Sci. Aging Knowl. Environ., 20 February 2002
Vol. 2002, Issue 7, p. pe3
[DOI: 10.1126/sageke.2002.7.pe3]


Is Postmenopausal Life-Span a Gift of Modern Health or a Product of Natural Selection?

Donna J. Holmes

The author is in the Department of Biological Sciences at the University of Idaho, Moscow, ID 83844-3051, USA. E-mail: electric{at};2002/7/pe3

Key Words: menopause • reproduction • female • feminist • estrogen • anthropology

Presidential Symposium with a Perimenopausal Perspective

This past November 2001, at the annual meeting of the Gerontological Society of America (GSA) in Chicago, I organized and chaired a Presidential Symposium entitled "Female Reproductive Aging: Biology and Ideology." Because I'm a middle-aged, perimenopausal biogerontologist myself, I'm interested in women's reproductive aging from both a personal and a scientific perspective. For several years now, I've been studying the way that scientists in very different disciplines view the origins of and mechanisms (evolutionary as well as physiological) underlying menopause in humans. At the same time, I've followed the growing coverage of menopause and women's mid-life health issues by the popular media.

Current scientific ideas about aging in women---whether well-accepted by the larger scientific community or not---can have a significant impact on the way we view our own reproductive aging and the possible solutions to any accompanying medical problems we may be confronting. With this in mind, I organized a symposium for a wide audience of professionals in the field of aging; my objective was to provide a broader, more interdisciplinary perspective on female reproductive aging than would have been available had I invited only speakers already familiar to biogerontology. Comparative zoology, primatology, and anthropology have much to contribute to our understanding of human menopause; hence I invited speakers representing these subdisciplines, as well as disciplines more typically covered at GSA, such as endocrinology and neurobiology.

The speakers were asked to address, at least in part, the following questions in their talks:

• Is reproductive aging in modern women abrupt, or does it take place gradually, over a decade or more?

• Is the pattern of reproductive aging in humans different from that in females of other animal species?

• Do the living conditions enjoyed by modern women have a bearing on the length of our postreproductive lives?

• Is midlife menopause an adaptive component of a larger human "life-history strategy"? Is it shaped by generations of natural selection to balance reproductive costs and benefits and maximize genetic payoffs to an individual woman and her closest kin?

Reproductive Declines: Gradual or Abrupt? Uniquely Human or Merely Mammalian?

The first two speakers were Roger Gosden and Phyllis Wise. Both are recognized experts in reproductive endocrinology and ovarian aging; however, Wise focuses particularly on the effects of estrogen on neural tissue. In his opening to the symposium, Gosden summarized a wealth of data from many studies (some by his laboratory, some by others) on reproductive aging in women and laboratory rodents.

These data clearly indicate that, in women and lab rodents alike, fertility, gonadotropin secretion, and follicle maturation all begin deteriorating well before the complete cessation of ovulation in midlife, with a great deal of individual variability evident in all of these phenomena within any group of individuals studied. For us human females, clear signs of reproductive aging can be detected over a decade before actual menopausal declines in estrogen and cessation of ovulation in our late 40s or 50s. In other words, human (and rodent) female reproductive aging is a very gradual physiological transition (including the newly newsworthy "perimenopausal" period) and not an abrupt drop in function. And, although the average age at menopause in Western women is around 50 years, there is a lot more variability around this age than is generally acknowledged; healthy women can finish cycling anywhere between their early 40s and late 50s. (The oldest woman ever recorded to have given birth was 58 years old.)

Moreover, aging lab rodents and women all undergo key hormonal changes before and during midlife that are remarkably similar (adjusting, of course, for the fact that rats and mice live only 3 to 4 years under the best of conditions). The secretion of follicle-stimulating hormone, for example, becomes increasingly variable and unpredictable between ovulatory cycles in these species well before the rapid drop in estrogen that characterizes the approaching end of fertility. Like humans, who can live over one-third of their lives in a postmenopausal state, healthy female lab rodents can enjoy a postreproductive life-span of a year or more. This kind of evidence, therefore, argues clearly against the idea that the pattern of reproductive aging in human females is unique or even unusual, and shows that the human pattern is shared by mammalian species that are only distantly related.

