Abstract

The association between educational attainment and health (i.e., the education-health gradient) among U.S. adults has strengthened in recent decades; however, explanations for this trend are incomplete. This study builds on recent work showing that the strength of the gradient differs across U.S. states.

We asked, how and why has the gradient changed over time in each U.S. state? Using 27 years of cross-sectional data from the 1993–2019 Behavioral Risk Factor Surveillance System on adults ages 30–64 years, we estimated how the association between educational attainment and self-rated health changed over time in each state. We also assessed the contribution of six individual-level factors, such as adults’ income, and three state-level contextual factors, such educational composition and policy context, to changes in the association in each state.

The gradient increased significantly in 49 of the 50 states during the period, but the magnitude of the increase differed notably by state. In most states, the largest single contributor to the increasing gradient was household income. Adults’ employment and obesity also contributed to the increasing gradient to varying degrees in many states, while smoking suppressed the increase in the gradient in over half of states. The state-level contextual factors contributed little to the increasing gradient.

The findings underscore the value of a subnational focus for explaining and potentially reversing the growth in the education-health gradient.