Safety of Antidepressants in Pregnancy
Once a decision is made that medication is needed, there are a number of available antidepressants from which to choose. Commonly used medications for the treatment of depression include selective serotonin reuptake inhibitors (SSRIs), norepinephrine/dopamine reuptake inhibitors (NDRIs), and serotonin/norepinephrine reuptake inhibitors (SNRIs). These drugs work by inhibiting the reuptake of the neurotransmitters in the synaptic cleft. Other drugs are available, but they are used less commonly; these include monoamine oxidase inhibitors (MAOIs) and tricyclic antidepressants (TCAs). MAOIs inhibit the degradation of monamine oxidase, while TCAs appear to increase the availability of norepinephrine by blocking the presynaptic reuptake of serotonin and by increasing the sensitivity of postsynaptic neurons to serotonin. These latter two classes of antidepressants are generally reserved for use by individuals with refractory depression because they have poorer safety and side effect profiles.
Unfortunately, there is limited evidence available to guide providers in choosing among the various antidepressants. Because of the dearth of quality information, it is tempting to consider delaying medication use until after birth. However, doing so may not be in the best interests of either the mother or the fetus, because untreated depression in pregnancy may pose risks for both mothers and infants. While being pregnant and having children in the household appears to be protective against suicide, other risk factors, such as poverty, lower educational attainment, and experiencing multiple stressful events are commonly seen in pregnant women, and these stressors may offset the benefits of having children in the household. The rate of completed suicide in women between 20 and 44 years of age has been estimated to be five to 10 deaths per 100,000 individuals. Therefore, while the suicide rate in pregnant women is likely to be even more rare, it is always a possibility.
Other problems associated with untreated depression reported in the literature include poor self-care, preterm birth, abnormal neurochemical changes in offspring, and poorer behavioral and mental health outcomes in children.[25,28–30] Of these, the evidence appears to be stronger and more consistently associated with preterm birth. However, it should be cautioned that interpretation of the literature is hampered by widespread methodologic issues, such as small sample sizes, the use of differing definitions of stress and depression, varying length of follow-up and, most importantly, the lack of a similarly impaired control group. This latter problem is particularly problematic when evaluating the safety of antidepressant use in pregnancy, because many studies compare the safety of antidepressants in treated depressed women to healthy, nondepressed women.
Only a few studies have included both treated and untreated women in their designs. A prospective longitudinal study of depressed pregnant women taking antidepressants (n = 49), pregnant women with depression who received either no or partial treatment with antidepressants (n = 22), and healthy, nondepressed pregnant women (n = 19) found significantly higher rates of preterm birth in those who received antidepressants (14.3%) compared to women with no or partial treatment (0%) and healthy women (5.3%; P = .5). Wisner et al. prospectively followed pregnant women who were and were not depressed (N = 279) from midpregnancy until birth. Women were assigned to five mutually exclusive groups: 1) no SSRI use, no depression (n = 131); 2) depressed and used SSRIs throughout pregnancy (n = 48); 3) depressed and no SSRI use (n = 14); 4) depressed with partial SSRI use (n = 23); and 5) partial depression with no SSRI use (n = 22). Maternal assessments were completed at four time points: 20, 30, and 36 weeks, and again at birth. Infants exposed to either the continuous use of SSRIs throughout pregnancy or untreated depression were at significantly higher risk of being born preterm than those born to pregnant women who were partially treated or healthy controls. The rate of preterm birth was more than 20% in these two groups, compared to 4% to 9% in the other groups (P = .009 after adjustment for maternal age and race). These findings suggest that medication use—and perhaps depression itself—may be associated with preterm birth.
The most recent study to date has a very large sample size relative to other studies and includes pregnant women with depression taking SSRIs (n = 329), women with a psychiatric history but no SSRI use in pregnancy (n = 4902) and pregnant healthy women (n = 51,770). Data were collected with a questionnaire, so the severity and duration of current or past depression was not measured, nor was information available about the duration or dose of the antidepressant being used by women in this study. Women used a variety of SSRIs, such as sertraline (Zoloft), fluoxetine (Prozac), and escitalopram (Lexapro), among others. Some also used paroxetine (Paxil), different antidepressant regimens, and/or combined the use of an SSRI with other medications, such as benzodiazepine, antipsychotics, and/or tricyclics. Women using SSRIs in pregnancy had a higher risk of giving birth before 37 weeks' gestation compared to the healthy control group (adjusted odds ratio [OR], 1.89; 95% CI, 1.29–3.16). The rate of preterm birth was reported to be 8.8% in women taking SSRIs, 5.0% in women with a psychiatric history, and 4.9% in healthy women. Of note is the fact that gestational age at birth was only 5 days shorter (95% CI, –6 to –3 days) in exposed pregnancies. This finding suggests that while the preterm birth rate appears to be significantly higher in SSRI-exposed pregnancies, it may only increase the risk by the matter of a few days. Unfortunately, the authors did not break down the risk of preterm birth by category, namely extreme prematurity (<28 weeks' gestation), very preterm (28–31 weeks' gestation), and mild preterm (32–36 weeks' gestation).This is important because one of the unanswered clinical questions is to what degree antidepressants increase the risk of preterm birth. If it is only slightly and the risk is concentrated in the mild preterm birth category, this has much different consequences for the fetus than if preterm birth occurs earlier in the pregnancy.
The evidence to date suggests that the safety of antidepressants may vary by trimester of exposure, class of antidepressant, specific agent, duration of use, and dose. Adverse effects can also be both short- and long-term. In general, authorities recommend use of the newer antidepressants, particularly the SSRIs, in pregnancy and lactation as first-line agents because more data are available about their safety than for many other classes of antidepressants.[36–38] Although bupropion (Wellbutrin) is commonly prescribed for smoking cessation and has been categorized as a pregnancy category B medication by the US Food and Drug Administration, very little data are available about its safety. The evidence to date does not suggest that bupropion is associated with any increased risk.
The use of antidepressants in the first trimester has been reported to be associated with increased risk of miscarriage in two meta-analyses. Rahimini et al. identified studies published between 1990 and 2005 that investigated birth outcomes following exposure to SSRIs in pregnancy; they excluded studies that did not include a control group. The summary OR for spontaneous abortion for the nine studies that met inclusion criteria was 1.7 (95% CI, 1.28–2.24) in SSRI-exposed pregnancies. In a large meta-analysis of six studies comprised of 3567 women (1534 exposed to antidepressants in pregnancy and 2033 with no exposure), Hemels et al. reported a miscarriage rate of 12.4% in those who used antidepressants compared to 8.7% in those who did not (relative risk, 1.45; 95% CI, 1.19–1.77). However, the authors noted that because the miscarriage rate did not vary by antidepressant class, it may be that underlying depression, not antidepressant exposure, increases the risk of miscarriage. Other studies have reported an association between untreated depression and higher risk of spontaneous abortion.
Evidence on whether the SSRIs increase the risk of congenital anomalies is mixed. Earlier studies found that exposure to SSRIs as a class during pregnancy did not increase the overall risk of congenital anomalies. However, among the SSRIs, use of paroxetine (Paxil) may be of concern. Several studies reported a higher risk of cardiac defects with the use of paroxetine, which resulted in a change in its status to a pregnancy category D medication according to the US Food and Drug Administration. Later studies also found an increased risk of cardiac defects associated with the use of paroxetine in the first trimester,[42–44] while others have not found an increased risk.[45,46] A recent meta-analysis of six studies reported that paroxetine use was significantly associated with increased risk for cardiac defects compared to other antidepressants (OR, 1.72; 95% CI, 1.22–2.42). However, the authors noted that their results may be influenced by detection error because of the widespread use of fetal echocardiograms in exposed pregnancies in these studies. In addition, while the rate of cardiac defects in exposed infants was significantly higher than in unexposed infants, the impact appears to be small: cardiac defects were reported in just under 2% of exposed infants compared to 1% in nonexposed infants. One of the largest studies to date found that the incidence of congenital anomalies in more than 3000 infants exposed to paroxetine (Paxil) during pregnancy was approximately 1%, the same as that found in the general population.
SSRI use has also been associated with other adverse neonatal complications, including pulmonary hypertension, preterm birth, and poor neonatal adaption after birth. Persistent pulmonary hypertension is a rare but potentially fatal complication. It is estimated to occur in less than 1% of births. Several studies have reported higher rates after exposure to SSRIs in late pregnancy,[48,49] but others have not reported higher rates. Even if the association between antidepressant exposure and persistent pulmonary hypertension is found to hold in future studies, the absolute risk is likely to be small. Chambers et al. estimate that even with the adjusted odds ratio (AOR) of 6.1 (95% CI, 2.2–16.8) reported in their study, 99% of infants exposed to SSRIs in pregnancy will be unaffected by persistent pulmonary hypertension.
Similarly, whether SSRI use in pregnancy increases the risk of prematurity or low birth weight is also controversial, although the results of the newest and largest study published to date are compelling.[35,51] Controlling for the level of severity of maternal depression; maternal self-care activities, such as smoking, drug and alcohol use, and nutrition; and exposure variables (specific medication[s] used, timing and duration of use in pregnancy, and dosage) can be difficult and may explain why studies have come to opposite conclusions. In contrast, the evidence seems to consistently show that newborns can have transient difficulties in the first few days after birth following in utero exposure to SSRIs. Neonates exposed to SSRIs in late pregnancy have been reported to experience jitteriness, irritability, feeding difficulties, and respiratory distress after birth. These symptoms are self-limited, of short duration, and respond to supportive care. The overall risk ratio of developing poor neonatal adaption after exposure to SSRIs in late pregnancy has been estimated to be 3.0 (95% CI, 2.0–4.4).
Whether the use of antidepressants in pregnancy has negative long-term health consequences for exposed infants is unclear. To date, the evidence is reassuring. In an extensive literature review published in 2007, Way concluded that the preponderance of the literature supports the conclusion that SSRI exposure in pregnancy does not adversely affect the health of children exposed in the future. Studies have not found significant differences in intelligence, language development, temperament, mood, distractibility, or behavior problems in children whose mothers used antidepressants during pregnancy. Only one study found subtle changes in fine motor control and slight delays in the psychomotor development in children exposed to SSRIs in utero, but the clinical implications of this finding are unknown.
Exposure to SSRIs in pregnancy and exposure to untreated maternal depression appear to affect the functioning of the HPA axis in infants. A small study comparing salivary cortisol levels in infants of mothers receiving SSRIs in pregnancy (n = 31) and nonexposed infants (n = 45) found that SSRI-exposed infants had significantly lower evening basal salivary cortisol levels, even when the results were controlled for both maternal mood and feeding status. Stress during pregnancy in untreated women also appears to affect cortisol levels; it has been found to be associated with higher cortisol levels and poorer performance on Bayley Scales of Infant Development in infants at 3 months of age. Therefore, further data are needed about the comparative impact of treated and untreated maternal depression on infant temperament, neurocognition, and stress reactivity before the safety of exposure to antidepressants in pregnancy and lactation can be determined, particularly because the studies to date are relatively few and of short duration.
J Midwifery Womens Health. 2010;55(2):90-100. © 2010 Elsevier Science, Inc.
Cite this: Antidepressant Medication Use in Pregnancy - Medscape - Mar 01, 2010.