Prenatal Air Pollution Exposures, Maternal Cytokine/Chemokines, and Risk of Autism Spectrum Disorder: The Early Markers for Autism (EMA) Study
Objectives: In this study, we examined the relationship between mid-gestational air pollution exposures and levels of maternal serum cytokines/chemokines to determine their potential combined associations with risk of ASD.
Methods: Study participants were from a large population-based, nested case-control study of archived maternal mid-pregnancy specimens from the same mother-infant pairs from Southern California. Study groups were identified through the linkage of California birth records, Department of Developmental Services (DDS) records, and the California Prenatal and Newborn Screening Program records. This study includes mothers of children with ASD (n=378), intellectual disability (ID) but not ASD (ID; n=165),or children with no known neurodevelopmental disability from the general population (GP; n=416). The ASD group was further divided into those with ID (DQ<70) (ASD+ID; n=181) and those without (DQ ≥70) (ASD-noID; n=197). To estimate air pollutant exposure during the second trimester of pregnancy, maternal residential addresses reported during prenatal screening visits and on the infant’s birth certificate were used. Regional air pollutant measures (NO2, PM10, PM2.5, Ozone, NO) based on the Environmental Protection Agency’s Air Quality System data were averaged for one month before the maternal serum sample collection date. Stored maternal serum samples, collected during weeks 15-19 of gestation for routine prenatal screening, were quantified for 22 cytokines/chemokines using Luminex multiplex analysis technology. Correlations between prenatal levels of air pollutants and maternal cytokines/chemokines were examined separately for each case group using Spearman’s correlation. Resulting correlations were then compared between each diagnostic group and GP controls using Fisher’s z-transformation to assess whether correlations between exposures and cytokine/chemokines differed across groups.
Results: The correlation between measured air pollutants and maternal cytokines/chemokines were generally weak (<0.2 in absolute value). Distinct patterns of correlation between cytokines/chemokines and air pollutants were observed for each diagnostic group. For example, the correlations between the allergy and asthma-associated cytokine IL-4 and NO2 (p=0.03) and PM10 (p=0.01) differed significantly between the ID and GP groups, and were greater in magnitude among ID, while a significant difference in the magnitude of correlation between ozone and IL-4 showed an inverse relationship in ID (p=0.01). The magnitude of association between IL-4 and ozone was significantly different for ASD+ID than for GP (p=0.04) and showed inverse association among ASD+ID. No significant differences in the magnitude of correlation between air pollutants and cytokines/chemokines were found between ASD-noID and GP groups. Additional analyses will quantify relationships between ASD risk and joint effects of air pollution and maternal immune markers, as well as explore potential mediation.
Conclusions: Correlations between mid-gestational maternal cytokines/chemokines and air pollutants varied in direction and magnitude across the different diagnostic gropups. The observed relationships between prenatal air pollutant exposures and maternal cytokines/chemokines may thus provide further insight of their potential role(s) in neurodevelopment, particularly in the context of risk of ASD and ID.