How Much Does Structural Racism Contribute to Diabetes Disparities Across American Communities?

Egede et al. (2026) examined whether African American or Black race, historic redlining, and contemporary structural racism are associated with diabetes prevalence in U.S. neighborhoods and what pathways explain those relationships. They analyzed data from 15,190 census tracts across 157 counties using CDC diabetes prevalence estimates, census data, historical redlining maps, and a multidimensional measure of contemporary structural racism. They found that contemporary structural racism was the strongest predictor of neighborhood diabetes prevalence, with an association nearly three times larger than the direct association of Black population share. Historic redlining also mattered, but much of its influence operated indirectly through present-day structural disadvantages. Their results suggest that current neighborhood conditions linked to structural racism are more strongly associated with diabetes prevalence than the legacy effects of redlining alone.

Why This Article Was Selected for The Policy Scientist

This article addresses an important question that extends well beyond diabetes. At its core, it examines how long-standing features of communities may be associated with contemporary health outcomes, contributing to a broader literature on the relationship between place, opportunity, and population well-being. The topic is particularly timely because policymakers, researchers, and health systems are increasingly interested in understanding why large differences in health outcomes persist across neighborhoods despite advances in medical care. Egede and his colleagues have been leading contributors to this field, and this study advances their earlier work by jointly examining historical and contemporary measures within a single analytical framework. The dataset is a major strength, encompassing more than 15,000 census tracts across the United States, which supports broad generalizability to many urban and suburban settings. The structural equation modeling approach is sophisticated and useful for evaluating potential pathways, but the cross-sectional design does not permit strong causal conclusions. Future research would benefit from longitudinal causal inference designs that can better identify whether specific changes in neighborhood conditions produce measurable changes in health outcomes over time.

Full Citation and Link to Article

Egede, L. E., Walker, R. J., Campbell, J. A., Ekwunife, O., & Linde, S. (2026). Race, structural racism, and prevalence of diabetes in US neighborhoods. JAMA Network Open, 9(4), e265122. https://doi.org/10.1001/jamanetworkopen.2026.5122

Central Research Question

This study investigates whether African American or Black race, historic structural racism, and contemporary structural racism are associated with neighborhood-level diabetes prevalence in the United States. More specifically, the authors seek to determine whether historic redlining and present-day structural conditions represent pathways through which racial differences in diabetes prevalence emerge across communities. Rather than focusing on individual behaviors or clinical factors, the study examines how characteristics of neighborhoods may be associated with population-level health outcomes. The authors hypothesize that both historical and contemporary forms of structural racism are directly associated with higher diabetes prevalence and that contemporary structural racism serves as an important mechanism linking historical discrimination to current health outcomes.

Previous Literature

The study builds upon a substantial body of research demonstrating persistent racial disparities in diabetes prevalence, complications, and mortality. Previous studies have documented that African American populations experience higher rates of diabetes and worse diabetes-related outcomes than many other demographic groups. Researchers increasingly attribute these differences not only to individual characteristics but also to broader social and environmental conditions.

The authors place their work within an expanding literature on structural racism as a determinant of health. Structural racism refers to the cumulative effects of institutions, policies, and social systems that influence access to resources and opportunities. Prior research has linked structural racism to numerous health outcomes, including chronic disease, mortality, and life expectancy. One commonly used measure of historical structural racism is residential redlining, a set of discriminatory lending and housing practices implemented during the twentieth century. Earlier studies have found associations between redlining and contemporary health outcomes, including diabetes prevalence.

The article also builds on recent efforts to measure contemporary structural racism. Traditional studies often examine a single domain such as housing, education, or income inequality. In contrast, the Structural Racism Effect Index (SREI) integrates multiple dimensions of neighborhood conditions into a comprehensive measure. The authors argue that examining both historical and contemporary measures simultaneously provides a more complete understanding of how neighborhood conditions are associated with diabetes prevalence.

Data

The analysis combines several large national datasets at the census-tract level. The final sample includes 15,190 census tracts located within 157 counties across all 50 states and Washington, DC. This broad geographic coverage represents one of the study’s principal strengths because it allows the authors to evaluate relationships across a diverse range of communities.

