Antidepressants during Pregnancy and Autism Risk: Update from the Stockholm Youth Cohort

Background:  It is still uncertain whether exposure to antidepressants during fetal life is causally linked with autism spectrum disorders (ASD). All seven epidemiologic studies published to date found crude associations between antidepressant use during pregnancy and offspring ASD. However, there was evidence of confounding, raising concern that the observed association reflected the risk of autism due to the underlying condition for which the medications were prescribed. One study reported that the risk of ASD with antidepressant use during pregnancy attenuated after adjustment for maternal depression, but an association with attention deficit hyperactivity disorder (ADHD) persisted. Further research is required to establish whether these associations are causal in order to inform clinical guidance.

Objectives:  To assess the association between maternal antidepressant use during pregnancy with offspring ASD (with and without intellectual disability), ADHD and intellectual disability; ii) to carry out analyses to strengthen causal inference, including using paternal antidepressant use during the period of pregnancy as negative control exposures, and propensity score matching methods to account for confounding by indication.

Methods:  Cohort study using the subset of the Stockholm Youth Cohort born between 1 January 1996 and 31 December 2010 (n=414,105). Data on parental depression and other potential confounders were prospectively recorded before the birth of the child. The sample was divided into two sub-cohorts: first the children with data on maternal report of medication use collected at the first antenatal interview (cohort born between 1996 and 2004, n=194,466); and second those with information on parental antidepressant use extracted from prescription records using the national patient register (cohort born after 2006, n=135,189). Children born in 2005 were excluded due to lower quality coverage of medication use in the registers. A total of 4,184 ASD cases were identified within these sub-cohorts using multisource case-ascertainment. Cluster robust logistic regression was used for primary analysis.

Results:  In both sub-cohorts, a maternal history of depression was associated with an increased risk of offspring ASD but this association was not present for paternal depression. Maternal use of antidepressants was associated with higher risk for ASD in both cohorts after adjustment for birth-year, sex, parity, parental ages, family income, parental country of birth, education and history of depression (adjusted OR sub-cohort 1: 2.07 (1.62, 2.65); OR for subcohort 2: 2.13 (1.24, 3.67)). A higher risk of ADHD was also observed with maternal antidepressant use during pregnancy in both cohorts (subcohort1: OR=1.89 (1.54, 2.33); subcohort2: OR=2.13 (1.01, 4.50)) after adjustment for the above covariates, but there was no heightened risk observed for individuals with intellectual disability. The negative control and propensity score matched analyses are underway.   

Conclusions:  The associations mirror previous findings of associations between antidepressant use during pregnancy and offspring ASD, and were observable using both maternal self report and prescription records data. The results of our ongoing analyses will be informative in relation to the causal meaning of these findings.