Background: Research in the past 20 years has shown that self-assessed health (SAH) is a consistent and reliable predictor of health outcomes. However, it is still unclear what factors are responsible for the association of SAH with mortality and other objective health indicators.
Purpose: This study examined the impact of trait positive affect, trait negative affect, and functional limitations (FL) on SAH cross-sectionally and longitudinally. We hypothesized that changes in SAH ratings would be mainly influenced by FL, whereas affective information would have a biasing effect on SAH ratings cross-sectionally.
Methods: We analyzed longitudinal data obtained from five successive annual interviews conducted with over 800 elderly respondents (mean age = 73 years) using latent growth curve modeling.
Results: Results revealed that SAH judgments were related to FL and to trait affects in each of the five waves (cross-sectional data), but only changes in FL were related to changes in SAH over time. These data also showed that changes in SAH were predictive of mortality above and beyond its initial levels.
Conclusions: The results suggest that the temporal trajectory of FL is a source of information that allows an individual's SAH to predict objective health measures of both morbidity and mortality.