1124
Correspondence
ing the point is defective. They examined thts numbers of births on each day of a period about 9 months after the event in question. In regard to the blackout, Borst suggested that there was a peak of births after an interval of 270 days; and in regard to Pearl Harbor, they noted an excess of births on Day 268 and a deficit on Days 269 and 270. They write, “The peak of 268 days is assumed to represent an unusual number of conceptions on December 7.” This seems to me an unwarranted inference. On the basis of 15,861 observations in Birmingham (England), Gibson and McKeown” concluded that the distribution of the interval between last menstrual period and live-birth confinement is unimodal and almost symmetrical, with a mean of 270 days and a variance of 225. Part of this variation would be due to variance of the interval from last menstrual period to conception. Potter, Burch, and Matsumoto4 summarize data from a number of studies which suggest that the variance of this preovulatory phase is about 16. It seems likely, therefore, that the interval from conception to confinement has a variance of about 200. So if there were an unusually large number of conceptions on December 7, one would expect an unusually large number of confinements over a whole range of days !I months later, not just on a single day. To judge from Gibson and McKeown’s data, about 75 per cent of the live-birth conceptions occurring on December 7 would result in live births over an interval of 28 days during the following summer. The remaining 25 per cent would be born outside this interval altogether. It, therefore, seems appropriate, when searching for associations between current events and subsequent natalit)-, to consider births during intervals not of one day but of one month (though not. of course, necessarily of one calendar month). William Department of Human Genetics University College London 4 Stephenson Way I.ondon N.W. 1, England
H. James,
Reply 7‘0 thz
to
Dr.
James
Editors
Dr. James underscores the intercstiq and cmexpected features of our statistical studies cli natality. Treloar, Behn, and Cowan’ confirm the data of Gibson and McKeown2 as measured from the last menstrual period. In contrast, houever, Clark and ZarrowJ have recently suggested coitus-induced ovulation. The blackout study of New York City is conTidered to be sound because a control population was available. The study of Pearl Harbor has been limited to published birth records in Buffalo, New York, and is therefore statistically less significant. The availability of New York City birth records would substantially improve the statistics. Efforts are now underway to find a dip in birth rate associated with a social event likely to reduce coitus temporarily. A negative effect has been identified in New York City data after the Cuban Missile Crisis. The Presidential announcements came late in the evening (on the East Coast). The effect is not large and must be checked by careful statistical analysis. Without question, births increase a few weeks before and after the optimum gestation period. Tf there were a 50 per cent increase in fertilization for one day, the increased birth rate spread over a month would be 50/30 = 1.6 per cent increase-unidentifiable in the statistical fluctuations. The existence of a one-day peak was unexpected but appears to be statistically sound. It is not possible with present data to tell whether the 30 per cent increase in birth rate for one day represents a small or large fraction of the fertilizations occurring on the day of the social perturbation. If small, then the principal number of births would be distributed in the conventional manner and the one-day peak would not be identified in published statistical rtudies based upon a few hundred cases.
Ph.D.
and Biometry
I,. B. Borst Professor of Physics and Astronomy State University of New York at Butialo Buffalo, New York 14214
REFERENCES
Borst, L. B.: AM. J. OBSTET. GYNECOL. 101: 422, 1968. 2. Trail, S.. and Borst, L. B.: AM. J. OBSTET. GYNECOL. 109: 1086, 1971. 3. Gibson, J. R., and McKeown, T.: Br. J. Sot. Med. 4: 221, 1950. 4. Potter, R. G., Burch, T. K., and Matsumoto. S.: Int. J. Fertil. 12: 127, 1967. 1.
REFERENCES
1. Treloar, A. E.: Behn, B. G., and Cowan, D. W.: .4~. J. OBSTET. GYNECOL. 9% 40. 1967. 2. Gibson, J. R., and McKeown, T.: Br. J. Sot. Med. 4: 221, 1950. 3. Clark, J. H., and Zarrow, M. X.: AM. J. OBSTET. GYNECOL. 10% 1083, 1971.