Pathways to Disability and Death in Middle Ages: Estimates from the Health and Retirement Study Data

Abstract

I review the microbiology literature on aging, finding that much of the process of health development over the lifespan is determined by early childhood epigenetic factors consisting of internal and external environments, health care use, and health related behavior, together with genetic factors. I introduce a statistical ‘disablement model’, specifically, a multi-state time-to-event life history model of health developments over the lifespan, incorporating childhood factors and health behaviors as regressors. I correct for unobserved heterogeneity biases of parameter estimates, extending a statistical technique from the literature of two-state models. Using the HRS (Health and Retirement Studies) dataset and taking the definition of disability as health states that qualify for the Social Security’s DI (Disability Insurance) program or the SSI (Supplemental Security Income) program, I study how childhood factors and health behaviors are associated with probability of following various transition paths through the health states of normal health, chronic diseases, disability, and death. The childhood factors include childhood socioeconomic status, health status and education level, and the health behaviors include smoking and exercising. The regressors also include biomarkers—body mass index (BMI), a measure of stress (CES-D) and a measure of cognitive health. I also carry out quantitative analysis of the effect of social policies improving childhood factors and health behaviors of various social groups on their risks of chronic illness, disability, or death in mid-ages.