Learning, menopause, and the human adaptive complexphooper/KaplanEtAl2010_Menopause.pdfin...

nyas˙5528 nyas2010v2.cls (1994/07/13 v1.2u Standard LaTeX document class) 3-22-2010 :1250

NYAS nyas˙5528-1489010 Dispatch: 3-22-2010 CE: VBF

Journal MSP No. No. of pages: 13 PE: Allie/Jen

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

Ann. N.Y. Acad. Sci. ISSN 0077-8923

ANNALS OF THE NEW YORK ACADEMY OF SCIENCESIssue: The Biodemography of Reproductive Aging

Learning, menopause, and the human adaptive complex

Hillard Kaplan,1 Michael Gurven,2 Jeffrey Winking,3 Paul L. Hooper,1 and Jonathan Stieglitz11Department of Anthropology, University of New Mexico, Albuquerque, New Mexico. 2Department of Anthropology, UC SantaBarbara, Santa Barbara, California. 3Department of Anthropology, Texas A&M, College Station, Texas

Address for correspondence: Hillard Kaplan, MSC01-1040 Department of Anthropology, University of New Mexico,Albuquerque, NM 87131. [email protected]

This paper presents a new two-sex learning- and skills-based theory for the evolution of human menopause. Thetheory proposes that the role of knowledge, skill acquisition, and transfers in determining economic productivityand resource distribution is the distinctive feature of the traditional human ecology that is responsible for theevolution of menopause. The theory also proposes that male reproductive cessation and postreproductive investmentin descendants is a fundamental characteristic of humans living in traditional foraging and simple horticulturaleconomies. We present evidence relevant to the theory. The data show that whereas reproductive decline is linked toincreasing risks of mortality in chimpanzees, human reproductive senescence precedes somatic senescence. Moreoverunder traditional conditions, most human males undergo reproductive cessation at the same time as their wives.We then present evidence that after ceasing to reproduce, both men and women provide net economic transfers tochildren and grandchildren. Given this pattern of economic productivity, delays in menopause would produce neteconomic deficits within families.

Keywords: menopause; fertility; senescence; intergenerational transfers; human life history

Introduction

Available demographic evidence from hunter-gatherers and forager-horticulturalists without ac-cess to modern medicine shows that men andwomen can expect to live an additional two decadesupon reaching age 45.1,2 This adult age-specificmortality profile is rather uniform across extant tra-ditional societies, and there is paleodemographicevidence suggesting the existence of older adultsthroughout the upper Paleolithic.3 This implies thatsurvival into old age is a fundamental feature ofhuman biology. Behavioral data also shows thatolder postreproductive adults of both sexes are quiteproductive,1,4–6 and tend to produce more energythan they consume until about age 70. Reproduc-tive senescence, however, occurs at much earlier agesin women and is largely complete by age 45. Thispattern is also rather uniform across human pop-ulations, and there is surprisingly little variation inage of menopause cross-culturally.7 The existenceof two nonreproductive decades of adult life raises

the fundamental evolutionary question: under whatconditions will organisms evolve for whom gen-eral somatic senescence proceeds much more slowlythan reproductive senescence?

Evolutionary theories of menopause that pro-pose an adaptive function for reproductive cessa-tion must show that the acceleration in reproductivesenescence relative to mortality-related senescenceresults in higher fitness than the standard simulta-neous decline in survival-related and reproductivefunctions (see Refs. 8–10 for reviews). Such theoriesneed to provide a reason why direct reproductionwill yield lower fitness than investing in alterna-tives, such as existing children and grandchildren.Special conditions must come into play for the fol-lowing reason. In a diploid sexually reproducingorganism, a female will be related to her offspringwith Wright’s coefficient of genetic relationship, r,of 0.5, whereas her grandchildren will only be halfas related to her (r = 0.25). Therefore, according toinclusive fitness theory,11 her investments in grand-children will have to produce twice the fitness effect

doi: 10.1111/j.1749-6632.2010.05528.xAnn. N.Y. Acad. Sci. xxxx (2010) 1–13 c© 2010 New York Academy of Sciences. 1


1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

Learning and human menopause Kaplan et al.

as in children for selection to favor investment ingrandchildren.

It is clear that the high dependence of human off-spring alone is not sufficient to explain menopause.Women undergo menopause about the time thatthey have reproducing daughters. If offspring needwere the sole driver, selection would more likely fa-vor “helping at the nest” by adult daughters and sons(a very common pattern among nonhumans) ratherthan reproductive cessation by the older female. Af-ter all, an individual is related to its sibling by anr of 0.5 if the two siblings share the same motherand father. Holding all else constant, an individualshould be indifferent between direct reproductionand helping her mother produce a sibling. Thus, ifchildren need additional investment, why is it thatyoung females (and males) do not defer reproduc-tion to help their mothers reproduce rather thanvice versa? Another way to frame the question aboutmenopause is to ask, “Why should women cease toreproduce and help descendants, instead of contin-uing to reproduce with the help of descendants?”An adaptive theory of menopause must specify theconditions that provide an answer to that question.