View larger version (124K):
[in this window]
[in a new window]
"Osun Magic" by Betty LaDuke, 1992. Credit:

Estrogen Replacement and the Nurture of Neurons

Phyllis Wise's talk also reviewed a substantial and varied body of work, primarily from her laboratory, illustrating the neuroprotective effects of estrogen on the brains of laboratory rats and mice. Although estrogen has long been recognized as the primary hormone responsible for the organization and activation of female reproductive organs and behaviors, its role as a neuroprotective or "trophic" agent in older animals, whose reproductive function is waning, has been established only recently. Wise's work shows that middle-aged female rats receiving estrogen replacement are protected against the neural damage associated with "strokes," or the occlusion of cerebral arteries. This is particularly true for damage resulting from apoptosis: the delayed, progressive cell death that follows oxygen deprivation from decreased blood flow. She also briefly discussed evidence that this neuroprotective effect involves selective up-regulation of the alpha estrogen receptor in rats and mice.

Wise's data not only add to the growing list of the potential benefits of judicious hormone replacement for many aging women, they are consistent with an argument that survival for many years beyond a mid-life menopausal transition is a relatively recent development for humans, rather than an ancestral trait shaped by natural selection. The logic for this argument runs as follows. Evolutionary theory predicts that natural selection will exert its strongest effects on traits with the most direct consequences for survival and reproduction of individuals in a population. If, then, the ancestors of modern women had been subject to many generations of evolutionary pressure to maximize their postreproductive health, we might expect to see a more finely tuned suite of physiological adaptations for coping with a hypoestrogenic environment. But if, instead, our postreproductive years are merely a fortunate by-product of enhanced survival under modern conditions, rather than a part of a rigorously selected ancestral life history, it stands to reason that we might be less well adapted to our postmenopausal hormonal profiles.

This latter, nonselective scenario for menopause could help to explain why so many women in industrialized societies who are otherwise in excellent health suffer from symptoms ranging from hot flashes to osteoporosis and increased risk of heart disease in the years following menopause. Under this scenario, we might well expect to need a medical boost here and there to maximize our health and comfort into our fifth decade and beyond! There is some recent clinical evidence corroborating Wise's studies and showing that replacement of estrogen, progesterone, or a combination thereof may help reduce the risk of cognitive declines or dementia in women. The case for the neuroprotective effects of hormone replacement remains controversial, however, and much more work is needed to clarify the possible benefits and trade-offs of hormone replacement for protecting against neurodegenerative conditions.

Natural Fertility and Pregnancy Loss: The Word from Bangladesh

In comparatively wealthy industrialized countries with good medical care, women's life-spans have increased dramatically over the past decade, from about 45 to over 80 years. But the advent of modern contraceptives, along with other medical innovations, makes it difficult to know how our life-long reproductive patterns, including aging-related changes in fertility, compare to those of our ancestors, and what the "natural" pattern of aging might have been. One way to address this problem is the study of so-called "natural fertility" populations in relatively unindustrialized societies where modern contraception and medical care are much less common. This is the approach used by Darryl Holman, the third speaker, who is an anthropologist specializing in age-related changes in human female reproduction.

Holman reported the results of his ambitious study of changing patterns of pregnancy loss with aging in a natural fertility population of up to 700 women in Bangladesh. As an indicator of pregnancy, he used levels of chorionic gonadotropin in the women's urine. In this way, he was able to begin to separate aging-related declines in overall "fecundability" (the probability of conception) from those affecting the chances of sustaining a successful pregnancy.

Holman's findings, too, run counter to the idea that reproductive aging in women in natural fertility or industrialized societies is an abrupt event. The increasing probability of losing a pregnancy or carrying an abnormal baby are but two components of gradual reproductive aging in women. Holman's data suggest that, although fecundability overall in Bangladeshi women, as in other societies, undergoes a relatively precipitous decline in the mid-30s, the frequency of pregnancy loss increases much more gradually with age, beginning in a woman's 20s and continuing until menopause.

