Author's Note: The following originally appeared at Carin Bondar's website (she now co-blogs at Scientific American's PsiVid). It was subsequently selected as a finalist in the Quark Science Writing Prize and appeared in the 2010 edition of  The Open Laboratory: The Best of Science Writing on the Web.

Imagine you are on the Serengeti Plateau and your children are hungry. For miles in every direction there's nothing but dry scrub grass with the occasional flat-topped acacia tree marking the landscape. Your oldest has found a spot to dig for tubers but he and your daughter aren't strong enough to scrape away the hard, baked earth by themselves. Your husband is tracking a wounded gazelle and could be gone for days.

Meanwhile, the infant slung to your hip has started screaming and the distinctive sound triggers a release of oxytocin that causes your breasts to swell and leak. You have to feed her but you can't do that and make sure your other children get enough to eat. There is a very real chance that some of them will be too weak to survive the next time fever breaks out unless you can get help.

You simply can't be everywhere at once. It is a desperate feeling but these are the daily realities among the East African Hadza. If it wasn't for your mother, already kneeled on the ground and using a stick to claw through several layers of tough sediment, it might have been your reality as well. While your baby makes soft cooing sounds as she suckles you can only feel grateful that you were the youngest child in your family, or else your mother might well have had an infant of her own to care for.

This scenario provides the backdrop behind a perplexing question about human evolution: the advent of female reproductive senescence. Between the ages of 45-50 all women undergo physiological changes commonly known as menopause that result in the cessation of ovarian function. Since most women live longer than 50, even in preindustrial and hunter-gatherer societies, this raises a profound evolutionary question: Why would a species "choose" to forego one-third (and sometimes as much as one-half) of their reproductive potential? As I've discussed previously, the leading explanation has become known as the "grandmother hypothesis."

The grandmother hypothesis posits that women who stopped ovulating in their golden years were freed from the costs of reproduction and were better able to invest in their existing children and grandchildren, thus helping to ensure that more individuals with their menopause inducing genes thrived and had children themselves.

To understand why humans are unique in this regards we must first examine what happens in nonhuman animals. All species make tradeoffs between reproductive and somatic investment. An individual animal is considered to have high reproductive fitness if they leave more offspring than others in their population. But if an individual focuses exclusively on reproduction (perhaps because of a mutation that causes a novel behavioral trait) and therefore doesn't invest in their physical health and growth, chances are they won't live long enough to achieve high fitness. Their genes will not be well represented in the next generation and the behavioral trait overemphasizing reproduction will be discarded into the waste-bin of evolutionary history.

While there are a few excellent examples of species adapted for this kind of reproduction-heavy focus (such as the male wasp spiders who offer themselves up as a meal to their hungry mate), the preferred strategy in most species is to invest in both reproductive and somatic interests. To put this in economic terms, it is of little use to keep up on your car payment if that means you will fall behind on your mortgage and end up on the streets.

Chimpanzees are a useful comparison since, along with bonobos, they're our closest evolutionary relatives and we shared a common ancestor with them between 4 and 6 million years ago. As Kristen Hawkes reported in the May 11, 2010 edition of Proceedings of the National Academy of Sciences, chimpanzees undergo reproductive senescence between the years of 45 and 50 just like humans do. However, that is also the extent of their lifespan.

As Figure 1 above highlights, humans have a longer lifespan than chimpanzees but made a different life history tradeoff to spend several nonreproductive decades during their golden years. One possibility to explain this could be that our hominin ancestors simply adapted to have a longer life after our two lineages diverged. However, this doesn't address why human fecundity wasn't also extended as well. Natural selection is a reproductive fitness engine; if humans are the only species that have such a lengthy period of nonreproductive life it suggests there must be another factor involved that compensates for this gap.

This is where the grandmother hypothesis comes in. Multiple studies have suggested that it is the maternal grandmother's assistance that is the most important. This is because, infidelity being what it is, grandmothers are more confident in their genetic relatedness to a daughter's child than to a son's. However, as it turns out, not all grandchildren are created equal and this can have a profound influence on how both maternal and paternal grandmothers offer their assistance.

As reported in the February 10, 2010 edition of Proceedings of the Royal Society, Cambridge biological anthropologist Molly Fox and colleagues show that, because of the way the X-chromosome is inherited by male and female offspring, grandmothers are more closely related to some grandchildren than others.

