Sibling Competition & Growth Tradeoffs. Biological vs. Statistical Significance
Sibling Competition & Growth Tradeoffs. Biological vs. Statistical Significance
Karen L. Kramer 0 1 2
Amanda Veile 0 1 2
Erik Otárola-Castillo 0 1 2
0 1 Department of Anthropology, University of Utah, Salt Lake City, Utah 84112, United States of America, 2 Department of Anthropology, Purdue University , West Lafayette, Indiana , United States of America
1 Data Availability Statement: Data are available from PURR-Purdue University Research Repository , DOI:10.4231/R7J964B4
2 Editor: Rebecca Sear, London School of Hygiene and Tropical Medicine, UNITED KINGDOM
Early childhood growth has many downstream effects on future health and reproduction and is an important measure of offspring quality. While a tradeoff between family size and child growth outcomes is theoretically predicted in high-fertility societies, empirical evidence is mixed. This is often attributed to phenotypic variation in parental condition. However, inconsistent study results may also arise because family size confounds the potentially differential effects that older and younger siblings can have on young children's growth. Additionally, inconsistent results might reflect that the biological significance associated with different growth trajectories is poorly understood. This paper addresses these concerns by tracking children's monthly gains in height and weight from weaning to age five in a high fertility Maya community. We predict that: 1) as an aggregate measure family size will not have a major impact on child growth during the post weaning period; 2) competition from young siblings will negatively impact child growth during the post weaning period; 3) however because of their economic value, older siblings will have a negligible effect on young children's growth. Accounting for parental condition, we use linear mixed models to evaluate the effects that family size, younger and older siblings have on children's growth. Congruent with our expectations, it is younger siblings who have the most detrimental effect on children's growth. While we find statistical evidence of a quantity/quality tradeoff effect, the biological significance of these results is negligible in early childhood. Our findings help to resolve why quantity/quality studies have had inconsistent results by showing that sibling competition varies with sibling age composition, not just family size, and that biological significance is distinct from statistical significance.
Competing Interests: The authors have declared
that no competing interests exist.
The effects of family size on offspring quality, often measured as offspring survival, growth,
health or reproductive outcomes, have been the focus of many human life history studies [
Because higher fertility can lower the availability of time, food and resources per offspring, a
negative relationship between the number and quality of offspring is theoretically predicted
]. Results, however, have been mixed. Negative associations between family size and child
quality have been demonstrated in a number of contemporary and historic populations [
Other studies demonstrate a tradeoff between the number and quality of children, but only
under circumstances where resources are limited [
]. Still other studies find either a positive
], or no relationship [
] between fertility and child quality.
Empirical results may be inconsistent for several reasons. First, intra-population phenotypic
variation in parental condition can pose conceptual and methodological challenges that
obscure quantity/quality tradeoffs [
]. For example, highly fit parents can produce large
numbers of offspring that are also of high quality. This is particularly problematic with
largescale national or regional-level data where within-population phenotypic variation is
substantial . Second, tradeoffs may be mediated by differences in sibling competition across
children’s developmental stages [
]. In high-fertility, traditional societies, younger and older
siblings, or male and female children, may have different, and even opposing effects on
postweaning childhood growth [
]. Consequently, family size as an aggregate variable may
obscure the source of sibling competition and confound tradeoff costs. Third, many study
designs fail to distinguish biological from statistical significance [
]. While family size
may be a statistically significant predictor of growth outcomes, the effect on growth may not be
biologically meaningful in terms of health and fitness. Indeed, substantial variation exists in
population growth trajectories and adult body sizes that may or may not correspond to
biological fitness ([
]see S1 Text).
In this paper we first outline the potentially different influences that younger and older
siblings may have on child quality. We then use linear mixed models to evaluate both the
statistical and biological significance that family size, younger siblings and older siblings have on
Maya children’s growth. We conclude by discussing the importance of disaggregating family
size, and considering biological significance and population-specific growth metrics when
comparing growth outcomes. Our goal is to highlight a number of methodological issues that
may help resolve why family size effects on child quality have had mixed empirical support.
Competition with younger siblings
Maternal investment is necessary for infant survival in all but the most modern of human
societies. Breast milk immunity, nutrients and hormones, and intensive direct maternal care buffer
nursing children from infection and nutritional disruption [
]. In most cases, maternal
lactation is not substitutable and the time mothers allocate to direct infant care is relatively
consistent cross culturally [
]. Mothers with both nursing and weaned children are challenged to
simultaneously care for infants while also spending time in economic activities that benefit
older children. In societies where maternal time allocation has been documented, mothers
balance these competing demands by maintaining direct childcare but reducing time spent in
either domestic activities, food production (foraging or field work) or caretaking weaned
]. While intensive maternal focus limits the infant’s exposure to sibling
competition, the weanling faces a variety of new challenges.
In natural fertility populations, children typically are weaned following a subsequent
maternal pregnancy or the birth of a younger sibling. Weanlings lose the protective nutritional and
immunological buffer of breast milk, and maternal attention shifts from the penultimate child
to the youngest child. Weanlings, who for the first time are relying only on their own immune
systems, are susceptible to new health-related growth challenges [
] as maternal care is
replaced or supplemented by others and they come into contact with an expanded social
sphere. Increased social contact and subsequent disease exposure may help explain why
stunting often occurs among recently weaned children [
]. Thus, because maternal care is
diverted away from weanlings, younger siblings may pose a unique threat to a young child. The
2 / 17
effects of young siblings should be particularly pronounced when fertility is high, birth
intervals are short and a child has been recently displaced by a newborn.
Competition with older siblings
Although it is clear why younger siblings may be a competitive hazard to a young growing
child, several opposing influences affect whether older siblings are a disadvantage or advantage.
Older siblings may present a negative influence if disease exposure and morbidity rates
correlate with family size [
]. If severe enough, high communicable disease loads can
compromise growth . Epidemiological risks of growing up in a large family, however, likely
attenuate with age as a child’s immune system matures. Older siblings also may have a negative
impact in societies where wealth is generated through land acquisition, herd size, wages, or
other divisible forms of income. Under these circumstances, a larger family may dilute
resources available per child and older siblings may be a source of competition.
