Prozac hit the market in 1988, the first selective serotonin reuptake inhibitor (SSRI) antidepressant approved by the FDA for the treatment of depression. Because it was safer and more tolerable than the antidepressants that preceded it, Prozac was soon the most commonly prescribed antidepressant in the United States.
In the late 1980s and early 90s, if a woman taking Prozac was planning a pregnancy or if she inadvertently conceived, she had an important decision to make. Would she continue to take the antidepressant during pregnancy, or would she stop it? At that point in time, we had little information regarding the course of psychiatric illness during pregnancy, and we had even fewer data regarding the reproductive safety of Prozac.
Nearly 25 years later, we stand in a very different place. Now we have far more information to guide us. We know that pregnancy is not protective against new onset or recurrent mood episodes and that women who stop their antidepressants during pregnancy are at high risk of depressive relapse. We also have accumulated data to indicate that untreated depression in the mother carries significant independent risks for the woman and her pregnancy.
Over the last two decades, numerous reports derived from the analyses of available data sources have addressed the reproductive safety of antidepressants; but ironically, it seems that having more data has actually made the decision-making process regarding the use of antidepressants during pregnancy even more complicated. Research regarding the reproductive safety of the SSRIs has yielded conflicting and often inconsistent results. While some studies suggest that SSRIs increase the risk of certain types of malformations – for example, cardiac septal defects – most studies do not demonstrate a link between SSRIs and congenital malformations. While some studies report serious adverse effects related to exposure to SSRIs late in pregnancy, such as pulmonary hypertension of the newborn (PPHN), most studies describe a more benign and self-limited pattern of symptoms, labeled as “poor neonatal adaptation”. Often the information yielded by these studies is so nuanced and complex that it is extremely difficult for patients and their doctors to fully understand the clinical implications, leaving them confused about their treatment options.
One guiding principle in the assessment of drug safety in pregnancy is that medications that cause birth defects (teratogens) do so consistently across studies that are adequately large, and the type of birth defects should also be consistent across studies. We do have known teratogens among some medications used to treat psychiatric disorder, most notably valproic acid (Depakote), a mood stabilizer, known to substantially increase the risk of neural tube defects. There are a relatively large number of studies of SSRIs in pregnancy; however, what we have seen is inconsistent findings of small risks of different problems varying across these studies. Interpretation of disparate findings is made even more complicated because across available studies, the contribution of underlying psychiatric disorder to the outcome is so often incompletely taken into account.
While we cannot say that SSRIs or any medications are risk-free during pregnancy, we can say that studies over decades have failed to demonstrate a consistent risk of teratogenicity. In fact, we have more data for SSRIs in pregnancy than for most medications that are commonly used in pregnancy,
Over the past few weeks we have received many calls and emails from patients and clinicians regarding a review article published in the journal Human Reproduction entitled “The risks of selective serotonin reuptake inhibitor use in infertile women: a review of the impact on fertility, pregnancy, neonatal health and beyond.” What has concerned, and confused, readers of this article is its stance that SSRIs should be avoided during pregnancy. Even more alarming than the review itself are the reiterations of the article in the lay press, many carrying alarmist messages: “Study suggests too much risk associated with SSRI usage and pregnancy.”
In the abstract of this article, Domar and colleagues stated that they “conducted a review of all published studies that evaluate females with depressive symptoms who are taking antidepressant medications and who are experiencing infertility.” That being said, the exclusion from discussion or review of many reputable reports and the consequent omission of references from peer-reviewed journals of articles which failed to demonstrate clinically significant effects on pregnancy outcomes is particularly striking. Furthermore, the authors’ failure to discuss many of the limitations inherent in the research studies included in the review is noteworthy and disconcerting.
For example, in the section on long-term neurobehavioral outcomes associated with fetal exposure to antidepressants, the authors cite three articles demonstrating the adverse a effects of antidepressant exposure on development. However, they fail to include the seminal research of Nulman and colleagues who prospectively followed over 200 children exposed to antidepressants in utero and found no differences in IQ, cognitive development, temperament, behavior, reactivity, mood, distractibility, or activity level between exposed and non-exposed children.
