UC Irvine

Environmental contamination is unevenly distributed across communities in the United States, reflecting historical and ongoing patterns of structural racism, industrial siting, infrastructure investment, and regulatory protection. Per- and polyfluoroalkyl substances (PFAS), lead, arsenic, and other persistent contaminants may therefore be encountered within broader place-based conditions that shape exposure and health. This dissertation used a riskscape-informed environmental justice framework to examine environmental burden across three analytic scales: U.S. counties, individuals residing in Orange County, California, and residential properties near a secondary lead-smelting facility in the City of Industry, California. The conceptual model situates structural racism as an upstream determinant of differential environmental conditions and exposures, while incorporating biological vulnerability as a pathway through which cumulative stress may alter chemical-health associations.Aim 1 examined associations between PFAS concentrations in public drinking water and all-cause and cause-specific mortality at the county level. All-cause mortality associations were generally null. Cause-specific estimates varied by PFAS compound, mortality definition, and exposure operationalization. PFHpA was positively associated with cerebrovascular mortality in continuous models, and PFHxS was positively associated with skin-cancer mortality in both continuous and detection-based models. Several inverse associations were also observed, particularly for PFHpA and cardiovascular, cancer, and metabolic outcomes. These inverse estimates were not interpreted as protective effects because residual geographic confounding, exposure and outcome misclassification, competing risks, and survival or selection processes could not be distinguished in an ecological design.Aim 2 used data from the Orange County site of the Centers for Disease Control and Prevention/Agency for Toxic Substances and Disease Registry Multi-Site Study to assess serum PFAS, lipid outcomes, and effect modification by the Environmental Justice Index. Most crude PFAS-lipid associations weakened after covariate adjustment. PFUnDA detection remained associated with higher high-density lipoprotein cholesterol after Bonferroni correction. MeFOSAA detection showed evidence of interaction with selected Environmental Justice Index measures: its association with prevalent high cholesterol was stronger in lower-burden settings and attenuated in higher-burden settings. This exploratory finding did not support a simple cumulative-burden amplification hypothesis and requires replication with larger samples and source-specific exposure data.Aim 3 partnered with Environmental Justice San Gabriel Valley to characterize metals in residential-yard soil near the Ecobat secondary lead-smelting facility. Sampling at 136 properties produced 369 soil samples. Lead concentrations declined with distance from the facility, and elevated concentrations clustered nearby. Although the study could not establish source attribution, 25.2% of samples exceeded the U.S. Environmental Protection Agency’s 2024 recommended residential soil-lead screening level of 200 mg/kg. Together, the findings show that environmental health research benefits from examining exposure within place-based social and environmental context. The dissertation supports cumulative-impact approaches that combine direct monitoring, attention to exposure pathways, and meaningful community participation.