 215 216 222 222 229 230 Figure 5.—-Percent distribution of 130 brands of cigarettes and 25 brands of little cigars by nicotine content List of Tables Table 1.—Percent distribution of U.S. males aged 21 and older by type of tobacco used for the years 1964, 1966, and 1970__ Table 2.—Percent distribution of U.S. males by type of tobacco used and age for 1970___2.--2 02 Table 3.—Percent distribution of British males aged 25 and older by type of tobacco used for the years 1965, 1968, and 1971 - Table 4.—Amounts of several components of 1 gram of par- ticulate material from mainstream smoke of tobacco prod- ucts. Table 5.—-A comparison of several chemical compounds found in the mainstream smoke of cigars, pipes, and cigarettes____ Table 6.—-Mortality ratios for total deaths by type of smoking (males only) -. 2-22-22. Table 7.—Mortality ratios for total deaths of cigar and pipe smokers by amount smoked—Hammond and Horn________ Table 8.—Mortality ratios for total deaths of cigar and pipe smokers by amount smoked—Best.......__...-6--. Table 9.—Mortality ratios for total deaths of cigar and pipe smokers by age and amount smoked—Kahn_____________. Table 10.—Mortality ratios for total deaths of cigar and pipe smokers by amount smoked—Hammond______.___._____. Table 11.—The extent of inhaling pipes, cigars, and cigarettes by British males aged 16 and over in 1968 and 1971______. Table 12.—Inhalation among cigar, pipe, and cigarette smokers by age—Doll and Hill.---2222-2 - Table 13.—Mortality ratios for total deaths of cigar and pipe smokers by age and inhalation—Hammond___...__._____. Table 14.—Percentage of British male cigar smokers who re- ported inhaling a lot or a fair amount by type of product smoked. _- 22-8 Table 15.—Percentage of individuals reporting inhalation of “almost every puff’ of tobacco smoke by current and pre- vious tobacco usage and tvpe of tobacco used___. Table 16.— Percentage of British males who reported inhaling a lot or fair amount of cigar smoke by current and previous tobacco usage and type of tobacco previously smoked (1968) - 168 Page 226 173 174 174 178 180 18} 181 188 188 Table 17.—Extent of reported inhalation of cigar smoke by British male cigar smokers who were ex-cigarette smokers in 1968, analyzed by extent of reported inhalation of cigarette smoke when previously smoking cigarettes__......._._____. Table 18.—Mortality ratios for total cancer deaths in cigar and pipe smokers. A summary of prospective epidemiological studies. - 22 Table 19.—Relative risk of lip cancer for men, comparing cigar, pipe, and cigarette smokers with nonsmokers. A summary of retrospective studies.__.___---2 2 Table 20.—Mortality ratios for oral cancer in cigar and pipe smokers. A summary of prospective epidemiological studies __ Table 21.—Relative risk of oral cancer for men, comparing cigar, pipe, and cigarette smokers with nonsmokers. A sum- mary of retrospective studies.__.____.________._...____. Table 22.—Mortality ratios for cancer of the larynx in cigar and pipe smokers. A summary of prospective epidemiological studies... Table 23.—Relative risk of cancer of the larynx for men, com- paring cigar, pipe, and cigarette smokers with nonsmokers. A summary of retrospective studies....._____.....______. Table 24.—Mortality ratios for cancer of the esophagus in cigar and pipe smokers. A summary of prospective epidemio- logical studies.--__._--_. 2-28 Table 25.—Relative risk of cancer of the esophagus for men, comparing cigar, pipe, and cigarette smokers with non- smokers. A summary of retrospective studies......___..___. Table 26.—Mortality ratios for lung cancer deaths in male cigar and pipe smokers. A summary of prospective studies____ Table 27.—Lung cancer death rates for cigar and pipe smokers by amount smoked—Doll and Hill___........... Table 28.—Iung cancer mortality ratios for cigar and pipe smokers by amount smoked—Kahn._______...__..._____. Table 29.—Relative risk of lung cancer for men, comparing cigar, pipe, and cigarette smokers with nonsmokers. A summary of retrospective studies....___.___....__._____. Table 30.