In other cases, however, older siblings may be an advantage. In many traditional foraging
and agricultural societies, wealth is mediated by the size of the household labor force, which
directly impacts the resources that it can produce [
]. In societies where older siblings
contribute to household production and have a positive economic value, they may add to the resources
available to young children. This is supported by the association between children’s economic
help, higher maternal fertility and improved sibling outcomes in a number of traditional
We take advantage of a large, longitudinal anthropometric dataset that tracks height and
weight measurements taken monthly from weaning to age five in a population of Maya
subsistence agriculturalists. We focus on post-weaning growth performance because 1) it is an
important proxy measure of biological fitness in traditional populations, 2) it is a life stage
when children are particularly vulnerable to sibling competition, and 3) early-life growth
deficits can have long-term health consequences [
]. 4) Lastly, focusing on early childhood in a
population with large families allows us to simultaneously evaluate the competitive effects of
younger and older siblings.
Because our test population is in the early stages of market integration and the subsistence
base is still largely agricultural, parents make relatively few cash investments in children [
In this context, we test three predictions. 1) While a tradeoff may be evident in very large
families, we expect that family size per se will not have a significant biological impact on young
children’s growth. 2) Because mothers focus time and energy on infant care, and weanlings are
inherently vulnerable, we expect that the number of younger siblings will have a negative effect
on growth outcomes. 3) Although Maya children were traditionally productive economic
contributors, because they currently spend more hours in school, we expect older siblings to
neither strongly add to nor detract from resources or time available to young children, and have a
neutral or negligible effect on children’s growth.
The study population
The Maya study community (n = 494) of subsistence maize agriculturalists is located in a
remote area of rural Campeche, Mexico [
]. Families make their living by small-scale farming,
and most food consumed is grown, although small amounts of cash may be generated through
3 / 17
maize and honey sales. Participation in wage labor is limited, and never by children, because of
distances needed to travel and the perceived low returns compared to agriculture.
Prolonged and intensive breastfeeding is the norm despite recent changes in health care
]. Supplemental foods are introduced at six months and children are fully weaned by
two and a half (median age = 2.58, n = 74, 95% CI = 2.46–2.69). As soon as children are able to
walk they are given great latitude to independently explore their environment. By age three or
four, children run errands, perform simple domestic tasks and take care of their younger
siblings. Compared to World Health Organization (WHO) standards, Maya children are short for
their age, but are well-nourished and healthy (S1 and S2 Figs; S1 Text for discussion). Schools
have been built in recent years, and most children ages six to twelve spend several hours a day
attending classes, with considerable recidivism at older ages.
The Maya study case is an ideal opportunity to evaluate sibling competition for several
reasons. 1) They are a high fertility population, and variation in numbers of older and younger
siblings is sufficient to evaluate the effects of competition. 2) Individual-level data on parental
anthropometrics and socioeconomic condition allow us to account for common measures of
phenotypic variation. 3) Longitudinal monthly data from weaning to age five permit us to
observe both short-term and long-term effects of sibling competition on child growth.
The height and weight of Maya children were collected at the beginning of each month as part
of a national child health surveillance program. Measurements were conducted in a clinic by a
community-based, physician-trained health promotor using government-provided standard
weigh scales and stadiometers. All community mothers participate in the program with few
missed monthly measurements. Children enter the program at birth and census out on their
fifth birthday. Seventy-five children ages 0 to 5 were measured monthly from 2007, when the
program was initiated, to 2011 (n = 1571 observations).
The children were measured an average of 20.9 times (SD = 9.2; Table 1). The health
promoter also keeps a record of births, and most children’s ages are accurate to the day. Children’s
birthdates and ages were cross-checked with annual censuses, including family size, the
number of older siblings and younger siblings, collected by Kramer and maintained in a
longitudinal database since 1992. Maternal heights were collected in 2010. This is a subsistence
agricultural economy, and wealth status is measured as the number of hectares a family has
under cultivation. During the same period that the children’s anthropometric data were
collected, each plot that a family has under cultivation was measured using GPS technology to
calculate total area under cultivation.
Written permits for research were secured from the local government and health promoter.
Consent on behalf of the children was obtained verbally from mothers (or fathers if mothers
were unavailable) during household visits. Written consent was waived because many parents
are illiterate and the research presented no more than minimal risk of harm to subjects and
involved no procedures for which written consent is normally required outside of the research
context. These research protocols and consent procedures were approved by Harvard
University’s Institutional Review Board and the University of Utah’s Institutional Review Board.
Sample considerations. Three children from one very large family (14 children total, 10
children living at home) were excluded from the analyses as an isolated case. Although the
young children were growing adequately (mean WAZ = 0.89, mean HAZ = 0.71 based on
measurements taken at age 3), the family’s size is not representative. When retained in the models,
the strength of the interactions between child size and age substantially increases; removing
them as outliers dampens the interaction effect.
4 / 17
19.3 ± 9.6 (1–30)
3.8 ± 0.7 (2.5–5)
91.1 ± 5.6 (76–105)
14.1 ± 1.9 (9.4–19.5)
3.7 ± 1.7 (1–8)
1.8 ± 1.9 (0–7)
0.82 ± 0.9 (0–3)
20.9 ± 9.2 (1–30)
3.8 ± 0.7 (2.5–5)
90.0 ± 5.7 (75–106)
13.6 ± 1.8 (9.4–19.5)
4.0 ± 2.1 (1–9)
2.2 ± 2.2 0–8)
0.76 ± 0.84 (0–3)
In the final analysis sample, family size, measured as number of offspring aged 15 and
younger living in the household including the child being measured, ranged from 1–9. We further
narrow the sample to observations of children between the age 2.5 (median observed age at
weaning) and 5.0 because breastfed infants tend to be buffered from the effects of sibling
competition, and variation in growth during the first two years of life is often a response to birth
weight rather than to exogenous factors [
]. The 2.5–5.0 age range allows us to focus on
family composition without these confounding influences.
We test our predictions by constructing a series of linear mixed models (LMM) fit by REML
(NLME package [
] in the R computing environment) [
]. Models were constructed for the
outcome variables height and weight. One set of models was constructed for the predictor
variable family size (the number of children 15 and younger living in the household, including the
child being measured). Because our primary interest is in the differential effects of family
composition, and because family size and older siblings are highly correlated, a second set of models
was constructed with predictors that disaggregate family size. We included both the number of
younger siblings and the number of older siblings in the same model, and hold one constant to
evaluate their respective effects separately. The value of the predictor variables is assessed each
time a child is measured. Family size, younger siblings and older siblings therefore change as
older siblings leave and new children are born into the family.
We accounted for the non-independent error across several factors by treating these models
as a nested repeated-measure LMM. Children were measured multiple times, therefore child ID
was treated as a random effect. In addition, many children have the same mother and their
growth responses will be correlated and not independent. To account for this, child’s id was
nested within mother’s id and treated as a random effect.
