If you think about it, you might realize that after an animal finishes reproducing, it should die, because genes that make you live on after you can no longer reproduce have no selective advantage: they are no better than genes that kill you off when you’ve had your last child. In principle, natural selection should keep you pumping out gametes and children until you die. But in some species, namely ours, some of our ape relatives, and, curiously, some toothed whales like killer whales, females continue to live for considerable periods after their reproduction ends. We call that end “menopause”. This leads to three questions:
a. Why do animals cease reproducing? That is, why don’t they continue to reproduce until they die?
b. Why in some cases do animals continue to live even after they cease reproducing?
c. Why don’t males undergo “manopause” in our species?
This new paper from PNAS (click title below to read, or find the pdf here) deals with the first two questions, but not with the third; and I’ll leave readers to ponder that one. The paper in fact, simply shows that in one population of mountain gorillas in Uganda, many females do show a form of menopause, living on for nearly a quarter of their adult lives as nonreproductives. While this phenomenon has been demonstrated in chimps, other studies of gorillas have not shown it. The authors posit that there may be different results in different wild populations of gorillas, though that’s hard to understand if you think the phenomenon involves natural selection. Why should such selection differ among populations of the same species? The hypotheses below don’t predict interpopulation variation.
Read on:
The first question above can be answered by realizing that menopause may be partly a cultural phenomenon. For the vast majority of our evolutionary history, humans probably died before the females stopped reproducing: probably between ages twenty and forty. There may have been no menopause in our species because nobody lived long enough to show it. And that may be one clue for why we show it now: any genes that cause women, at least, to lose reproductive ability when older were simply not expressed, and thus not selected against. This may also be the reason why earlier studies of chimps showed menopause: they were taken care of in zoos or reserves in a way that allowed them to live longer than they did during most of their evolutionary history.
Further, this population of gorillas, though living in a reserve, were not given special food or treatment (some were given vet care, but those were omitted from the study), and still showed not only menopause, but long lives after menopause. The “evolutionary history” phenomenon can’t easily explain that. Nor can it explain postreproductive life in toothed whales—unless that was seen only in aquaria where they lived longer than they would have during much of their evolution, including now when living in the open ocean. (One would have to look at the studies to determine that.)
But regardless of the cause, one can say that one population of mountain gorillas under natural conditions—probably similar to those that obtained during most of their evolution—often show not only a cessation of reproduction but also considerable years of life beyond that. And the behavior of gorillas makes some of the evolutionary hypotheses for menopause seem unlikely.
Results. This will be short. The authors studied 25 adult female mountain gorillas (Gorilla beringei beringei, one of two subspecies of the Eastern Gorilla) in Bwindi Impenetrable National Park in Uganda. The females came from four groups, and their life histories were known, presumably through intense observation. (Mountain gorillas hsve indeed been studied intensively, most famously by George Schaller and later by Dian Fossey (who was murdered during her studies.) The study populations are fairly easily habituated to human presence, which allows this study.
Here’s how they define “post-reproductive females” and how many of them showed menopause:
According to a commonly used definition, “postreproductive females” are those who live past the age of their last reproduction for longer than the mean plus two SD of successful interbirth intervals (2). We calculated this value as 7.7 y [5.1 (2 1.3)] in our study population, suggesting that seven out of the 25 study females qualified as postreproductive. Six of these seven females have been conservatively estimated (based on the ages of genetically identified offspring, body condition and hair loss) to be older than 35 y old, which is the maximum age of observed reproduction (Figs. 1 and 2). All the seven postreproductive females exhibited a postreproductive lifespan of at least 10 y (Fig. 1), minimizing the possibility to be “mistakenly” classified as postreproductive. These females were not observed mating for an average of 7.4 5.8 y before they exit the study
And a summary:
Our study shows that wild Bwindi mountain gorillas can exhibit long postreproductive lifespans. Given that female gorillas rarely reach 50 y of age in the wild (6), the 10 postreproductive years lived by one third of the study females represents at least 25% of their adult lifespan (adults: 10 y old). More generally, the standardized population measure of PrR suggested that females spend 10% of their adult lifespan as postreproductive. Importantly, neither of the two methods we used to derive postreproductive lifespan can distinguish menopause from other causes of sterility, such as an increased fetal loss probability in old females. Nevertheless, the extensive duration of postreproductive lifespan, the reduced or lack of mating activity, and previous endocrine analyses of old females (8, 9) suggest that menopause is a highly plausible cause for the reproductive patterns we observed. The selective pressure(s) which might have favored the evolution of this trait in gorillas remain unclear.