—Changes in bronchial epithelium of male cigar, pipe, and cigarette smokers as compared to nonsmokers____ Table 31.—Tumorigenic activity of cigar, pipe, and cigarette smoke condensates in skin painting experiments on animals__ Table 32.—Mortality ratios for cardiovascular deaths in male cigar and pipe smokers. A summary of prospective epi- demiological studies Page 189 189 192 193 194 196 198 200 201 204 204 205 206 209 213 216 169 Table 33.—Mortality ratios for chronic obstructive pulmonary deaths in male cigar and pipe smokers. A summary of pros- pective epidemiological studies......-.2.------ Table 34.—Prevalence of respiratory symptoms and illness by type of smoking_..---- 222 Table 35.—Pulmonary function values for cigar and pipe smokers as compared to nonsmokers..._..________._.____ Table 36.—Mortality ratios for peptic ulcer disease in male cigar and pipe smokers. Summary of prospective studies___ Table 37.—Shipment of small and large cigars destined for domestic consumption (1970, 1971, 1972)__-__. Table 38.—Selected compounds in mainstream smoke________ Table 39.—The pH of the mainstream smoke of selected tobacco products 170 Pag Introduction This chapter is a review of the epidemiological, pathological, and experimental data on the health consequences of smoking cigars and pipes, alone, together, and in various combinations with cigarettes. Previous reviews on the health consequences of smoking have dealt primarily with cigarette smoking. Although some of the material on pipes and cigars presented in this chapter has been presented in previ- ous reports of the Surgeon General, this is the first attempt to summa- nize what is known about the health effects of pipe and cigar smoking. Since the use of pipes and cigars is limited almost exclusively to men inthe United States, only data on men are included in this review, The influence of pipe and cigar smoking on health is determined by examining the overall and specific mortality and morbidity ex- Perienced by users of these forms of tobacco compared to nonsmokers. Epidemilogical evidence Suggests that individuals who limit their smoking to only pipes or cigars have overal] mortality rates that are slightly higher than nonsmokers. For certain specific causes of death, however, Pipe and cigar smokers experience mortality rates that are *S great as or exceed those experienced by cigarette smokers, This analysis becomes more complex when combinations of smoking forms ire examined, The overall mortality rates of those who smoke pipes, ‘gars, or both in combination with cigarettes appear to be inter- mediate between the high mortality rates of cigarette smokers and the lower rates of those who smoke only pipes or cigars. This might Seem to suggest that smoking pipes or cigars in combination with ciga- rettes diminishes the harmful effects of cigarette smoking. However, ‘nanalysis of mortality associated with smoking combinations of ciga- rettes, pipes, and cigars should be standardized for the level of con- sumption of each of the products smoked in terms of the amount ‘moked, duration of smoking, and the depth and degree of inhalation. or example, cigar smokers who also smoke a pack of cigarettes a day might be expected to have mortality rates somewhat higher than those tho smoke only cigarettes at the level of a pack a day, assuming that th groups smoke their cigarettes in the same way. Mixed smokers ho inhale pipe or cigar smoke in a manner similar to the way they ‘noke cigarettes might be expected to have higher mortality rates ‘han mixed smokers who do not inhale their cigars and pipes and also wl 171 resist inhaling their cigarettes, Unfortunately, little of the publisheq material on mixed cigarette, pipe, and cigar smoking contains these types of analyses or controls. A paradox seems to exist between the mortality rates of ex-smokers of pipes and cigars and ex-smokers of cigarettes, Ex-cigarette smokers experience a relative decline in overall and certain specific causes of mortality following cessation. This decline is important but indirect evidence that cigarette smoking is a major cause of the elevated mor. tality rates experienced by current cigarette