Per- and polyfluoroalkyl substances in drinking water and cancer prevalence in the United States
Menée à l'aide des données d'analyses publiques concernant 76 606 zones desservies en eau potable, cette étude analyse l'association entre la prévalence d'antécédents de cancer chez les adultes et les concentrations en substances per- et polyfluoroalkylées (PFAS) dans l'eau
Background: Research concerning the adverse health effects of per- and polyfluoroalkyl substances (PFAS) continues to grow. With the recent releases of nationwide data on PFAS in drinking water and public drinking water system service boundaries, it is now possible to conduct nationwide geospatial analyses on the relationships between PFAS in drinking water and aspects of health.
Objective: To examine associations between PFAS in drinking water and cancer history prevalence in the United States.
Methods: We examined cancer history prevalence, at the census tract level, among adults aged ≥ 18 years diagnosed with any cancer in or prior to 2022 using the United States Population Level Analysis and Community EStimates dataset. We obtained data for PFAS in public drinking water from the Fifth Unregulated Contaminant Monitoring Rule (UCMR 5, 2023-ongoing). We used geographic information systems to spatially join water system boundaries (n = 9,733) with census tracts applying population-weighted areal interpolation. We calculated prevalence ratios (PRs) and 95% confidence intervals (CIs) for the associations between PFAS in drinking water and prevalence of cancer history, adjusted for census tract-level sociodemographics, health conditions and behaviors, and environmental factors.
Results: This analysis included 76,606 census tracts with an average cancer history prevalence of 7.8%. We observed positive associations of cancer history prevalence with PFAS levels in drinking water for 6:2-FTS, PFBA, PFBS, PFHpA, PFHxA, PFHxS, PFNA, PFOA, PFPeA, and PFPeS (p < 0.01). For example, for 6:2-FTS, the adjusted PR comparing the highest quintile (0.0182–0.663 µg/L, population-weighted) to samples below the minimum reporting level (< 0.005 µg/L) was 1.04 (95% CI 1.02–1.07, p < 0.001). No associations were observed for HFPO-DA and PFOS. Models mutually adjusted for correlated PFAS showed generally similar results.
Significance: Higher levels of certain PFAS in drinking water were independently associated with higher cancer history prevalence. Future research should examine the relationships between individual-level cancer outcomes and individual-level exposure to PFAS in drinking water.
Environmental Health , article en libre accès, 2026