Evolution

Putting Off Growing Old

+ See all authors and affiliations

Science  08 May 2009:
Vol. 324, Issue 5928, pp. 693
DOI: 10.1126/science.324_693a
CREDIT: ISTOCKPHOTO

In a classic 1957 paper on the evolution of senescence, Williams argued that when extrinsic mortality (death due to predation, infectious disease, or accident) is high, natural selection favors investment in early reproduction. When it is low, the increased return from allocating resources to maintain and repair the soma should lead to longer life spans. In the absence of precise information on causes of death, researchers have used hazard models to partition mortality into age-independent (interpreted as extrinsic) and age-dependent components.

Taking this approach, Gurven and Fenelon analyze mortality data from 13 remote, small-scale societies and in historical cohorts from Sweden and England over the past 250 years. They explore two statistical models (Weibull and Gompertz-Makeham) of adult mortality patterns and consider three measures of actuarial aging (mortality rate doubling time, Ricklefs's ω, and slope of the mortality function between ages 60 and 70). The variation in results across these two estimation procedures and three measures complicates the interpretation of the data. Nonetheless, some patterns are robust: The subsistence groups and the early Swedish cohorts exhibit similar actuarial aging, but more recent European cohorts show progressively slower aging. In the longitudinal samples, slower aging and reduced extrinsic mortality are linked. Women have lower rates of senescence than men, a difference that has increased over time. These “modest but nontrivial” changes support Williams's claims, and the authors discuss individual-level mechanisms that could underlie them.

Evolution63, 1017 (2009).

Navigate This Article