I wanted to share this letter with you, created by Dr. Laura Miller, a reproductive psychiatrist and the director of Women’s Mental Health at Brigham and Women’s Hospital in Boston. Here’s the information she offered to her own patients and other clinicians regarding the treatment of depression in pregnancy and the recent articles about it:
A recent review article (Domar et al. 2012) about the risks of selective serotonin reuptake inhibitor (SSRI) antidepressants during pregnancy and infertility treatment has generated considerable media attention. In essence, it concludes that there is no credible evidence for the efficacy of SSRIs in treating mild to moderate depression in pregnancy and that exposure to SSRIs during pregnancy causes significant harm to fetus and newborn. This has caused consternation and confusion among many patients and health care providers. For this reason, leaders in the BWH Department of Psychiatry would like to clarify the evidence related to the article’s key points, as a guide for clinicians whose patients have questions.
Decisions about treatment with antidepressants during pregnancy are made on an individual basis. The authors assert at the outset that it is a “standard recommendation” that “the benefit of antidepressant use outweighs the risk of depression during the gestational and post-partum period”. This is not the case. The standard recommendation, repeated in the conclusions of numerous studies of antidepressant use during pregnancy, is to carefully weigh the risks of prescribing medication against the risks of withholding medication in each individual case. This is well expressed in the words of a Food and Drug Administration (FDA) advisory regarding antidepressant use during pregnancy: “Women who are pregnant or thinking about becoming pregnant should not stop any antidepressant without first consulting their physician…The decision to continue medication or not should be made only after there has been careful consideration of the potential benefits and risks of the medication for each individual pregnant patient.”
Untreated depression in pregnancy is associated with adverse outcomes. The authors state that “There is an assumption in the psychiatric community that the risks to a fetus are greater if the mother has untreated symptoms of depression”. This is not an assumption; it is a finding from numerous studies. For example, untreated maternal depression during pregnancy is associated with reduced prenatal care (Marcus 2009), preterm birth (Li et al. 2009; Bansil et al. 2010), reduced birth weight (Henrichs et al. 2010), altered behavior at birth (Zuckerman et al. 1990), increased risk of infection (Rahman et al. 2004; Traviss et al. 2012), and more difficult temperaments (Huot et al. 2004). There is increasing evidence that some of these effects are due to epigenetic influences on fetal development that are mediated by elevated cortisol levels (Glover et al. 2009; Oberlander et al. 2008).
Antidepressants, when appropriately indicated, are effective in the treatment of depression. The authors state that “The best available evidence suggests that antidepressants do not provide clinically meaningful benefit for most women with depression.” While they raise the important point that publication bias has marred appraisal of the efficacy of many medications, including antidepressants, recent research using methodology to correct for publication bias (Duval and Tweedie 2000) continues to substantiate the efficacy of antidepressants.
Both depression and SSRI may have direct and indirect, positive and negative effects on fertility. Regarding the impact of antidepressants on fertility, the authors correctly state that systematic data are sparse. A meta-analysis of 14 studies of women using assisted reproductive technologies found that depressive symptoms had no significant effect on the likelihood of becoming pregnant (Bolvin et al. 2011), although depression can contribute to a decision to stop infertility treatment earlier (Verhaak et al. 2010). Retrospective data comparing women undergoing in vitro fertilization found no significant difference in pregnancy rates or live birth rates with or without SSRI use. By contrast, some study findings suggest the possibility that SSRIs can reduce fertility in men (Safarinejad 2008; Tanrikut and Schlegel 2007; Tanrikut et al. 2010).
Obstetric and neonatal risks of SSRIs
The authors report a number of obstetric and neonatal risks of SSRI use during pregnancy as definitive findings. Here are some clarifications:
Miscarriage: The authors are correct that this is a consistent finding in prospective, controlled studies. However, the authors fail to highlight the magnitude of the effect or explore its context to rates of miscarriage (spontaneous abortion) in the general population. In a prospective study (Einarson et al ,2009) the rates of spontaneous abortion in the group exposed to antidepressants were found to be (13%) as compared to (8%) in the non-exposed group. Similar rates were found in a meta-analysis (Hemels et al, 2005). In weighing these findings, it is important to keep in mind that the rate of spontaneous abortions in the general population is believed to be between 12-15% (Robinson et al, 2012). In a review of literature of antidepressants in pregnancy (Lorenzo et al, 2011), the degree of risk was qualified as “a small increased risk for spontaneous abortions” in women who are exposed to antidepressants.
