Serum per- and polyfluoroalkyl substances and prevalent heart disease

Abstract

Previous research investigating exposure to per- and polyfluoroalkyl substances (PFAS) and their association with heart disease has found conflicting results, with recent studies reporting null or inverse associations. We estimated associations between serum PFAS and prevalent heart disease in a large, cross-sectional, multi-site study of adults residing in communities in the United States with historic PFAS drinking water contamination. Participants were recruited from communities in eight states (2019-2023), where they completed questionnaires and provided blood samples. PFAS concentrations were measured in serum. Four PFAS detected in > 70% of participants were modeled as continuous variables, and three less-frequently detected PFAS were modeled as binary variables (above/below the limit of detection). Logistic regression models estimated associations between each PFAS and prevalent heart disease based on self-report, adjusting for demographics, smoking and alcohol consumption, waist-to-hip ratio, estimated glomerular filtration rate (eGFR), and study site. Quantile g-computation models were used to estimate the joint effects of PFAS as a mixture. A total of 5,761 adult participants were included in the analysis, with a mean age of 56 years and 60.2% female. The prevalence of self-reported heart disease was 9.2%. Adjusted analyses showed no statistically significant associations between all PFAS and self-reported heart disease. Our mixture analysis revealed no association between PFAS as a mixture and prevalent self-reported heart disease (OR: 0.99, 95% CI: 0.95-1.03). Our findings suggest largely null associations between serum PFAS and self-reported heart disease.