Sage Journals: Discover world-class research

Abstract

Part of three systematic reviews on the effects of psychotropic medication exposure in pregnancy, this paper critically reviews the literature on adverse effects of antidepressant use during pregnancy, and derives recommendations for clinical practice. Electronic databases were searched for original research studies examining the effects of gestational exposure to antidepressants on pregnancy, neonatal and longer-term developmental outcomes. Most results were derived from cohort (prospective and retrospective) and casecontrol studies. There were no randomized controlled trials. Congenital malformations: 35 studies identified, 12 demonstrated a significant association between antidepressant use in early pregnancy and congenital malformations. Pregnancy outcomes: 35 articles identified, outcomes measured rates of spontaneous abortion (4 out of 7 studies reporting elevated risk), preterm birth (15 out of 19 reporting elevated risk) and abnormal birth weight (8 out of 23 reporting elevated risk). Neonatal outcomes: 17 controlled studies including one meta-analysis were identified concerning neonatal adaptation. 15 studies showed an association between gestational exposure to antidepressants and neonatal adaptation difficulties. Three studies examined an association between selective serotonin reuptake inhibitor (SSRI) exposure and persistent pulmonary hypertension in the neonate with conflicting results. Longer-term developmental outcomes: 6 of 7 studies comparing developmental outcomes of children exposed to antidepressants in utero with non-exposed children reported no significant differences. Most of these medications remain relatively safe in pregnancy, but some significant areas of concern exist, particularly some evidence of higher risk of preterm birth, neonatal adaptation difficulties and congenital cardiac malformations (with paroxetine). The impact of these findings on the risk-benefit analysis when treating pregnant women with antidepressants is discussed.

Keywords

antidepressant agents congenital malformations developmental outcomes neonatal outcomes pregnancy outcomes

It is now widely recognized that women are at increased risk of depression in both pregnancy and the post-partum period. About 12% of women report symptoms of depression during pregnancy. Further, of women who experience significant depression in the post-partum period, 40% of these developed symptoms during pregnancy [1]. Studies have shown that in women with a pre-existing depression who are on antidepressants and discontinue medication during pregnancy the relapse rate is significantly higher than if they continue treatment [2].

There are considerable risks involved if depression is left untreated during pregnancy. Suicide has been identified as a major cause of maternal mortality in a number of studies [3,4]. Maternal depression is associated with poor attendance at antenatal clinics, poor nutrition, increased risk of smoking, alcohol and other drug abuse all of which have negative impacts on foetal well-being [5]. Some studies have suggested a relationship between obstetric complications such as preeclampsia and depression [6–8], whereas others suggest a relationship between ‘chronic stress’ and preterm delivery and low birth weight [8], perhaps due to the effect of chronic anxiety and depression on adrenocortical regulatory functions that influence growth, birth weight and gestation. Finally, maternal depression is associated with poor maternal–infant attachment and poorer long-term developmental outcomes for the child [9].

Clinicians have thus become aware of the importance of active treatment of perinatal depression. Several studies have established foetal exposure to antidepressants during pregnancy via both placental (measured in cord blood) [10–13] and amniotic fluid passage [14]. Variable cord blood-to-maternal serum ratios have been found for both tricyclic depressants (TCAs) [15] and selective serotonin reuptake inhibitors (SSRIs) [10,11,13]. Amniotic fluid levels of drugs have been documented in at least one study [14], which point to another potential avenue of foetal exposure that bypasses first-pass hepatic metabolism. It is therefore clear that exposure in utero must be a consideration for any pregnant woman using antidepressants.

Prescribers need to be aware of the current state of knowledge of the risks to the developing foetus and neonate of exposure to antidepressants. This article evaluates the most recent literature on the use and risks of antidepressants during pregnancy. This systematic review of antidepressants in pregnancy forms part of three separate systematic reviews covering antipsychotics, mood stabilizers and antidepressants in pregnancy, which was undertaken as a joint initiative across the perinatal mental health services in Melbourne.

