Methodological Aspects of Studying Human Aging, Health and Mortality

Anatoliy I. Yashin, Duke University
Igor Akushevich, Duke University
Alexander Kulminski, Duke University
Svetlana V. Ukraintseva, Duke University

In this paper we develop a methodology for studying human aging, health and mortality using a mathematical model describing age-trajectories of human biomarkers, changes in health status, and mortality in a cohort of individuals participated in longitudinal study. The model contains jumping and continuous components. The jumping component represents changes in health/well-being status. It is described in terms of a finite state continuous time (non-Markov) random process with transition intensities depending on current values of continuous component. The continuous component, represents gradual aging related changes in physiological state. It is described by stochastic differential equations, where coefficients depend on current value of jumping component. Gaussian approximation of conditional distributions of continuously changing component is described. The approaches to analyses of longitudinal data on aging, health and longevity using this model and three different observational plans are developed. The use of developed approaches in the joint analyses of different datasets is discussed.