the other is not. This can be seen by reframing the problem of the postreproductive life span. One could ask, ''Given that humans cease to reproduce in their early to late forties, why do they live so long?" This focuses the question on the evolution of the long life span and shows that both menopause and longevity require explanation (Hawkes et al., 1996, have independently come to a similar conclusion).

Several possible ancestral conditions may be considered. One possibility is that the ancestral condition is a somewhat longer reproductive period and a shorter life span. Selection, due to an increased value of grandparenting, could direct additional resources toward longevity and investment in descendant kin, at the expense of shortening the reproductive period. If this were true, the grandmother hypothesis could account for both menopause and the long life span.

Another possible ancestral condition is the same life span as is found currently but no cessation of reproduction. In that case, the tradeoff is solely between production of offspring and investment in existing kin. Selection due to increased benefits from grandparenting could produce menopause. The long life span of humans for their body size would then require another explanation.

A third possible ancestral condition is a short life span, ending at about the same time as, or before, menopause occurs today but with no menopause. In this case, selection could favor the increase in life span without a concomitant increase in the length of the reproductive period. One possible scenario for such a selection regime is that different tradeoffs are involved in the evolution of the reproductive period than in the life span.

With respect to the reproductive period, it appears that the physiological cause of menopause is the depletion of oocytes due to the process of follicle decay, known as atresia (vom Saal et al., 1994).8 Mammalian females begin life with their full complement of germ cells, and this process of follicle decay seems to be a very general feature of mammalian reproductive physiology (ibid., Finch, 1994). Follicle decay appears to exhibit a constant exponential decline through life, with an acceleration just before menopause—with menopausal women having essentially no viable oocytes left (Richardson et al., 1987: figure 3). The main difference between humans and other mammals, except for some whales, is that total loss of oocytes occurs in human females well before most have died and before other organs senesce (Austad, in this volume, and Hill, 1993).

It is possible that the main constraints on the reproductive life span are the number of germ cells at the outset and the rate of follicle decay. There is


There is some confusion in discussions of menopause, due to definitional issues. Many mammals, including primates, show evidence of decreasing fecundity and ovarian function late in life (Caro et al., 1995; Gould et al., 1981). If such evidence is considered to be indicative of menopause, then it is fairly common. On the other hand, other authors have been interested in the evolution of menopause as a postreproductive period of significant duration. If this latter definition is used, then it is quite rare (see Austad, in this volume, for a related discussion).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement