Over the last decade, attention in the medical literature has gathered logarithmically to focus on potentially efficacious treatments for perinatal depression. Studies of relevant databases, editorials, and various reviews have addressed the reproductive safety concerns of antidepressant treatments, particularly selective serotonin reuptake inhibitors (SSRIs) on one hand, and the impact of untreated maternal psychiatric illness on fetal and maternal well-being on the other.
Depression clusters in women across the childbearing years, and growing numbers of women in this group are being treated with antidepressants. Historically, pregnancy had been considered a time of emotional well-being, but it now is generally accepted that pregnancy does not confer protection against psychiatric disorder. This is particularly the case for women with histories of depression; for women with histories of depression who are on maintenance therapy with antidepressants to sustain emotional well-being, discontinuation of these medications proximate to or during pregnancy is associated with a high risk of depressive relapse. This appears to be the case particularly among women with histories of severe depression.
So what is the most appropriate treatment of depression during pregnancy? Navigating the growing literature on relative risks of antidepressant treatment during pregnancy can be particularly burdensome for clinicians. While clearly the amount of data is vast, the quality of these data is highly variable. Randomized controlled clinical trials of antidepressant use during pregnancy are not performed for obvious ethical reasons, and reports of outcomes of fetal exposure to psychiatric medications including antidepressants frequently derive from analyses of data from small cohort studies or from administrative databases, which are limited by serious methodologic difficulties – small numbers of exposed cases, problems ascertaining reliable information regarding actual medication exposure, comorbid substance use, or the presence of active psychiatric disorder across pregnancy. These factors make reliable assessment of risk that much more difficult, and determination of clinical implications of such findings even more challenging. Variable interpretations of the same literature make it even more difficult for clinicians in the community to know how to proceed with the patient presenting for guidance about appropriate use of antidepressants during pregnancy or alternative evidence-based management strategies for perinatal depression.
A recently published article addressing the risks of SSRIs in infertile women, which discusses, among other issues, the risk of SSRIs during pregnancy across outcomes of miscarriage, birth defects, neonatal side effects (so-called neonatal adaptation syndrome), and persistent pulmonary hypertension of the newborn (PPHN), is an example of how published reports and reviews frequently confuse more than they inform. The authors offer the following conclusions: Antidepressants are not effective for most women with depression; a strong signal for teratogenicity is evident for SSRIs; and outcomes such as preterm birth, PPHN, and poor neonatal adaptation are common after prenatal exposure to SSRIs. (The review was published online in October 2012 in Human Reproduction [doi: 10.1093/humrep/des383], and no external funding or conflicts of interest of the authors are noted.) The review, described as "systematic," appears to be relatively selective with respect to data presented, and as such, many of the conclusions are not uniformly supported by the data in the literature; hence its conclusions should be put in perspective.
As an example, more data support the absence of a strong teratogenic signal with first-trimester exposure to SSRIs than for almost any other medicine taken by pregnant women. This was the main point made in an editorial published in the New England Journal of Medicine in 2007, titled "Teratogenicity of SSRIs – serious concern or much ado about little?" The editorial refers to two studies published in the same issue, and states that the reports "together with other available information, do suggest that any increased risks of these malformations in association with the use of SSRIs are likely to be small in terms of absolute risks" (N. Engl. J. Med. 2007;356:2732-3).
Certainly, over the last decade, efforts have been made to refine the risk-benefit decision regarding antidepressant use during pregnancy by a better understanding of the vast data on the impact of SSRIs on fetal well-being and outcomes such as malformations, as well as the effects of untreated depression on fetal and neonatal well-being. Across the review noted above, the numerous studies citing adverse effects of untreated depression on obstetrical and neonatal well-being are not discussed. Most critical is the omission of a meta-analysis of adverse effects of depression during pregnancy, which identified an increased risk of preterm delivery and low birth weight associated with prenatal depression (Arch. Gen. Psych. 2010;67:1012-24).
Also omitted is what is one the most pivotal studies published to date on the long-term neurobehavioral follow-up of children exposed to antidepressants during pregnancy. This study followed children exposed to fluoxetine (Prozac) or a tricyclic antidepressant for a period of up to 72 months and compared outcomes including IQ, reactivity, temperament, mood, or distractibility to a matched group of nonexposed children; no differences were noted between the groups (N. Engl. J. Med. 1997;336:258-62). Other studies of long-term neurocognitive sequelae of fetal exposure to antidepressants will lay the groundwork for the most critical refinement of guidelines for the use of antidepressants during pregnancy.