LAUSR

OBJECTIVE The association between structural racism, as captured by historic residential redlining practices under the Home Owners' Loan Corporation (HOLC), and present-day diabetes mortality, and years of life lost (YLL), remains unknown. RESEARCH DESIGN AND METHODS Age-standardized mortality and YLL data were combined with historic HOLC redlining data for the city of Seattle, WA (a sample of 109 census tract-level observations) for each of the years 1990 through 2014 (25 years). Spatial autoregressive regression analyses were used for assessment of the association between an area's historic HOLC redlining score and diabetes (and all-cause) mortality and YLL. RESULTS Spatial autoregressive model estimates indicate that an area's HOLC redlining score explains 45%-56% of the variation in the census tract-level diabetes mortality rate and 51%-60% of the variation in the census tract diabetes YLL rate between the years of 1990 and 2014. For 2014, estimates indicate that areas with a unit-higher HOLC grade are associated with 53.7% (95% CI 43.3-64.9; P < 0.01) higher diabetes mortality rates and 66.5% (53.7-80.4; P < 0.01) higher diabetes YLL rate. Magnitudes of marginal effects were consistently larger for diabetes than for all-cause outcomes. CONCLUSIONS Results indicate sizable, and statistically significant, associations between historic redlining practices and present-day diabetes mortality and YLL rates. In addition, the persistence of these associations across the 1990-2014 period highlight a need for targeted action to undo the impact of historical redlining on current health.