Association Between SSRI Exposure During Pregnancy with Behaviors and Conditions Among Children with ASD

Background:  ASDs are heterogeneous disorders that have numerous etiologies, making risk factor identification difficult. Case-group stratification may help researchers identify more homogeneous subgroups that would aid in the assessment of causal mechanisms for ASD. Different symptom profiles or behaviors among children with ASD who were or were not prenatally exposed to selective serotonin reuptake inhibitors (SSRIs) could indicate a potential role of serotonin in ASD development among some children.

Objectives:  To examine associations between prenatal SSRI exposure and the occurrence of developmental and behavioral characteristics, as well as clinical features, in a population-based sample of young children with confirmed ASD.

Methods:  Cases of ASD from the Childhood Autism Risks from Genetics and the Environment (CHARGE) study, a large population-based case-control study, were analyzed to evaluate differences between those who were and were not prenatally exposed to SSRIs. Children were assessed on the ADOS, ADI-R, Vineland Adaptive Behavior Scales (VABS), and Mullen Scales of Early Learning (MSEL). Questionnaires collected extensive comorbidity and medication data for both the mother and child. Mother-child pairs for which self-reported maternal medication history was available totaled 492 children with ASD, 29 exposed and 463 unexposed. Exposure was defined as mothers who reported taking an SSRI between conception and birth. Multivariate linear and logistic regression models were used to assess the relationship between SSRI exposure during pregnancy and child characteristics.

Results:  Cognitive functioning and adaptive behaviors were below average overall, but there was a tendency for exposed children to have higher scores. Generally, exposed children had a greater percentage of co-occurring clinical symptoms than did unexposed children. In multivariate analyses that adjusted for child’s age, maternal age at child’s birth, and maternal birthplace, SSRI exposed children had significantly higher overall MSEL composite scores (β: 6.64; 95% CI: 0.02, 13.26; +0.44 SDs) and VABS communication scores (β: 6.06; 95% CI: 0.51, 11.61; +0.40 SDs), suggesting better cognitive and adaptive communication functioning. In nonverbal children, communication scores were significantly lower as measured by the ADI-R (β: -1.32; 95% CI: -2.50, -0.14; -0.55 SDs), indicating less abnormality. Stereotyped/repetitive behaviors were significantly less abnormal in exposed children on the ADOS (β: -0.88; 95% CI: -1.51, -0.25; -0.24 SDs). The likelihood of having frequent gastrointestinal symptoms was higher in children exposed to SSRIs (AOR: 2.40; 95% CI: 1.05, 5.50), as were the odds of using over-the-counter or prescription medications indicated for any of the following: allergy, the central nervous system, gastrointestinal, or respiratory problems (AOR: 3.49; 95% CI: 1.50, 8.11).

Conclusions:  Children with ASD who were prenatally exposed to SSRIs had better cognitive functioning, as well as less abnormal communication and stereotyped behaviors relative to unexposed children with ASD. GI symptom frequency and medication use were higher in exposed children. Unexplored familial socioeconomic characteristics may have confounding effects on these results. Future studies addressing how prenatal SSRI exposure affects serotonin levels and the functioning of the serotonergic system in a developing child would be invaluable to help understand what may be underlying these findings, and may point to potential etiologic mechanisms for ASD.