This paper presents a new learning- and skills-based theory for the joint evolution of humanmenopause and extended postreproductive life. Thetheory proposes that the role of knowledge, skill ac-quisition and transfers in determining economic pro-ductivity, and resource distribution is the distinctivefeature of the traditional human ecology that is respon-sible for the evolution of menopause. Moreover, we ar-gue that the traditional hunter-gatherer pattern ofproduction, reproduction, and parental investmentdepends fundamentally on a cooperative division oflabor between men and women. The theory there-fore proposes that in addition to female menopause,male reproductive cessation and postreproductiveinvestment in descendants is a fundamental charac-teristic of humans living in traditional foraging andsimple forager-horticultural economies. The theorybuilds on existing ideas—specifically the Grand-mother and Mother hypotheses12,13—in proposingthat menopause and the decrease in fertility withage that precedes it are evolved human traits thathave been maintained by selection because womenwill leave more descendants by ceasing to reproduceand investing in existing descendants. However, thespecific causal hypotheses that the theory integratesare new.

The paper begins with a brief presentation of thetheory, followed by a discussion of the evidenceupon which the theory is built. We begin with acomparative analysis of chimpanzee and human fe-male reproductive senescence. We then examine theage-specific fertility of men and the likelihood ofreproducing following menopause of wives. This isfollowed by behavioral evidence concerning foodproduction and resource transfers across genera-tions by women and men. The next section exam-ines the total expected net caloric consumption offamilies as it varies over the life cycle, then simulatesthe caloric effects of adjusting the age schedule ofwomen’s fertility, delaying the onset of menopause.The paper concludes by linking these observationsto the theory, and discussing directions for futuretheoretical and empirical research.

A “learning” theory of human reproductivedecline and cessation

Although human foragers have lived in virtually allthe world’s terrestrial habitats, they always occupyone extreme feeding niche, eating the highest qual-ity, most nutrient dense, and difficult to acquireplant and animal foods in their environment.1,14

More than any other organisms, humans rely onbrain-based skills and knowledge to acquire foodfrom the environment. Those mental abilities com-bine with physical abilities—such as strength, co-ordination, and balance—to determine the rate ofenergy acquisition per unit time. In a series of pa-pers, we have shown that peak physical condition inhumans occurs in the early to mid-twenties, but thatpeak economic productivity does not occur until af-ter age 40. This is due to the fact that skill acquisitionand learning continue to increase after peak physicalcondition is reached. Thus, peak economic produc-tivity between 40 and 50 years of age can be morethan four times as high as at age 20. After age 50,however, declines in physical condition begin to out-pace gains from learning and economic productiv-ity, and people cease to be net producers by aroundage 70 (see Refs. 1 and 5 and section “Evidence,”for supporting data for these claims). Our theoryproposes that the nature of the high-skill humanforaging niche has a series of implications which,taken together, disfavor old-age reproduction andfavor old-age production and kin investment forboth women and men, and thus drive the evolutionof human reproductive decline and menopause.

2 Ann. N.Y. Acad. Sci. xxxx (2010) 1–13 c© 2010 New York Academy of Sciences.


1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

Kaplan et al. Learning and human menopause

Most theories of menopause and the empiri-cal tests they stimulate estimate the age-specificcost of reproduction by the probability of dyingin childbirth.15–18 We propose that the cost of re-producing at advanced ages also includes increasedrisks of future mortality and reduced expected fu-ture productivity due to maternal depletion.19 Forexample, maternal immune responses are loweredduring pregnancy;20 as women age and experienceimmunosenescence, the costs of immunosuppres-sion are likely to increase. The energetic costs of lac-tation also probably occupy a greater proportion of awoman’s physiological reserves as she ages. For thesereasons, the cost of reproduction, both in terms offuture mortality and future economic production,is likely to be higher for a 45-year-old woman thanfor her 20-year-old daughter.

Although most species are likely to evidence in-creasing costs of reproduction with age, late-age re-production may be particularly costly for humans.Because human productivity is determined by bothphysical condition and long-developing skills, it ismore important to survive long enough and main-tain good enough condition to reap the rewards ofearlier investments in skill development. This is ac-complished by favoring somatic maintenance (andthus future production) over reproduction as thebody begins to age.

The payoff to late-age reproduction in humans isalso reduced by declining oocyte quality with age.There is significant evidence that oocyte quality de-clines with age in most mammals.21,22 Because hu-man offspring require extraordinary levels of invest-ment to reach independence, the cost of continuingto reproduce from a deteriorating stock of oocytesshould weigh more heavily in the human case thanfor most other species. An older mother produc-ing highly dependent offspring may either: (a) riskinvestment in particularly low-quality offspring; or(b) ensure that she produces only sufficiently high-quality offspring, either by investing more energyin maintaining the quality of her oocytes, or by be-ing more selective in allowing oocytes to implantor come to term. All of these options entail ener-getic costs, lost investments, or reduced fertility forolder females that should be greater in species withheavier parental investment. Evidence presented byEllison23 and Haig24 suggests that much of the bur-den in maintaining pregnancy prior to implanta-tion depends on chemical signals produced by the

embryo to maintain the corpus luteum and spurprogesterone production. They suggest that mater-nal physiology utilizes these signals to detect qualitydifferences in embryos and terminate low-qualitypregnancies. We propose that human reproductivephysiology may be particularly sensitive to embryoquality, and employ a more stringent selective sieveto prevent inferior embryos from implantation. Par-ticularly long-lived animals may additionally facegreater relative declines in oocyte quality over thelifespan (see Refs. 21 and 22 for reviews), whichwould also lower the returns to direct reproductionat advanced ages in humans.