What's So Special About Infertile Grandmothers?

All of the presentations of speakers in the symposium that I've discussed thus far suggest first, that reproductive aging in women is gradual, rather than abrupt; and second, that aging-related changes in reproduction in women are by no means unique to humans but are shared at least by domestic rodents living under healthy laboratory conditions. These findings both argue strongly against the idea that human menopause is a special adaptation.

There is, however, an important alternative view often discussed in the literature. In this view, menopause is an unusually sudden, precisely timed event, suggesting strong selection over the course of human evolution for timing in mid-life (about two-thirds of the way through the average modern woman's life-span of 80-plus years). Respected and influential anthropologists and primatologists, including Kristin Hawkes at the University of Utah and her associates, are among those currently holding an adaptive view of menopause. Data they have gathered on the Hadza, a hunter-gatherer tribe in Tanzania, show that postreproductive women contribute as much or more to their social groups in terms of food provisioning than either reproducing women or men of any age. According to these researchers, these data are consistent with a hypothesis that natural selection has historically favored a shift away from personal reproduction by females reaching middle age, when the risks of childbirth increase dramatically, and has instead channeled these females' energy into supporting their close kin.

The adaptive menopause hypothesis, often called the "Grandmother Hypothesis," has by no means been firmly proven, and remains very controversial among scientists. Data consistent with a model of adaptive evolution are insufficient to reject a null hypothesis of no adaptation. Nonetheless, it has received a great deal of attention from the popular media, as well as being embraced by large numbers of women in our society for its obvious feminist appeal.

A rigorous evolutionary approach demands that special adaptations require special kinds of evidence to support or refute their existence. A compelling argument against the menopause-as-adaptation view can be made if other species, either in the wild or living under conditions similar to those enjoyed by modern women, exhibit similar patterns of reproductive aging. I have already mentioned that this type of evidence is widely available for laboratory rats and mice. The second anthropologist in the symposium, Tim Gage, presented data that characterize age-related changes in mortality and fecundity in our closest relatives, the nonhuman primates.

Menopause: Not Just for Humans Any More

Gage's presentation focused on demographic data gathered from zoos and research primate colonies of a wide variety of species, as well as primate populations in the wild. He included species such as the golden lion tamarins at the National Zoo in Washington, DC: small monkeys with relatively short (15-plus years) life-spans and much less intensive parental care than our closer cousins, the great apes. As medical care, diet, and other aspects of the husbandry of captive primates continue to improve, their life-spans grow longer, just as modern humans have become longer lived over the past century.

In his data analysis, Gage used the kind of state-of-the-art, stringent demographic techniques now accepted by the biogerontological research community (see, for example, Carey, 1999). In a nutshell, the data he presented fail to support an adaptive view of mid-life human aging, or the contention that women's reproductive aging is abrupt or otherwise unusual. On the contrary, declines in fertility in nonhuman primate species, including chimps, Old World monkeys (macaques and baboons), New World monkeys, and prosimians are, like those in modern women, gradual rather than abrupt and are followed by long postreproductive life-spans (at least 20 to 30% of the maximum life-span for each species). This is particularly clear when examining the growing amount of available data from captive animals, which are most like humans in terms of the health care and protection they receive from natural sources of mortality. The oldest documented age of reproduction by a captive female golden lion tamarin, for example, is about 12 years; some females have lived up to 17 years. This means that these little monkeys can spend one-quarter to one-third of their lives in a postreproductive state, a period that will probably lengthen as our captive husbandry improves.

In my own, concluding talk, I emphasized further the lack of support for an adaptive scenario for the evolution of menopause and summarized the comparative evidence that refutes it. Patterns of reproductive aging similar to those in women can readily be seen in females in a whole potpourri of species, wild and captive, short- and long-lived, with or without extended parental care; these include opossums, rodents, quail, chickens, and even some whales--which all exhibit postreproductive life-spans of up to one-third their total longevity, preceded by a period of gradual reproductive decline. Japanese quail, for instance, have survived in the laboratory for just over 5 years. By the end of their first year, however, quail hens are already becoming infertile; a 3-year-old hen is quite aged, reproductively speaking.

