Issues of Selection in Human Survivorship: A Theory of Mortality Change from the Mid-Eighteenth to the Early Twenty First Century

Abstract

Is variation in empirical mortality across populations consistent with a hypothesis of selection? To examine this proposition an extended frailty mortality model is put forward; incorporating biological frailty; a common non-parametric hazard, joint for men and women, rep-resenting endogenous mortality in terms of degenerative aging (senescence); and environ-mental influence on survivorship. As the model is fitted to empirical cohort mortality exhibiting extreme variation, biological aging is identified up to a multiplicative factor. Mortality of elected cohorts born in Sweden, Denmark, and Iceland during the past 250 years and in Japan any ten years between 1950 and 1990 is approached appropriately by the model. Reduced natural selection may account for a substantial part of the empirical mortality change in the course of the demographic transition. Survivorship in the late nineteenth and the twentieth century ties selection to major medical advances and rapid recent mortality decline, probably with consequences for future health and survivorship.