Prenatal exposure to per- and polyfluoroalkyl substances (PFAS) and incidence of asthma and wheeze in childhood: A register-based cohort study in Ronneby, Sweden.

BACKGROUND

Early-life exposure to per- and polyfluoroalkyl substances (PFAS) may impact the developing lungs and immune system and increase the risk of childhood asthma, but no studies have been conducted in a high-exposed population. The objective of this study was to estimate associations between prenatal PFAS exposure and childhood incidence of asthma and wheeze in Blekinge County, Sweden, where a subset of residents in the city of Ronneby was exposed to PFAS from drinking water contaminated by aqueous film-forming foam (AFFF).

METHODS AND FINDINGS: We constructed a register-based open cohort of 11,488 children born in Blekinge county between 2006 and 2013 and followed each individual from birth until age 12 or December 31, 2022. Maternal address history was linked to water distribution records to create a categorical proxy variable for prenatal PFAS exposure from drinking water. We identified incident cases of wheeze and asthma from administrative health records and estimated hazard ratios (HRs) using Cox proportional hazards models adjusted for individual-level confounders, including maternal smoking in early pregnancy, maternal age at delivery, parity, child sex, parental asthma, and socioeconomic factors. As a secondary analysis, we applied a Rubin Causal Model (RCM) analysis to estimate the average marginal effect of prenatal PFAS exposure on wheeze and asthma among the very highly-exposed population, using a matched dataset of very-high and background-exposed individuals balanced on measured confounders. Overall, 18% of children were diagnosed with wheeze and 17% with asthma during follow-up. Very high prenatal PFAS exposure was associated with incidence of asthma (HR: 1.44, 95% CI [1.08, 1.92]), whereas no associations were observed for the high or intermediate exposure groups or for wheeze. In the RCM analysis, the estimated cumulative incidence of asthma was 16.1% in the background-exposed group and 26.7% in the very highly exposed group (Fisherian p < 0.001). Study limitations include reliance on an address-based categorical proxy for prenatal PFAS exposure, which likely results in non-differential exposure misclassification and limits the ability to distinguish prenatal from early-childhood exposure effects.

CONCLUSIONS

In this study, very high prenatal PFAS exposure was associated with a higher incidence of childhood asthma. Although these results should be replicated, they suggest an important public health impact of AFFF-associated PFAS contamination.