Abstract
Per- and polyfluoroalkyl substances (PFAS) are persistent environmental chemicals that have been linked to pancreatic cancer in animal studies, but evidence in humans remains limited. We evaluated associations between pre-diagnostic circulating concentrations of 11 PFAS and pancreatic cancer risk in a nested case-control study within the New York University Women's Health Study, a prospective cohort of 14,273 women. Serum samples collected prior to diagnosis from 116 pancreatic cancer cases and 346 matched controls were analyzed for PFAS levels. Associations between individual PFAS and pancreatic cancer risk were assessed using multivariable conditional logistic regression, and mixture effects were examined using repeated holdout weighted quantile sum (WQSRH) regression and Bayesian kernel machine regression (BKMR). Ten PFAS met the inclusion criteria for analysis. In single-chemical models, N-methyl perfluorooctane sulfonamidoacetic acid (NMFOSAA) and perfluorooctanoate (PFOA) were inversely associated with pancreatic cancer risk, but associations did not remain significant after false discovery rate correction. WQSRH models with negative constraints suggested a borderline inverse association with the PFAS mixture, primarily driven by NMFOSAA, PFOA, and perfluoroundecanoate (PFUNDA), while no significant associations were observed in positively constrained WQSRH or BKMR models. Overall, these findings do not support a positive association between circulating PFAS and pancreatic cancer risk in women. Although some compounds showed suggestive inverse associations, the overall findings indicate no clear link between PFAS exposure and pancreatic cancer risk.