Modeling Midlife Mortality: New Insights into Sex Specific Patterns of Human Mortality

Oliver Wisser, Max Planck Institute for Demographic Research
James W. Vaupel, Max Planck Institute for Demographic Research

A noticeable literature gap exists between the frequency of models dealing with infant and senescent mortality and models which try to capture the mortality in the middle part of human life course. Diminishing this gap we define midlife mortality μm(x) as the difference between a Gamma-Gompertz model for old ages, and an exponential decay model for young age mortality. By modeling the ratio of μm(x) and the total mortality μ(x) in the next step we provide a powerful tool (relative Midlife Mortality (rMM)) for describing differences in midlife mortality between time, sex and region. Applying this tool to several countries in the “Human Mortality Database” we uncover some interesting patterns: The rMM was high and nearly the same for both sexes until the late 1940s when it started to decrease. However, since the 1960s it follows a positive trend for both sexes, but higher and parallel shifted for men.