Sunday, 27 January 2013

Decidualization and menstruation

Karoo rock sengi (Elephantulus pilicaudus) © Galen Rathbun

Menstruation is a distinctive feature of human reproduction shared by apes, monkeys, a few bats – and sengis (also known as elephant-shrews).

How is this relevant to the evolving placenta? In addition to trophoblast and other fetal tissues, most placentas incorporate part of the endometrium (the non-muscular part of the uterine wall). It comprises connective tissue, glands and blood vessels. In preparation for pregnancy, the endometrium undergoes a process called decidualization. This involves a change in the size, shape and properties of the connective tissue cells. It is a necessary prerequisite for implantation of the blastocyst – an early stage in embryonic development.

In most mammals, decidualization does not occur until there is an embryonic signal. So there is a good chance the decidua will come in useful and help build a placenta. But in humans decidualization is spontaneous – in response to a maternal signal in the second half of the menstrual cycle. The decidua will be useful if there is a pregnancy, but otherwise must be shed by menstruation.

Why do women menstruate when mice do not? A recent paper by Emera, Romero & Wagner suggests there is an advantage to spontaneous decidualization. They believe early development of the decidua makes it easier for the mother to detect and reject defective embryos. Therefore natural selection has favoured the evolution of spontaneous decidualization. Menstruation is of no value in itself – just an inevitable corollary. The original paper develops the idea in much greater detail.

It makes even better sense if we consider the natural history of our species. Roger Short once took a careful look at the anthropological data, including studies of the Kung hunter-gatherers of the Kalahari. The first thing he noted was that women married young but it took a couple of years before they became pregnant. They started married life with infertile cycles. Teenage pregnancy was not a problem in primitive societies because it was a biological impossibility. At the other end of the scale, life span was briefer than in more advanced societies. Short reckoned a woman would manage about five pregnancies. Importantly, each of these children would be nursed for about three years. There are no fertile cycles during pregnancy and lactation and thus no menstruation. In total Short reckoned a female hunter-gatherer would experience 15 years of lactational amennorhoea, four years of pregnancy and just four years of menstruation. So menstruation was not such a big price to pay for the perceived advantages of spontaneous decidualization.

And sengis? They make a decidua with a cosy little chamber to welcome the embryo. If no pregnancy occurs, it is shed by a process akin to menstruation. The South African embryologist C. J. van der Horst documented this in the 1940’s. His work has been largely forgotten so it was nice to see it cited by Emera and colleagues.


  1. I am currently preparing a talk on the evolutionary factors involved in the development of the human species and one can see the evolutionary advantages of the menopause in a social species where there is a long childhood period. There is clearly a point where it is more beneficial (in terms of passing on ones genes) to act as a grandmother helping to being up grandchildren that having a child when you are likely to die before it reaches maturity.

    While I am not particularly concerned with the detailed mechanism – this discussion of mensuration introduces a factor which I had not thought of. If you have one child at a time, and, like the Kung hunter-gatherers of the Kalahari, only have an average of 5 children, you need to rear 40% of your children to maturity, and if you can squeeze in an additional pregnancy so much the better. If the menopause mechanism means that defective embryos can be aborted more rapidly it increases the chances of an extra child during the fertile period.

    I am unaware of the relevant literature in bats – but having worked with Greater and Lesser Horseshoe bats in Devon some 50 years ago I am aware that they normally only have one young a year in the summer – all at about the same time – so they don't want to miss a whole year because of a defective embryo. It would make evolutionary sense for them to start pregnancies in the autumn but delay development during the early part of the hibernation period. A mechanism such as the one you describe would give them a chance to reject the embryo, or “restart” the process if there is no embryo, with the opportunity to have another go during the hibernation period. While I did not see it myself, I know that bats were sometimes found apparently hibernating in the intercourse position. Do you know which species of bats show mensuration, and whether some aspect of their life cycle made this of particular value?

  2. Robert D. Martin of The Field Museum Chicago writes insightfully on human reproduction (e.g. Am J Phys Anthropol 2007;Suppl 45:59-84) and it is worth going back to the paper by R.V. Short (Proc R Soc Lond B Biol Sci 1976;195:3-24. I cannot answer your question about horseshoe bats. Studies of menstruation in bats are limited to a few species such as short-tailed bats (Carollia sp.) and the common vampire bat (Desmodus rotundus).

  3. how long does the elephant shrews menstrual cycle last?

    1. The only paper i can find is Lumpkin et al. J Reprod Fertil 1982; 66:671-3 (Open Access on the Reproduction web site). They obtained a value of 13 days but with considerable variation. The detailed description of the menstrual cycle by van der Horst was based on specimens caught in the wild so an accurate estimate of the length of the cycle was not possible.

  4. Very interesting. Some authors talk about menses in bats that fail to get pregnant during the reproductive station, i.e those bats undergo estral cycles. I think that in a near future menstruation (i mean menses) will be regarded as a phenomenon not neccessarily related to menstrual cycles.

    I find that the study of Emera fails to explain why women have so copious menses. In this sense it is very interesting the aportation of Short, telling that some tribal teens undergo a couple of years of not fertile menstrual cycles. The same, by the way, has been reported by Goodall and other authors in chimps.

    Coming back to the ideas of Emera et al, hemochorial placentas happen in mice and they don't menstruate, while bats have different types of placentas and they menstruate, on the other hand. The hypothesys of the invasive fetus in humans fails to explain these differences, as well as why chimps and bonobos have so much lighter menses, when they have (menses have only been observed in captive chimps).

    All in all, it seems that enviromental factors, food?, has many things to do with copious and hurting menses in women.