The primary outcome measure is diabetes prevalence from the Centers for Disease Control and Prevention’s PLACES database. These estimates are generated using survey data combined with demographic information to create small-area measures of health conditions. The mean diabetes prevalence across census tracts was 11.8%.

Race is measured as the proportion of residents within each census tract who identify as African American or Black using American Community Survey data. Historic structural racism is measured through Home Owners’ Loan Corporation (HOLC) grades derived from digitized redlining maps. Higher scores indicate greater historical exposure to redlining.

Contemporary structural racism is measured using the Structural Racism Effect Index. This index incorporates 42 variables across nine domains: built environment, criminal justice, education, employment, housing, income and poverty, social cohesion, transportation, and wealth. Together, these data provide a multidimensional portrait of neighborhood conditions and allow the authors to examine how multiple forms of structural disadvantage may be associated with diabetes prevalence.

Methods

The authors employ structural equation modeling (SEM) to examine direct and indirect associations among race, historic redlining, contemporary structural racism, and diabetes prevalence. SEM is particularly useful when researchers seek to evaluate multiple pathways simultaneously because it allows direct and indirect relationships to be estimated within a single analytical framework.

The study tests three primary hypotheses. First, both historic and contemporary structural racism are expected to be directly associated with higher diabetes prevalence. Second, contemporary structural racism is expected to mediate the relationship between historic redlining and diabetes prevalence. Third, both historic and contemporary structural racism are expected to serve as pathways linking race and diabetes prevalence.

The authors report standardized coefficients, which allow comparison of the relative strength of associations. Model fit statistics indicate excellent statistical fit. However, the authors appropriately acknowledge an important limitation: because the data are cross-sectional, the analysis cannot establish causality. Structural equation modeling can identify patterns consistent with hypothesized pathways, but it cannot determine whether one factor directly causes another. As a result, the findings should be interpreted as associations rather than causal effects.

Findings/Size Effects

The study finds statistically significant associations between race, structural racism, and diabetes prevalence. Contemporary structural racism emerged as the strongest direct correlate of diabetes prevalence. The standardized coefficient for contemporary structural racism was β = 0.56, indicating a substantially stronger association than other explanatory variables included in the model.

African American or Black population share was also directly associated with higher diabetes prevalence (β = 0.37). By comparison, the direct association between historic redlining and diabetes prevalence was relatively small (β = 0.02), although it remained statistically significant. These results suggest that contemporary neighborhood conditions exhibit a stronger relationship with diabetes prevalence than historical redlining alone.

The pathway analysis provides additional insight. Historic redlining was indirectly associated with diabetes prevalence through contemporary structural racism (β = 0.17). This finding suggests that some of the association between historical discrimination and current health outcomes may operate through present-day neighborhood conditions.

The strongest pathway in the model linked African American or Black population share to contemporary structural racism (β = 0.58). This coefficient was larger than the pathway connecting Black population share to historic redlining (β = 0.18). The results therefore indicate that contemporary structural conditions play a more substantial role than historical redlining in explaining observed associations between race and neighborhood diabetes prevalence.

From a practical perspective, contemporary structural racism had an association approximately 50% larger than the direct association of Black population share and nearly 28 times larger than the direct association of historic redlining. These comparisons illustrate the central role contemporary neighborhood conditions played within the authors’ analytical framework.

Conclusion

The study concludes that both historical and contemporary measures of structural racism are associated with diabetes prevalence across U.S. neighborhoods. However, contemporary structural racism appears to represent the most influential pathway identified in the analysis. The findings suggest that current neighborhood conditions may be an important mechanism through which historical patterns continue to be associated with contemporary health outcomes.

The article contributes to the literature by integrating historical and contemporary measures within a single model and by utilizing a large national sample of census tracts. The breadth of the dataset supports substantial external validity across many urban and suburban settings in the United States. At the same time, the authors appropriately recognize that cross-sectional analyses cannot establish causal relationships. Future research using longitudinal data, natural experiments, or other causal inference approaches would strengthen understanding of how changes in neighborhood conditions influence diabetes prevalence over time.

Overall, the study provides evidence that both historical and contemporary neighborhood characteristics are associated with diabetes prevalence, while highlighting the comparatively strong relationship between contemporary structural conditions and population health outcomes.