Predicting Blood Lead Levels among Children in Ohio using Spatial Analysis Techniques

Keywords: Blood Lead Levels, Children, Spatial, Hot spot, Predict
Abstract Type: Poster Abstract

Abstract

The heavy metal lead has numerous health impacts, particularly on younger children, who are especially vulnerable due to their hand-to-mouth behaviors. This study aims to predict Blood Lead Levels (BLLs) in children aged 0-6 across Ohio to identify census tracts with high concentrations of BLLs among this age group. Housing, racial and socioeconomic data were used to predict BLLs, including variables such as BLLs in children aged 0-6, the number of households built before 1980, vacant households, traffic count data, etc. These data were collected from the Ohio Department of Health, Ohio Department of Transportation, American Community Survey 5-year estimates, etc. To determine the spatial pattern of BLLs, Global Moran’s I technique was applied. Hot spot and cold spot analysis were conducted to identify clusters of high and low BLLs. Additionally, a spatial lag model was used to predict BLLs for 2022. This study found spatial dependence in childhood BLLs indicates that census tracts with high BLLs are clustered. In the presence of spatial dependence, the spatial lag model performed better, as evidenced by AIC values. Overall, this study provides valuable insights into high-risk census tracts for elevated BLLs, enabling targeted actions to reduce BLLs.