The full model for the predictor family size and the outcome variables height and weight
includes the random effects child’s id nested within mother’s id, the control variables age
(recorded at each measurement time point) and sex, and the covariates maternal height and
family wealth status to account for phenotypic variation (S2 Text). Three interaction terms—
age sex, age predictor, and sex predictor—were also added. The full model for the predictor
variables younger siblings and older siblings and the outcome variables height and weight
includes the same random effects and control variables as above, and sex and age interaction
terms for both predictors. Best-fit models for each predictor variable were selected by
comparing AIC values when variables were dropped from the full model using backward selection. If
5 / 17
the change in AIC between two models differed by 2, the model with fewer predictor
variables was selected.
WHO-defined stunting and the use of population-specific Z-Scores
The WHO defines stunting as children’s whose height is <-2 HAZ-scores below the mean for a
sample of global populations. This guideline is used to identify children who are potentially at
risk of failing to thrive. In several populations children below this threshold also have been
linked to negative health and fitness outcomes [
], including greater childhood morbidity
and mortality [
]. Short adult stature is also associated with reduced reproductive success;
shorter women have smaller babies and more obstetric complications [
], whereas shorter
men have fewer reproductive partners [
]. The detrimental downstream effects of stunting
are so pronounced that in 2012 the WHO adopted a resolution to reduce the number of
stunted under-five children by 40% by 2025 .
Mean WHO HAZ-scores for the 75 Maya boys and girls in the sample are -2.95 and -2.72,
respectively (S1 Text). Adult height for 20–40 year olds averages 143.2 cm for females and
156.1 cm for males. Although Maya children and adults would be classified as stunted by
WHO criteria, this does not meaningfully reflect Maya health or fitness for a number of
reasons. Children and adults are well nourished; BMIs for all children ages 10–20 are within the
50th to 90th percentiles, depending on age and sex (average male BMI for 10–20 year
olds = 20.4, SD = 3.0, n = 61; average female BMI = 21.9, SD = 3.5, n = 51). Surviving fertility
for women 40 and older is 6.4 (SD = 2.8, n = 60), and has remained unchanged over the last 20
years (t = 1.38, p = .1741, n = 52). Ninety-eight percent of children born survive to age 16 [
] (S1 Text). Birthweights documented since 2002 are within the WHO range of normal
(mean 3.04 kg, SD = 0.49, n = 109). Of this sample, 9% are low birth weight (LBW) babies, the
same percent of LBW reported for Mexico . Although Maya stature might increase with
different dietary inputs and life styles [
], these characteristics strongly suggest that while
short by WHO standards, their stature is not critically compromising to their health and
fitness. Consequently, we use Maya population-specific Z-scores as a more biologically relevant
metric of within-population comparisons of children’s growth in families of varied
] (see S1 Text for calculation of population-specific Z-scores). We display WHO
Zscores (S1 Fig, S2a and S2b Table), but only for the purpose of situating the Maya within a
Biological significance criteria
We emphasize that the magnitude of a parameter estimate needs to have a biological, not just a
statistical impact on growth. We differentiate between these measures of significance because
there is tremendous variation in population growth trajectories and size, not all of which may
have biologically meaningful fitness impacts [
]. We expect that if a parameter estimate is
statistically significant, but very small, it may be of little consequence to early childhood health or
fitness. We define biological significance using the following two criteria. 1) If the Maya
population-specific Z-score is <-2 for any given number of siblings, we consider the sibling effect to
be biologically significant. 2) If the relative change in Maya population-specific Z-scores is
associated with a >2 decrease as sibling number increases, we consider this to be biologically
We retain the WHO threshold of -<2 Z-scores but apply it to Maya-specific body size
distributions. We do this because children below the -2 Z-score threshold represent a substantial
deviation from mean body size within their population (the smallest 2.5%). We expect that
individuals falling below this threshold would be at increased health risks and longer-term
6 / 17
fitness compromises. Although a population-specific threshold for stunting is not often
employed in the growth literature (but see [
]), the WHO standard is not appropriate to
the Maya. Under such circumstances it is common practice for authors to assign a reasonable
effect size in the absence of established biological significance criteria . Finally, although
our criteria do not directly assess long-term fitness outcomes in Maya children, body size and
growth are common proxies of fitness used in life history analyses [
4, 23, 76
To establish biological significance, we use parameter estimates from the best-fit models (S1
Table) to calculate the predicted height (cm) and weight (kg) of Maya children at age 2.5 and
5.0. For the first criteria we computed the difference in predicted height and weight with each
additional increase in family size, older or younger sibling. Predicted height and weight values
were then converted into Maya population-specific Z-scores (see S1 Text). For the second
criteria we calculate the relative change in Maya population-specific Z-scores with each increase in
family size or addition in the number of younger and older siblings.
Family size effects on young children’s growth
The addition of a family member is slightly negatively associated with child height and weight
at 2.5 years of age, and the negative effects of additional family members on growth increase
with age (Fig 1, S1 Table Models 1a and 1b). The best-fit model’s parameter estimates predict
that the height and weight of a 2.5-year-old Maya child decreases by 0.5 cm and 0.14 kg (or
-0.17 HAZ and -0.11 WAZ), respectively, for each increase in family size (Fig 1, Table 2). The
predicted height and weight of a 5-year-old Maya child decreases by 0.7 cm and 0.32 kg, (or
-0.19 HAZ and -0.19 WAZ), respectively, for each increase in family size (Fig 1, Table 2, S2a
and S2b Table).
Younger sibling effects on young children’s growth
The addition of a younger sibling (holding older siblings constant) is positively associated with
child height at 2.5, but negatively associated with child height by age 5 (Fig 2, S2a Table). The
best-fit model’s parameter estimates predict that at 2.5, the height of a Maya boy and girl
increases by 1.3 cm and 0.7 cm (or 0.57 HAZ and 0.21 HAZ), respectively, per younger sibling.
Younger siblings become negatively associated with height by age 5, and the predicted height
of a Maya boy and girl decreases by 0.2 cm and 0.8 cm (or -0.06 HAZ and -0.22 HAZ),
respectively, per younger sibling (Fig 2, Table 2, S1 Table Model 2a).