Good Scientists or Granny-Bashers: What's a Woman to Do?

As a whole, the research summarized in this symposium strengthened my own conviction that the comparative evidence refutes rather than supports an adaptive scenario for the evolution of mid-life menopause in humans. Although older women, including hard-working grandmothers, certainly deserve to be valued by our society in their own right, the postmenopausal life-span, like that seen in many other animals living in a healthy setting protected from disease, injury, and other natural sources of mortality, is more likely to be a gift of modern health and longevity than a product of natural selection.

Whether supported by the comparative biogerontological evidence or not, the adaptive menopause or Grandmother Hypothesis has been received enthusiastically by many of my female peers: well-educated, middle-aged, baby-boomer women, veterans of the second wave of the women's movement. In our youth- and physical perfection-oriented society, women of a certain age (if that age is visible) are perceived as less attractive or important than our younger sisters. The idea that postmenopausal females might possess our own special kind of evolutionary fitness, "valued" and shaped by centuries of natural selection for a particular kind of life history, is profoundly appealing to those of us looking for meaning in this mid-life "passage," as popular author Gail Sheehy calls it.

Many of us suffer a great deal of discomfort during the years of menopausal transition, and some must address even more serious health issues. We have come a long way over the past century toward taking control of our own health and medical care, and many of us still harbor a deep resentment of the medical "establishment" and its history of inequities, both real and perceived. Although we face increasingly complex and difficult questions about the desirability and risks of treatment for the effects of declining hormone levels, women are still woefully underrepresented in the sciences and untrained in (or even suspicious of) the analytical approaches that are so useful for sorting out complex medical issues.

To some of my contemporaries, the suggestion that the adaptive menopause hypothesis lacks strong scientific support is tantamount to demonstrating an affiliation with the "dark side" of the sexist medical establishment, the same folks that coined the term "hysteria" for mental ailments thought to have their roots in a women's uterus. Not only does it feel politically incorrect, it sounds downright anti-grandmother!

As every scientist knows, negative evidence is never as newsworthy as data that seem consistent, at least on the surface, with an attractive or provocative new idea. Comparative data refuting an adaptive basis for menopause are unlikely to be featured on the cover of Newsweek. But clear thinking about these issues is essential for understanding women's attitudes toward an important health transition with serious medical, psychological, social, and spiritual consequences. Whether a woman considers the change from reproductive to postreproductive status to be "natural," "special," or just an incidental by-product of the longer, healthier life-spans we now enjoy in societies, her views of menopause will undoubtedly affect the medical and life choices she makes.

February 20, 2002

Suggested ReadingBack to Top

  • S. N. Austad, Menopause: an evolutionary perspective. Exp. Gerontol. 29, 255-263 (1994).
  • J. R. Carey, in Methods in Aging Research, B. P. Yu, Ed. (CRC Press, Boca Raton, FL, 1999) pp. 3-24.
  • T. B. Gage, The comparative demography of primates: with some comments on the evolution of life histories. Ann. Rev. Anthropol. 27, 197-221 (1998).
  • R. G. Gosden, Biology of Menopause: The Causes and Consequences of Ovarian Aging (Academic Press, London, 1985).
  • K. Hawkes, J. F. O'Connell, N. G. Blurton Jones, Hadza women's time allocation, offspring provisioning, and the evolution of long postmenopausal life spans. Curr. Anthropol. 38, 551-577 (1997).
  • K. A. O'Connor, D. J. Holman, J. W. Wood, Declining fertility and ovarian ageing in natural fertility populations. Maturitas 30, 127-136 (1998).
  • C. Packer, M. Tatar, A. Collins, Reproductive cessation in female mammals. Nature 392, 807-811 (1998).
  • L. Partridge, Menopause for thought. Nature 364, 386 (1993).
  • G. Sheehy, The Silent Passage (Random House, New York, 1992).
  • P. M. Wise, K. M. Krajnak, M. L. Kashon. Menopause: the aging of multiple pacemakers. Science 273, 67-70 (1996).

Science of Aging Knowledge Environment. ISSN 1539-6150