Maternal Depression and Antidepressant Use During Pregnancy and Risk of Childhood Autism Spectrum Disorders (ASD) in Offspring: Population-based Cohort and Bidirectional Case-Crossover Sibling Study

Study type
Protocol
Date of Approval
Study reference ID
15_256
Lay Summary

Depression is a common condition in women of childbearing age and it has been suggested that both depression and antidepressant use during pregnancy increase the chances of having an adverse outcome in the child. Rates of autism spectrum disorders (ASD), a disorder where people have trouble with social interaction, communicating with others, and behavioural challenges, have been increasing over the last few decades. It is possible that antidepressant use during pregnancy is associated with the development of ASD. Using the CPRD, we will evaluate maternal exposure to antidepressant drugs and/or depression in pregnancy and the risk of ASD in the child.

Technical Summary

Using the Mom-Baby linkage, we will conduct a cohort study of maternal depression and prenatal antidepressant medication use during pregnancy and the risk of ASD in children. We will consider three exposure categories; a) Antidepressant and Depression Exposed, b) Depression Only Exposed, and c) Antidepressant Only Exposed, each compared to a matched (4:1) unexposed cohort of women who have neither depression nor use of antidepressant medications. Defining the exposed cohort in this way will allow for independent evaluation of the effects of depression, the combined effect of antidepressants and depression, and the effect of antidepressants in the absence of depression on the risk of ASD in offspring. We will calculate cumulative incidence rates (IRs) of ASD and corresponding 95% confidence intervals (CIs). We will also conduct a ‘within-mother-between-pregnancy’ nested case-crossover sibling analysis to estimate odds ratios (ORs) and corresponding 95% CIs for the association between maternal prenatal antidepressant use and risk of ASD in the offspring. Mother’s age, sex, calendar time, and other potential confounders identified in the main analysis will be included in the regression model. We will examine confounding for a wide variety of additional medical and behavioral covariates using matching, stratification, and multivariable analysis.

Collaborators

Susan Jick - Chief Investigator - BCDSP - Boston Collaborative Drug Surveillance Program
Katrina Hagberg - Corresponding Applicant - BCDSP - Boston Collaborative Drug Surveillance Program
Lin Li - Collaborator - BCDSP - Boston Collaborative Drug Surveillance Program

Linkages

CPRD Mother-Baby Link