John Wagner

Publication Date

Spring 7-1-2012


This work provides an integrative approach for assessing population health among a group of indigenous South American forager-farmers, the northern Aché of eastern Paraguay. The Aché were full-time hunting and gathering nomads up until the time of first peaceful contact in the early 1970s when they experienced a devastating virgin-soil population epidemic that killed approximately 40% of the population, where all age and sex groups were affected more or less equally with the exception of a higher survival rate among reproductive-aged women. The Aché are now settled on several reservations and have fully recovered their population numbers. The Aché have experienced a secular increase in body size in making the transition from forest to reservation living. The more abundant nutritional resources following contact as well as the extremely stressful conditions of the contact period provide a natural means for examining how environmental conditions interact with developmental sex differences. Whereas both males and females are exquisitely sensitive to environmental conditions, each sex responds differently in ways that are organized around intrinsic physiological differences. For example, with respect to body size and particularly height, males tend to be both taller and exhibit greater variation in adult body size than females. This pattern arises from principles of energy allocation towards reproduction such that males invest greater energetic resources towards physical growth, such as building bone and muscle mass, while females invest greater energetic resources towards other facets necessary for obligate reproduction, such as building lipid stores to meet offspring energetic requirements. It follows that resource variation will interact with each sex in ways that reflect these differences; for example, stature is expected to fluctuate more closely with resource availability in males than females, whereas lipid deposition, ovulation and conception rates are expected to vary as a function of resource availability among females and, with the possible exception of pre-pubescent lipid deposition, are non-issues among males. Previous studies of developmental sex differences in response to environmental conditions among human populations have proven inconclusive due to numerous mitigating factors; the results presented in this study represent a well-controlled natural experiment by which to probe the effects of changing dietary status and psychosocial stress on several developmental parameters. The parameters assessed in this study, and particularly the anthropomorphic variables, are slanted towards observing developmental effects in males versus females. However, this should not be taken as evidence that males are in general more sensitive to environmental conditions than females; but simply that the data collected here were more informative of male than female physiological responses. A main finding presented here is that Aché males were more responsive in terms of physical growth to dietary conditions such that adult males made relatively greater gains in height and weight in comparison to adult females across the transition from forest to reservation living. Independent sample T-tests between the cohort of adults that completed growth before the contact period, and the cohort of fully grown adults that were born after the contact period, revealed more significant gains in height among males (p = 0.000) versus females (p = 0.029). The same pattern was observed in terms of weight where males made more significant gains (p = 0.001) versus females (p = 0.005). Concomitantly, linear regression equations using age as the independent variable indicated steeper slopes for male (-0.21) versus female (-0.10) height across adulthood (cross-sectional), and male (-0.22) versus female (-0.18) weight across adulthood, where height and weight were dependent variables. The relative differences between the sexes amounted to an increase of 5.1% in male height versus a 2.4% increase in female height; and a 15.5% increase in male weight versus a 13.7% increase in female weight. A second finding of this study concerned the effect of the contact experience on peri-adolescent growth. Aché males who experienced the onset of the adolescent growth spurt, about 11 years of age, during one of the more severe years of the epidemic showed evidence of permanent developmental perturbations in terms of being significantly shorter, lighter, and somewhat smaller-headed than either adjacent male cohort. Regression analyses for the entire adult sample using a dummy coded variable for the contact cohort and controlling for age revealed a significant effect of contact on male height (p = 0.002) and weight (p = 0.016). Similar effects were not found for females of this same cohort, who, along with the preceding female cohort, or those who initiated take-off growth before contact but experienced at least two years of adolescent growth during contact, exhibited increases in height and weight, although the sample sizes within these cohorts were too limited to draw firm conclusions. Similarly, the male cohort who preceded the contact cohort, or those who initiated take-off growth before the epidemic but experienced at least two years of adolescent growth during contact, were also taller and heavier although, again, the sample size was limited. One possible interpretation for the overall pattern found among these peri-adolescent cohorts of both sexes is that the contact period may have exerted a mortality bias towards the elimination of smaller individuals, such that larger bodied individuals survived to be measured in the current study. This