Plasma levels of per- and polyfluoroalkyl substances (PFAS) are associated with altered levels of proteins previously linked to inflammation, metabolism and cardiovascular disease.

Background: Per- and polyfluoroalkyl substances (PFAS) have been linked to immunotoxic and cardiometabolic effects in both experimental and epidemiological studies, but with conflicting results.

Aim: The aim of the present study was to investigate potential associations between plasma PFAS levels and plasma levels of preselected proteomic biomarkers previously linked to inflammation, metabolism and cardiovascular disease.

Methods: Three PFAS (perfluorooctane sulfonic acid (PFOS), perfluorooctanoic acid (PFOA) and perfluorohexane sulfonic acid (PFHxS)) were measured by non-targeted metabolomics and 249 proteomic biomarkers were measured by the proximity extension assay (PEA) in plasma from 2,342 individuals within the Epidemiology for Health (EpiHealth) study from Sweden (45-75 years old, 50.6 % men).

Results: After adjustment for age and sex, 92% of the significant associations between PFOS concentrations and proteins were inverse (p < 0.0002, Bonferroni-adjusted). The results were not as clear for PFOA and PFHxS, but still with 80% and 64 % of the significant associations with proteins being inverse. After adjustment for age, sex, smoking, education, exercise habits and alcohol consumption, levels of epidermal growth factor receptor (EGFR), and paraoxonase type 3 (PON3) remained positively associated with all three PFAS, while resistin (RETN) and urokinase plasminogen activator surface receptor (uPAR) showed inverse associations with all three PFAS.

Conclusions: Our findings imply that PFAS exposure is cross-sectionally linked to altered levels of proteins previously linked to inflammation, metabolism and cardiovascular disease in middle-aged humans.