The addition of a younger sibling (holding older siblings constant) is positively associated
with child weight at age 2.5, but negatively associated with weight by age 5 (Fig 2c, S1 Table
Model 2b). Though girls are lighter than boys, the magnitude of the interaction does not differ
between them. The best-fit model’s parameter estimates predict that at age 2.5, the weight of a
Maya child increases by 0.06 kg (0.05 WAZ) for additional younger sibling, while by age 5,
their weight decreases by 0.5 kg (-0.33 WAZ) with each additional younger sibling (Fig 2c,
Table 2). This is the greatest per-sibling effect we find on child growth, and accounts for a large
portion of the decrease in weight observed for family size (-1.9% per family member).
Older sibling effects on young children’s growth
The addition of an older sibling (holding younger siblings constant) is negatively associated
with child height at 2.5 years of age and the negative effect increases with age (Fig 3, S1 Table
Model 2a). Height of a 2.5-year-old Maya child decreases by 0.3 cm (-0.13 HAZ), and the
7 / 17
Fig 1. Growth performance stratified by family size (FS) for Maya children ages 2.5–5.0. Lines plot predicted (A) height (cm) and (B) weight (kg) by age
for boys and girls in a family with 1, 5 and 9 children (drawn from S1 Table Models 1a-b).
height of a 5-year old decreases by 0.6 cm (-0.14 HAZ) with each additional older sibling (Fig
3, Table 2).
The addition an older sibling (holding the number of younger siblings constant) is positively
associated with child weight at age 2.5, but starting at 3.0 years of age, younger siblings are
negatively associated with weight (Fig 3, S1 Table Model 2b). Although girls were lighter than
boys, the magnitude of the interaction did not differ between them. The best-fit model’s
parameter estimates predict that at 2.5 years of age, the weight of a Maya child increases by
0.03 kg (0.02 WAZ) for additional younger sibling, while by 5 years of age, weight of a Maya
child decreases by 0.15 kg (-0.09 WAZ) for each additional younger sibling (Fig 3, Table 2).
As per our criteria, detrimental effects of sibling composition on children’s growth is
considered biologically significant if: 1) the Maya population-specific Z-score is <-2; or 2) the relative
Boys & Girls
-0.6% (-0.5 cm)-0.7% (-0.7 cm)
1.5% (1.3 cm)-0.2% (-0.2
0.8% (0.7 cm)-0.8%
-0.4% (-0.3 cm)-0.6% (-0.6 cm)
Boys & Girls
-1.2% (-0.14 kg)-1.9% (-0.32
0.5% (.06 kg)-3.0% (-0.5 kg)
0.25% (.03 kg)-0.9% (-.15
8 / 17
Fig 2. Growth performance stratified by the number of younger siblings (YS) for Maya children ages 2.5–5.0. Lines plot predicted height (cm) in (A)
boys and (B) girls and predicted weight (kg) for (C) girls and boyswith 0, 1, 2, and 3 younger siblings (drawn from S1 Table Models 2a-d).
Fig 3. Growth performance stratified by the number of older siblings (OS) for Maya children ages 2.5–
5.0. Lines plot (A) predicted height (cm) for boys and girls, and predicted weight (kg) for (B) boys and girls
with 0, 4 and 8 older siblings (drawn from S1 Table Models 3a-c).
9 / 17
change in Maya population-specific Z-scores is associated with a >2 decrease as sibling
number increases. Based on these criteria, Maya children’s population-specific mean height and
weight Z-scores do not reach thresholds of biological significance for any value of family size,
nor for any number of younger or number of older siblings at age 2.5 or 5.0 (Table 3; S2a and
S2b Table). According to these criteria, siblings do not have a biologically meaningful
detrimental effect on young children’s growth in the Maya, even in very large families.
The quantity/quality tradeoff predicts that siblings in large families compete for limited
parental resources and consequently children’s growth is expected to be negatively associated with
family size. While several studies have found a negative family size effect on growth [
], others have not [
]. Using a large, longitudinal panel study and accounting for
differences in parental condition, we focus our analysis on young children where the negative
impacts of sibling competition are potentially most concentrated. While we find statistical
evidence of a quantity/quality effect, the biological significance of these results appears minimal
during early childhood, even at large family sizes. Maya children never fall below the <-2
criteria for population-specific Z-scores, nor do they lose >2 Z-scores at any number of siblings.
These results raise questions about methodological approaches to sibling competition, the
meaning of statistical versus biological significance in growth studies, and the use of standard
references when comparing within population growth variation.
We have suggested that in high fertility societies, family size can conflate the potentially
differential effects that younger and older children have on sibling competition. Nursing siblings,
who monopolize much of a mother’s time, may directly compete with recently weaned
children. This follows with our finding that competition with younger siblings poses the greatest
per-sibling compromise to young Maya child’s growth. This is consistent with the Godoy et al.
] study in which stunted Bolivian Tsimane children’s catch-up growth decreased with each
additional younger sibling. Similar results have also been reported for the Yanomamo, the
Hadza and the Ngandu [
11, 80, 81
]. In the Maya case, younger siblings have a more
pronounced effect on growth at age 5.0 than at age 2.5, suggesting that the buffering advantages of
breastfeeding persist for some time after weaning or that allocare is preferentially directed
We have proposed that older siblings may have either a positive or negative effect depending
on their economic value, the level of market integration and cash outlays or other divisible
forms of wealth that parents invest in children. As predicted, despite growing up in large
families, older Maya children negligibly affect their younger sibling’s growth performance. The
Maya are in the earliest stages of market integration, and older children neither draw down nor
augment household wealth. Here we parsed family size into younger and older siblings to
disaggregate different kinds of pressure on parental investment and competition among siblings.
In other ethnographic contexts, the potentially differential influence of siblings may be
meaningful disaggregated in other ways.
Biological vs. statistical significance
We emphasize that while family composition, or other exogenous factors, may be statistically
associated with growth, the question has to answered, is it biologically meaningful? While
biological significance is not addressed in most growth model results [
12, 25, 77
under10 / 17
reported aspect of growth analyses is critical to determine whether family size and sibling
competition actually compromise future health and reproductive outcomes.
We have used a cutoff of <-2 population-specific Z-scores and a decrease of >2
populationspecific Z-scores to assess biological significance. Given these criteria, our findings suggest that,
although the Maya children in our sample are small, this is not largely attributable to sibling
competition in early childhood. Further, their size appears not to be fitness compromising.
Both fertility and child survivorship are high compared to many traditional populations [
], and birth weights, the proportion of LBW births and BMI performance in older children
are all within normal ranges (see above). Although short maternal stature is associated with
obstetric complications in other populations [
], surviving fertility, detailed reproductive
histories and structured interviews with the community’s midwife and older women
corroborate that rates of maternal and infant mortality (one woman died in childbirth over the last 30
years)were low even before western biomedical care was available .
