Lucia Zanotto, Università di Padova
Vladimir Canudas-Romo, Max-Planck Odense Center on the Biodemography of Aging
Stefano Mazzuco, Università di Padova
We study the changes of premature mortality over time, using a parametric model with response variable the life table distribution at deaths. The model is a mixture of three distributions: one for the infant and child mortality, another for accidental and premature mortality and the last for adult mortality. The main advantages of the model are: the possibility to compute, in explicit form, the three component contributions of life expectancy; the identification of the three modes (one for each function), which helps to split the overall area distribution in the different stages of life. Moreover all parameters have a demographic interpretation. The mixture distribution model is tested using the Swedish raw data from the Human Mortality Database. Our results show that, over time, the premature mortality function becomes flatter and more symmetric, and its mode shifts progressively. This indicates that the accidental mortality has disappeared, while the premature mortality cannot be neglected. We also show that its contribution, both to explain life expectancy and the area of the distribution, decreases in the last century, but in recent years it starts to grow slightly.
Presented in Session 86. Modelling mortality