Prenatal exposure to mixtures of per- and polyfluoroalkyl substances and organochlorines affects cognition in adolescence independent of postnatal exposure

Background

Studies on cognitive and neurodevelopmental outcomes have shown inconsistent results regarding the association with prenatal exposure to perfluoroalkyl substance (PFAS) and organochlorines. Assessment of mixture effects of correlated chemical exposures that persist in later life may contribute to the unbiased evaluation and understanding of dose-response associations in real-life exposures.

Methods

For a subset of the 4th Flemish Environment and Health Study (FLEHS), concentrations of four PFAS and six organochlorines were measured in respectively 99 and 153–160 cord plasma samples and 15 years later in adolescents' peripheral serum by Ultra Performance Liquid Chromatography-Tandem Mass Spectrometry (UPLC-MS/MS). Sustained and selective attention were measured at 14–15 years with the Continuous Performance Test (CPT) and Stroop Test as indicators of potential neurodevelopmental deficits. Quantile g-computation was applied to assess the joint associations between prenatal exposure to separate and combined groups of PFAS and organochlorines and performance in the CPT and Stroop Test at adolescence. Subsequently, individual effects of each chemical compound were analyzed in mixed effects models with two sets of covariates. Analytical data at birth and at the time of cognitive assessment allowed for off-setting postnatal exposure.

Results

In mixtures analysis, a simultaneous one-quantile increase in the natural log-transformed values of PFAS and organochlorines combined was associated with a decrease in the mean reaction time (RT) and the reaction time variability (RTV) in the CPT (β = -15.54, 95% CI:-29.64, -1.45, and β = -7.82, 95% CI: -14.97, −0.67 respectively) and for the mixture of PFAS alone with RT (β = -11.94, 95% CI: -23.29, -0.60). In the single pollutant models, these results were confirmed for the association between perfluorohexanesulfonate (PFHxS) with RT (β = -17.95, 95% CI = -33.35, -2.69) and hexachlorobenzene with RTV in the CPT (β = -5.78, 95% CI: -10.39, -0.76). Furthermore, the participants with prenatal exposure above the limit of quantification for perfluorononanoic acid (PFNA) had a significantly shorter RT and RTV in the CPT (β = -23.38, 95% CI: -41.55, -5.94, and β = -9.54, 95% CI: -19.75, -0.43, respectively).

Conclusion

Higher prenatal exposure to a PFAS mixture and a mixture of PFAS and organochlorines combined was associated with better sustained and selective attention during adolescence. The associations seemed to be driven by PFHxS and were not linked to exposure levels at the time of assessment.