Antenatal depression affects 10%–20% of pregnant women (2, 3). It is associated with a greater risk of negative pregnancy outcomes, such as prematurity, fetal distress, neonatal behavioral differences (4), and a greater risk of postpartum depression (5). Recent data emphasize that relapse rates for depression are high during pregnancy. Cohen and colleagues (6) found that 43% of a sample of pregnant women with a history of major depression experienced a major depressive episode during pregnancy; among women who discontinued antidepressant medication at the start of the pregnancy, the rate was 68%. Recent studies have raised concerns about teratogenicity and neonatal complications associated with commonly used antidepressants. However, there is substantial confusion and controversy about how to integrate new information regarding the possible risks of antidepressant medication into risk/benefit analyses for the treatment of antenatal depression. Moreover, in practice, risk/benefit assessments commonly overlook or minimize the risks associated with untreated maternal depression.

The selection of treatment for depression in a pregnant woman must take into account the well-being of both mother and baby in a context of limited risk-benefit data. The patient’s depressive symptoms themselves may also affect her decision making. In our women’s mental health program, we find that most mothers are exceedingly anxious about accepting antidepressant medication, and despite moderate to severe depressive symptoms and impaired functioning, they tend to focus on the risks of in utero exposure to medication rather than on the risks of untreated depression. Less frequently, we find that depressive symptoms have a negative impact on the mother’s ability to emotionally attach to her baby, which may interfere with her risk/benefit assessment in a different way.

Treatment decisions should be collaborative, involving the mother, the baby’s father when appropriate, and health care providers. The risks and benefits of treatment are complex, and the weight given to the factors involved varies among individuals.

Treatment decisions should take into account the severity of symptoms, past history of depression and treatment response, and patient preferences. In cases of mild depression, nonpharmacological treatments may be considered first. For pregnant women with moderate to severe depression, antidepressants are a reasonable part of a treatment plan. A history of previous postpartum depression or recurrent major depression is a risk factor for perinatal depression and may support the decision to use antidepressants during pregnancy.

Antidepressants

Antidepressants are generally assumed to be efficacious for the treatment of depression in pregnant women, although a small amount of data suggest that higher doses might be necessary to maintain therapeutic benefit (7). Many women are concerned about possible in utero effects of antidepressants. Although most older data on tricyclic antidepressants and SSRIs do not suggest that antidepressants increase the risk of malformation or miscarriage (4), some more recent data have contributed to increased concerns. An association has been reported between first-trimester use of paroxetine and cardiovascular malformations (8). One recent study (9) suggested an association between maternal SSRI use after 20 weeks’ gestation and persistent pulmonary hypertension of the newborn, a rare but serious condition with significant morbidity and mortality. There is also growing concern about a neonatal withdrawal or adaptation syndrome and medical complications associated with use of antidepressants in late pregnancy (10, 11). Additionally, Suri and colleagues demonstrate in an article in this issue (12) that antidepressant use during pregnancy in women with a history of major depression is associated with a higher risk of premature delivery and lower gestational age at birth compared with women with a history of major depression who elected to discontinue medication during pregnancy as well as healthy comparison women. Suri et al. also found that the effect of antidepressants remained after the analyses controlled for severity and duration of depressive symptoms during pregnancy. Information about the long-term effects of in utero exposure to antidepressants is limited, although the available literature is generally reassuring about the lack of negative consequences, particularly with tricyclic antidepressants and fluoxetine (4).

Since antenatal depression is a risk factor for postpartum depression, the safety of treatment during breastfeeding should also be considered. Despite a few case reports of possible adverse effects in infants breastfed by mothers taking antidepressants, most data on antidepressant levels in breast milk and infant serum suggest a low level of infant exposure to antidepressants via breastfeeding (13). Relative to other antidepressants studied, use of fluoxetine during breastfeeding, perhaps particularly if it was also used during the pregnancy, may be more likely to result in higher infant blood levels and adverse neonatal effects (13).

Nonpharmacological Treatments

Psychotherapy, including interpersonal psychotherapy, appears to be efficacious for antenatal depression and prevention of postpartum depression in women at risk (14–16). Unfortunately, many women have financial or time constraints that limit participation in psychotherapy. Bright light therapy also appears to be a promising treatment (17, 18). In cases of severe depression, ECT is considered a safe and effective treatment option during pregnancy, provided that appropriate precautions are taken (19). Exercise has been noted to have at least modest antidepressant effects, and the American College of Obstetricians and Gynecologists recommends 30 minutes of exercise daily during pregnancy (20, 21). In pilot studies (1, 22), our group has found omega-3 fatty acids for treatment of depression to be well accepted and well tolerated in pregnant and postpartum women.

