Explaining variation in elevated blood lead levels among children in Minnesota using neighborhood socioeconomic variables

Abstract Background Childhood lead exposure is linked to numerous adverse health effects and exposure in the United States is highest among people living in substandard housing, which is disproportionately inhabited by socioeconomically disadvantaged individuals. In this paper, we compared the Vox lead exposure risk score and concentrated disadvantage based on principal component analysis (PCA) to weighted quantile sum (WQS) regression to determine which method was best able to explain variation in elevated blood lead levels (EBLLs). Methods We constructed indices for census tracts in Minnesota and used them in Poisson regression models to identify the best socioeconomic measure for explaining EBLL risk. Results All indices had a significant association with EBLL in separate models. The WQS index had the best goodness-of-fit, followed next by the Vox index, and then the concentrated disadvantage index. Among the most important variables in the WQS index were percent of houses built before 1940, percent renter occupied housing, percent unemployed, and percent African American population. Conclusions The WQS approach was best able to explain variation in EBLL risk and identify census tracts where targeted interventions should be focused to reduce lead exposure.