LAUSR

INTRODUCTION The objective of this study was to validate the hypothesis that increased serum concentrations of perfluoroalkyl substances (PFASs) cause kidney damage. A causal interpretative study was designed using the US 2003-2018 National Health and Nutrition Examination Survey (NHANES) datasets. METHODS Three statistical models, including multivariable linear regression, generalized additive model, and regression discontinuity model (RDM), were applied to the US 2003-2018 NHANES datasets to evaluate the causal relationship between the four PFAS agents and estimated glomerular filtration rate (eGFR). Directed acyclic graphs were plotted for a more valid causal inference. RESULTS AND DISCUSSION In the RDM, when the natural logarithm of each PFAS agent increases by 1 ng/mL after each cut-off value, eGFR decreased 4.63 mL/min/1.73 m2 for perfluorooctanoic acid, 3.42 mL/min/1.73 m2 for perfluorooctane sulfonic acid, 2.37 mL/min/1.73 m2 for perfluorohexane sulfonic acid, and 2.87 mL/min/1.73 m2 for perfluorononanoic acid. The possibility of reverse causation that increased serum PFAS concentration is the consequence of reduced eGFR, not the cause, was low, and an additional adjustment of potential confounders was not needed. CONCLUSION This study contributes to the understanding of PFAS-induced kidney damage. Further longitudinal epidemiological and toxicological studies are recommended.