Victor Ronget & Jean-Michel Gaillard
Actuarial senescence (i.e. the increase of mortality with age) is widespread throughout life but its strength varies a lot across species and even among populations. An increasing number of studies currently investigate the reasons for those differences in actuarial senescence. However, before trying to explain such differences, a reliable picture of patterns is required. Most studies performed to date relied only on a single metric to measure actuarial senescence: maximum longevity (i.e. the lifespan of the oldest individual in a population). In this work, we test whether maximum longevity is an appropriate metric to assess mortality patterns and then provide alternative metrics for future studies of aging.
We demonstrate that maximum longevity is not an appropriate metric to use when studying variation in actuarial senescence patterns because it is highly dependent on the sample size of the focal population. We thus propose two reliable alternative metrics, life expectancy (i.e. the mean age at death) and 90% longevity, which are easy to compute and do not depend on sample size. Furthermore, we argue that researchers on aging should not only consider metrics involving the timing of specific events in mortality patterns (i.e. pace metrics) but also investigate metrics that describe the magnitude and form of the mortality increase with age (i.e. shape metrics). One easy way to obtain those shape metrics is to calculate the ratio of two pace metrics. The use of a broader range of metrics to study mortality patterns should shed new light on the process of actuarial senescence.