To address whether statistical effects are accompanied by biologically meaningful effect
sizes, we used a -2 Z-score threshold as a common indictor of substantial deviation from the
population mean for body size (the smallest 2.5%). We note however, that for other study
populations or research questions, a more or less conservative value may be appropriate. Indeed,
our longitudinal Maya life history project will help to determine a potentially more appropriate
population-specific effect size in the future.
While the WHO reference standards were recently adjusted to reflect broad patterns of global
child growth variation and are often used to gauge nutritional status [
], we use
populationspecific Z-scores for several reasons. The short stature of the Maya is not an indicator that they
have limited caloric intake, are in poor health or have compromised fertility as adults. Standard
references may not be sensitive to the range of healthy growth or genetic constraints in some
]. We would expect that the reaction norm for healthy growth to express
variation and be sensitive to ecological context. We include Z-scores derived from the WHO
standards for cross-cultural comparative purposes, but emphasize that a population-specific metric
11 / 17
is more appropriate for within group comparisons, in this case the effects of sibling
competition on growth. This approach has gained traction in recent years [
Lastly, our results raise a question about why we find minimal evidence for a
quantity/quality tradeoff in early childhood. Reasonably, the more offspring a mother has, the less time, food
or resources are available per individual. Among the Maya, as in most societies, investment in
children is not limited to parents, but is distributed across many helpers [
24, 25, 46, 83–87
Consequently the relationship between sibling competition, family composition and growth
outcomes is expected to be mediated by the availability of nonmaternal help [
]. We also
expect it to be mediated by how wealth is produced. Wealth in Maya society was traditionally
generated through household labor and the agricultural production of parents and children.
Beside their economic contributions, Maya children spend substantial time providing
childcare, especially to weanlings [
]. For these reasons, we expect that the economic value of
children and the time they spend in childcare are important determinants of whether older
siblings, and consequently family size, limit the per capita investment available for other
Among the Maya and other modernizing populations, traditional factors that affect the
quality of offspring (children’s economic value, living in extended families, the availability of
nonmaternal help and patterns of disease transmission) are in transition. Because parents can
distribute their time and resources among children in novel ways (e.g., education, cash inputs),
it is not surprising that quantity/quality tradeoff studies across diverse societies that differ in
their level of market integration are divided in their findings. Family size may well have a
negative impact on child growth under many circumstances. However, the relationship between the
number and size of offspring is likely context-specific. In those societies where wealth is
generated by family labor, children’s economic value is high and allocare is common, we would
predict that family size and older siblings are not significant predictors of negative child outcomes.
However, in societies where wealth is divisible (generated through wages, land or herd size),
the economic value of children is low or non-parental sources of help are limited, we predict a
quantity/quality tradeoff to be more evident, and older and younger siblings both a source of
Limitations. This study has a limited age focus on early childhood from the post-weaning
period to five years old. During this period we find minimal evidence of a biologically
meaningful quantity-quality tradeoff. Because of the short duration studied, we cannot directly compare
our results with studies that cover a broader range of child development stages. The negative
interaction of siblings and child age suggests that the negative relationship might increase over
time. In future studies, continued measurements of child growth in this population will allow
us monitor the longer term effects of sibling competition.
Our results show that young Maya children’s growth is not compromised in a biologically
meaningful way by sibling competition and growing up in large families. The quantity/quality
tradeoff is complex in humans because cooperation and the exchange of resources and
childcare necessary for growth and survival extend well beyond what parents provide. Family size
and sibling competition rather than having a universally negative effect on child quality, are
likely context-specific. Our findings suggest that 1) siblings of different ages can have different
effects on children’s growth, and family size as an aggregate variable may not capture this
distinction or the source of sibling competition. 2) Statistical significance may not reflect a
biologically meaningful effect size. In the case of this study and the criteria used, sibling competition
was not found to have biologically significant effects on young children’s growth. 3) This lead
12 / 17
to our final point that population-specific Z-scores for many traditional and transitioning
societies may be a more appropriate metric for research questions about determinants of
withinpopulation growth performance.
S1 Fig. WHO calculated Z-scores for Maya boys (n = 39) and girls (n = 36) ages 2.5–5.0 for
(A) height and (B) weight. Graph plots monthly measurements taken from 2007–2011.
S2 Fig. Population-specific Z-scores for Maya boys (n = 39) and girls (n = 36) ages 2.5–5.0
for (A) height and (B) weight. Graph plots monthly measurements taken from 2007–2011.
S1 Table. Best-fit models for height and weight and predictor variables. See main text for
explanation of how models were calculated.
S2 Table. (A) Predicted estimates from best-fit models for height (cm). Includes
population-specific Z-scores (Maya HAZ), WHO Z-scores (WHO HAZ), and Z-score changes (Maya
HAZ Δ and Who HAZ Δ) at age 2.5 and age 5.0 for each predictor variable (see main text and
S1 Text for explanation of calculation). For predictor younger siblings, boys and girls are
calculated separately because of the sex younger siblings interaction effect. (B) Predicted estimates
from best-fit models for weight (kg). Includes population-specific Z-scores, WHO Z-scores,
and Z-score changes at age 2.5 and age 5.0 for each predictor variable (see text and S1 Text for
explanation of calculation).
S1 Text. Constructing WHO and population-level Z-scores.
S2 Text. Fertility, maternal age, height and wealth status.
Much appreciation to the Maya for their ongoing willingness and patience to participate in this
study. We are particularly grateful to Maximiliano Moo Moo and Dra. Ada Fuentes for
diligently facilitating the local anthropometry program. We thank Russell Greaves for his
assistance in the field and Jeffrey Winking for giving us input on the manuscript. We appreciate the
very helpful comments from Ed Hagen and Emily Emmott. This research was supported by
NSF award #0964031.
Conceived and designed the experiments: KLK AV. Analyzed the data: AV EOC. Wrote the
paper: KLK AV. Managed data retrieval: KLK. Provided substantial editorial input: KLK AV
13 / 17
14 / 17
15 / 17
16 / 17
1. Borgerhoff Mulder M. Optimizing offspring: the quantity-quality tradeoff in agropastoral Kipsigis . Evolution and Human Behavior . 2000 ; 21 : 391 - 410 . PMID: 11146305
2. Kaplan H , Lancaster JB . The evolutionary economics and psychology of the demographic transition to low fertility . In: Cronk L, Chagnon N , Irons W , editors. Adaptation and Human Behavior: An Anthropological Perspective . New York: Aldine de Gruyter; 2000 . p. 283 - 322 .
