Educational differences in age -related patterns of disease: An empirical test of cumulative disadvantage theory

This dissertation investigates the relationship between education and health and extends our understanding of how this relationship varies across age. Some evidence suggests that educational differences in health increase over the life course according to a cumulative disadvantage perspective. Other studies find that health inequalities diminish in late adulthood and support the age-as-leveler argument. I argue that both hypotheses are valid and that existing research remains fragmented because it inadequately distinguishes individual health-decline from aggregate changes occurring in the population. This research tests whether both arguments support the same process from different units of analysis. In a departure from previous research that focuses almost exclusively on population-level indicators, I examine measures of disease prevalence, incidence, and survival to differentiate individual health-decline from aggregate changes in health. A secondary objective of the research explores how several behavioral, social, and economic disadvantages affect the onset and duration of disease over the life course. Four waves of the National Health and Nutrition Examination Survey-I (NHANES I) Epidemiologic Follow-up Study (NHEFS) are used for analysis. Results from age-specific prevalence rates and pooled logistic regression models support the leveling hypothesis by showing educational differences in disease prevalence are largest at mid-life and decline at older ages. Results from discrete-time and semi-parametric hazard models support the cumulative disadvantage hypothesis by showing that education is positively related to the age of onset of disease and/or the duration of survival after onset. The results are generally consistent across cohorts with somewhat stronger effects in the two youngest cohorts. I conclude from the findings that cumulative disadvantage operates at the individual level by increasing age-specific rates of illness and mortality; leveling occurs at the population level as a reflection of the inequalities in health decline and mortality selection. The implications of these findings are discussed.