Birth defects: The authors report a “consistent ‘signal’ implicating SSRI use during pregnancy to various congenital anomalies”. In fact, prospective, controlled studies (collectively including more than 20,000 exposures to antidepressants) and meta-analyses do not show an association between antenatal SSRI exposure and any congenital anomaly. Large, retrospective database studies have had highly inconsistent findings, ranging from no association to association with a variety of different anomalies, with no specific pattern. These studies are unable to disentangle the effects of key compounds, and most have not corrected for doing multiple statistical tests. The only exception is that first trimester paroxetine exposure has been associated with an increased risk of cardiac anomalies in several retrospective (although no prospective) studies. Although this association remains controversial, a resultant FDA advisory appropriately focuses on individual risk/benefit analyses: “Physicians who are caring for women receiving paroxetine should alert them to the potential risk to the fetus if they plan to become pregnant or are currently in their 1st trimester of pregnancy. Discontinuing paroxetine should be considered for these patients. In individual cases the benefits of continuing paroxetine may outweigh the potential risk to the fetus.”
Preterm birth: SSRI reduce gestational age by less than a week similarly to untreated antenatal depression. The authors are correct that studies have found an association between antenatal SSRI use and preterm birth. What they failed to report is that the most careful head-to-head study found no significant difference between the effects of untreated antenatal depression and the effects of SSRI antidepressants on preterm birth (Wisner et al. 2009). They also failed to report the magnitude of the effect: both depression and antidepressants reduce gestational age by less than a week (Einarson et al. 2011; Marroun et al. 2012, Yonkers et al. 2012).
Neonatal side effects: side effects are found in 7-8% of neonates and are typically mild and transient. The authors correctly report that “up to 30%” of SSRI-exposed newborns develop neonatal side effects, however, they fail to note that only one study (the one they cite) found that 30% of babies had side effects; other studies have typically found about 7 – 8% of babies are affected. In addition, they described these effects as “persistent” despite consistent findings that it is rare for these effects to last more than 1 – 2 days.
Persistent pulmonary hypertension (PPHN): The authors correctly state that antenatal SSRI use is associated with an increased risk of PPHN in several, but not all studies. However, they fail to report the magnitude of this effect. The estimated prevalence of PPNH in the general population is 1.9/1000; the combined reported cases after SSRI exposure show a prevalence of 2.0/1000 (Occhiogrosso et al. 2012).
Pre-eclampsia: The authors report studies suggesting an association between SSRI use and increased risk of hypertension during pregnancy in this section, although those findings are preliminary and did not include findings of increased risk of pre-eclampsia. They failed to report studies showing increased risk of pre-eclampsia from untreated depressive symptoms.
Long-term neurobehavioral effects: The authors do not cite the most methodologically sound prospective, controlled studies, which showed no significant neurodevelopmental differences between exposed and control children (Nulman 2012). They fail to report the considerable methodologic limitations of some of the studies whose conclusions they cite.
The Brigham Women’s Mental Health Service remains committed to helping women achieve and maintain optimal health for themselves and their babies during pregnancy and the postpartum period. Our individual, thorough evaluations help to identify factors that render women vulnerable to depression in pregnancy and factors that promote resilience. Our treatment plans are multifaceted, encompassing evidence-based psychotherapies, nutritional modifications, aerobic exercise and phototherapy when indicated. We conduct careful risk/benefit analyses about the use of antidepressant medication, and use prescribing practice which reduces risks. The sooner a woman comes to see us, the more we can help her reduce the risks of both untreated symptoms and medications. The optimal time to refer is preconception. Preconception consultation for women with pre-existing psychiatric disorders, or risk factors such as premenstrual dysphoria or family history, can help women and their obstetric providers form a proactive treatment plan that will promote a healthy pregnancy.