3. Lack D. The significance of clutch size . Ibis . 1947 ; 89 : 302 - 52 .
Roff DA . The Evolution of Life Histories: Theory and Analysis . New York: T. Chapman Hall; 1992 .
5. Smith CC , Fretwell SD . The optimal balance between size and number of offspring . American Naturalist . 1974 ; 108 : 499 - 506 .
6. Walker RS , Gurven M , Burger O , Hamilton MJ . The trade-off between number and size of offspring in humans and other primates . Proceedings of the Royal Society , B. 2008 ; 275 : 827 - 33 .
7. Lawson DW , Alvergne A , Gibson MA . The life-history trade-off between fertility and child survival . Proceedings of the Royal Society , B. 2012 ; 279 : 4755 - 64 .
8. Meij JJ , van Bodegom D , Ziem JB , Amankwa J , Polderman AM , Kirkwood TBL , et al. Quality-quantity tradeoff of human offspring under adverse environmental conditions . Journal of Evolutionary Biology . 2009 ; 22 : 1014 - 23 . doi: 10 .1111/j.1420- 9101 . 2009 . 01713 . x PMID : 19298492
9. Penn DJ , Smith KR . Differential fitness costs or reproduction between the sexes . Proceedings of the National Academy of Sciences . 2007 ; 104 : 553 - 8 .
10. Strassmann BI , Gillespie B . Life-history theory, fertility and reproductive success in humans . Proceedings of the Royal Society of London B . 2002 ; 269 : 553 - 62 .
11. Hagen EH , Hames RB , Craig NM , Lauer MT , Price ME . Parental investment and child health in a Yanomamo village suffering short-term food stress . Journal of Biosocial Science . 2001 ; 33 : 503 - 28 . PMID: 11683222
12. Lawson DW , Mace R . Sibling configuration and childhood growth in contemporary British families . International Journal of Epidemiology . 2008 ; 37 : 1408 - 21 . doi: 10 .1093/ije/dyn116 PMID: 18577530
13. Gillespie D , Russell AF , Lummaa V . When fecundity does not equal fitness: evidence of an offspring quantity versus quality trade-off in pre-industrial humans . Proceedings of the Royal Society , B. 2008 ; 275 : 713 - 22 .
14. Borgerhoff Mulder M. The demographic transition: are we any closer to an evolutionary explanation? Trends in Ecology and Evolution . 1998 ; 13 : 266 - 70 . PMID: 21238295
15. Hill K , Hurtado AM . Ache Life History . New York: Aldine de Gruyter; 1996 .
16. Kaplan H , Lancaster JB , Bock JA , Johnson SE . Does observed fertility maximize fitness among New Mexican men? A test of an optimality model and a new theory of parental investment in the embodied capital of offspring . Human Nature . 1995 ; 6 ( 4 ): 325 - 60 . doi: 10 .1007/BF02734205 PMID: 24203123
17. Draper P , Hames R . Birth order, sibling investment and fertility amng Ju/Hoansi (!Kung) . Human Nature . 2000 ; 11 ( 2 ): 117 - 56 . doi: 10 .1007/s12110-000 -1016-0 PMID: 26193364
18. Jamison CS , Cornell LL , Jamison PL , Nakazato H . Are all grandmothers equal? A review and a preliminary test of the "Grandmother Hypothesis" in Tokugawa Japan . American Journal of Physical Anthropology . 2002 ; 119 : 67 - 76 . PMID: 12209574
19. Pennington R , Harpending H. Fitness and fertility among Kalahari! Kung. American Journal of Physical Anthropology . 1988 ; 77 : 303 - 19 . PMID: 3228170
20. Hesketh T , Qu JD , Tomkin A . Health effects of family size: cross-sectional survey in Chinese adolescents . Archives of Disease in Childhood . 2003 ; 886 : 467 - 71 .
21. Tada Y , Keiwkarnka B , Pancharuniti N , Chamroonsawaski K. Nutritional status of preschool children of the Klong Tbey slum , Bangkok. Southeast Asian Journal of Tropical Medicine and Public Health . 2002 ; 33 : 628 - 37 . PMID: 12693602
22. Pelto GH , Urgello J , Allen LH , Chavez A , Martinez H , Meneses L , et al. Household size, food intake and anthropometric status of school-age children in a highland Mexican area . Social Science & Medicine . 1991 ; 33 ( 10 ): 1135 - 40 .
23. Stearns SC . The Evolution of Life Histories . Oxford: Oxford University Press; 1992 .
24. Kramer K. The evolution of human parental care and recruitment of juvenile help . Trends in Ecology and Evolution . 2011 ; 26 ( 10 ): 533 - 40 . doi: 10 .1016/j.tree. 2011 . 06 .002 PMID: 21784548
25. Sear R , Allal N , Mace R . Family matters: kin, demography and child health in a rural Gambian community . In: Bentley GR , Mace R , editors. Substitute Parents: Alloparenting in Human Societies . New York: Berghahn Books; 2009 . p. 50 - 76
26. Johnson DH . The insignificance of statistical significance testing . The Journal of Wildlife Management . 1999 ; 63 ( 3 ): 763 - 72 .
27. Stephens PA , Buskirk SW , Martínez del Rio C. Inference in ecology and evolution . Trends in Ecology and Evolution . 2007 ; 22 ( 4 ): 192 - 7 . PMID: 17174005
28. Yoccoz NG . Use, overuse, and misuse of significance tests in evolutionary biology and ecology . Bulletin of the Ecological Society of America 1991 ; 72 ( 2 ): 106 - 11 .
29. Godoy R , Magvanjav O , Nyberg C , Eisenberg DT , McDade TW , Leonard WR , et al. Why no adult stunting penalty or height premium?: Estimates from native Amazonians in Bolivia . Economics and Human Biology . 2010 ; 8 : 88 - 99 . doi: 10 .1016/j.ehb. 2009 . 08 .002 PMID: 19766067
30. Kirchengast S. Differential reproductive success and body size in! Kung San people from northern Namibia . Collegium Antropologicum . 2000 ; 24 : 121 - 32 . PMID: 10895538
31. Sear R , Allal N , Mace R . Height, marriage and reproductive success in Gambian women . Research in Economic Anthropology. 2004 ; 23 : 203 - 24 .
32. Stulp G , Barrett L. Evolutionary perspectives on human height variation . Biological Reviews . 2016 ; 91 ( 1 ): 206 - 34 . doi: 10 .1111/brv.12165 PMID: 25530478
33. McDade TW , Worthman CM . The weanling's dilemma reconsidered: a biocultural analysis of breastfeeding ecology . Journal of Developmental & Behavioral Pediatrics . 1998 ; 19 ( 4 ): 286 - 99 .
