Post-reproductive life

Why do humans live for decades after they can no longer have children?

Selection works through descendants, so a trait expressed only after reproduction ends should be nearly invisible to it. Yet human females routinely stop reproducing around fifty and then live on for two or three decades in good working order. That is not a brief coda; it can be a third of a life. Why has selection tolerated it — or, more pointedly, why does it look less like tolerance and more like design?

First check that the puzzle is real, because most people's version of it is not. The usual objection is that pre-modern people simply died before fifty, so there was nothing to explain. That misreads what life expectancy at birth is: an average over everyone, dragged down brutally by infant and child mortality. In a population where a quarter of children die before their fifth birthday, the average can sit near thirty while those who cleared childhood commonly reach their sixties and seventies. A review of forager demography put the most common adult age at death at around seven decades. Grandmothers are not a modern invention; they are the ordinary case for anyone who survived childhood.

So the puzzle stands. What could keep selection interested in a female who has stopped bearing children? Only her effect on the reproduction of others who carry copies of her genes. A grandmother shares a quarter of her genes with each grandchild. If her presence makes enough grandchildren survive who otherwise would not, that is a return in the same currency as having children herself.

Does she? The evidence is real. Among the Hadza, older women forage intensively and provision weaned grandchildren not yet efficient foragers themselves — the observation that started this line of thought. And in historical parish records from Finland and Canada, women who lived past fifty had more surviving grandchildren than those who did not, with the effect concentrated where the grandmother lived close by. That is the grandmother hypothesis: post-reproductive lifespan persists because helping raise existing descendants outperforms continued childbearing, especially in a species with a long, dependent, expensively-taught childhood.

But notice what that explains and what it does not. It explains a long life. It does not explain why fertility should stop — why not keep breeding and help as well? One partial reason: childbirth risk rises with maternal age, and a woman with several dependent children who dies bearing another may lose more than she gains.

The sharper rival is reproductive conflict. Consider a mother and her adult daughter-in-law who could breed at the same time. If the pattern of who moves between groups means the younger incoming female is less related to the older female's kin than the older female is to hers, the two are not symmetric competitors: the older loses more from the conflict. Models of this predict the older female should cede breeding, and historical data showing poor outcomes for children born to overlapping generations of co-resident women support it.

There is also a deflationary possibility worth stating honestly: that menopause is not adapted at all but a byproduct. Females are born with a fixed stock of oocytes that depletes on a schedule, and if human lifespan was extended for other reasons, ovarian shutdown may simply be the first system to fail. This account is weakened by the comparative evidence but not eliminated by it.

That comparative evidence is the strongest constraint on all these stories. Menopause — a long post-reproductive life, not merely declining fertility — is very rare among mammals, well documented in humans and in a small number of toothed whales including killer whales and short-finned pilot whales. In killer whales, post-reproductive females lead the group disproportionately when salmon are scarce, and their sons' survival drops sharply when they die — so both accounts have support from a lineage separated from ours by tens of millions of years. Any correct explanation has to say why these particular species and almost no others.

First, "post-reproductive" here is strictly about direct reproduction. In the currency selection actually counts — copies of genes in future generations — a grandmother provisioning grandchildren is reproducing, just indirectly. Nothing about this requires selection to have gone soft.

Second, these explanations are not mutually exclusive and are not settled between. Grandmother effects, mother effects and reproductive conflict can all operate at once, and the live argument is about weighting, not about which single story is true. The honest summary: the grandmother hypothesis is the best-known account, genuinely supported by forager and historical demographic data, and contested — with conflict models currently doing much of the work on why fertility ends rather than merely tapering.

One asymmetry is worth naming. Male fertility declines with age but does not stop, so post-reproductive life in this strict sense is a female phenomenon, and any explanation resting on general longevity alone has to account for that too.

• Why killer whale menopause and human menopause arose independently, and what their social structures share.
• How life history theory predicts the trade-off between number of offspring and investment per offspring.

Every term the collection defines is gathered in the glossary.