Inferring mortality deceleration patterns from a gamma-Gompertz-Makeham framework

Filipe Ribeiro, Universidade de Évora
Trifon I. Missov, Max Planck Institute for Demographic Research

We calculated life-table aging rates (LARs) for overall and cause-specific mortality by estimating a gamma-Gompertz-Makeham (GGM) model and taking advantage of LAR’s parametric representation by Vaupel and Zhang (2010). For different HMD countries we study how the evolution of estimated LAR patterns (for all or specific causes of death) could explain observed 1) life expectancy dynamics, and 2) mortality improvement or deterioration at different ages. We compare our findings across countries (data from HMD) and major causes of death (data from WHO). Preliminary results reveal that GGM model-based LAR fits well the observed LAR, capturing simultaneously, a shift in the age of mortality deceleration with time. Across the studied countries and between sexes, it’s also identifiable different ages of mortality deceleration, what suggests a connection between the rate of life expectancy increase and the estimated LARs. Summing up, we intend with this study answering the following questions: 1) can we identify a similar pattern across countries, sexes and CODs? 2) can ages of mortality deceleration be associated with the pace of life expectancy increase? 3) speaking statistically, does the LAR approximation of Vaupel and Zhang (2010) provides a good fit to the observed LAR?, and 4) can we find a more pronounced pattern according to different CODs?