Indeed; menopause remains a mystery in all species that show it. We have hypotheses but no substantive answers.
So the question arises of what, if any, selective pressures could have promoted female longevity beyond reproduction. This assumes—which we don’t know—that postreproductive survival was an adaptation. If it was, and not just a “spandral” here are a few hypotheses. The bold headings are mine, and indented text is from the paper:
a. Reproductive conflict:
The “reproductive conflict hypothesis,” posits that old females cease reproduction to avoid competition for limited reproductive opportunities with young (related) individuals (12); e.g., their daughters or the mates of their sons]. Female gorillas disperse from their natal groups and often disperse again from groups where they have reproduced (13), meaning that they have low relatedness to their groupmates. Hence, the benefits of reproduction for female gorillas at an old age may be greater than that for chimpanzees or humans, where female local relatedness increases with age and females reproduce simultaneously with their offspring (12, 14).
Avoiding conflict with individuals is advantageous only if they’re related, for this would be a form of “kin selection”. Since gorillas’ dispersal take them away from their kin, that makes this hypothesis less likely but not completely unlikely.
b. Intergenerational help, one form of which is the “grandmother hypothesis”.
Another relevant set of hypotheses, also relatively unlikely to apply to gorillas, posit that intergenerational help, and its positive influence in grandoffspring fitness, may drive the evolution of postreproductive lifespan through two not mutually exclusive evolutionary pathways [see also “grandmother hypothesis”; (1)]: by selecting for longer female lifespan to allow females overlap with grandoffspring and help them increase their fitness (e.g., by offering their ecological knowledge, or by defending them). . .
. . . The associated “mother hypothesis” (15) might have greater predictive power in gorillas. This hypothesis posits that old females cease reproduction to minimize energy expenditure or other reproductive costs, and maximize investment to existing offspring and their fitness. Consistent with this hypothesis, maternal presence, care, and support is critical even for adults in gorillas and other hominids (16).
This too is a form of kin selection (as is parental care), for genes that help you take care of your grand-apes, or your offspring when you’re old, will still be helping copies of those genes in their still-reproductive descendants. This is feasible for taking care of offspring, but given the dispersal of female gorillas, the “grand-ape” hypothesis is less likely.
And here’s a nonadaptive hypothesis, but one that is popular:
c. Menopause is a nonadaptive byproduct of gorillas’ life history.
A final hypothesis posits that postreproductive lifespan is a nonadaptive by-product of life-history patterns. Given that many wild animals die from predation, disease, or starvation, genes whose deleterious effects appear only in advanced ages, may not be purged (15). When “favorable” conditions allow individuals to survive at these ages, deleterious effects that prevent reproduction may appear (4, 11). Accordingly, greater food abundance and potentially lower predation pressure in comparison to the evolutionary history of chimpanzees, may allow Ngogo chimpanzees to live longer and exhibit menopause (4). Similarly, Bwindi gorillas currently do not face any predation risk from leopards, their main potential nonhuman predators,
A version of this hypothesis is that some genes have the effect of promoting reproduction early in life, but at the price of inhibiting reproduction later in life. Under many conditions, such “early reproducing genes” will be more adaptive than genes promoting later reproduction, because the former leave more copies of themselves earlier. (Those genes, for example, would be heavily favored in a growing population.). Thus senescence and menopause could simply be the result of the accumulation of adaptive “early-reproducing genes.”
Which, if any, of these hypotheses are right? We don’t really know for primates or toothed whales, and though there may be evidence for “senescing” genes in some laboratory species, I’m not aware of it.
The question remains why don’t male chimps, gorillas, and humans show “manopause”. Some human males, for example, can father offspring even at the age of 80, but you’ll never find a woman reproducing at that age And we have no data from chimps or gorillas on males, at least as far as I know.
So, as always, “more work needs to be done”. But at least we now know that gorillas and chimps have menopause in females, which might make you a big hit if you bring it up at a cocktail party. And don’t forget to mention those toothed whales!
See: Monaghan, P. and Ivimey-Cook, E.R. (2023), No time to die: Evolution of a post-reproductive life stage. J Zool, 321: 1-21. https://doi.org/10.1111/jzo.13096
It would help if you told us why you are sending us to this paper. Thanks.