Although the returns to late-age reproduction arereduced, the returns to old-age kin investment areincreased for humans relative to other animals. Be-cause the skills required for efficient food produc-tion take time to learn, children in foraging societiesdo not produce as much food as they consume un-til they are 18–20 years of age.1 This means thatthey must rely on subsidies from other individu-als. As the number of overlapping dependents ina young mother’s household grows with each birth,total caloric need is expected to outpace a single cou-ple’s combined productivity, creating a demand forcalories from sources outside the immediate house-hold (see Ref. 25 and section “Evidence”). Olderkin enjoying high levels of learning-based produc-tivity and facing increasing costs of direct reproduc-tion are in a prime position to meet this demand.This is true to a greater extent for humans thanfor most other mammals as a result of the life his-tory characteristics—high productivity late in life,high offspring need, and the simultaneous depen-dency of multiple offspring—which coevolved withthe skills-based human foraging strategy.

Finally, the skills-based foraging niche also pro-vides the conditions that lead most men in forag-ing societies to undergo reproductive cessation atthe same time as their wives. The skills-based econ-omy of humans is associated with unusually highmale energetic investment in offspring. In fact, menprovide the majority of energetic support for repro-duction in most hunting and gathering groups.26,27

Despite the high need for protein and lipids tosupport brain growth during development, themobility, danger, and long-term skill investmentsinvolved in human hunting make it largely incom-patible with the primate female’s evolved commit-ment to carrying (rather than caching) infants and

Ann. N.Y. Acad. Sci. xxxx (2010) 1–13 c© 2010 New York Academy of Sciences. 3


1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


lactation-on-demand. This generates a complemen-tarity between male and female inputs into offspringsuccess, a sex-specific specialization in hunting bymen, and high returns to male parental investment.Woman, in turn, specialize in a mix of childcare andforaging for plant resources.

The returns to male parental investment and thelong overlapping dependence of children interact inproducing a dominant pattern of long-term pair-bonding and male reproductive cessation in tradi-tional foraging societies. Given that children remaindependent after their younger siblings are born, menand women in foraging societies face higher costsfrom switching mates than in many other species.A mother who begins a new union often suffers re-duced paternal investment from the father of herprevious children. Conversely, for a woman who isabout to initiate reproduction, a man who has chil-dren from a previous union is less attractive becausehe already has vessels in which to invest. Considera 20-year-old woman who is about to begin repro-ducing. For her, a 50-year-old man is less attractivethan a 25-year-old, even though the older man maycurrently be more economically productive. The 50-year-old has two disadvantages: first, he already hashis peak dependency load of existing children; sec-ond, his food production will decrease in the futureand his mortality risk will increase. If the 20-year-old prefers to have all her children with one man, theyounger man is preferable, because of his expectedfuture contributions. This also implies that oldermen, who also face a trade-off between investingin existing children and grandchildren and seekinga new mate, most often “choose” to remain mar-ried and cease reproducing when their wives reachmenopause.28

Our theory is that these altered age-specificbenefits and costs of fertility, production andkin-investment—which derive from the special-ized skills-based foraging niche and its attendantshift in economic productivity toward older ages—combine to favor “early” reproductive cessation inboth men and women. This pattern has only evolvedonce. Even in those toothed whales that evidencefemale menopause, there is no such equivalent inmales. In those species, males typically have muchshorter lifespans than females and do not invest inoffspring.29 Humans are an outlier species in manysenses, from brain size to lifespan to menopauseto male parental investment. Special conditions are

necessary to produce such an outlier. The com-bination of a brain-based, knowledge-, and skill-intensive foraging niche with a primate heritage se-lected for this complex of traits (large brains, longlifespan, long offspring dependence, high selectiv-ity of oocyte quality, high male parental investment,and bisexual reproductive cessation).

In a recent paper, Kaplan and Robson30 presenta formal bioeconomic model for the evolution ofaging. They show that reproductive cessation can beoptimal prior to the optimal time to cease invest-ing in mortality reduction and future longevity. TheKaplan–Robson model does not include all the con-siderations elaborated earlier, but provides an an-alytical result demonstrating the conditions underwhich menopause can evolve by natural selection.It shows that if (a) the energetic costs of repro-duction increase with declining physical conditiondue to senescence, (b) economic transfers can allowsurplus productivity at one point in the life courseto be utilized at another point in the life course,and (c) individuals remain economically produc-tive after reproductive cessation, then there is anage at which fitness—measured in terms of the in-stantaneous growth rate, r, of the lineage—can bemaximized by reproductive cessation and the alloca-tion of remaining resources to mortality reduction,physical maintenance, and intergenerational trans-fers. The present theory is based partially on theinsights derived from that formal model.

The remaining sections of the paper will focus onthe empirical evidence related to the theory.

Evidence

Chimpanzee and human reproductive declineand its link to somatic senescenceThis section provides a comparative analysis ofchimpanzee and Tsimane fertility. A recent analy-sis by Emery Thompson et al.31 showed that whilemean chimpanzee fertility rates decline toward theend of life, females in good physical condition showno significant fertility decline with age (Fig. 1, panelA, adapted from Ref. 31). Among females aged25 and older, healthy individuals have significantlyhigher fertility than females who died within 5 yearsof the birth or risk year considered. Their findingssuggest that chimpanzee reproductive senescence istightly linked to somatic senescence and vulner-ability to mortality. Using a similar approach for



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


Figure 1. Impact of physical condition on chimpanzee andhuman fertility rates. (A) Probability of giving birth among fe-male chimpanzees, stratified by those who died within 5 yearsof giving birth, and those robust enough to live at least fiveadditional years following a birth. Adapted from Ref. 31. (B)Probability that a Tsimane woman gives birth using prospectivedata collected from 2002 to 2008, stratifying women into threegroups based on their baseline body mass index (BMI). (Sampleincludes 1267 females between the ages of 5 and 59 and rep-resents a total of 3121 observation years. Because of the rapidchange in BMI across adolescence, females under age 20 wereseparated into BMI terciles within 1-year age intervals, whereasolder women were separated into BMI terciles within 5-year ageintervals.)

traditional humans, we expect to see a decouplingof somatic and reproductive senescence.

