To Örebro University

Perfluoroalkyl substances (PFAS) have been linked to immunotoxicity in experimental studies. Although PFAS exposure is associated with altered immune response in epidemiological studies of children, it is less known whether this is observed also in elderly adults. Eight PFAS and 86 proteins were measured in plasma from 965 elderly individuals from Sweden (all aged 70, 50% women). PFAS were measured using isotope-dilution ultra-pressure liquid chromatography coupled to tandem mass spectrometry. Proteins were measured using a multiplex proximity extension assay (PEA) and covered among others inflammatory marker proteins such as monocyte chemoattractant proteins, tumor necrosis factors, and interleukins. We examined cross-sectional associations using multivariable linear regression at two levels of adjustment. We observed significant decreases in levels of 24 proteins in relation to a ln-unit increase in PFAS concentrations following adjustment for sex, sample storage time in freezer, and correction for multiple testing. Associations of PFAS and hepatocyte growth factor (HGF) and macrophage colony-stimulating factor 1 (CSF-1) remained significant (p-value < 0.05) following full covariate adjustment for smoking, exercise habits, education, energy, and alcohol intake, body mass index (BMI), glomular filtration rate (GFR) as well as corticoid- and COX-inhibitor treatment. CSF-1 was inversely associated with perfluorohexane sulfonic acid (PFHxS) beta: -0.08: 95% confidence interval (CI); -0.13, -0.02), perfluorooctanoic acid (PFOA) beta: -0.04: 95% CI; -0.07, -0.006, perfluorononanoic acid (PFNA) beta: -0.04: 95% CI; -0.08, -0.003, perfluorodecanoic acid (PFDA) beta: -0.03: 95% CI; -0.06, -0.003, and perfluoroundecanoic acid (PFUnDA) beta: -0.05: 95% CI; -0.08, -0.02. The magnitude and direction of PFAS vs protein relationships were similar also for HGF. Our findings implicate PFAS exposure with decreased levels of proteomic markers of inflammation in elderly humans.