To treat or not to treat major depression during pregnancy with antidepressant medication is a critical question to clinicians concerned with the welfare of the mother, on the one hand, and the healthy development of the infant, on the other. The patient and her psychiatrist face a dilemma: untreated depression in the mother can impair the neurocognitive development of the infant and result in preterm birth; medication use during pregnancy also can impact the fetus and has been associated with an increased risk of preterm birth.

The landmark study of Wisner et al. (1) takes a first step toward systematically addressing this question. The authors investigated the impact of major depression versus antidepressant treatment on pregnancy and neonatal outcomes. In particular, they examined the effects of selective serotonin uptake inhibitor (SSRI) use and major depressive disorder on 1) minor physical anomalies, 2) maternal weight gain and infant birth weight, 3) rate of preterm birth, and 4) neonatal adaptation in 238 pregnant women who had no SSRI exposure or major depressive disorder, had either continuous or partial SSRI treatment, or had either continuous or partial depression. They found that infants exposed continuously to either SSRIs or major depression were more likely to be born preterm (>20% rate) than were unexposed or partially exposed infants. Other outcomes were similar in the two groups. In addition to its prospective, observational nature, the study’s particular strengths include the investigators’ division of the SSRI and depression groups into partial versus continuous exposure, confirmation of SSRI exposure by maternal serum assays, blind ratings of outcomes, and controlling for other potentially modifying influences on outcome measures, such as smoking, alcohol abuse or dependence, and other drug use.

As a necessary first step, the study by Wisner et al. focused on critically important effects of major depressive disorder versus SSRI exposure during pregnancy on neonatal outcomes at 2 weeks postpartum. Another consideration for the clinician weighing the risks and benefits of antidepressant treatment during pregnancy is long-term outcomes. Although long-term outcomes were beyond the scope of this study, as the authors state in the discussion, "The underlying depressive disorder and its sequelae may constitute a long-term disease risk factor independent of current symptom level." The investigators found that the groups with continuous exposure (major depression or SSRIs) differed from the nonexposed infants primarily in the risk of preterm birth, which itself can have long-term medical and social consequences (2). As other short-term outcomes did not differ between groups, the clinician needs to be able to inform the mother with regard to other potential long-term consequences of treating versus not treating major depression pharmacologically during pregnancy.

From the evidence available to date, the risks of an untreated maternal depression are far greater than the risks of serious adverse sequelae from antidepressant medication. As now demonstrated in multiple other studies, major depression during pregnancy may impair the neurocognitive and socioemotional development of the child, predict sleep problems in infancy and toddlerhood, alter neuroendocrine function, and increase the risks of mental and medical disorders in the offspring later in life (3—8). Depression during pregnancy is a risk factor for the development of postpartum depression, and women with postpartum depression are at increased risk for recurrent depressive illness, all of which may further impair the healthy development of the child (9). In their ongoing longitudinal study of the effects of maternal depression on long-term outcomes in offspring, Murray et al. reported on cognitive (7) and socioemotional (8) impairments in children of depressed mothers at 5 years. More recently, after 13 years of follow-up, these investigators (10) reported that maternal depression was associated with higher rates of affective disorders in adolescent offspring. Mothers who developed puerperal depression were more likely to experience depression subsequently, which contributed to the development of depressive disorders in their offspring. Anxiety disorders were elevated in offspring exposed to maternal depression regardless of subsequent depressive episodes in the mother. On the basis of an extension of this study, in which the investigators examined morning and evening salivary cortisol measures in the adolescents who had or who had not been exposed to maternal depression, they deduced that alteration in cortisol secretion was a possible mechanism by which risk for depression was transmitted from mother to offspring (6). As Nulman et al. (3) found in following children of depressed pregnant mothers up to 71 months, the child’s IQ and language achievement were affected adversely by duration and number, respectively, of depressive episodes after delivery. In a study by Weissman et al. (4) of the offspring of depressed parents 20 years later, the risks for anxiety disorders, major depression, and substance dependence were three times higher in the offspring of depressed parents than in the offspring of nondepressed parents. Social impairment, medical problems, and mortality also were higher in the offspring of depressed parents. Remission of maternal depression after 3 months of medication treatment in STAR*D (5) was associated with reductions in the children’s diagnoses and symptoms, whereas continued depression in the mothers was associated with increased rates of disorders among the children.

The risks of untreated depression in the mother must be weighed against the potential long-term effects of pharmacological treatment. As Nulman et al. (3) found in following children up to 71 months who had been exposed to tricyclic antidepressants or fluoxetine throughout fetal life, neither tricyclic antidepressants nor fluoxetine adversely affected the child’s global IQ, language development, or behavior. Wisner et al. (1) discuss studies that showed increased risks of neonatal syndrome for paroxetine and fluoxetine, in particular, when used in pregnancy, although some more recent studies do not support these findings (11, 12). As the authors emphasize in the discussion, the results do not support a recommendation for partial SSRI treatment during pregnancy to reduce preterm birth. Investigators from the STAR*D-Child study observed that it was treatment of maternal depression until remission that was associated with decreased psychiatric symptoms and improved functioning in the offspring. In the interim, we need to investigate further nonpharmacological treatment strategies (e.g., psychotherapies and sleep and light treatments).

Thus, all things considered, on the basis of the findings from the methodologically sound and rigorous study of Wisner et al. and the evidence available from long-term studies, this author thinks that the risk of untreated major depression outweighs the risk of effects of SSRI treatment on neonatal outcomes. We need to consider not just short-term, but also long-term, consequences of our decisions. In addition to focusing on the child, the clinician needs to consider the risk of untreated major depression in the mother. These risks include exacerbation or recurrence of her underlying psychiatric illness, which can have adverse effects on her morbidity and mortality and can impair not only her functioning, but that of her family and other children under her care. The field eagerly awaits further prospective trials in larger samples that would allow for delineation of demographic characteristics (e.g., age, personal and family history, race, marital status) to confirm initial findings from this important study. The ultimate value of these studies is that they will provide the necessary information to allow women to make more informed choices about their reproductive health care.