Geographic inequalities in COVID-19 diagnosis are now well documented. However, we do not sufficiently know whether inequalities are related to social characteristics of communities, such as collective engagement. We tested whether neighborhood social cohesion is associated with inequalities in COVID-19 diagnosis rate and the extent the association varies across neighborhood racial composition. We calculated COVID-19 diagnosis rates in Philadelphia, PA, per 10,000 general population across 46 ZIP codes, as of April 2020. Social cohesion measures were from the Southeastern Pennsylvania Household Health Survey, 2018. We estimated Poisson regressions to quantify associations between social cohesion and COVID-19 diagnosis rate, testing a multiplicative interaction with Black racial composition in the neighborhood, which we operationalize via a binary indicator of ZIP codes above vs. below the city-wide average (41%) Black population. Two social cohesion indicators were significantly associated with COVID-19 diagnosis. Associations varied across Black neighborhood racial composition (p <0.05 for the interaction test). In ZIP codes with ≥41% of Black people, higher collective engagement was associated with an 18% higher COVID-19 diagnosis rate (IRR=1.18, 95%CI=1.11, 1.26). In contrast, areas with <41% of Black people, higher engagement was associated with a 26% lower diagnosis rate (IRR=0.74, 95%CI=0.67, 0.82). Neighborhood social cohesion is associated with both higher and lower COVID-19 diagnosis rates, and the extent of associations varies across Black neighborhood racial composition. We recommend some strategies for reducing inequalities based on the segmentation model within the social cohesion and public health intervention framework.
Keywords: Black/African American; COVID-19; Ecological; Inequality; Racial disparities; Social capital; Social cohesion; Spatial; Structural determinants.