Methodology

This study was designed with the objective of reviewing the literature on available antidepressants in Australia and outcomes of their use in pregnancy in terms of teratogenicity, pregnancy complications, neonatal complications and longer-term neurodevelopmental outcomes. A computerized search of the following databases was completed: MEDLINE, PsycINFO and EMBASE, for the period of 1950 to May 2009. The key terms used were: ‘pregnancy’, ‘depression’, ‘antidepressants’, ‘neonate’, ‘SSRI’, as well as individually named antidepressants. These were: fluoxetine, fluvoxamine, sertraline, paroxetine, citalopram, escitalopram, reboxetine, venlafaxine, mirtazapine, duloxetine, amitriptyline, imipramine, nortriptyline, clomipramine, doxepin, dothiepin, moclobemide, tranylcypromine and phenelzine. Papers were limited to English language and studies in humans. There was extensive cross-referencing and manual searching through journal articles, particularly review papers for relevant articles not identified through the initial searches. Where available, papers using a scientific, experimental methodology were preferentially reviewed. Only in their absence were case reports and case series considered for discussion.

Results

Congenital malformations

Thirty-five studies of three study types; meta-analyses, cohort and case–control studies were identified. See Table 1. Of these, only twelve (N = 18 451 out of 159 618) demonstrated a statistically significant association between antidepressant use in early pregnancy and congenital malformations.

Table 1.