One of the most glaring omissions of this review is the long list of studies which indicate that depression, when left untreated, carries certain risks. While we must advise our patients regarding the risks of exposure to medications, we cannot ignore the impact of untreated depression in the mother. Untreated depression is associated with poor prenatal care, increased risk of self-harm and behaviors which may negatively affect outcomes, including smoking and the use of alcohol and illicit drugs. There have also been reports which indicate that depressive symptoms and/or anxiety may increase the risk of various outcomes including preterm birth and low birthweight. Rather than reviewing those reports, the authors simply declare that evidence supporting the independent association of depression with these outcomes is weak.
Domar and colleagues point out that there are alternatives to the use of antidepressants during pregnancy. Cognitive-behavioral therapy (CBT) has been shown to be effective for the treatment of major depression. They also cite various complementary and alternative treatments which might be effective for treating depression during pregnancy, including physical exercise, yoga, acupuncture, omega-3 fatty acids, and myo-inositol.
While there is strong evidence to support the use of CBT in cases of mild to moderate depression, we are concerned that some readers may be led to believe that these other alternative treatments are more attractive and/or safer than using antidepressants. However, the efficacy of these alternative interventions has not been well-established in pregnant or non-pregnant samples of women. Yoga is currently under study for the treatment of depression in pregnancy but at this time has not been shown to be efficacious for major depression. Exercise alone has not been established as an effective treatment for depression, in gravid or non-gravid women, although it is a reasonable part of a treatment plan considering it does have mood elevating properties and confers other benefits.
To date, the studies of omega-3 fatty acids indicate that omega-3 fatty acids seem to augment the response to an antidepressant; the efficacy of omega-3 fatty acids as monotherapy has not been established. Inositol remains largely unstudied, and its safety in pregnancy is not established.
The use of alternative treatments with unproven efficacy when effective treatments are available and when the clinical situation demands definitive treatment places the woman – and her pregnancy – at risk.
Unfortunately, the article by Domar and colleagues fails to emphasize that treatment chosen is dependent upon the severity of the depression. For some women, depression may be a mild and circumscribed illness, amenable to treatment with a variety of interventions and may not require long-term treatment with an antidepressant. We suspect that this may be the case for many women experiencing infertility, the population specifically addressed in this article.
However, for many women, particularly those taking antidepressants on a long-term basis, depression is a severe, recurrent illness that impacts all areas of life and health. Antidepressants, particularly SSRIs, are first-line treatments for moderate to severe depression. In our clinic, most of the women come to the prospect of planning to conceive or an actual pregnancy with knowledge of the disorder for which they are treated. Often they have tried many different types of interventions, including medication and psychotherapy. And for so many of these women, the road to euthymia has been an arduous and painful one.
In our clinical experience, women who can safety stop antidepressants during pregnancy usually do. Many women attempt to discontinue antidepressants and relapse, often becoming acutely ill, and some ultimately require more medication than they were initially taking in order to restore emotional well-being. Women in this situation often struggle with the treatment decisions they are forced to make, decisions that are all the more difficult due to stigma against psychiatric illness and psychiatric medications and the confusion surrounding psychiatric illness and treatment among health care providers and the lay public. Quite simply, women do not decide in a cavalier fashion to use medications during pregnancy. To suggest otherwise is at best glib, and at worst disrespectful and cruel.
It is in such a climate that thoughtfully written and carefully researched review articles and consensus guidelines are important, if not essential, to the clinical decision-making process. A review article takes all of the available data, evaluates it, and presents a well-balanced summary of the information. However, an article that fails to effectively synthesize the available data makes thoughtful decision-making around a critical issue, such as use of psychiatric medications during pregnancy, that much more challenging for the women and their clinicians who wrestle with wishes for a healthy pregnancy and the need for treatment of underlying psychiatric disorder
An article that is not balanced, that discredits the use of psychotropic medication regardless of illness severity, unfortunately serves to create an environment in which women feel judged and guilty for pursuing treatment for illnesses that are associated with substantial morbidity and even mortality. As clinicians and researchers, it should be our goal to provide women in this difficult situation with clinically relevant and carefully researched information regarding their illness and treatment options and help them to make informed, well-considered decisions regarding their care.
Marlene Freeman, MD
Lee Cohen, MD