Novak, Beatriz
Working paper no. 2011-10
Abstract
Purpose: The main objective of this study was to investigate whether subjective survival expectations in a developing country were as good as predictors of mortality as they have shown to be in other contexts.
Design and Methods: Data for this study was drawn from the first wave of the Costa Rican Longevity and Healthy Aging Study (2004-2006), a nationally representative sample of non-institutionalized adults over 60 years of age (N = 3000). Parametric Gompertz regressions were used to model mortality stratifying by gender and age (60-69 and 70-89). The first of the six models estimated included subjective survival expectations but not self-rated health status. The second included self-assessed health status but not subjective survival expectations. The third model included both variables. The last three models were similar to the first three, but each of them included four body functioning biomarkers. All models were adjusted by sociodemographic, health-related and health-related behavior variables.
Results: Only among males aged 60 to 69, subjective survival expectations were independent mortality predictors in models that may or not include body functioning biomarkers (p < 0.05 and p < 0.1 respectively). Results also suggested that subjective survival may be mediating the effects of objective measures of health status on mortality, and not just mediating the effects of self-assessed health status on mortality.
Implications: Differences between underlying mortality risks and those perceived by individuals can be employed to establish effect of external health shocks on updating self reported health status as well as potentially significant behavioral changes.