To compare a traditional human case with EmeryThompson et al.’s results, we performed a prospec-tive analysis of the effect of physical condition, rep-resented by body mass index (BMI), on age-specificfertility among Tsimane women. We examined theprobability of a live birth occurring in each fullcalendar year following a woman’s first nonpreg-nant BMI measure based on census data collectedbetween 2002 and 2008. Panel B shows the meanfertility of Tsimane women by age divided into low,

Table 1. GEE logistic model of older women’s likelihoodof giving birth by BMI tercile (N = 537 person-yearsacross 224 women aged 35–54 years)

Variable B SE Wald � 2 P

Intercept 4.723 0.9990 22.348


1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


society is 0–4%, and in most societies fewer than10% of marriages are polygynous.33 One of theconsequences of monogamy is reproductive cessa-tion among men after their wives reach menopause.For example, Ache foragers have high initial divorcerates when they are young; nevertheless, 90% of menwho had more than one child with a woman did notreproduce after their wives reached menopause.34

Tsimane demographic data show that 90% of Tsi-mane men whose wives reached menopause did notreproduce again after their wife’s last birth. Of the10% who did reproduce, half (5.2%) were polyg-ynously married and had a child with a youngerco-wife, still within the bonds of marriage. The re-mainder had affairs outside of marriage (3.1%) orreproduced after the wife’s death (1.5%). Given thatsome men at risk of reproducing after their partnerreached menopause are still alive and may repro-duce in the future, we conducted a survival analysisof male reproduction following menopause. Fromthe survival curve (Fig. 2, panel A) it is evident thatthe greatest chance of reproduction is in the first 5years after the wife’s last child, consistent with thepattern of polygynous men reproducing with theyounger co-wife. Because the younger co-wives wereoften reaching middle age as well, most of these menonly reproduced once after their first wife reachedmenopause.

The linkage of men’s reproductive schedules withwomen’s can also be seen from the age-specific fer-tilities of the two sexes. The male curve is shiftedto the right of the female curve by about 5 years,consistent with the age differences among spouses(Fig. 2, panel B). The tail of the male curve stretchesout a bit from the female curve due to some menbeing more than 5 years older than their spouseand the few men who reproduce after their wife’smenopause. The male and female curves for ex-pected future fertility (i.e., reproductive value) arestrikingly similar, after age differences in marriageare taken into account (Fig. 2, panel C).

Physical condition, age profiles of productivityand intergenerational transfersAge profiles of productivity and intergenera-tional transfers among human hunter-gatherers andforager-horticulturalists have been documented ina series of publications.1,5,6,35,36 Those data showthat children remain dependent on their parentsuntil 18–20 years of age, with a peak dependency in

Figure 2. Male and female reproductive cessation among theTsimane. (A) Probability that a Tsimane man did not reproduceafter his wife had her last birth (see text for details). (Samplebased on retrospective reproductive histories including 188 finalfemale births, and 182 husbands; 6 were married polygynouslyto two wives.) (B) Age-specific fertility rates for Tsimane men andwomen, given in 5-year intervals. (Sample based on retrospectivereproductive histories of 431 women and 391 men coveringthe period 1950–2002; this includes 12,394 risk years and 2238births for women, and 12,514 risk years and 1943 births for menaged 15–64.) (C) Expected future fertility by age considers thecumulative sum of remaining future reproduction discountedby the probability of surviving to those ages. Survivorship dataare from Ref. 46.



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


early adolescence (from birth, caloric requirementsgrow faster than productivity until about age 12 or13 years).

Peak productivity in adulthood for both men andwomen occurs well after strength and physical con-dition peak. For example, among Ache foragers ofParaguay, men’s strength peaks at around 25 years ofage but both meat acquired and hunting return rates(amount acquired per hour spent hunting) peak be-tween 40 and 50 years of age.37 Data on strengthand hunting ability among Tsimane men show thesame pattern;5 moreover, skill in successfully pur-suing prey is the most important determinant ofhunting success.

Figure 3 shows the age-profiles of food produc-tion and consumption among Tsimane men andwomen. Both male and female production peaks af-ter age 40. Males produce as much as they consumeby about age 20 years, and females by age 28 years.The caloric deficit in childhood is compensated forby a caloric surplus in adulthood. The increase intotal food production is driven by two effects. First,there is an increase in efficiency (production perunit time) until the mid-40s. Second, there is a cor-responding increase in work effort, probably reflect-ing the increase in dependency load. The decrease inproduction with age is driven primarily by declinesin efficiency.

Figure 4 examines physical decline with age. PanelA shows the decline in strength with age for bothmen and women, and panel B shows pain-relatedfatigue among women while they pound rice. Bothfigures show considerable declines before peak pro-ductivity is reached.