Trying to answer your question about the nonexistence of the equivalent of menopause in human males, in my opinion, given the parental investment of the human male in reproduction and eventual subsequent upbringing, probably there has not been in the evolutionary history a very important selection pressure to eliminate reproductive capacity, since it does not compete so significantly for the energy resources of individuals.
Could it be that the genes that improve your reproductive fitness when you are young, say extending your fertile period, simply have the side-effect that they extend your lifespan after your reproductive years (you are fit when you are young, so you remain fit when you are old). In the wild, older individuals die from predation, etc. so the effect isn’t noticeable unless you manage to live longer. Fertility may be terminated because deleterious mutations in eggs increase as the female ages to ensure the eggs that are available during her reproductive period are higher quality.
This is absolutely fascinating. I used to be satisfied that the grandmother hypothesis explained menopause in humans. All of our friends who are grandparents participate actively in raising their grandchildren. (That may be a very recent innovation, however.)
But the fact that female Gorillas disperse and do not help raise their grandchildren leads me to question the grandmother hypothesis. If menopause in Gorillas is not for purposes of (e.g., selected for) caring for second generation offspring, then why would it be in humans? Assuming that menopause in Gorillas and menopause in Humans are homologous, parsimony would lead one to hypothesize that menopause arose in the common ancestor of both. If so, it seems unlikely that it arose for purposes of grandparental care.
Now, it may be that menopause in Humans was inherited from a common ancestor and not specifically selected for at the dawn of Homo sapiens. But by having the effect of enabling postreproductives to help raise grandchildren, menopause was later co-opted by selection and enhanced. Humans females can spend up to half their lives in menopause. So it seems possible that the grandmother hypothesis doesn’t explain the origin of menopause in Human women, but that it does explain how long women live after menopause. Once grandmotherly care started enhancing the survival chances of grandchildren, selection captured menopause and extended it.
Raising human children is much more time consuming than other species.
The grandmothers are a big help. Accidental? Like you say, it could have reinforced the already existing menopausal grandmother.
I’m thinking that menopause was originally just a shared derived character, derived from the common ancestor of Gorillas, Chimpanzees, and Humans, and that it was not originally present because of its incipient contribution to childrearing.
However, in Humans, which have long childhoods and long lifespans, grandmothers do have a strong impact on the success of their grandchildren. Hence, in humans, the menopausal period was co-opted by natural selection. The period was extended and the grandmother-grandchild bond was strengthened. Natural selection grabbed onto menopause as a trait that increased the reproductive success of grandchildren. Hence, the genes for both menopause and those for loving grandmotherhood were preserved and became ubiquitous in Humans.
I adored my grandmothers and feel privileged to have lived close to them as a child and to have known them.
Very interesting to learn that the grandmother hypothesis isn’t necessarily the quick answer to “why menopause”. Still, in those several other species where members live in extended multi-generational families, it seems to me that getting help from mom and from grandma would increase fitness and so menopause would be selected for.
So now I’m wondering if the bigger picture could be still supportive of the grandmother hypothesis. Suppose that getting help from non-reproductive grandma is the primitive state, and it evolved independently in multiple lineages (in apes and toothed whales in particular), and then for some reason in gorillas it was secondarily lost. It could be lost under circumstances where the fitness effects of out-crossing by females dispersing was greater than the fitness effects of helpful grandmothers.
I’m reminded of ‘A New Eusocial Vertebrate?’ (Foster, Ratnieks, 2005), which relied heavily on the existence of menopause and the grandmother hypothesis for its argument.
Granted, that paper was also clearly written rather cheekily and aimed more at demystifying/un-specialing eusociality as a concept than as a serious hypothesis, but still.
Nevertheless, it’s certainly interesting to see a case against the grandmother hypothesis.
How well do young gorillas whose mothers die before they’re mature fare?
Kristen Hawkes has shown in several papers that the presence of older females in sufficient numbers (proportion of the population) for the Grandmother hypothesis to work IS NOT MERELY A FEATURE OF modern demography. Her Google Scholar pg is here:https://scholar.google.com/citations?hl=en&user=YqBhezoAAAAJ&view_op=list_works&sortby=pubdate.
I would urge folks to read some of her more recent papers where she develops many implications of the GMH for the human life history and other aspects of being human. Much has been added since our 1997 PNAS paper, which is available for free. https://www.pnas.org/doi/full/10.1073/pnas.95.3.1336…. or See link on her GS page.
