Black-white difference in long-term predictive power of self-rated health on all-cause mortality in United States

Ann Epidemiol. 2016 Feb;26(2):106-114. doi: 10.1016/j.annepidem.2015.11.006. Epub 2015 Dec 11.


Purpose: Despite the well-established association between self-rated health (SRH) and mortality, limited information exists on Black-White differences in this link. Using a nationally representative sample of adults in the United States, the present study had four aims: (1) to assess whether the association between baseline SRH and all-cause mortality over a long follow-up differs for blacks and whites, (2) to test whether any race difference in the SRH-mortality link depends on how the SRH variable is treated (e.g., nominal, dichotomous, continuous), (3) to test if the SRH-mortality link or any differences in the association by race are explained by differences in objective health measures (chronic medical conditions [CMC]), and (4) to assess whether these associations vary by gender.

Methods: Data came from the Americans' Changing Lives Study, a nationally representative longitudinal cohort of U.S. adults 25 years and older with up to 25 years of follow-up. The study followed 3361 blacks or whites for all-cause mortality between 1986 and 2011. The predictor of interest was a single-item measure of SRH in 1986, treated as a nominal, dichotomous (fair/poor vs. excellent/very good/good), and continuous variable. Confounders included baseline age, education, income, depressive symptoms, and CMC. Race (black vs. white) was the focal effect modifier. We ran Cox proportional hazard models for the pooled sample and also stratified by race and gender, before and after adjusting for CMC.

Results: Regardless of how SRH was treated and for both men and women, we found significant interactions between race and SRH, indicating a stronger predictive role of SRH for all-cause mortality among whites compared to blacks. Before adjustment for chronic medical conditions, lower SRH was associated with higher risk of mortality among blacks and whites, but after adjustment, the SRH-mortality association was no longer significant among blacks.

Conclusions: Baseline SRH continues to predict long-term mortality among white but not black Americans after adjustment for chronic medical conditions at baseline, and these patterns are similar for men and women. Future research should test whether the differential predictive validity of SRH across race groups arises because SRH reflects different aspects of health of black and white Americans.

Keywords: African Americans; Chronic medical conditions; Ethnic groups; Mortality; Self-rated health.

Publication types

  • Research Support, N.I.H., Extramural
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • African Americans / statistics & numerical data*
  • Age Distribution
  • Aged
  • Chronic Disease / ethnology
  • Depression / ethnology
  • Female
  • Health Status*
  • Humans
  • Male
  • Middle Aged
  • Mortality / ethnology*
  • Self Report*
  • Sex Distribution
  • Socioeconomic Factors
  • Whites / statistics & numerical data*