Implications for practice and research

• Second trimester exposure to antidepressants was associated with preterm birth.

• Third trimester exposure was associated with infant convulsions.

• The absolute risk for both outcomes is extremely low and most infants will not have these complications.

• Decisions about whether to use antidepressants during pregnancy should include the severity of the mother's psychiatric illness.

• Large studies that include severity of the mother's illness need to be conducted.

Context

Approximately 10–20% of women will meet the criteria for major depressive disorder (MDD) during their lifetime, including during pregnancy.1 Rates of antidepressant exposure during pregnancy are now estimated to be at least 10% of all pregnancies.2 The use of antidepressants during pregnancy has risen in part due to research that demonstrates a high relapse rate (>60%) of MDD in women who discontinue their antidepressants for pregnancy.3 Major depressive episodes (MDE) during pregnancy are associated with poor outcomes for the child including preterm birth and elevated cortisol levels.4 ,5 Women who are depressed during pregnancy have a higher rate of postpartum depression, which can impact language development and IQ in the exposed child.6 However, a number of studies have associated poor outcomes for the child with antidepressant exposure in utero. Preterm birth and reduction in birth weight have been associated with use of selective serotonin reuptake inhibitors (SSRI), even after controlling for maternal psychiatric illness.4 Furthermore, Poor Neonatal Adaptation (PNA), a cluster of mostly transient symptoms including tachypnoea, hypoglycaemia, temperature instability, irritability and rarely, seizures4 ,7 has been detected in approximately 30% of infants exposed to SSRIs during the third trimester.7 Whether PNA leads to persistent infant developmental problems is unknown.4 ,7

Findings

Of the 23 280 (10.2% of the sample), who had filled at least one prescription prior to pregnancy, most (75%) had discontinued their antidepressants either before or during the first trimester. Second-trimester antidepressant use was significantly associated with preterm birth with a shortened gestation of 1.7, 3.7 and 4.9 days for filling 1, 2 and ≥3 antidepressant prescriptions. Third-trimester exposure was associated with infant convulsions with adjusted ORs of 1.4, 2.8 and 4.9 for filling 1, 2 and ≥3 antidepressant prescriptions. Both preterm birth and infant convulsions were rare events in this large cohort. Antidepressant use was also not associated with low birth weight when the authors controlled for confounders.

Commentary

The study's strengths include large sample size and control for potential confounders. It also demonstrates the difficulty of separating out the influence of maternal psychiatric illness and health behaviours that may influence neonatal outcomes. Though there was a positive association with preterm birth and infant convulsions with antidepressant use, these results are reassuring in that the absolute risk for each is small with 98% of exposed infants not undergoing early preterm labour and 99% of exposed infants not experiencing convulsions. The results were also reassuring in that antidepressant use was not associated with low birth weight. For those who treat pregnant women with MDD, the ultimate conclusion is that although there are risks with neonatal exposure to antidepressants, those risks are small and must be balanced with the potentially serious risks associated with untreated maternal psychiatric illness.