So, we now have demonstrated post-reproductive survival in all 4 of the extant great-ape lineages (though in orangs it has only been observed in captivity where life spans tend to be longer than in the wild). And it may be relevant that the longevity of early human populations and the reproductive life span were about the same; few seemed to live beyond 50 (though some seemed to).
Idealized models of the grandmother hypothesis show that human females can increase the total proportion of their genes in future generations by switching from caring and providing for their OWN daughters to a focus on their granddaughters. Of course granddaughters have only 1/4 of her genes on average, BUT the grandma’s reproductive value (estimated future reproductive contributions) diminish a lot after age 40, so the switch to granddaughters with 1/4 of her genes can actually result in a HIGHER percentage than for (many fewer) future daughters with 1/2 of her genes each.
And—should grandma survive this long—switching to her great-granddaughters as her own daughters approach menopause increases this benefit as well. Some of this is direct care, some of it is provisioning, and some of this is accumulated knowledge that puts these descendants at an advantage.
So the long grandma lives after menopause, the more benefit accrues to her non-reproductive contributions to the preservation of her genes in future generations.
There IS a decline in male reproductive function with age. It begins later than with females, but its decline is comparable. By age 65 in humans, male reproductive function averages about 80% of the average for a 25-yo male. Between 65 and 85 yrs, it declines to about 10%! That’s about 3.5% per year (in a linear model) compared to decline in females from 25–50 years of about 4% per year. So, the difference seems to be more about the age of onset than the rate of change.
Regarding ALL the anatomic, physiologic, and environmentally influenced changes:
Gunes S, Hekim GN, Arslan MA, Asci R. Effects of aging on the male reproductive system. J Assist Reprod Genet. 2016;33(4):441-454. doi:10.1007/s10815-016-0663-y
There are also comparable changes in other mammalian males:
Solène C, Victor R, Florentin R, Jean-Michel G, Jean-François L. Male Reproductive Senescence in Mammals Is Pervasive and Aligned With the Slow-Fast Continuum. Ecol Lett. 2025;28(9):e70194. doi:10.1111/ele.70194
The problem is (as always) establishing PAternity. We can be 95% sure of who our mothers are, but it is harder to prove the same about our fathers!
Caveat emptor: almost everything I know about evolutionary biology I have learned from this blog. But I do know that female humans are born with all the eggs they will ever have — unlike males, who keep making more gametes. So can menopause be explained simply by the egg supply running out, and the limited number of original eggs explained by some evolutionary advantage to (1) not having to keep making eggs, and (2) not starting with a huge supply of eggs? That still doesn’t explain why (in a few species) females live beyond the age of reproduction. But the difference in whether one continues to make new gametes does explain why there’s no direct male counterpart to menopause, right?
For an overview of human menopause, in all its aspects [ evolution, historical, medical] I recommend Susan Mattern’s book:
https://press.princeton.edu/books/hardcover/9780691171630/the-slow-moon-climbs
A detailed discussion of the Grandmother hypothesis.
Is the explanation that all primate females over ~45 years old have egg quality that is so poor that the risk of miscarriage or birth defects (e.g. Downs Syndrome in humans) outweighs the benefits of having additional offspring? I don’t know if this is group selection and verboten but the consequences could be 1) taking grandmothers away from helping daughters with healthy offspring care for their young while preoccupied with raising her own offspring with much higher mortality, reducing her own net genetic longevity 2) mating/pairing with healthy males whose offspring have higher mortality (bad for them and resource investment) and for troop survival so selected against by females’ bodies (i.e. those clans die out due to reduced numbers while menopause restores generational longevity).
We know that females are born with all of their eggs that can accumulate mutations throughout theirs lives and manifest as lethal genetic defects and increased miscarriages and birth defects. It’s why many infertile women seek egg donors. Males make new sperm throughout their lives so they don’t accumulate as many mutations and their offspring don’t carry as many genetic defects and high incidence of child mortality. I thiught that’s why there is no manopause. Also, if many males die in combat or acquiring food, it only takes a few males to impregnate all the females so the clan survives instead of dwindles and why they produce viable sperm until old, though granted sperm quality and count decrease with their age as well.
FWIW: New study of changes in sperm as men age. Increase in “danger” as the headline says. As usual in these reports, the headline is overwrought, but the message and reporting are reliable if one reads the rest:
The hidden evolution making men’s sperm more dangerous with age.
Aging amplifies hidden evolutionary forces in sperm, boosting the odds of harmful mutations being passed to future generations.
https://www.sciencedaily.com/releases/2025/10/251019120513.htm