The X-chromosome contains an estimated 1,529 genes and translates to roughly 8% of the total number of genes that humans have. During reproduction, paternal grandmothers will pass one of their X-chromosomes to her son (the Y-chromosome, of course, being supplied from the paternal grandfather). If her son later has a daughter he will pass on this same X-chromosome because it is the only one he has. This means that 50% of a paternal grandmother's X-chromosomal genes will be represented in her granddaughter (see Figure 2).

However, if she instead has a grandson, none of her X-chromosomal genes will be passed on because her son will only pass on the Y-chromosome. Paternal grandmothers will therefore share more genes overall with granddaughters than with grandsons (the authors calculate an overall genetic relatedness of 31% with granddaughters but only 23% with grandsons).

Maternal grandmothers have a somewhat different genetic relationship with their grandchildren. They will also pass along one X-chromosome to their daughter but both granddaughters and grandsons will have a 50:50 chance that this same chromosome will be passed to them in turn. This means that maternal grandmothers share about 25% of their overall genes with grandsons and granddaughters equally. While this genetic variability may not seem very significant in the abstract, it could have very real implications for child survival.

If the grandmother hypothesis is correct we should expect to find that those children who receive grandmother assistance will have survival rates consistent with this genetic variability. The authors therefore make the bold prediction that boys will survive better when receiving investment from maternal grandmothers while girls will survive better with investment from paternal grandmothers.

To test this prediction the authors used data on child survival rates in the presence of grandmothers for seven populations of rural farmers separated in both geography and time: Japan (1671-1871), Germany (1720-1874), England (1770-1790), Ethiopia (1998-2003), Gambia (1950-1975), Malawi (1994-1997), and Canada (1680-1750). All of these populations lived without the benefits of modern scientific medical care and in an environment where a grandmother's assistance could play an important role in child survival.

The results of this study matched the predictions beautifully. In all seven populations grandsons survived better in the presence of maternal rather than paternal grandmothers (p = 0.0081) while granddaughters survived better in six of the seven populations when they received assistance from paternal instead of maternal grandmothers (p = 0.0046). Culture certainly plays a role in how grandmothers interact with their grandchildren but, because these results were found in such diverse populations, it makes a strong argument that fitness benefits were the ultimate cause of these differences in grandmother investment. It also provides further support for the hypothesis that menopause is an evolutionary adaptation that allowed women to pass on more of their genes by helping their children have greater fitness instead of reproducing themselves.

It is important to point out, however, that this does not mean that all maternal grandmothers will prefer grandsons to granddaughters (or the opposite on the paternal side). There is never a one-to-one correspondence between statistical probabilities and everyday experience. However, on average, the general trend for Homo sapiens is to follow the same evolutionary forces that exist for any other species. Those traits that allowed more of an individual's genes to be passed on to the next generation were selected for and those that did not were ultimately discarded. The evidence has been steadily advancing that menopause is not simply the result of women living longer than their chimpanzee counterparts but that it was an evolutionarily successful strategy that allowed more members of our species to survive and reproduce.

When living on the African savanna a grandmother's assistance can literally be a matter of life or death. For several million years our ancestors lived an experience very much like the Hadza today and the struggles they faced have left their mark inscribed on our bodies. Those who are about to experience the hot flashes and emotional ups and downs that come with the physiological changes of menopause may take solace in the fact that their present discomfort is the very thing that helped our distant relatives survive.

If it wasn't for the cessation of reproductive function our ancestors would have likely seen many more of their children and grandchildren face a bitter end. Thanks to the assistance of grandmothers our species has thrived to the point where many of us now no longer need their help. We owe them a debt of gratitude and, at the very least, periodic phone calls to thank them for everything they have done for us (even if they're not really sure why).


Hawkes, K. (2010). How grandmother effects plus individual variation in frailty shape fertility and mortality: Guidance from human-chimpanzee comparisons, Proceedings of the National Academy of Sciences, 107 (Supp. 2), 8977-8984. DOI: 10.1073/pnas.0914627107

Fox, M., Sear, R., Beise, J., Ragsdale, G., Voland, E., & Knapp, L. (2010). Grandma plays favourites: X-chromosome relatedness and sex-specific childhood mortality, Proceedings of the Royal Society B: Biological Sciences, 277 (1681), 567-573. DOI: 10.1098/rspb.2009.1660