34. Kramer KL . Cooperative breeding and its significance to the demographic success of humans . Annual Review of Anthropology . 2010 ; 39 : 414 - 36 .
35. Hames R. The allocation of parental care among the Ye'Kawana . In: Betzig L, Borgerhoff Mulder M , Turke P , editors. Human Reproductive Behavior . Cambridge: Cambridge University Press; 1988 . p. 237 - 51 .
36. Hurtado M , Hawkes K , Hill K , Kaplan H. Trade-offs between female food acquisition and child care among Hiwi and Ache foragers . Human Nature . 1992 ; 3 ( 3 ): 1 - 28 .
37. Marlowe F. A critical period for provisioning by Hadza men: Implications for pair bonding . Evolution and Human Behavior . 2003 ; 24 : 217 - 29 .
38. McDade TW . Life history theory and the immune system: steps toward a human ecological immunology . Yearbook of Physical Anthropology . 2003 ; 46 : 100 - 25 .
39. Gurven M. Infant and fetal mortality among a high fertility and mortality population in the Bolivian Amazon . Social Science & Medicine . 2012 ; 75 ( 12 ): 2493 - 502 .
40. Bodner C , Godden D , Seaton A . Family size, childhood infections and atopic diseases . The Aberdeen WHEASE Group. Thorax . 1998 ; 53 ( 1 ): 28 - 32 . PMID: 9577518
41. Ford AC , Forman D , Bailey AG , Goodman KJ , Axon AT , Moayyedi P. Effect of sibling number in the household and birth order on prevalence of Helicobacter pylori: a cross-sectional study . International Journal of Epidemiology 2007 ; 36 ( 6 ): 1327 - 33 . PMID: 17905807
42. McDade TW , Reyes-Garcia V , Tanner S , Huanca T , Leonard WR . Maintenance versus growth: investigating the costs of immune activation among children in lowland Bolivia . American Journal of Physical Anthropology . 2008 ; 136 : 478 - 84 . doi: 10 .1002/ajpa.20831 PMID: 18383156
43. Kramer KL . Maya Children: Helpers at the Farm . Cambridge: Harvard University Press; 2005 .
44. Bereczkei T. Kinship network, direct childcare, and fertility among Hungarians and Gypsies . Evolution and Human Behavior . 1998 ; 19 ( 5 ): 283 - 98 .
45. Bove RB , Valeggia CR , Ellison PT . Girl helpers and time allocation of nursing women among the Toba of Argentina . Human Nature . 2002 ; 13 ( 4 ): 457 - 72 . doi: 10 .1007/s12110-002 -1003-8 PMID: 26193090
46. Kramer K. Does it take a family to raise a child ? In: Bentley G , Mace R , editors. Substitute ParentsBiological and Social Perspectives on Alloparenting in Human Societies. Studies of Biosocial Society , vol 3 . New York: Berghahn Books; 2009 . p. 77 - 99 .
47. Lee RD , Kramer KL . Children's economic roles in the Maya family life cycle: Cain, Caldwell and Chayanov revisited . Population and Development Review . 2002 ; 28 ( 3 ): 475 - 99 .
48. Sear R , Mace R . Who keeps children alive? A review of the effects of kin on child survival . Evolution and Human Behavior . 2008 ; 29 : 1 - 18 .
49. Hagen EH , Barrett HC , Price ME . Do human parents face a quantity-quality tradeoff? Evidence from a Shuar community . American Journal of Physical Anthropology 2006 ; 130 : 405 - 18 . PMID: 16365856
50. Veile A , Kramer KL . Birth and breastfeeding dynamics in a modernizing indigenous community . Journal of Human Lactation 2014 ; 5 : 1 - 11 . doi: 10 .1177/0890334414557177
51. Espo M , Kulmala T , Maleta K , Cullinan T , Salin ML , Ashorn P . Determinants of linear growth and predictors of severe stunting during infancy in rural Malawi . Acta Paediatrica . 2002 ; 91 ( 12 ): 1364 - 70 . PMID: 12578296
52. Marins VMR , Almeida RMVR . Undernutrition prevalence and social determinants in children aged 0-59 months , Niteroi, Brazil. Annals of Human Biology . 2002 ; 29 ( 6 ): 609 - 18 . PMID: 12573077
53. Pinheiro J , Bates D , DebRoy S , Sarkar D. nlme: Linear and Nonlinear Mixed Effects Models . See http:// cran.r-project.org/web/packages/nlme/nlme.pdf. 2012 .
54. R Development Core Team. R: a language and environment for statistical computing: R Foundation for Statistical Computing; 2012 .
55. Adair LS , Fall CHD , Osmond C , Stein AD , Martorell R , Ramirez-Zea M , et al. Associations of linear growth and relative weight gain during early life with adult health and human capital in countries of low and middle income: findings from five birth cohort studies The Lancet . 2013 ; 382 : 525 - 34 .
56. Black RE , Allen LH , Bhutta ZA , Caulfield LE , De Onis M , Ezzati M , et al. Maternal and child undernutrition: global and regional exposures and health consequences . The Lancet 2008 ; 371 ( 9608 ): 243 - 60 .
57. Black RE , Victora CG , Walker SP , Bhutta ZA , Christian P , de Onis M. Maternal and child undernutrition and overweight in low-income and middle-income countries . The Lancet . 2013 371( 243 - 60 ). doi: 10 . 1016/S0140- 6736 ( 13 ) 60937 -X
58. Martorell R , Horta BL , Adair LS , Stein AD , Richter L , Fall CH , et al. Weight gain in the first two years of life is an important predictor of schooling outcomes in pooled analyses from five birth cohorts from lowand middle-income countries . The Journal of Nutrition . 2010 ; 140 : 348 - 54 . doi: 10 .3945/jn.109.112300 PMID: 20007336
59. Rice AL , Sacco L , Hyder A , Black RE . Malnutrition as an underlying cause of childhood deaths associated with infectious diseases in developing countries . Bulletin of the World Health Organization 2000 ; 78 : 1207 - 21 . PMID: 11100616
60. Walker SP , Chang SM , Powell CA , Simonoff E , Grantham-McGregor SM . Early childhood stunting is associated with poor psychological functioning in late adolescence and effects are reduced by psychosocial stimulation . Journal of Nutrition . 2007 ; 137 : 2464 - 9 . PMID: 17951486
61. Mahmood TA , Campbell DM , Wilson AW. Maternal height, shoe size, and outcome of labour in white primigravidas: a prospective anthropometric study . British Medical Journal . 1988 ; 297 : 515 - 7 . PMID: 3139180
62. Smith GC , White IR , Pell JP , Dobbie R . Predicting cesarean section and uterine rupture among women attempting vaginal birth after prior cesarean section . PLoS Medicine . 2005 ; 2 (9):e252 . PMID: 16146414
63. Wells JCK , DeSilva JM , Stock JT . The obstetric dilemma: an ancient game of Russian roulette, or a variable dilemma sensitive to ecology? American Journal of Physical Anthropology . 2012 ; 149 ( S55 ): 40 - 71 .
