Gestational Endocrine Disruption and Autism Spectrum Disorders

Abstract

Background Autism Spectrum Disorder (ASD) is a neurodevelopmental communication disorder with increasing prevalence. Emerging yet conflicting evidence suggests that alterations in the fetal endocrine environment, and specifically in thyroid and androgen signaling, play important roles in the development of the disorder. This question is of high importance, as endocrine problems among women of reproductive-age are relatively common and potentially treatable, and exposures to endocrine-disrupting chemicals are ubiquitous. Better understanding of the involvement of endocrine-related factors in ASD will advance the etiological understanding of the disorder, including the possible contribution of environmental risk factors. To this aim, we evaluated the associations between atypical in-utero thyroid and androgen exposures, and ASD risk. Methods We used a population-based cohort of nearly 500,000 births from a large Israeli health fund, with data abstracted from medical and birth records. Complete validation of all suspected ASD cases was performed. Since androgen levels are not routinely measured clinically, maternal diagnoses of Polycystic Ovarian Syndrome (PCOS) and related conditions, and male newborns’ diagnoses of cryptorchidism and hypospadias, were used as proxy indicators for children presumed to be gestationally exposed to hyper- or hypoandrogenemia, respectively. Associations of ASD risk with maternal thyroid dysfunction were evaluated, and specific effects of gestational thyroid hormone levels were explored. Analyses were performed using generalized estimating equations (GEE) and additive models (GAM), with additional models to assess mediation by androgen-related comorbidities. Results Overall, we observed that children presumed to be exposed to either hyper- or hypo-androgenic in-utero environment had elevated ASD risks. The effect of hyperandrogenemia as indicated by PCOS on ASD risk was largely independent of androgen-related metabolic and cardiovascular comorbidities. The associations with hypoandrogenemia-related male congenital malformations were pregnancy-specific, despite familial aggregation of each disorder individually. Children of mothers with thyroid dysfunction had higher ASD risks, although analyses of gestational thyroid hormone levels suggested that the effects were not directly caused by dysregulation of thyroid hormone concentrations. Conclusions Overall, the results indicate the involvement of endocrine-related factors in ASD etiology, with stronger evidence for androgen hormones. The results additionally suggest that exposure to exogenous factors may drive some of the observed associations.