In the selection of an antidepressant for a pregnant patient, previous response to a particular antidepressant rather than possible differences in risk among antidepressants should be strongly considered; the likelihood of a response with the same antidepressant may be preferable to the risk of exposure to multiple medications in the event of nonresponse. As suggested earlier, it is difficult to make definitive statements about the relative safety of the various antidepressants, although paroxetine should not be considered as a first-line medication during the first trimester because of its possible association with fetal cardiovascular malformations.

If a pregnant woman with major depression is unable to refrain from smoking, bupropion may be considered among the antidepressant options. Although safety information for bupropion in pregnancy is limited, cigarette smoking is known to have teratogenic effects, and bupropion is efficacious in smoking cessation as well as in treatment of depression.

Women who experience antenatal depression have a high risk of postpartum depression, and this risk should be considered during pregnancy. The possible effects on the fetus of maternal use of medication late in pregnancy must be balanced against the risk of postpartum depression and its effects on the mother and the baby. While some studies demonstrate low levels of medication exposure for breastfeeding infants during maternal antidepressant treatment, there have been some case reports of suspected adverse events (13). Fluoxetine has been observed at relatively high blood levels, which may have clinical adverse effects in breastfeeding infants. Since an antidepressant started during pregnancy will often be continued during the postpartum period, this should be a consideration for mothers who plan to breastfeed their infants (13).

After a woman and her health care provider have determined that an antidepressant will be part of her treatment plan, her other health care providers, such as her obstetrician and pediatrician, should be informed about the treatment and its rationale. This allows the patient to receive support for her treatment decisions, helping to minimize the potential for mixed messages to the patient and anxiety about her treatment decisions.

Sorting out the risks and benefits of treatment for depression, particularly those of antidepressant medications, is complicated in pregnancy. The selection of treatment should be based on the severity of symptoms, the patient’s depression history (including past response to medication), and the patient’s preferences. In women with mild depression, nonpharmacological approaches may be useful first-choice treatments. In women with moderate to severe depression or a history of previous postpartum depression or recurrent major depression, antidepressants should be strongly considered, alone or in combination with nonpharmacological treatment. The risks of antidepressant exposure to the baby are unclear, although some studies have suggested the potential for risks such as cardiac teratogenicity with paroxetine; persistent pulmonary hypertension of the newborn and other adverse outcomes with SSRIs; and preterm birth and lower gestational age at birth with antidepressants in general.

In the first case presented earlier, Ms. A was concerned about the safety of medication exposure for her baby and said she was “too busy” for psychotherapy. In fact, however, she felt that engaging in psychotherapy was too stigmatizing—it would mean that “something is really wrong with me.” Two months after delivery of a healthy baby, Ms. A decided to stop breastfeeding and to begin treatment with sertraline, which was selected because she had a good response to it in prior treatment. Her symptoms responded to sertraline within 6 weeks, at which time she accepted and followed through with a referral for psychotherapy. Within 3 months, she experienced a full remission of her major depressive episode. When asked if she wished she had accepted treatment with an antidepressant sooner, Ms. A stated emphatically that she did not regret her treatment decisions and under no circumstances would have accepted treatment with medication during pregnancy or while breastfeeding. Four years later, Ms. A has maintained remission from major depression while taking sertraline, except during a brief period when she tried a decreased dose of medication to reduce sexual side effects; she achieved remission again when the dose was increased, although the sexual side effects persist. Her child is healthy and is undergoing normal development.

On the other hand, Ms. B had difficulty integrating the risks pertaining to her baby into her decision-making process. She accepted an invitation to bring her husband into the decision-making process, and he participated in interviews by telephone and later in person. The patient and her husband were presented with treatment options and their risks and benefits. Ms. B and her husband came to the decision that the risks of treatment with an antidepressant were acceptable relative to the risks of untreated antenatal depression. Ms. B was started on a course of escitalopram in addition to psychotherapy. Escitalopram was selected because of a family history of a good response to this agent in the treatment of depression. Within 2 months, Ms. B noted improvement in her depressive symptoms and began to feel excitement about her baby.