Bayesian deprivation index models for explaining variation in elevated blood lead levels among children in Maryland

Lead exposure adversely affects children's health. Exposure in the United States is highest among socioeconomically disadvantaged individuals who disproportionately live in substandard housing. We used Bayesian binomial regression models to estimate a neighborhood deprivation index and its association with elevated blood lead level (EBLL) risk using blood lead level testing data in Maryland census tracts. Our results show the probability of EBLL was spatially structured with high values in Baltimore city and low values in the District of Columbia suburbs and Baltimore suburbs. The association between the neighborhood deprivation index and EBLL risk was statistically significant after accounting for spatial dependence in probability of EBLL. The percent of houses built before 1940, African Americans, and renter occupied housing were the most important variables in the index. Bayesian models provide a flexible one-step approach to modeling risk associated with neighborhood deprivation while accounting for spatially structured and unstructured heterogeneity in risk.