Psychostimulant Use During Pregnancy and the Risk of Neurodevelopmental Disorders in Children

Background: The potential for long-term neurodevelopmental impact of in utero exposure to prescription psychostimulants has not been widely studied despite rising prevalence of use during pregnancy.

Objectives: To assess the association between amphetamine/dextroamphetamine and methylphenidate use in pregnancy and neurodevelopmental disorders (NDD) in children, specifically attention-deficit/ hyperactivity disorder (ADHD) and autism spectrum disorder (ASD).

Methods: Pregnant women linked to children in the Medicaid Analytic eXtract (2000-2014) and the IBM MarketScan Research Database (2003-2015) were included. Two primary exposure groups were defined based on ≥1 fill of amphetamine/dextroamphetamine or methylphenidate in the 2nd half of pregnancy, excluding women with fills for both drugs, and compared to unexposed (no psychostimulant fills from 90 days pre-pregnancy to delivery). A secondary definition was ≥1 fill in the 1st half of pregnancy. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated with Cox proportional hazards models for ASD, ADHD, and a composite of any NDD. Confounding by indications, medications, comorbidities, demographics, and lifestyle factors was controlled using propensity score fine stratification weighting. In sensitivity analyses, exposed women were compared to those discontinuing psychostimulants prior to pregnancy to further control for confounding.

Results: There were 3824 amphetamine/dextroamphetamine users and 901 methylphenidate users in the 2nd half of pregnancy, and 12,121 and 3439 in the 1st half, respectively, compared to 3,351,900 unexposed. Unadjusted HR ranged from 1.73-3.08 for ASD, 2.36-3.17 for ADHD, and 1.65-2.25 for any NDD. After adjustment, there was no association between late amphetamine/dextroamphetamine use and ASD (HR 0.96, 95% CI 0.55-1.68), ADHD (1.14, 0.86-1.53), or any NDD (0.90, 0.75-1.07). Late methylphenidate use was associated with ADHD (1.39, 1.01-1.91) but not ASD (1.14, 0.55-2.36) or any NDD (1.09, 0.86-1.37). Early amphetamine/dextroamphetamine use was not associated with any outcome, but early use of methylphenidate was associated with ASD (1.67, 1.16-2.39), ADHD (1.35, 1.13-1.62), and any NDD (1.21, 1.07-1.37). Comparing to psychostimulant discontinuers (N=4446), no association was observed for methylphenidate exposure and any outcome.

Conclusions: Amphetamine/dextroamphetamine exposure in pregnancy was not associated with NDDs in children. Possible associations between methylphenidate and NDDs did not persist in analyses comparing to psychostimulant discontinuers, suggesting unmeasured confounding in the main analysis. These results do not support an association between prescription psychostimulant exposure in utero and NDDs in children.