Food sharing data allow for a more direct under-standing of inter-generational wealth flows. Figure 5plots the net transfers between pairs of related in-dividuals. Net transfers are calculated by taking thetotal amount of food given from individual A toindividual B and then subtracting the total amountgiven from B to A. Those amounts are derived fromdata on the consumers of food acquired by all familymembers. In the figure we present those nets fromfathers to children, mothers to children, grandfa-thers to grandchildren, and grandmothers to grand-children. Even though food is transferred in bothdirections between these pairs of individuals, thefigure shows that net transfers flow downward acrossgenerations. The downward flow from both mothersand fathers to their children continues into adult-

Figure 3. Age-specific caloric production and consumptionprofiles for Tsimane. Daily production was estimated for non-rice foods from interviews covering the previous 2 days of foodproduction. These data covered 43,656 sample days over 749 in-dividuals. Rice production was estimated from interviews con-cerning the amount of rice harvested in the previous year. Thesedata covered 589 individuals from the non-rice sample. Creditfor rice production was based on the proportional time spent infield labor from the 2-day production interviews. Loess curveswere fit over the daily non-rice and rice production rates byage and sex. The loess prediction curves were then summed toproduce the final curves. Consumption was estimated by firstcalculating the total energy expenditure (TEE) based on the age,sex, and weight of individuals.47 These were plotted by age andthe maximal consumption level was estimated to be 2770 calo-ries per day for the Tsimane. The TEE of each individual wasdivided by this to determine the proportion of consumer (POC).The number of production days sampled was multiplied by eachindividual’s POC and these were then summed to determine thetotal number of consumer days. The total production duringthe sampling period was then divided by this sum to determinethe true caloric intake of the maximal consumer, which equaled2661 calories per day. Each individual’s POC was multiplied by2661 to determine their consumption level. We then fit a loesscurve to the consumption levels by age and sex.

hood, even when their children become adults andhave children of their own. During the postrepro-ductive period of life (after age 45), transfers to ex-isting children dominate during middle age, withan increasing proportion of resources being trans-ferred to grandchildren, especially in the 60s. Nettransfers approach zero after age 70.

From Figure 5, it can also be seen that men trans-fer more calories to descendants than do women.However, women’s work in childcare, food pro-cessing, and household maintenance exceeds thatof men, and both sexes spend similar amounts oftotal time in work.38 This division of labor appearsto be universal in foraging societies, although the



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


Figure 4. Physical decline with age among Tsimane adults. (A)Strength is the sum of chest, shoulder, thigh, leg, and hand gripstrength, measured using the Lafayette Manual Muscle Testerand Smedley III Analog Grip Strength Tester. (Sample includes416 women and 428 men.) (B) Proportion of Tsimane womenthat report experiencing physical problems during rice pound-ing. Problems include arm and hand pain and poor eyesight.(Sample includes 104 women.)

relative energetic contributions of the two sexes varyaccording to local ecology. For the sample of 10 for-ager societies for which quantitative data exist, men,on average, acquired 68% of the calories and almost88% of the protein; women acquired the remaining32% of calories and 12% of protein.26

Transfers, calories, and menopauseJust as food production increases with age dur-ing the reproductive period, so too do the caloricdemands of dependents. In fact, the caloric demandson parents increase faster than does their produc-tivity.25 Figure 6, panel A, shows data from the Tsi-mane on the net productivity of parents, the netcaloric demands of children and the net surplus ordeficit of families as a function of a woman’s age.This figure shows that as families grow, their netdeficit increases, even though parental productiv-

Figure 5. Net caloric transfers between kin groups across threegenerations. Transfers were calculated using data from 3850 con-sumption events by 674 individuals during instantaneous scanobservations. The number of events in which individual A wasthe acquirer of food consumed by individual B divided by thetotal number of times individual A was named an acquirer wasinterpreted as the proportion of individual A’s production thatwent to individual B. (For foods with multiple acquirers, eachacquirer was assigned a proportion of credit, and these creditswere the values actually tallied.) The proportional distributionto each kin member was then calculated for each aggregated age-sex group, as the number of observations per individual was low.To capture the observed population age structure, each individ-ual alive in a 25-community census was assigned their age- andsex-specific daily production and proportional distribution lev-els. Daily production (represented in Fig. 3) was multiplied byproportional distribution to determine gross transfers. Thesewere summed in both directions for each kin dyad to determinenet transfers. Averages were then calculated for each age-sexgroup.

ity is increasing as well. Most importantly, it canbe seen that the deficit of growing families is com-pensated for by the net surplus of postreproductiveindividuals, who provision descendant kin (see alsoFig. 5).

Figure 6, panel B, simulates the caloric effectsof a delay in reproductive decline and menopause.The average net caloric demand of children infamilies headed by mothers in their 30s was ex-tended throughout the 40s; the net caloric demandof children beyond the 40s then continued 10 yearsbehind schedule (so that a 60-year-old was experi-encing the typical progeny dependency of a 50-year-old). In this case, the surplus provided by older peo-ple continues to be consumed by their dependentchildren. Figure 7 shows the cumulative net caloric



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


Figure 6. Parental production, children’s demands, and netfamily production. (A) Observed Tsimane pattern. These cal-culations are based on rice and non-rice production data from106 families, including 561 individuals who were sampled for anaverage of 67 days. (B) Simulation based on delayed menopause(see text). Daily caloric production and estimated consump-tion levels were summed for parents and children within fam-ilies. Parental production and child demand levels were thenaggregated over 5-year age intervals to calculate overall familybalances. Because cumulative mortality risk leads to a largernumber of families headed by younger parents than familiesheaded by older parents, those families that do survive mustproduce surpluses that more than make up for their previousdeficits. To take into account the effects of mortality, summednet balances for all age intervals were divided by the numberof families in the 15–19 age interval in an attempt to includein the denominator those families that were lost to mortality.Third-order polynomial curves were fit to the mean values ofeach age interval.

