Perfluoroalkyl and polyfluoroalkyl substances and incidence of uterine leiomyomata: a prospective ultrasound study

Abstract

Background: Per- and polyfluoroalkyl substances (PFAS) are persistent synthetic endocrine-disrupting chemicals used in commercial and consumer products, including non-stick coatings, carpeting, and food packaging. We hypothesized that PFAS exposure would influence the incidence of uterine leiomyomata (UL), hormone-dependent neoplasms that are a major source of reproductive morbidity. Methods: We examined the association between PFAS exposure and UL incidence among 1,158 participants from the Study of Environment, Lifestyle, and Fibroids (SELF), a Detroit-based prospective cohort study of Black women aged 23-35 years. At baseline, we collected demographic, behavioral, and medical data via self-administered questionnaires, telephone interviews, and in-person clinic visits. We also measured seven PFAS in non-fasting baseline plasma using liquid chromatography-tandem mass spectrometry (MeFOSAA, PFDA, PFHxS, PFNA, PFUnDA, PFOA, PFOS). Participants underwent transvaginal ultrasound to detect UL at baseline and follow-up (20, 40, 60 months). We estimated hazard ratios (HRs) and 95% confidence intervals (CIs) from Cox models adjusted for age, SES, lifestyle, anthropometrics, and reproductive history. We stratified by parity, which was inversely associated with UL and PFAS. Results: During 60 months of follow-up, we identified 269 incident UL cases. PFHxS, PFNA, PFOA, and PFOS were detected in ≥97% of women; PFDA in 85%, MeFOSAA in 69%; and PFUnDA in 47%. UL incidence was inversely associated with MeFOSAA (≥0.5 ng/ml vs. non-detected: HR=0.56, CI=0.32-0.97), PFDA (≥0.5 ng/ml vs. non-detected: HR=0.59, CI=0.34-1.03), and PFUnDA (detected vs. non-detected: HR=0.70, CI=0.54-0.90), and positively associated with PFHxS (≥1.5 vs. <0.4 ng/ml: HR=1.20, CI=0.75-1.90). PFOA, PFOS, or PFNA showed little association. Parity-stratified results were similar, with the exception of PFHxS, for which a stronger association was observed among parous women (<0.4 vs. ≥1.5 ng/ml: HR=1.67, CI=0.89-3.16). Conclusions: We provide the first epidemiologic evidence on PFAS exposure and UL incidence, indicating positive associations with PFHxS and inverse associations with MeFOSAA, PFDA, and PFUnDA.