Until the 1930s physicians blithely told women that they conceived during menstruation and could avoid pregnancy by restricting coitus to other days. This, of course, conflicts directly with subsequent advice that conception is most likely between menstruations at mid-cycle, when one ovary releases an egg (ovulation). Counting two days for sperm survival and a day for egg viability yields a three-day fertile window. Adding a few days to allow for variable ovulation times, the rest of the cycle is seen as a conception-free safe period. This “egg-timer” model ushered in the rhythm method of birth control, subsequently refined to include basal body temperature and mucus secreted by the neck of the womb (cervix) as indicators of ovulation.
Another myth of human uniqueness
Conception close to mid-cycle is generally accepted as typical for monkeys and apes as well as humans. But copulation outside the fertile window is widely touted as uniquely human, despite decades of research showing that monkeys and apes also commonly mate on supposedly conception-free days. Many ingenious explanations have been proposed for human coitus outside the fertile window. Yet a fundamental problem generally passes unmentioned: Logically, the egg-timer model has “danger zones” at the edges of the fertile window, where sperms or eggs are over-ripe.
Laboratory mammal studies showed ages ago that fertilization with aged sex cells leads to pregnancy loss or fetal deformity. So how could natural selection ever favour copulation outside the fertile window? Other mammals typically mate only during a clearly defined female heat period (oestrus), usually lasting about three days. But higher primates typically lack oestrus and mating is not tightly restricted to a few days.
The largely unrecognized problem of danger zones bordering the supposed fertile window puzzled me for years. Constantly seeking clarification, I found an unexpected source: variation in pregnancy lengths. In 1982 Richard Kiltie reported that variation in monkeys and apes is double that in other mammals. I saw two possible explanations: Either pregnancy lengths in monkeys, apes and humans really do vary more than in other mammals, or greater variation is an artifact arising from how we identify conception time. But mammal pregnancy timing is typically unusually precise, so why would higher primates be an exception? Kiltie did not mention lower primates (lemurs, lorises, tarsiers), so I collected additional data. In 2007, I confirmed that reported pregnancy lengths for all higher primates are unusually variable, whereas in lower primates (which show clear oestrus) variation is low as in other mammals.
Published pregnancy lengths for higher primates are mostly based on copulations. In mammals with a clear oestrus, mating generally coincides closely with conception time. So pregnancy length variation generally reflects real differences in conception-to birth intervals. But monkeys and apes differ greatly: Copulation across the cycle cannot accurately indicate conception. For humans the problem is even greater because coitus can occur on any cycle day without external signs of ovulation. So physicians time pregnancy from the last menstrual period (LMP).
Pregnancy length variation for monkeys and apes should diminish when conception is inferred from ovulation itself. Hormonal monitoring of female cycles in higher primates has shown that calculated pregnancy lengths are indeed less variable. So the doubled variation in pregnancy lengths reported for higher primates results from wrongly equating copulation with conception.
Shattering the fertile window
In the “egg-timer” model both ovulation and coitus resulting in conception occur at mid-cycle. Yet I gradually uncovered a large body of little-cited evidence indicating that single copulations resulting in conception extend across the human cycle. Information, predominantly in German journals, came from brief home visits by soldiers on military leave, legal decisions on rapes and paternity, and gynecological clinic records. Such data extend back at least to 1869 when Johann Ahlfeld reviewed some 200 clinical records. At least 20 similar studies were published subsequently. Eventually, I compiled over 4,000 cases yielding a remarkably smooth curve. Single acts of coitus resulting in conception occur on virtually every cycle day, although the probability of conception is notably higher before mid-cycle. Moreover, the peak is 5-6 days before mid-cycle ovulation.
Fertility scientists have generally ignored those early reports. They are dismissed as circumstantial evidence before hormonal monitoring became readily available, and women’s recollections of dates for menstruation and coitus have been seen as “unreliable”. But a 2000 paper by Allen Wilcox and colleagues using modern methods reached a very similar conclusion. Over 120 women who became pregnant after stopping birth control recorded dates for coitus and menstruation, while ovulation times were determined from urinary hormones. The resulting curve indicated that conception probability peaked on days 12 and 13, but any woman had at least a 10% chance of conceiving on cycle days 6 through 21.
The joker in the pack
Halfway through writing my book How We Do It I realized that my long-standing puzzlement about human conception times disappeared if sperms survive longer than two days. All available evidence indicates that ovulation typically occurs close to mid-cycle, but sperm storage after ejaculation would have major implications for everything. It has long been known that human sperms can survive intact for several days in mucus from the cervix. John Gould and colleagues reported in 1984 that they recovered normally motile sperms from cervical mucus up to 5 days after insemination. Ironically, cervical mucus is a much-discussed guide to ovulation time, but precious little is known about any sperms it contains. Yet it is widely accepted that hordes of sperms are stored in crypts in the cervix from which the mucus flows. Many review papers and gynaecological textbooks mention this in a single sentence.
In 1980 gynecologist Vaclav Insler and colleagues published results of research in which 25 women scheduled for womb removal (hysterectomy) valiantly volunteered to be artificial inseminated beforehand. After removal and serial sectioning of the womb, sperms present in the crypts were examined microscopically. Within two hours after insemination, sperms colonized the entire length of the cervix. Insler and his team calculated numbers and densities for sperm-containing crypts. Up to 200,000 sperms were found in crypts of a single womb. Moreover, it was noted that live sperms have been found in cervical mucus up to day 9 after insemination. The cervix apparently serves as a sperm reservoir for a week or more, gradually releasing viable sperms that can migrate into the oviduct.
Findings from artificial insemination provide supporting evidence for extended survival of human sperm. In a gem of a paper published in 1960, Yutaka Yoshida reported results from single inseminations, taking basal body temperature and cervical mucus condition as indicators of ovulation. Successful inseminations were performed on cycle days 8-22. More importantly, insemination times ranged from 10 days before to 4 days after estimated ovulation time. This alone indicated survival of human sperms for up to 10 days.
Dogs do it too
It seems barely conceivable that no further medical research into sperm storage in the human womb was conducted after that seminal 1980 paper by Insler and colleagues. But this omission seems less surprising once it is realized (as I did only recently) that veterinary science was almost as slow to identify a very similar condition in dogs. An initial account of sperm storage was actually published back in 1967; but the full story emerged only recently. One key pointer is the fact that pregnancy lengths for dogs calculated from dates of copulation or artificial insemination are unusually variable. Variation is more than double the level typical for mammals and comparable to that found with human pregnancy lengths calculated from LMP. Yet when dog pregnancy lengths are calculated using hormonal evidence of ovulation, variation is greatly reduced and well within the normal mammal range. Stored sperms have now been clearly documented in dogs, but the site is different. Whereas storage occurs in cervical crypts in women, in dogs sperms are stored in gland openings in the wall of the womb itself. A 2004 paper by Tom Rijsselaere and colleagues included a striking image showing sperms with heads buried and tails protruding.
Storage of viable sperms in the human womb for 10 days or more has game-changing implications. It effectively eliminates the “danger zone” before ovulation. Moreover, artificial insemination procedures could benefit from appropriate modification, with multiple shots over several days rather than one or two around expected ovulation time. And perhaps natural selection of ancestral higher primates for mating at times well away from ovulation was driven by a need for “sexual priming” by semen to counteract immunological problems with a highly invasive placenta. (See my previous blog post of February 2, 2016: A Biological Function for Oral Sex?)
For illustrations of human sperm storage see the section on the cervical canal in the online course in embryology for medical students developed by the Swiss universities of Fribourg, Lausanne & Bern: http://www.embryology.ch/indexen.html
Concannon, P.W. (2000) Canine pregnancy: Predicting parturition and timing events of gestation. In: Recent Advances in Small Animal Reproduction (eds. Concannon, P. England, W.E. & Verstegen, J.): International Veterinary Information Service.
Doak, R.L., Hall, A. & Dale, H.E. (1967) Longevity of spermatozoa in the reproductive tract of the bitch. Journal of Reproduction & Fertility 13:51-58.
France, J.T. (1981) Overview of the biological aspects of the fertile period. International Journal of Fertility 26:143-152.
Gould, J.E., Overstreet, J.W. & Hanson, F.W. (1984) Assessment of human sperm function after recovery from the female reproductive tract. Biology of Reproduction 31:888-894.
Insler, V., Glezerman, M., Zeidel, L., Bernstein, D. & Misgav, N. (1980) Sperm storage in the human cervix: a quantitative study. Fertility & Sterility 33:288-293.
Karre, I., Meyer-Lindenberg, A., Urhausen, C., Beineke, A., Meinecke, B., Piechotta, M., Beyerbach,M. & Günzel-Apel, A.-R. (2012) Distribution and viability of spermatozoa in the canine female genital tract during post-ovulatory oocyte maturation. Acta Veterinaria Scandinavica 54:49:1-9.
Martin, R.D. (2007) The evolution of human reproduction: A primatological perspective. Yearbook of Physical Anthropology 50:59-84.
Martin, R.D. (2013) How We Do It: The Evolution and Future of Human Reproduction. New York: Basic Books.
Rijsselaere, T., Van Soom, A., Van Cruchten, S., Coryn, M., Görtz, K., Maes, D. & de Kruif, A. (2004) Sperm distribution in the genital tract of the bitch following artificial insemination in relation to the time of ovulation. Reproduction 128:801-811.
Wilcox, A.J., Dunson, D. & Baird, D.D. (2000) The timing of the "fertile window" in the menstrual cycle: day specific estimates from a prospective study. British Medical Journal 321:1259-1262.
Yoshida, Y. (1960) Studies on single insemination with donor's semen. Journal of the Japanese Obstetric and Gynecological Society 7:19-34.