Figure 7. Cumulative net caloric balance of families given theTsimane sample and the delayed menopause simulation fromFigure.

balance, given the Tsimane sample and the delayedmenopause simulation. The “contrary to fact” de-layed menopause simulation shows genetic lineageswith a fertility and economic transfer regime thatwould be in net economic deficit, and thereforecould not support itself.

Discussion and conclusions

The theory presented in this paper builds on ex-isting adaptive hypotheses for the evolution ofmenopause. Most adaptive explanations have fo-cused on women’s roles as mothers and grandmoth-ers.4,12,13,15,16,39–41 The mother version emphasizesthe long period of juvenile dependence in humans,and its possible links to brain development.13,39 Ac-cording to this view, women stop reproducing at theexpected age at which they will be able to raise theirlast child to maturity before dying. If children re-quire 20 years of parental investment, then ceasingto reproduce at age 45 would make sense with anexpected age of death of 65, given survival to thatage. The grandmother version proposes that womencease reproducing in order to invest in grandchil-dren and help their daughters reproduce.12,40,41 Ac-cording to Hawkes et al., the strength-intensive na-ture of human foraging means that grandmotherscan acquire more than children and help provisionthem.



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


The present theory extends and modifies thoseideas in three important ways. First, it specifiesthe unique ecological conditions responsible for theevolution of the temporal separation of reproduc-tive and somatic senescence in humans. Second, itidentifies the important role that men play in thehuman life history strategy, and highlights that pre-mature reproductive cessation occurs in men as wellas women. Third, it addresses the question of why re-productive cessation and downward kin-investmentby elders should be favored over the alternativeof continued reproduction supported by younger,nonreproductive “helpers at the nest.”

The fundamental premise of our theory is thatthe role of brain-based skills and learning in eco-nomic production for both men and women dur-ing our evolutionary past is at the far end of theevolutionary continuum. Skills and knowledge areaccumulated throughout life, but physical condi-tion, from strength to immune function, declinesthroughout adulthood. As a result, human eco-nomic productivity—which is a function of bothcumulatively learned abilities and physical strengthand endurance—continues to increase even af-ter physical condition begins to decline. We pro-pose that this disjunction between economic andphysiological aging is the ecological key to humanmenopause. It simultaneously generates two condi-tions: (1) the physiological cost of later reproductionis rendered high for women, but their economic pro-ductivity and that of their husbands remains high;and (2) infants, juveniles, and adolescents produceless food than their growing bodies require.

A key feature of our theory is that it incorporatesdeclining oocyte quality and increasing physiolog-ical costs of reproduction. We argue that decliningoocyte quality with age has a larger impact on thetrade-off between reproducing and investing in de-scendants for humans than it does for chimpanzeesand most other mammals. Here we base our argu-ment on evidence showing that human reproductivephysiology is replete with mechanisms designed toensure that investment is curtailed in low-qualityoocytes and embryos from the follicular develop-ment phase through implantation and placental de-velopment.23,42 We further propose that it is thelength of human parental investment that selectedfor more stringent mechanisms of quality control, toallocate investment in high-quality offspring ratherthan “waste” it on low-quality offspring. The logical

extension of this argument is that across species, op-timal levels of selectivity with respect to oocyte qual-ity will increase as parental investment increases.

To this logic, we add the observation that thephysiological cost of reproduction increases withmaternal age, not only due to increased risks ofdeath in childbirth, but also due to maternal deple-tion that should affect survival and productivity atfuture ages. Given the disjunction between physi-ological aging and economic aging in humans andgiven the low productivity of children who have yetto learn, those physiological costs of reproductionshould weigh more heavily on women than on fe-males of other species. Human females have more togive by living longer, and thus should be less willingto risk death than other species.

Finally, our theory is two-sex, in that it proposesthat reproductive cessation occurs regularly amonghuman males as well as females. We argue that hu-man males also face a similar trade-off betweeninvestment in existing descendants and continuedreproduction. However, instead of facing increasedphysiological costs of reproduction with age, malesbecome less attractive as mates as they age. This isdue to two reasons. First, the importance of maleinvestment in offspring and the long term depen-dence of young in humans have resulted in long-term monogamous pair-bonds between men andwomen. Marriage to an older man is less attractiveto a young woman, because he is likely to die beforeshe completes her reproductive career. Second, giventhat older men are likely to have existing dependentyoung, their investment in children produced by anew marriage will likely be lower.

We compare chimpanzee decline in fertility withage to that of Tsimane females. The data com-piled by Emery Thompson et al. show that ag-ing chimpanzees in relatively good condition donot reduce fertility with age (or reduce fertility atlater ages).31 In contrast, the Tsimane data showthat while women with higher BMI, one measureof condition, do have higher fertility late in life,the decline in fertility with age in women is moredramatic than among healthy chimpanzees. This iseven more striking given that chimpanzee femalesin relatively good condition in their 30s are stillin worse condition than most women at that age,showing much more advanced signs of aging. Thissuggests that in response to declining oocyte quality,chimpanzee female reproductive physiology is less



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


selective than that of human females. We used datafrom the Ache and Tsimane data to show that menhave a low probability of reproducing after theirwives reach menopause, and that their age-relatedfertility decline is very similar to that of their wives.

We then presented evidence for both menand women that caloric production in traditionaleconomies does not exceed consumption untiladulthood, and that middle- and old-age adults pro-duce a caloric surplus. The data also show that in-tergenerational wealth flows are downward withinfamilies, and that both men and women invest inexisting children and then grandchildren after theycease reproducing. We then examined the joint eco-nomic and reproductive life histories of families byplotting the expected caloric demands of children,the net productivity of parents, and the resultinghousehold net caloric surplus (or deficit) as func-tions of a woman’s age. Those results showed thatenergetic burden of reproducing families producesa caloric deficit, which is compensated for by thecaloric surpluses of postreproductive individuals.

