Understanding lead exposure through data and domain expertise: Insights from New Jersey with a geographically weighted regression analysis

Lead contamination remains a persistent and dangerous threat, particularly affecting young children and vulnerable communities. This study aims to develop a comprehensive lead exposure risk map for New Jersey municipalities by integrating diverse lead contamination data and analyzing the spatial distribution and magnitude of lead exposure risks. Utilizing both data-driven and participatory approaches, we employed Principal Component Analysis (PCA) and Analytical Hierarchy Process (AHP) to create distinct multi-criteria lead exposure indices. Then, a Geographically Weighted Regression (GWR) analysis was conducted to explore the local variations in lead exposure and the factors influencing these risks using both indices. Our linear models indicate that both PCA and AHP-based indices effectively capture the essence of lead exposure in urban areas, with significant correlations (Adjusted R² = 0.225 for PCA and 0.466 for AHP, p < 0.01) observed between the indices and socioeconomic factors, with poverty, percentage of people of color, and housing tenure consistently identified as critical local predictors. The GWR analysis revealed not only the local variability of these factors' influence on lead exposure, but also that incorporating stakeholder knowledge and expert input provides valuable insights that pure data-driven methods may overlook. The study revealed significant spatial variations in lead exposure across New Jersey, identifying localized risk hotspots in urban areas such as Newark. Socioeconomic disparities, particularly poverty levels, percentage of people of color, and rental housing rates, though having spatially varying influences on lead exposure, were found to have significantly influence nonetheless, highlighting important environmental justice concerns. Furthermore, combining expert-driven (AHP) and data-driven (PCA) indices provided complementary insights, emphasizing the value of integrating stakeholder expertise with empirical data for targeted public health interventions.