Can PFAS exposure alter male sex hormones?

PFAS are persistent environmental chemicals with endocrine-active signals in experimental systems. In men, the strongest human evidence supports association—not proof of causation—between particular compounds and circulating hormone measures.

Clinical evidence

• A 2024 meta-analysis of 12 studies (7,506 males) found higher perfluorononanoic acid (PFNA) associated with lower testosterone (β = −0.05; 95% CI, −0.09 to −0.02). Adolescent NHANES analyses also linked PFNA and PFHxA with lower testosterone.
• This is not a uniform “PFAS effect.” In other NHANES analyses, PFDA, PFOS, and PFHxS were associated with higher testosterone; PFOA and PFHxS have been linked to higher estradiol, while PFNA was not associated with estradiol. PFOS results are particularly conflicting.

Mechanistic evidence

• In cell and animal models, PFAS can alter steroid output and expression of steroidogenic genes, including CYP19A1, HSD3B2, and CYP17A1. Human granulosa-cell studies found mixture-dependent changes in progesterone, estradiol, and STAR protein.
• Several PFAS show weak ERα/ERβ activation or ERα binding and, in some reporter assays, antagonize androgen-receptor activity or modify testosterone-induced receptor translocation. PPAR-mediated effects on lipid and cholesterol handling are another plausible route to altered steroid synthesis.
• These effects are compound-, concentration-, and cell-type-dependent; many occur above typical serum concentrations.

Interpretation

• Most human evidence is cross-sectional and heterogeneous by PFAS, age, assay, exposure matrix, and confounder adjustment. Hormonal status may also affect PFAS distribution and clearance, creating credible reverse causation.
• A PFAS blood concentration does not establish the exposure source, predict endocrine disease, or show that PFAS caused an individual’s symptoms.

Bottom line

• PFAS—especially PFNA—are associated with altered male hormone measures, and experimental evidence supports plausible interference with steroidogenesis and receptor signaling. However, human findings remain inconsistent and do not establish a consistent, clinically actionable causal effect in an individual man.