The impacts of reproductive cessation on calo-rie balance were then illustrated by simulating thecontinued reproduction of women at their 30-year-old rate until age 50. That simulation revealed thatall of the caloric surplus of older people would beconsumed by the extension of the reproductive pe-riod, and the whole family lineage would remain incaloric deficit.

The evidence presented in this paper cannot beconsidered a test of the theory, because the theorywas developed in response to the evidence. In addi-tion, most of the evidence is “circumstantial” in thatit is consistent with the theory, but does not demon-strate that the relative importance of foraging forhigh-quality resources using learning-intensive ac-quisition strategies is the primary ecological driverof menopause. Given that menopause has evolvedso infrequently and its particular two-sex formin humans is unique, ecological tests may proveelusive.

Nevertheless, individual components of the the-ory may be testable with comparative data. Forexample, there is a growing corpus of data onwhales that should allow for comparative tests. Sometoothed-whales show clear evidence of menopauseand a long postmenopausal lifespan in females.29,43

It is interesting to note that this branch of thecetacean line shows some broad similarities in its

foraging niche to humans. Killer whales, for ex-ample, demonstrate ecologically diverse foragingstrategies, strongly based on cultural traditionspassed through matrilineal kin from old to young(see Ref. 44 for a detailed review of learned culturaltraditions in cetaceans). Their foraging strategiesand brains also reflect complex cognitive pro-cesses.45 Similarly, comparative research on comple-mentarity, male parental investment and the link-age between male and female reproductive strategiescould test other components of the theory.

Future research should focus on investigatingthe costs of reproduction, selectivity with respectto oocyte quality, and economic transfers. We stillknow very little about maternal depletion in tra-ditional natural fertility societies, and how agingaffects the costs of reproduction in terms of fu-ture longevity, health, and productivity. Anotherarea for investigation is species differences in oocytequality control. Do humans and chimpanzees dif-fer in the selectivity of oocytes prior to ovulation,during fertilization and the completion of meiosis,or during embryogenesis? Food sharing in tradi-tional societies is also very complex. There are bothwithin and between family transfers, and the mix ofkinship, reciprocity and other factors determiningthose transfers is still poorly understood. A clearerunderstanding of those phenomena will help eval-uate the present theory and provide insight into theevolution of human reproductive cessation.

Acknowledgments

This research supported by the National ScienceFoundation (BCS-0422690) and the National In-stitute on Aging (R01AG024119–01). The authorsthank Sam Bowles, Ken Wachter, Melissa EmeryThompson, and Benjamin Hanowell for their veryhelpful comments, as well as collaborators on pre-vious papers, Jane Lancaster, Arthur Robson, KimHill, and Magdalena Hurtado, for their help in de-veloping the ideas and data presented here.

Conflict of Interest

The authors declare no conflicts of interest.

References

1. Kaplan, H., K. Hill, J. Lancaster & A.M. Hurtado. 2000. Atheory of human life history evolution: diet, intelligence, andlongevity. Evol. Anthropol. 9: 156–185.



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


2. Gurven, M. & H. Kaplan. 2007. Longevity among hunter-gatherers: a cross-cultural comparison. Popul. Dev. Rev. 33:321–365.

3. Konigsberg, L.W. & N.P. Herrmann. 2006. The osteologicalevidence for human longevity in the recent past. In TheEvolution of Human Life History. K. Hawkes & R. Paine,Eds.: 267–306. School of American Research Press. Santa Fe.

4. Hawkes, K. 2003. Grandmothers and the evolution of humanlongevity. Am. J. Hum. Biol. 15: 380–400.

5. Gurven, M., H. Kaplan & M. Gutierrez. 2006. How long doesit take to become a proficient hunter? Implications on theevolution of delayed growth. J. Hum. Evol. 51: 454–470.

6. Kaplan, H. 1994. Evolutionary and wealth flows theories offertility: empirical tests and new models. Popul. Dev. Rev. 20:753–791.

7. Thomas, F. et al. 2001. International variability of ages atmenarche and menopause: patterns and main determinants.Hum. Biol. 73: 271–290.

8. Austad, S. N. 1997. Why We Age. Wiley. New York.9. Armstrong, E. & D. Falk. 1982. Primate Brain Evolution:

Methods and Concepts. Plenum Press. New York.10. Sherman, P.W. 1998. The evolution of menopause. Nature

392: 759–761.11. Hamilton, W.D. 1964. The genetical evolution of social be-

havior. J. Theor. Biol. 7: 1–52.12. Hawkes, K. et al. 1998. Grandmothering, menopause, and

the evolution of human life histories. Proc. Natl. Acad. Sci.USA 95: 1336–1339.

13. Peccei, J.S. 2001. Menopause: adaptation or epiphe-nomenon? Evol. Anthropol. 10: 43–57.

14. Kaplan, H.S. 1997. The evolution of the human life course.In Between Zeus and the Salmon: The Biodemography ofLongevity. K. Wachter & C. Finch, Eds.: 175–211. NationalAcademy of Sciences. Washington, DC.

15. Hill, K. & A.M. Hurtado. 1991. The evolution of reproductivesenescence and menopause in human females. Hum. Nat. 2:315–350.

16. Rogers, A. 1993. Why menopause? Evol. Ecol. 7: 406–420.17. Sear, R., F. Steele, I. McGregor & R. Mace. 2002. The effects

of kin on child mortality in rural Gambia. Demography. 39:43–63.

