December 26, 1967

Dr. Per Dalén

Psychiatric Research Centre
Géteborg University

St. Jorgen Hospital
Lillhagen, Sweden

Dear Dr. Dalen:

I am happy to have the comments of your letter of October 18, and the subse-
quent reprints relating to seasonality of birth and schizophrenia.

There are extensive statistics connecting season of birth with maternal age
in the vital statistics of the United States. In the data I have reviewed G
I do not see any drastic shift in the main trends with age, but the intensity fe
of the seasonal fluctuation seems to diminish with births to older women.
Unfortunately, we have no analyses with a sufficient data base in which
variables like maternal age, parity, socio-economic status, and ethhéc and
geographic diversity have been independently sorted out. So I am afraid that =~
maternal age is going to be confounded with many other variables. :

I have in hand a report written by F. H. Godley of the Division of Vital
Statistics, National Center for Health Statistics, Public Health Service,of
the United States on "Fertility and Educational Attainment - Puerto Rico,
1962". These data show that the secular trend tawards a peak incidence of
births in September has been much more rapid for hicher educational classes.
You may be also acquainted with the work of Dr. Ursula Cowgill of Yale Univer-
sity on similar trends in Puerto Rico. (She has published in the Proceedings
of the National Academy of Sciences, United States, November, 1964) Again
there are many variables confounded here.

I was astonished at the congruence of the curves for total live births in
Figure 2 of the materials you sent me with that for rural live births in
Figure 4. Is this correct, or a mislabeling? May I also ask whether the
total population of Sweden for the appropriate interval has a proportion of
rural births corresponding to the 10,791/16,238 among your schizophrenics?

There must be important secular changes with time in the seasonality of births
in Sweden as there have been in many other countries. In the presentation of
the material it would be helpful if a curve could be given for adjustment for
age specificity of the live birth distribution in the control population.

That is to say, each year's births should be weighted by the representation
of that cohort among your schizophrenics. Otherwise the discrepancy between
seasonality of births among your schizophrenic patients may simply reflect

the age distribution of your patient populations.
Dr. Per Dalen December 26, 1967
Page 2

We do not at the present time have any papers in preparation in this field,
but will be glad to see that you receive copies of anything we do publish.

Sincerely yours,

Joshua Lederberg
Professor of Genetics

LAL DAMA H ¢ | Sky SL

j -
dtpurrlele.