effect could account for the relatively larger body masses observed in these cohorts with the addition that there appeared to be a specific and permanent developmental effect on only those males who initiated take-off growth under conditions of severe physical and psychosocial stress. Note that under this scenario, mortality would have removed even smaller males from the middle male cohort during the contact period. The perturbational effect on the development of this cohort gains some support from other developmental parameters described below. Evidence from visual evoked potentials (VEP) and bioelectric impedance analysis (BIA) were generally consistent with the pattern observed in the male contact cohort although the statistical results were not as strong as they were for the anthropomorphic measures. Visual evoked potentials (VEP) assess the integrity of the neurovisual pathway by recording the latency between a visual stimulus and occipital processing where the major marker is referred to as the P100 latency. This latency can be used in conjunction with head length to calculate an estimate of nerve speed, referred to as Central Nerve Conduction Velocity (CNCV). Phase angle (PA) is derived from BIA parameters and is believed to reflect cellular health and membrane function. P100, CNCV and PA evinced characteristic ontogenetic and senescent patterns among the population as a whole, and all were consistent with a developmental perturbation occurring among the male adolescent cohort that experienced take-off growth during one of the severe contact years, although, again, these trends did not reach the statistical strength that the anthropomorphic variables showed. Senescent trends in VEP and PA measures complicate identification of secular trends although senescence itself can be compared between the sexes and here males were found to senesce more significantly across the lifespan than females for both VEP and PA, although some of this pattern may be attributable to males having higher values in these measures during early adulthood than females. Linear regression analyses using a dummy coded variable for the male contact cohort and controlling for age were not significant for VEP measures or PA, although all of the coefficients were in the expected direction. Conducting independent sample T-tests for these measures would require that age be controlled; however, the male contact cohort was significantly different than the younger adjacent cohort for CNCV (p = 0.041) and PA (p = 0.048), but not the older adjacent cohort despite similar means for each of the adjacent cohorts, which can, in part, be attributed to the low sample size of the older adjacent cohort. However, evaluating significance at p-values of 0.05 or less increases the risk of Type II error, or rejecting a truth, and does not appear to be ideal for detecting a real effect when comparing groups with small sample sizes. In general, the proposed effect of contact on male take-off growth was most profound for body size parameters, less for BIA parameters, and less still for VEP parameters, indicating that visual maturation\u2014as measured by the VEP parameters used in this study\u2014was largely complete and/or immune to the psychosocial stress that obtained during the contact period. To probe for potential correlations among the anthroporphic, visual evoked potential, and bioelectrical impedance variables, correlation analyses were performed on all variables and subjects by calculating individual residual values from non-parametric fit lines (LOWESS) within each sex such that age was controlled. The most pronounced and unanticipated findings concerned BIA variables and body size. Phase angle (PA) residuals were inversely correlated with body size residuals among children and this pattern was stronger for weight than height, stronger in females versus males, and stronger at younger versus older childhood ages. Reactance, which reflects the capacitance of cells or their ability to hold an electrical charge and is used in the calculation of PA, was significantly and inversely correlated with weight residuals for the entire Aché population. These findings were novel and unexpected and suggest that being smaller for age was associated with better health and/or capacitance among the Aché. The general agreement among the anthropomorphic and physiological measures with respect to the effect of contact on the male cohort that initiated take-off growth is of interest because there is no obvious mechanism linking these various measures. While highly informative, BIA measurements and particularly PA are not well understood biologically; potential explanations for this state of affairs are briefly explored in conjunction with the properties of electricity in biological systems that might account for the pattern observed in this research. The overarching goal of this project was to glean information from traditional populations experiencing unique ecological conditions such that it may better inform the understanding of the relation between ecology and health. By extension, factors relevant to the unique trajectory of human emergence may be illuminated. The evidence presented here adds to the body of knowledge concerning human physiological adaptation in response to variable ecological conditions, including the occurrence of an acute, stress episode in the form of a virgin soil contact.

Project Sponsors

University of New Mexico, Sigma Xi, Louis Leakey Foundation

Document Type




Degree Name


Level of Degree


Department Name


First Committee Member (Chair)

Kaplan, Hillard

Second Committee Member

Gangestad, Steve

Third Committee Member

Glew, Robert

Included in

Anthropology Commons