Background Per- and poly-fluoroalkyl substances (PFAS) are a class of emerging persistent organic pollutants, and their negative health impacts have been widely concerned. There is a lack of epidemiological studies on the associations of PFAS exposure with lipid homeostasis.

Objective To investigate the associations of perfluorooctanesulfonic acid (PFOS) and perfluorooctanoic acid (PFOA) exposure with blood lipid levels and dyslipidemia in middle-aged and elderly women.

Methods This study was based on 795 middle-aged and elderly women from a female sub-cohort of the Dongfeng-Tongji cohort study, excluding the participants without blood lipid measurements and/or reported use of lipid-lowering drugs at baseline. The concentrations of plasma PFOS and PFOA were measured by ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). The concentrations of serum total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C) were measured using an automatic analyzer. The normally distributed continuous variables were presented as mean ± standard deviation ( \barx\pm s) , while the skewedly distributed continuous variables were presented as median (M) and the 25^th and 75^th percentiles (P₂₅, P₇₅). Generalized linear models (GLM), generalized additive models (GAM), and logistic regression models were applied to evaluate the associations of PFOS and PFOA exposure with blood lipid levels and the risk of dyslipidemia. Stratified analyses were also implemented to explore potential modification effects of sociodemographic characteristics, lifestyles, and menopause on associations of PFOS and PFOA exposure and blood lipids.

Results The \barx\pm s of baseline age for the study participants was (59.4±8.6) years old, and their baseline body mass index (BMI) was (24.3±3.4) kg·m⁻². The M (P₂₅, P₇₅) of baseline plasma concentrations for PFOS and PFOA were 9.96 (6.24, 15.09) μg·L⁻¹ and 1.20 (0.84, 1.80) μg·L⁻¹ respectively. The GLM analysis showed that plasma concentration of PFOS was positively associated with serum HDL-C b (95%CI): 0.04 (0.01, 0.07). The plasma concentration of PFOA was also positively associated with serum TC b (95%CI): 0.05 (0.02, 0.08) and serum LDL-C b (95%CI): 0.05 (0.01, 0.09). No significant association was observed between plasma PFOS and serum TC, TG, or LDL-C, nor between plasma PFOA and serum TG or HDL-C. The stratified analyses showed that the association between PFOA and LDL-C was significant among the participants aged <60 years old b (95%CI): 0.06 (0.01, 0.11), P=0.014. A modification effect was observed for age on the association of plasma PFOA with serum LDL-C, with P_interaction=0.046. The analysis of the associations between PFOS/PFOA exposure and the risk of dyslipidemia showed that an increased plasma PFOA was significantly associated with an increased risk of hypercholesterolemia, with an OR (95%CI) of 1.69 (1.23, 2.15). No association was observed between PFOS exposure and the risk of dyslipidemia.

Conclusion This cross-sectional study reveals that common PFAS exposure could affect the homeostasis of blood lipids based on the female sub-cohort of the Dongfeng-Tongji cohort, which provides new evidence for the negative health impact of PFAS.