Maternal exposure to per- and polyfluoroalkyl substances and epitope level antibody response to vaccines against measles and rubella in children from the Boston birth cohort

Abstract

Background: Previous studies suggest that per- and polyfluoroalkyl substances (PFAS) may act as immune suppressants. However, research about the impact of PFAS exposure on antibody responses to the measles, mumps, rubella (MMR) vaccine is limited and inconsistent.

Methods: This report includes 748 mother-child pairs from the Boston Birth Cohort, with 8 PFAS compounds measured in maternal plasma shortly after delivery. IgG reactivities to measles and rubella were profiled in cord blood and venous blood plasma during early childhood, using Phage ImmunoPrecipitation Sequencing. Linear regression models were applied to assess the relationships between log2-transformed PFAS and IgG reactivities as measured by Viral Aggregate Reactivity score (VARscore, with inverse normal transformation) for measles and rubella. Quantile g-computation was applied to evaluate the PFAS mixture - VARscore associations.

Results: The detection rate for 8 PFAS compounds ranged from 90 % to 100 % in maternal plasma. Maternal PFAS burden score (P = 0.01), but not individual PFAS compounds, was associated with lower VARscore for measles in cord blood. In 348 children after receiving the MMR vaccine, three maternal PFAS compounds (Me-PFOSA-AcOH, PFHpS and PFHxS) were significantly associated with lower measles VARscore (P < 0.05). Me-PFOSA-AcOH and PFHxS were significantly associated with higher risk of having low reactivity to measles defined as VARscore < 25th percentile. PFAS mixture analysis revealed a significant inverse association between quantile of the PFAS mixture and measles VARscore (P = 0.025) in children after vaccination, with PFHxS as an important contributor to this association. These inverse associations were more pronounced in Black children (compared to non-Black children) and in preterm children (compared to term children). In comparison, no associations were found for rubella VARscore.

Conclusions: This prospective birth cohort study provides suggestive evidence that maternal PFAS exposure is associated with a reduced immune response to the measles vaccine, especially, among Black or preterm children.

Keywords: Effect modifier; Immune Response; Measles; Per- and polyfluoroalkyl substances; PhIP-Seq; Vaccination.

Fig. 1.

Flow chart of participant enrollment in this study.

Fig. 2.

Violin plots for the distribution of VARscore at birth, during early childhood from pre-vaccination children, and during early childhood from post-vaccination children, respectively. The dotted horizontal line represents VARscore of 1.

Fig. 3.

Forest plots for the associations between maternal PFAS levels after delivery and inversely normalized VARscores for measles and rubella during early childhood in 348 post-vaccination children, adjusted for race and ethnicity, maternal age at delivery, maternal smoking during pregnancy, delivery type, preterm birth, feeding modality, infant’s sex and years since last vaccination.

Fig. 4.

Forest plots for the associations between maternal PFAS levels after delivery and risk of having low IgG reactivity to measles and rubella during early childhood in 348 post-vaccination children, adjusting for race and ethnicity, maternal age at delivery, maternal smoking during pregnancy, delivery type, preterm birth, feeding modality, infant’s sex and years since last vaccination. Low IgG reactivity is defined as VARscore < 25th percentile.

Fig. 5.

Quantile-based g-computation mixture effect sizes and positive or negative weights for each maternal PFAS compound in association with inversely normalized measles VARscore or with risk of having low IgG reactivity to measles during early childhood in 348 post-vaccination children.