Despite the well-established literature on the effects of race and socioeconomic status (SES) on mortality, limited information exists on mediators of these effects. Taking a life-course epidemiology approach, and using a nationally representative sample of adults in the USA, the current study has two aims: (1) to assess the effects of race and SES at baseline on all-cause mortality over a 25-year follow-up and (2) to test whether the number of chronic medical conditions (CMCs) as a time-varying covariate mediates the effects of race and SES on all-cause mortality.
Data came from the Americans’ Changing Lives (ACL) Study, a nationally representative longitudinal cohort of US adults 25 and older. The study followed 3361 Blacks or Whites for all-cause mortality for up to 25 years from 1986 to 2011. The predictors of interest were race and SES (education and family income) at baseline measured in 1986. Confounders included baseline age and gender. CMC was the potential time-varying mediator measured in 1986, 1989, 1991, 2001, and 2011. We ran Cox proportional hazard models with and without CMC as time-varying covariates.
In separate models, race and SES were predictors of all-cause mortality. In the model that tested the combined effect of race and SES, SES but not race was predictive of all-cause mortality. We also found evidence suggesting that CMC fully mediates the effect of race on all-cause mortality. Number of CMC only partially mediated the effect of SES on mortality.
The number of CMC fully mediates the effects of race and partially mediates the effects of SES on all-cause mortality in the USA. Mortality prevention for minority populations will benefit tremendously from elimination of CMC disparities as well as enhancement of CMC management by minority populations. Elimination of the gap due to SES may be more challenging than the elimination of the racial gap in mortality.