Age-related physiological dysregulation progresses slowly in semi-free-ranging chimpanzees

Abstract

Background and objectives: Lifestyle has widespread effects on human health and aging. Prior results from chimpanzees (Pan troglodytes), one of humans' closest evolutionary relatives, indicate that these lifestyle effects may also be shared with other species, as semi-free-ranging chimpanzees fed a naturalistic diet show healthier values in several specific health biomarkers, compared with their sedentary, captive counterparts. Here, we examined how lifestyle factors associated with different environments affect rates of physiological aging in closely related chimpanzees.

Methodology: We compared physiological dysregulation, an index of biological aging, in semi-free-ranging chimpanzees in an African sanctuary versus captive chimpanzees in US laboratories. If the rate of aging is accelerated by high-calorie diet and sedentism, we predicted greater age-related dysregulation in the laboratory populations. Conversely, if costs of a wild lifestyle accelerate aging, then semi-free-ranging chimpanzees at the sanctuary, whose environment better approximates the wild, should show greater age-related dysregulation. We further tested whether dysregulation differed based on sex or body system, as in humans.

Results: We found that semi-free-ranging chimpanzees showed lower overall dysregulation, as well as lower age-related change in dysregulation, than laboratory chimpanzees. Males experienced lower dysregulation than females in both contexts, and the two populations exhibited distinct aging patterns based on body system.

Conclusions and implications: Our results support the conclusion that naturalistic living conditions result in healthier aging in chimpanzees. These data provide support for the proposal that lifestyle effects on human health and aging are conserved from deeper into our evolutionary history.

Keywords: aging; health; human evolution; lifestyle; primates.

Figure 1.

Age-related change in D_M, split by facility. Predicted values are estimates of D_M (log-transformed and z-scored within each site) from model accounting for subject identity, age, facility, age × facility and number of biomarkers. Lines indicate linear fit; ribbons indicate 95% confidence intervals. See Supplementary Fig. S3 for raw dysregulation scores, which increase in both groups

Figure 2.

Age-related change in D_M, split by sex at the (A) sanctuary and (B) laboratories. Predicted values are estimates of D_M (log-transformed and z-scored within each site) from models accounting for subject identity, age, sex, age × sex and number of biomarkers (and site for the laboratory model). Lines indicate linear fit; ribbons indicate 95% confidence intervals. See Supplementary Fig. S4 for raw dysregulation scores

Figure 3.

Age-related change in D_M, split by body system at the (A) sanctuary and (B) laboratories. Predicted values are estimates of D_M (log-transformed and z-scored within each site) from models accounting for subject identity, age, body system, age × body system and number of biomarkers (and site for the laboratory model). Lines indicate linear fit; ribbons indicate 95% confidence intervals. Cardiometabolic biomarkers were removed from body system analysis due to low sample size. See Supplementary Fig. S5 for raw dysregulation scores, which increase across body systems at Ngamba