How many lives do we live?. . . and some surprising corollaries thereof.

Adrien Remund , Rijksuniversiteit Groningen

How many times did Jeanne Calment cheat death before becoming the oldest person to have ever lived? How exceptional was this feat, compared, for instance, to becoming a centenarian in th 1700s? Usual measures of longevity only give partial answers to these questions. As human longevity increased steadily in most countries during the last century, more and more people reached ages that used to be considered exceptional. For instance, the proportion of person-years lived above 100 years of age has been multiplied by more than 100 in the Netherlands since the 1950s. This banalization of (very) old age is challenging our definitions and measures of longevity. This paper introduces a new measure of longevity that builds on the dominant theories of ageing. This measure consists in computing the number of times that an average individual needs to outlive his or her expected age at death in order to reach a given target age. It turns out that this measure can be closely approximated as a logarithm transform of the probability to reach this age, which makes its computation trivial given a classic life table. Moreover, I show that this direct relationship rests on the surprisingly universal pattern of the probability to repeatedly survive until one’s expected age at death, which systematically tends to 1/e at (very) old ages across vastly different historical and hypothetical mortality schedules. This approach introduces "lives lived" as a universal measure of time that standardizes survival probabilities across human and possibly non-human populations.

See paper

 Presented in Session P1. Postercafe