smokers. In contrast to this finding, several prospective epidemiological investigations, Hammond and Horn (40), Best (9), Kahn (40), and Hammond (38), have reported higher death rates for ex-pipe and ex-cigar smokers than for current pipe and cigar smokers. This phenomenon was ana- lyzed by Hammond and Garfinkel (39). The development of ill health often results in a cigarette smoker giving up the habit, reducing his daily tobacco consumption, switching to pipes or cigars, or choosing a cigarette low in tar and nicotine. In many instances, a smoking. related disease is the cause of ill health, Thus, the group of ex-smokers includes some people who.are ill from smoking-related diseases, and death rates are high among persons in ill health. As a result, ex-cigarette smokers initially have higher overall and specific mortality rates than continuing cigarette smokers, but be- cause of the relative decrease in mortality that occurs in those who quit smoking for reasons other than ill health, and because of the dwindling number of ill ex-smokers, a relative decrease in mortality is observed (within a few years) following cessation of cigarette smoking. The beneficial effects of cessation would be obvious sooner were it not for the high mortality rates of those who quit smoking for reasons of illness. A similar principle operates for ex-pipe and ex- cigar smokers, but because of the lower initial risk of smoking these forms and therefore the smaller margin of benefit following cessation, the effect produced by the ill ex-smokers creates a larger and more persistent impact on the mortality rates than is seen in cigarette smoking. For the above reasons a bias is introduced into the mortality rates of current smokers and ex-smokers of pipes and cigars, so that a more accurate picture of mortality might be obtained by combining the ex-smokers with the current smokers and looking at the resultant mortality experience. Because of a lack of data that would allow a precise analysis of mortality among ex-pipe and ex-cigar smokers, a detailed analysis of these groups could not be undertaken in this review. For each specific cause of death, tables have been prepared which summarize the mortality and relative risk ratios reported in the major 172 prospective and retrospective studies which contained information about pipe and cigar smokers. The smoking categories used include: cigar only, pipe only, total pipe and cigar, cigarette only, and mixed. The total pipe and cigar category includes: those who smoke pipes only, cigars only, and pipes and cigars. The mixed category includes: those who smoke cigarettes and cigars; cigarettes and pipes; and cigarettes, pipes, and cigars. Mortality and relative risk ratios were calculated relative to nonsmokers, The Prevalence of Pipe, Cigar, and Cigarette Usage The prevalence of pipe, cigar, and cigarette smoking in the United States was estimated by the National Clearinghouse for Smoking and Health from population surveys conducted in 1964, 1966, and 1970 (98, 99, 100). In each survey, about 2,500 interviews were conducted ona national probability sample stratified by type of population and geographic area. The use of these products among adults aged 21 and older is summarized in tables 1 and 2. The prevalence of pipe, cigar, and cigarette smoking in Great Britain for the years 1965, 1968, and 1971 is presented in table 3. Taste 1.—Percent distribution of U.S. male smokers aged 21 and older by type of tobacco used for the years 1964, 1966, and 1970 Forms used 1964 1966 1970 (percent) (percent) (percent) 1. Cigar only... 2-22-28 6.8 5.5 5. 6 2, Pipe only____.-_.______.. 1.7 3.0 3.6 3. Pipe and cigar__..._...-__...... 3.9 4.9 4.4 4. Cigarette only.....-.__-______.___...___. 28. 6 31.2 25. 9 5. Cigarette and cigar_...____....._._______ 11.3 9.9 6. 6 6. Cigarette and pipe.._...._._.__.....___. 5.3 4.9 5.3 7. Cigarette, pipe, and cigar_____--_-2_2 7.7 6. 3 4.6 8 Nonsmoker_....._._.._._.__........_._. 34. 7 34. 3 44.0 Total__. 22 100. 0 100. 0 100. 0 Number of persons in sample_____._________ 2, 389 2, 679 2, 861 Total pipe users (2+3+6+7)............._- 18.7 19.2 17.9 Total cigar users (1+34+5+7)_........