Associations of perfluoroalkyl substances (PFAS) with non-targeted metabolomics across two independent cohorts of elderly individuals

Abstract

BACKGROUND AND AIM: Perfluoroalkyl substances (PFAS) have been linked to adverse health effects including increased blood cholesterol levels and hepatic toxicity. Alterations in metabolism and oxidative stress are among the hypothesized mechanisms by which PFAS exposure may exert effects; however, studies of associations between PFAS and metabolome are scarce. The main aim of this study was to assess associations of PFAS exposure and metabolome by using non-targeted metabolomics in two independent cohorts of elderly individuals. METHODS: In two population-based cohorts of elderly individuals from Sweden and California (the Prospective Investigation of the Vasculature in Uppsala Seniors [PIVUS, n = 901, median age 70 years]; and the California Teacher Study [CTS, n = 325, median age 69 years]), plasma/serum PFAS concentrations were determined using isotope-dilution ultra-pressure liquid chromatography (UHPLC) coupled to tandem mass spectrometry. Non-targeted metabolomics was performed using UHPLC high-resolution mass spectrometry (QTOFMS). Associations of PFAS and metabolomics signatures was assessed using a discovery-validation approach with multivariable linear regression. RESULTS: In the discovery cohort 89 annotated metabolites out of 204 could be associated with levels of five PFAS following adjustment for age, sex, and correction for multiple testing. Eleven associations could then be replicated in the CTS. The metabolites being replicated in the CTS were primarily lipid related metabolites such as lysophosphatidylcholine, phosphatidylcholine, sphingomyelin as well as amino acids such as betaine. CONCLUSIONS: Using a non-targeted metabolomics discovery-validation approach, we the previously reported associations between PFAS and lipid related metabolites and identified and replicated novel associations. Our results encourage additional studies investigating the underlying mechanisms of our findings. KEYWORDS: PFAS, PFOS, PFOA, Metabolomics, Lipid metabolism