64. Nettle D. Height and reproductive success in a cohort of British men . Human Nature . 2002 ; 13 ( 4 ): 473 - 91 . doi: 10 .1007/s12110-002 -1004-7 PMID: 26193091
65. Pawlowski B , Dunbar RIM , Lipowicz A . Evolutionary fitness: tall men have more reproductive success . Nature 2000 ; 403 ( 6766 ): 156 .
66. de Onis M , Dewey KG , Borghi E , Onyango AW , Blössner M , Daelmans B , et al. The World Health Organization's global target for reducing childhood stunting by 2025: rationale and proposed actions . Maternal and Child Nutrition . 2013 ; 9 : 6 - 26 .
67. Kramer KL , McMillan GP . The effect of labor saving technology on longitudinal fertility changes . Current Anthropology . 2006 ; 47 ( 1 ): 165 - 72 .
68. World Health Organization U. Low birthweight. Country, regional and global estimates . http:// wwwuniceforg/publications/files/low_birthweight_from_EYpdf, accessed 2/5/ 2016 . 2004 .
69. Bogin B , Smith P , Orden AB , Varela Silva MI , Loucky J . Rapid change in height and body proportions of Maya American children . American Journal of Human Biology 2002 ; 14 : 753 - 61 . PMID: 12400036
70. Goldstein H , Tanner JM . Ecological considerations in the creation and the use of child growth standards . Lancet 1980 ; 315 : 582 - 5 .
71. Bustamante A , Freitas D , Pan H , Katzmarzyk PT , Maia J . Centile curves and reference values for height, body mass, body mass index and waist circumference of Peruvian children and adolescents . International Journal of Environmental Research and Public Health . 2015 ; 12 2905 - 22 . doi: 10 .3390/ ijerph120302905 PMID: 25761169
72. Kramer KL , Greaves RD . Synchrony between growth and reproductive patterns in human females: rapid juvenile growth among Pumé foragers . American Journal of Physical Anthropology . 2010 ; 141 ( 2 ): 235 - 44 . doi: 10 .1002/ajpa.21139 PMID: 19844999
73. Silva S , Maia J , Claessens AL , Beunen G , Pan H . Growth references for Brazilian children and adolescents: healthy growth in Cariri study . Annals of Human Biology . 2012 ; 39 ( 1 ): 11 -8 doi: 10.3109/ 03014460. 2011 .632646 PMID: 22092114
74. Urlacher SS , Blackwell AD , Liebert MA , Madimenos FC , Cepon-Robins TJ , Gildner TE , et al. Physical growth of the Shuar: height, weight and BMI references for an Indigenous Amazonian population . American Journal of Human Biology . 2015 ; published online June 30, 2015 . doi: 10 .1002/ajhb.22747
75. Sokal RR , Rohlf FJ . Biometry: The Principles and Practice of Statistics in Biological Research . 3rd ed. New York: W.H. Freeman and Company; 1995 .
76. Brown JH , Marquet PA , Taper ML . Evolution of body size: consequences of an energetic definition of fitness . American Naturalist 1993 ; 142 : 573 - 84 . doi: 10 .1086/285558 PMID: 19425961
77. Magvanjav O , Undurraga EA , Eisenberg DTA , Zeng W , Dorjgochoo T , Leonard WR , et al. Sibling composition and children's anthropometric indicators of nutritional status: evidence from native Amazonians in Bolivia . Annals of Human Biology . 2012 ; 40 ( 1 ): 23 - 34 . doi: 10 .3109/03014460. 2012 .728621 PMID: 23095024
78. Taha SA . Ecological factors underlying protein-calorie malnutrition in an irrigaed area of the Sudan . Ecology, Food and Nutrition 1979 ; 7 : 193 - 201 .
79. Godoy R , Nyberg C , Eisenberg DT , Magvanjav O , Shinnar E , Leonard WR , et al. Short but catching up: statural growth among native Amazonian Bolivian children . American Journal of Human Biology . 2010 ; 22 ( 3 ): 336 - 47 . doi: 10 .1002/ajhb.20996 PMID: 19844899
80. Hawkes K , O'Connell J , Blurton Jones N. Hadza women's time allocation, offspring provisioning and the evolution of long postmenopausal life spans . Current Anthropology . 1997 ; 38 ( 4 ): 551 - 77 .
81. Helfrecht C , Meehan CL . Sibling effects on nutritional status: Intersections of cooperation and competition across development . American Journal of Human Biology . 2015 : n/a-n/a . doi: 10 .1002/ajhb.22763
82. World Health Organization Multicentre Growth Reference Study Group. WHO child growth standards based on length/height, weight and age . Acta paediatrica . 2006;Supplement 450 : : 76 .
83. Crittenden AN , Marlowe FW . Allomaternal care among the Hadza of Tanzania . Human Nature . 2008 ; 19 : 249 - 62 . doi: 10 .1007/s12110-008 -9043-3 PMID: 26181616
84. Hawkes K , O'Connell J , Blurton Jones N , Alvarez H , Charnov E . Grandmothering, menopause and the evolution of human life histories . Proceedings of the National Academy of Sciences . 1998 ; 95 : 1336 - 9 .
85. Hrdy SB . Mothers and Others: Belknap Press; 2009 .
86. Kaplan HS , Hooper PL , Gurven M. The evolutionary and ecological roots of human social organization . Philosophical Transactions of the Royal Society of London, B . 2009 ; 364 : 3289 - 99 .
87. Russell AF , Lummaa V . Maternal effects in cooperative breeders: from hymenopterans to humans. Proceeding of the Royal Society , B. 2009 ; 364 ( 1520 ): 1143 - 67 .
88. Hagen EH , Barrett HC . Cooperative breeding and adolescent siblings . Current Anthropology . 2009 ; 50 ( 5 ): 727 - 37 .