18. Sear, R., R. Mace & I.A. McGregor. 2000. Maternal grand-mothers improve the nutritional status and survival ofchildren in rural Gambia. Proc. Biol. Sci. 267: 1641–1647.

19. Tracer, D.P. 1991. Fertility-related changes in maternal bodycomposition among the Au of Papua New Guinea. Am. J.Phys. Anthropol. 85: 393–405.

20. Heyborne, K. & R.M. Silver. 1996. Immunology of postim-plantation pregnancy. In Reproductive Immunology. R.A.Bronson, N.J. Alexander & D. Anderson, Eds.: 383–417.Blackwell Science. Oxford.

21. Ottolenghi, C. et al. 2004. Aging of oocyte, ovary, and humanreproduction. Ann. N.Y. Acad. Sci. 1034: 117–131.

22. vom Saal, F.S., C.E. Finch & J.F. Nelson. 1994. Natural historyand mechanisms of aging in humans, laboratory rodents andother selected vertebrates. In Physiology of Reproduction, Vol.2. E. Knobil, J. Neill & D. Pfaff, Eds.: 1213–1314. Raven Press.New York.

23. Ellison, P.T. 2001. On Fertile Ground: A Natural History ofHuman Reproduction. Harvard University Press. Cambridge,MA.

24. Haig, D. 1993. Genetic conflicts in human pregnancy. Q.Rev. Biol. 68: 495–532.

25. Gurven, M. & R. Walker. 2006. Energetic demand of multipledependents and the evolution of slow human growth. Proc.Biol. Sci. 273: 835–841.

26. Kaplan, H.S., K. Hill, A.M. Hurtado & J. Lancaster. 2001.The embodied capital theory of human evolution. In Re-productive Ecology and Human Evolution. P.T. Ellison, Ed.:293–317. Aldine de Gruyter. Hawthorne, NY.

27. Marlowe, F. 2001. Male contribution to diet and femalereproductive success among foragers. Curr. Anthropol. 42:755–760.

28. Winking, J., H. Kaplan, M. Gurven & S. Rucas. 2007. Whydo men marry and why do they stray? Proc. Biol. Sci. 274:1643–1649.

29. Foote, A. 2008. Mortality rate acceleration and post-reproductive lifespan in matrilineal whale species. Biol. Lett.4: 189–191.

30. Kaplan, H. & A. Robson. 2009. We age because we grow.Proc. Biol. Sci. 276: 1837–1844.

31. Emery Thompson, M. et al. 2007. Aging and fertility patternsin wild chimpanzees provide insights into the evolution ofmenopause. Curr. Biol. 17: 2150–2156.

32. Marlowe, F.W. 2004. The mating system of foragers in thestandard cross-cultural sample. Cross-Cult. Res. 37: 282–306.

33. Binford, L.R. 2001. Constructing Frames of Reference. Uni-versity of California Press. Berkeley, CA.

34. Hill, K. & A.M. Hurtado. 1996. Ache Life History: The Ecologyand Demography of a Foraging People. Aldine. Hawthorne,NY.

35. Gurven, M.D. & H.S. Kaplan. 2006. Determinants of timeallocation to production across the lifespan among theMachiguenga and Piro Indians of Peru. Hum. Nat. 17: 1–49.

36. Kaplan, H. & A. Robson. 2002. The emergence of humans:the coevolution of intelligence and longevity with intergen-erational transfers. Proc. Natl. Acad. Sci. USA 99: 10221–10226.

37. Walker, R., K. Hill, H. Kaplan & G. McMillan. 2002. Age-dependency in skill, strength and hunting ability amongthe Ache of eastern Paraguay. J. Hum. Evol. 42: 639–657.

38. Gurven, M. et al. 2009. A bioeconomic approach to mar-riage and the sexual division of labor. Hum. Nat. 20: 151–183.

39. Lancaster, J.B. & B.J. King. 1985. An evolutionary perspectiveon menopause. In In Her Prime: A View of Middle AgedWomen. V. Kerns & J.K. Brown, Eds.: 13–20. Bergen andGarvey. Garden City, NJ.

40. Williams, G.C. 1957. Pleiotropy, natural selection and theevolution of senescence. Evolution. 11: 398–411.

41. Hawkes, K., J.F. O’Connell & N. Blurton Jones.1989. Hardworking Hadza grandmothers. In ComparativeSocioecology of Humans and Other Mammals. V. Standen &R.A. Foley, Eds.: 341–366. Basil Blackwell. London.



1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52


42. Hunt, P. & T. Hassold. 2008. Human female meiosis: Whatmakes a good egg go bad? Trends Genet. 24: 86–93.

43. Ward, E.J. et al. 2009. The role of menopause and reproduc-tive senescence in a long-lived social mammal. Front. Zool.6: 1–10.

44. Rendell, L. & H. Whitehead. 2001. Culture in whales anddolphins. Behav. Brain Sci. 24: 309–382.

45. Marino, L. et al. 2007. Cetaceans have complex brains forcomplex cognition. PLoS Biol. 5: 0966–0972.

46. Gurven, M., H. Kaplan & A. Zelada Supa. 2007. Mortalityexperience of Tsimane Amerindians: regional variation andtemporal trends. Am. J. Hum. Biol. 19: 376–398.

47. FAO. 2001. Human energy requirements. Food and Nutri-tion Technical Report Series.


Date post:	21-Oct-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

Learning, menopause, and the human adaptive complexphooper/KaplanEtAl2010_Menopause.pdfin...

Documents