The biology of grandmothers

Evolutionary biologist Nichola Raihani introduces an exclusive extract from her book The Social Instinct, exploring why women experience a physiological menopause where fertility ceases long before death.

                                                                                                                                                     Extract from The Social Instinct by Nichola Raihani 

Since having children of my own, I have become aware of slipping a few places down the hierarchy of people that my own parents want to hang out with. Like a smash hit from a new pop star, the grandchildren have entered the charts at number one. But as rewarding as grandparenting may be, their existence is an evolutionary conundrum: why does human female reproduction cease so long before death?

Hardly any other species on Earth has such a prolonged post-reproductive lifespan. In most species – including all of our primate cousins – individuals continue breeding (or try to) until they drop dead. Humans do things differently. Unlike any of the other great-ape species, we have almost no overlap in breeding careers between mothers and daughters. Instead, the period when daughters become reproductively active coincides with the time when their mothers undergo a major physiological transition: menopause. Though we sometimes bemoan this period of life as signifying the start of old age, perhaps feeling as though we are becoming decrepit and defunct, I want to offer an alternative perspective. Menopause is an important switch-point in a woman’s life that serves a specific purpose: this is when we change reproductive lanes, going from being breeders to being helpers.

The menopause is not simply an artefact of longer lifespans due to recent improvements in health care and lifestyles. Across most – if not all – societies, menopause occurs at around age fifty and is accompanied by an extended post-reproductive lifespan, even among people who lack access to technology or modern medicine. This includes contemporary hunter-gatherers and even historically high-mortality populations, such as plantation slaves living in Trinidad in the eighteenth century. Age of menopause is also heritable and, as more and more women living in industrialised societies delay having children until later in life, menopause also seems to be getting later.

Taking a closer look at the underlying physiology shows us that menopause is not just part of the normal ageing process. Human females are born with anywhere from 35,000 to 2.5 million follicles in situ, each capable of producing an egg. This supply declines steadily throughout life: by the time she is twenty, the average female has around 100,000 follicles remaining, and around 50,000 by the time she is thirty-five. Extrapolating forwards, even with this rate of decline, the average woman should be able to continue reproducing until she is well into her sixties and maybe even seventies. But something strange happens when she is around thirty-eight years old. Now, the number of follicles nose-dives, going into a much steeper rate of decline. As a consequence, by the time she’s fifty or thereabouts, her follicle levels drop below the minimum threshold required for regular menstrual cycles.

This highlights the mechanics of menopause. But it doesn’t answer the question of why. Why do women experience this sharp, non-linear decrease in our fertility in our late thirties? And from a purely biological point of view, why do we then persist as sterile vessels when we have become reproductive dead ends?

Answering these questions requires an evolutionary perspective. Through this lens, we see that menopause is the outcome of an evolutionary battle, played out over millennia, between grandmothers and their daughters-in-law. Poor old mothers-in-law are the butt of a thousand schoolboy jokes but, as the saying goes, a joke is the truth wrapped in a smile.

The science isn’t totally resolved on this, but there are good reasons to believe that dispersal in ancestral humans was female-biased. In other words, reproductive-age females tended to move to live with their reproductive partner and his family, rather than the other way around. An important consequence of female-biased dispersal is that the younger females (the wives) are potentially competing with their mothers-in-law over the limited resources needed to successfully raise children.

We can make use of historical data sets of pre-industrial humans to get a feel for the effects of this competition. In Finland, the Lutheran church kept meticulous records of marriages, births and deaths from the 1700s until the early 1950s. Originally kept for tax purposes, these records now have a far greater value in helping us understand how selection might have acted upon historical human populations, in a time before inventions like the contraceptive pill and modern medicine could complicate estimates of fitness too much. This data shows that when a grandmother had children alongside her daughter-in-law, all of the children suffered. The costs were heavy: children were less than half as likely to survive to the age of fifteen when there was competition between breeding females.

Nevertheless, co-breeding was also exceedingly uncommon, with just thirty or so grandmothers (out of more than 500) being reproductively active at the same time as their daughters-in-law. In most cases, we see a case of what looks like altruism: the older females concede to the younger ones in these reproductive battles. But how might grandmothers possibly benefit from curtailing their own reproduction and allowing younger females to breed unhindered?

This puzzle can be solved by considering the ways in which the younger and older females are related to one another’s offspring. The mother-in-law has a vested genetic interest in any children produced by her son’s wife so long as they are definitely his children (Probably quite a safe bet in the Finnish data set, where monogamous marriage was strictly enforced and adultery was severely punished. On average, the median estimate for misattributed paternity worldwide is low, estimated to be about 1–2% – the Finnish data set is likely to be comparable.).

The wife, on the other hand, has no genetic interest whatsoever in any offspring produced by her mother-in-law. This is what’s known as a relatedness asymmetry – and it weakens the mother-in-law’s hand. A grandmother is disincentivised to breed, if doing so harms her grandchildren. The reverse is not true: the younger female’s genes simply don’t care about any costs they might impose on those residing in the mother-in-law and her children. Due to this relatedness asymmetry, the grandmother is more likely to concede in any battle over reproduction; her pay-off coming, instead, in the form of grandchildren. Once she is committed physiologically to sterility, she can make the best of it by helping to raise her grandchildren.

The benefits that grandmothers confer are well documented and can provide the selective impetus needed to favour the increased post-reproductive lifespan. From the ashes of an evolutionary conflict, grandmothers rise up.

When all we have to go on are records of births, deaths and marriages, it is very difficult to infer how, exactly, grandmothers helped their grandchildren to survive. It is likely that these ancient grandmothers acted as repositories of knowledge, passing vital information on everything from breastfeeding to dealing with infants’ illnesses. In some cultures, grandmothers even breastfeed their grandchildren, and are able to produce milk for the child even when their own breeding attempts ended many years earlier. Grandmothers are also an extra pair of hands, someone who is available to help care for any dependent children, which allows mothers to undertake other jobs (foraging and paid labour, for example) that contribute to infant survival.

Another broad pattern emerging from these data sets is that not all grandmothers are equal: data from twenty-six historical and contemporary natural fertility populations has shown that that maternal, rather than paternal, grandmothers make the most difference to the survival of their grandchildren. This is a bit confusing – we know women tended to have their babies in their husband’s household, which might seem to imply that it would be the paternal grandmothers who would be doing the heavy lifting when it came to the childcare. The answer to this evolutionary riddle comes from yet another church database – Canadian this time – documenting the lives of French settlers in and around Quebec in the seventeenth and eighteenth centuries. The data shows that, even when daughters leave home to have their babies, the maternal grandmothers are still able to help out – so long as the daughters don’t move too far away. Increased distance between mothers and their daughters corresponded with decreased survivorship of the daughter’s offspring, likely because the maternal grandmother was less able to help out from afar. The general pattern therefore seems to be that conflict among mothers-in-law and daughters-in-law explains the evolution of menopause, but post-reproductive females direct their investments to the grandchildren they are most certain that they are related to: to their daughter’s rather than to their son’s children.

If women’s post-reproductive lifespan has been extended by the benefits they could bestow upon their grandchildren, we might ask why grandmothers don’t live for even longer. In fact, why do they have to die at all? Before answering this question, it is important to dispel the seemingly intuitive explanation that people die because they become old and decrepit. Senescence – the process of ageing – is not just a biological inevitability. Instead, it is something that is under the control of natural selection. If there were sufficient fitness advantage to living a bit longer and not being riddled with the ailments of old age, then we probably would have longer healthy lifespans. Ageing is what happens when evolution no longer sees a future for us, and selection becomes less assiduous in maintaining and overseeing basic physiological processes, like cell division. There’s no point proofreading a document that no one is going to read.

So why don’t grandmothers live forever? A recent analysis of the same Finnish data set indicates that grandmothers are only useful (in an evolutionary sense) for the child’s first few years of life. Most of the children a grandmother can expect to help with will have been born by the time she is about seventy-five. Beyond this point, not only are grandmothers unhelpful for child survival, but they actually become a liability: living with one means that any child is less likely to survive into adulthood. This detrimental effect acts as a counterweight to the selective force favouring increased lifespan: in the end, grandmothers are no longer selected to live, but to die.

At this point, you might be wondering about men. There is some evidence that men experience an age-related decline in testosterone and are less likely to successfully attract female mating partners in old age – but men don’t undergo a distinct menopause in the same binary fashion as women. So why do grandfathers get to keep their reproductive options open when grandmothers don’t? What role do grandfathers play in the survival of their grandchildren? This is a burgeoning area of research and one where there is little real consensus, yet. A general pattern is that the presence of a grandfather is, on average, less decisive for the long-term survival of his grandchildren than is the presence of a grandmother (although there are a handful of studies which report exceptions to this general rule). If we take this as read for now, then the fact that men live so long starts to look a bit puzzling. In a monogamous society, a man’s reproductive career ought to end when his wife hits the menopause. If grandfathers aren’t especially pivotal for the survival of their grandchildren, then what keeps them alive beyond this age?

One possibility is that men might live longer because they can still find new breeding partners, even when their wives undergo menopause. In the Finnish data set, although divorce was not permitted, people could remarry if their current spouse died. Widowed men were three times as likely to remarry, compared to widowed women and, when they did so, they invariably took a much younger wife. More than 90% of the remarried men subsequently had more children, unlike the remarried women. Cross-cultural studies also suggest that, at least in some societies, a male’s reproductive market value remains reasonably high, even as he enters his autumn and winter years, especially when men can accumulate wealth during their lives (for example in pastoralist and horticulturalist societies, like the Turkana and the Tsimane). In such societies, men can attract younger wives and continue having children when they are well into their sixties and even their seventies. The fact that men can reproduce at relatively low cost,and can potentially continue to accrue fitness benefits by having their own children late into life, might therefore explain why men also live long lives and don’t go through a physiological ‘andropause’".

Nichola Raihani FRSB is Professor of Evolution and Behaviour at University College London.

The Social Instinct, published by Penguin Random House, is available in paperback from Thursday 6th October.