Association of per-and polyfluoroalkyl substances (PFAS) exposure with cognitive function in firefighters
By Xi Pan, Krystal J. Godri Pollitt, Sheng Liu, Felipe Armstrong, and Bonny Amin
Nature
March 25, 2026
DOI: 10.1038/s41370-026-00861-y
Background
Evidence on cognitive function in relation to per- and polyfluoroalkyl substances (PFAS) exposure in firefighters is scarce.
Objective
We conducted a pilot study aiming to explore the associations of individual PFAS analyte and PFAS mixture with domain-specific cognitive performance in firefighters.
Methods
Firefighters (n = 65), who were actively serving four cities in Central Texas 2003–2025, were assessed for global cognition and processing speed measured by the NIH Toolbox Cognition Battery. Whole blood samples were collected from these participants and analyzed by liquid chromatography-tandem mass spectrometry for 24 PFAS analytes. Linear regression models were used to evaluate associations between cognitive performance and both individual PFAS analytes and the overall 24-PFAS mixture. Additionally, quantile g-computation and Bayesian Kernel Machine Regression (BKMR) were applied to investigate associations between 6-PFAS mixture and cognitive performance, with adjustment for potential covariates.
Results
Six PFAS analytes including potassium perfluoro-1-octanesulfonate (PFOS), perfluoro-n-octanoic acid (PFOA), perfluoro-n-heptanoic acid (PFHpA), perfluoro-n-nonanoic acid (PFNA), perfluoro-n-decanoic acid (PFDA), and potassium perfluoro-1-hexanesulfonate (PFHxS) were detected in the least 70% of participants. PFOS was negatively associated with scores of processing speed (ß = –0.58 p < 0.001). Neither the 6-PFAS mixture nor the 24-PFAS mixture was associated with either cognitive domain.
Significance
These preliminary findings may inform future research seeking to clarify the biological pathways linking PFAS exposure to cognitive outcomes and to explore possible modifiable factors related to cognitive health in firefighters. They might also underscore the potential value of continued effort to minimize workplace PFAS exposure for this occupational group.
Impact
Our study provides the first data on PFAS exposure and cognitive performance for firefighters. Our analysis revealed that PFOS was the primary contributor among the PFAS mixture (consisting of PFOS, PFOA, PFHxS, PFHpA, PFNA, and PFDA) linked to reduced processing speed scores in firefighters. Our findings pave the way for future, larger studies that observe firefighters for longer duration to develop confident estimates of cognitive health in relation to PFAS exposure. These data could inform analyses in support of workplace exposure reduction efforts and revision of preventive health care for first responders.
Topics: