Race, Structural Racism, and Prevalence of Diabetes in US Neighborhoods

Importance: Race and measures of structural racism are associated with higher prevalence of diabetes at the census-tract level, but pathways and targets for intervention are unclear. Objective: To investigate the direct and indirect associations of race and 2 measures of structural racism (historic and contemporary) with prevalence of diabetes at the census-tract level. Design, Setting, and Participants: Cross-sectional study of 15 190 US census tracts across 157 counties within 50 states and Washington, DC conducted September 2025 to February 2026. Exposures: Historic structural racism was defined as historic redlining using the Home Owners' Loan Corporation neighborhood grades (score between 1 = best and 4 = redlined) overlayed on present-day census tracts. Contemporary structural racism was defined using the structural racism effect index (SREI), a summary score of 9 domains (built environment, criminal justice, education, employment, housing, income and/or poverty, social cohesion, transportation, and wealth). African American or Black race was defined as proportion within a census tract based on American Community Survey 2019 data. Main Outcomes and Measures: Diabetes prevalence from the US Centers for Disease Control and Prevention Population Level Analysis and Community Estimates 2019 dataset was the main outcome. Structural equation modeling was used to investigate direct and indirect associations between race and both measures of structural racism. Analyses were completed using standardized estimates (β), which are interpreted as the change in SD associated with 1 SD increase in the variable. Results: Among 15 190 census tracts, mean (SD) diabetes prevalence across census tracts was 11.8% (4.9%), and a mean (SD) of 26.7% (31.4%) of the population within census tracts was African American or Black. Both historic (β = 0.02; 95% CI, -0.05 to 0.08; P = .001) and contemporary (β = 0.56; 95% CI, 0.51 to 0.61; P < .001) measures of structural racism were directly associated with higher diabetes prevalence. Historic structural racism was indirectly associated with higher diabetes prevalence via the pathway of contemporary structural racism (β = 0.17; 95% CI, 0.15 to 0.20; P < .001). Contemporary structural racism showed a stronger association compared with historic structural racism for the pathway through which African American or Black race was associated with higher diabetes prevalence (β = 0.58; 95% CI, 0.52 to 0.63 vs β = 0.18; 95% CI, 0.14 to 0.22; both P < .001). Conclusions and Relevance: African American or Black race, historic structural racism, and contemporary structural racism were associated with higher diabetes prevalence at the census-tract level in the US. Domains of contemporary structural racism as measured by the SREI are potential targets for future interventions.