Perfluoroalkyl substances and cognitive function in older adults: Should we consider non-monotonic dose-responses and chronic kidney disease?

Although potential neurotoxicity of perfluoroalkyl and polyfluoroalkyl substances (PFAS) is suggested, previous epidemiologic studies have reported a 'protective' association between serum PFAS concentration and cognition function. Poor outcome assessment, residual confounding, non-monotonic dose-responses (NMDRs), and the role of reduced kidney function in PFAS excretion may be alternative explanations of these findings.We examined the association of perfluoroalkyls with cognitive functions assessed using the Consortium to Establish a Registry for Alzheimer's Disease word learning and recall; the Animal Fluency; and the Digit Symbol Substitution tests.We included 903 adults aged ≥60 years from the National Health and Nutrition Examination Survey (NHANES) 2011-2014. We computed a composite z-score as an average of four individual cognitive z-scores and used it as the outcome. Linear and generalized additive models were used to evaluate linear and non-linear associations.With the linearity assumption, perfluorooctanoate (PFOA) and perfluorononanoate (PFNA) were significantly positively associated with composite z-score after adjustment for age, sex, race/ethnicity, education, smoking, poverty-income ratio, health insurance, food security, alcohol, and physical activity. Smoothing plots suggested NMDRs, especially for perfluorooctane sulfonate (PFOS) with a U-shape dose-response. When restricting to participants without chronic kidney disease (CKD) (n = 613), the positive associations for PFOA and PFNA observed in the whole population diminished, whereas PFOS was inversely and significantly associated with composite z-score. Also, negative confounding effects of fish/seafood consumption seem to be substantial. Effect estimates of composite z-score were -0.055 (95% CI: -0.097, -0.012, P = 0.01) for a doubling increase in PFOS.These findings suggest that the previous epidemiologic findings of a 'protective' association between PFAS and cognition may be explained by CKD, NMDRs and confounding by fish consumption. PFOS at the current population exposure level in the U.S. may be a risk factor for cognitive decline in older adults with normal kidney function.