Is Self-Rated Health a Valid Outcome Measure for Studying Educational Differences in Health?

Anna Zajacova, Princeton University

Researchers frequently use self-rated health as outcome measure when analyzing health inequalities by education. The common use of SRH is justified by its strong association with future mortality, even controlling for ‘objective’ health characteristics. Recent research, however, found gender and race differences in the relationship between SRH and mortality. We ask whether SRH captures the same aspects of health status at varying levels of educational attainment – that is, does the effect of SRH on mortality differ by educational attainment? We develop and test two opposing hypotheses describing the effect of education on the health evaluation process. We use data from a non-public version of the National Health Interview Survey-Multiple Cause of Death (NHIS-MCD) with almost a million adults, to test the predictive power of SRH for all-cause, as well as cause-specific mortality by education. The results have important implications for the use of SRH in examining educational health disparities.