_.._ 29. 9 26.7 21.2 Total cigarette users (4454647)... 52.9 52.4 42. 3 Source: U.S. Department of Health, Education, and Welfare (98, 99, 100). 173 TABLE 2.—Percent distribution of U.S. male s co used and age for 1970 Forms used Age groups 21 to 34 35 to 44 45 to 54 mokers by type of tobae. 55 to 64 65 to 75 1, Cigar only___._.._2 888. 3.7 6.5 4,7 6.7 93 2. Pipe only____-----2 4.3 3.5 3.0 3. 2 3.6 3. Pipe and cigar_._.._._.___. 3.8 3.3 5. 2 4. 4 6.9 4. Cigarette only__.__. 28. 8 29. 0 27. 1 24, 3 13.6 5. Cigarette and cigar__________ 6. 8 10. 4 5. 5 5. 2 49 6. Cigarette and pipe..________ 6.6 4.4 5.6 4.0 3 § 7. Cigarette, pipe, and cigar. ___ 5. 8 4.8 5.0 4.0 Ls & Nonsmoker_.__-._-.--- 40, 2 38. 1 43.9 48. 2 37,9 ee Total. 2-22 100.0 100.0 100.0 100.0 100. 9 ns Number of persons in sample___ 1, 009 528 523 405 388 Total pipe users...._--_________ 20. 5 16. 0 18. 8 15. 6 15.7 Total cigar users_.____________ 20. 1 25. 0 20. 4 20. 3 21.6 Total cigarette users__________ 48.1 48. 6 43. 3 37.5 23. 6 Souree: U.S. Department of Health, Education, and Welfare (100). TaBLe 3.—-Percent distribution of British male smokers aged 25 ani) older by type of tobacco used for the years 1965, 1968, and 1971 Forms used 1965 1968 1971 1. Cigars only._...222 2-8 1.9 2.8 3.3 2. Pipe only_.-__-.2 8 5. 1 5. 6 5.9 3. Cigarettes only._--_22 288 46. 8 45. 7 40.8 4. Cigarettes and pipe._.....---.. 8.0 7.0 6.1 5. Mixed smokers... 7.5 9. 1 & 4 6. Nonsmokers.-_.-_..--_. 68. 30. 7 29.9 35. 4 a Total. 22. 100. 0 100. 0 100. 0 Number of persons in sample.___.__..______ 3, 576 3, 566 3, 594 Total pipe users... 13.9 14.3 13.3 Total cigar-_____=8 ee 9. 0 11.7 11.3 Total cigarette._.-.---- 67.6 67.6 61. 6 Source: Todd, G. F. (94). 174 The Definition and Processing of Cigars, Cigarettes, and Pipe Tobaccos Cigareties The U.S. Government has defined tobacco products for tax pur- poses. Cigarettes are defined as “( 1) Any roll of tobacco wrapped in paper or in any substance not containing tobacco, and (2) any roll of tobacco wrapped in any substance containing tobacco which, because of its appearance, the type of tobacco used in the filler, or its packaging ind labeling, is likely to be offered to, or purchased by, consumers as a cigarette described in subparagraph (1).” Cigarettes are further classified by size, but virtually all cigarettes sold in the United States are “small cigarettes” which by definition weigh “not more than 3 pounds per thousand” which is not more than 1.361 grams per cigarette (96). American brands of cigarettes contain blends of different grades of Virginia, Burley, Maryland, and oriental tobaccos. Several varieties of cigarette tobaccos are flue-cured. In this process, tobacco leaves are ‘ured in closed barns where the temperature is progressively raised over a period of several days. This results in “color setting,” fixing, ind drying of the leaf. The most conspicuous change is the conversion of starch into simpler sugars and suppression of oxidative reactions. Flue-cured tobaccos produce an acidic smoke of light aroma (35, 112). Cigars Cigars have been defined for tax purposes as: “Any roll of tobacco ‘rapped in leaf tobacco or in any substance containing tobacco (other than any roll of tobacco which is a cigarette within the meaning of ‘ubparagraph (2) of the definition for cigarette)” (772). In order to clarify the meaning of “substance containing tobacco” the Treasury lepartment has stated that, “The wrapper must (1) contain a signifi- “ant proportion of natural tobacco; (2) be within the range of colors Normally found in natural leaf tobacco; (3) have some of the other tharacteristics of the tobaccos from which produced; e.g., nicotine “ontent, pH, taste. and aroma: and (4) not be so changed in the constitution process that it loses all the tobacco characteristics” 2). Further, “To be a cigar, the filler must be substantially of ‘obaccos unlike those in ordinary cigarettes and must not have any added flavoring which would cause the product to have the taste or ‘toma generally attributed to cigarettes. The fact that a product does 75