Abstract
INTRODUCTION: The purpose of this study was to use the US National Health and Nutrition Examination Survey (NHANES) datasets to examine potential relationships between four per- and polyfluoroalkyl substance (PFAS) exposures and each type of cancer, specifically perfluorooctanoic acid (PFOA), perfluorooctane sulfonic acid (PFOS), perfluorohexane sulfonic acid (PFHxS), and perfluorononanoic acid (PFNA). METHODS: A logistic regression analysis was performed. A directed acyclic graph was plotted to adjust for the potential confounders. RESULTS: The odds ratio (OR) of brain cancer for a one-unit increase in ln (PFHxS) was 8.16 (95 % confidence interval [CI] 2.98-68.89). The OR of esophageal cancer for one unit increase of ln (PFOA) and ln (PFOS) was 5.10 (95 % CI 1.18-17.34) and 3.97 (95 % CI 1.24-11.42), respectively. The OR of melanoma for one unit increase of ln (PFOA) and ln (PFHxS) was 1.65 (95 % CI 1.07-2.58) and 1.55 (95 % CI 1.07-2.25), respectively. The OR of prostate cancer for one unit increase of ln (PFOS) and ln (PFNA) was 1.21 (95 % CI 1.00-1.48) and 1.27 (95 % CI 1.00-1.62), respectively. The OR of lung cancer for one unit increase of ln (PFOS) and ln (PFNA) was 2.62 (95 % CI 1.24-5.83) and 2.38 (95 % CI 1.00-5.52), respectively. DISCUSSION: Considering that brain, esophageal, and melanomatous skin cancers have not been targets of epidemiologic studies regarding PFAS exposure, future studies could target these cancers as outcomes of interest.