SSRIs have been the subject of substantial pregnancy exposure investigation compared with other drug classes. Two recent articles in the New England Journal of Medicine provided large-scale case-control data.28,29 The investigators from both studies concluded that the association of SSRI exposures with birth defects (if any) carries a small absolute risk, and in his editorial on the articles, Greene30 agreed.
Alwan and colleagues28 observed no association between overall exposure and defects previously related to SSRIs (particularly congenital heart defects). The investigators did find an association between maternal SSRI use and defects not previously reported, which suggests the possibility of association by chance. It is notable that only a small number of infants with defects had been exposed to an SSRI.
In the second article, Louik and colleagues29 reported that overall SSRI use was not associated with the defects reported by Alwan’s group nor with congenital heart defects. However, they found a doubling of the risk of septal defects associated specifically with sertraline(Drug information on sertraline) (13 exposed patients) and a tripling of the risk of right ventricular outflow tract obstruction defects associated with paroxetine(Drug information on paroxetine) (6 exposed patients).
The baseline rate of right ventricular outflow tract obstruction defects is about 5.5 cases per 10,000 live births in the general population, so the absolute risk remains low. For example, even a 4-fold increase would raise the rate of malformation to just 2 per 1000 births (or less than 0.2%).29 The only association that overlapped with the findings of Alwan and colleagues was that between sertraline and omphalocele (odds ratio [OR], 5.7; 95% confidence interval [CI], 1.6 - 20.7, only 3 exposed patients). No significantly increased risks were associated with fluoxetine(Drug information on fluoxetine) or non-SSRI antidepressants for any assessed birth defects.28
Einarson and colleagues31 evaluated the association of paroxetine with cardiovascular defects in infants who had been exposed during the first trimester. The investigators collected 1174 unpublished cases of exposed infants; an additional 2061 cases were identified from 5 previously published database studies. The rates of cardiovascular defects in the 1174 unpublished cases were 0.7% in the exposed group and 0.7% in the unexposed group. In the database group, the rate was 1.5%. The combined mean rate of cardiovascular defects was 1.2% (95% CI, 1.1 - 2.1). The authors concluded that paroxetine was not associated with an increased risk of cardiovascular defects.
Negative reproductive outcomes associated with SSRI exposure have been reported, however. Chambers and colleagues32 observed several adverse outcomes related to later gestational SSRI exposure. Their data showed a 3-fold increase in the rate of preterm birth (14.3%) in infants whose mothers were treated with fluoxetine after 25 weeks’ gestation compared with infants whose mothers stopped the drug earlier in pregnancy (4.1%) or those who had not been exposed (5.9%). Infants exposed to fluoxetine during late gestation had significantly lower birth weights than infants whose mothers discontinued the drug earlier or who served as controls.
In evaluating the association of reproductive outcomes related to drug exposure, a critical issue is the contribution of the underlying illness being treated. Suri and colleagues33 addressed this issue in their study of pregnant women in 3 exposure groups: those treated with an antidepressant for most of the pregnancy; those with MDD with no or brief drug exposure; and controls with neither exposure. The rates of preterm birth were 14.3%, 0%, 5.3%, respectively. The findings suggest that SSRI treatment rather than MDD is associated with an increased risk of preterm birth.
Oberlander and colleagues34 used population health data to link records of births with records of maternal health and prenatal maternal prescriptions for SSRI. A total of 119,547 neonates were identified and divided into 3 groups: those whose mothers had received a diagnosis of depression and had been treated with an antidepressant; those whose mothers had received a diagnosis of depression but had not been treated for the depression; and nonexposed controls—neonates whose mothers had neither received a diagnosis of depression nor been treated with antidepressant medication. To account for the expected differences between depressed wom-en who did and those who did not access drug therapy during pregnancy, propensity score matching was used to identify a comparison subgroup from the depressed mothers who had not been treated with antidepressants but otherwise had similar characteristics to those of the SSRI-treated mothers.
SSRI-exposed infants had lower birth weights, decreased mean gestational ages, and longer hospital stays. Findings also showed more neonatal respiratory distress, jaundice, and feeding problems in exposed infants than in infants of depressed unmedicated mothers. However, when compared with propensity score–matched neonates of depressed women, SSRI exposure was associated only with increased incidence of birth weight below the 10th percentile and respiratory distress.