Congenital malformations

Study	Study design	Medication and n	Findings
Positive	Meta-analysis
Bar-Oz et al., 2007 [19]	Meta-analysis 2 Case–control 4 Retrospective cohort	Paroxetine Other SSRIs Other antidepressants (TCAs, mirtazapine, venlafaxine, nefazodone, trazodone, bupropion, MAOIs, moclobemide) n = 2621	Increased rates of congenital malformations OR 1.72 (95%CI 1.22–2.42) Increased rates of major congenital malformations OR 1.31 (95% CI 1.22–2.42)
Negative studies
Einarson et al., 2005 [18]	Meta-analysis 7 Retrospective cohort	SSRIs, reboxetine, venlafaxine, mirtazapine, nefazodone, trazodone, bupropion n = 1174	No increased risk of congenital malformations RR 1.01 (95%CI 0.57–1.8)
Rahimi et al., 2006 [17]	Meta-analysis 1 Prospective cohort 6 Retrospective cohorts 1 Case–control	SSRIs n = 1782	No increased rates of major congenital malformations OR 1.39 (95%CI 0.91–2.15) p = 0.1554 or congenital cardiac malformations OR 1.19 (95%CI 0.53–2.68) p = 0.8236
O'Brien et al., 2008 [16]	Meta-analysis 3 Case–control 6 Cohort	Paroxetine n = 30 247	No increased rate of congenital malformations OR 1.18 (95%CI 0.88–1.59)
Positive studies	Prospective cohort
Chambers et al., 1996 [20]	Prospective cohort Teratology Information Service	Fluoxetine n = 228 Controls n = 254	Increased rates of more than 3 minor defects 15.5 exposed vs. 6.5% controls p = 0.03
Diav-Citrin et al., 2008 [21]	Prospective cohort Teratology Information Service	Paroxetine n = 410 Fluoxetine n = 314 Controls n = 1467	Increased rates of congenital cardiac malformations for fluoxetine, OR 4.47 (95%CI 1.31–15.27) but not for paroxetine, OR 2.66 (95%CI 0.8–8.9)
Negative studies
Pastuszak et al., 1993 [22]	Prospective cohort Teratology Information Service ‘Motherisk Program’	Fluoxetine n = 128 TCAs n = 74 Controls n = 128	No increased rates of congenital malformations compared with controls, p = 0.3, or within population baseline 1–3%
Goldstein et al., 1997 [31]	Prospective cohort from Eli Lilly Fluoxetine Pregnancy Registry	Fluoxetine n = 796 Historic controls	No increased rates of congenital malformations 3.5% (95%CI 2.1–4.9)
Nulman et al., 1996 [23]	Prospective cohort Teratology Information Service ‘Motherisk Program’	TCAs n = 80 Fluoxetine n = 55 Controls n = 84	No increased rates of congenital malformations TCA n = 3 VSD, hypospadius, pyloric stenosis Fluoxetine n = 2 VSD, PDA Controls n = 2 cyanotic heart disease, VSD
Kulin et al., 1998 [24]	Prospective cohort Teratology Information Service ‘Motherisk Program’	SSRIs n = 267 Controls n = 267	No increased risk of congenital malformations RR 1.06 (95%CI 0.43–2.62)
Einarson et al., 2001 [25]	Prospective cohort Teratology Information Service ‘Motherisk Program’	Venlafaxine n = 150 SSRI n = 150 Controls n = 150	No increased rates of malformations with venlafaxine 2 cases (1.6%) with SSRIs 3 cases (2.4%) within population baseline 1–3%
Casper et al., 2003 [32]	Prospective cohort	SSRIs n = 31 Controls n = 13	No increased rates of major congenital malformations or congenital cardiac malformations VSD, n = 1, p = 0.72 or minor malformations, 54% unexposed 76% exposed, p = 0.17, or More than 3 minor congenital malformations15% unexposed 29% exposed, p = 0.37
Einarson et al., 2003 [26]	Prospective cohort Teratology Information Service ‘Motherisk Program’	Nefazodone/trazedone n = 147 Other antidepressants n = 147 Controls n = 147	No increased rates of congenital malformations 1.6% within population baseline 1–3%
Hendrick et al., 2003 [33]	Prospective cohort	SSRIs n = 138 Controls n =	No increased rate of congenital malformations Rate 1.4% within populations baseline
Chua-Fai-Chan et al., 2005 [27]	Prospective cohort Teratology Information Service ‘Motherisk Program’	Bupropion n = 136 Controls n = 133	No increased risk p = 0.46
Sivojelezova et al., 2005 [28]	Prospective cohort Teratology Information Service ‘Motherisk Program’	Citalopram n = 132 Controls n = 132	No increased rates: 1 case umbilical and scrotal hernia. 0.9% congenital malformation rate
Djulus et al., 2006 [29]	Prospective cohort Teratology Information Service ‘Motherisk Program’	Mirtazapine n = 104 Other antidepressant n = 104 Controls n = 104	No increased rates of major congenital malformation with mirtazapine exposure 1.9% (95%CI 0.5%-3.6%) or other antidepressant exposure 1% (95%CI 0.2%-2.7%)
Einarson et al., 2008 [30]	Prospective cohort pooled results from Teratology Information Service ‘Motherisk Program’	Paroxetine n = 3435	No increased rates of congenital cardiac malformations TIS cohort 0.7% in exposed 0.7% in unexposed OR 1.1 (95%CI 0.36–2.78) Databased studies 1.5% Pooled TIS and Database 1.2% (95%CI 1.1–2.1)
Wisner et al., 2009 [34]	Prospective cohort with untreated depressed control arm	SSRIs venlafaxine, bupropion, TCADs (confirmed with maternal serum levels) n = 238 categorized into 5 groups. SSRI exposure continuous n = 48, partial n = 23 Major depressive disorder continuous n = 14, partial n = 22 Controls non-depressed, non-exposed SSRI n = 131	No increased rate of minor congenital malformations with partial or continuous antidepressant exposure or depression No major malformations observed 203 infants examined, 30 (15%) more than 3 minor congenital malformations
Positive studies	Retrospective cohort birth registries
Kallen et al., 2003 [35]	Retrospective cohort Swedish Birth Registry	TCAs/tetracyclics SSRIs n = 5015 Controls n = 577 730	Increased rates of congenital cardiac malformations with TCA/tetracyclics exposure OR 1.77 (95%CI 1.07–2.91), congenital cardiac malformations with Clomipramine exposure OR 2.03 (95%CI 1.22–3.4) No increased rates with SSRIs in general
Kallen et al., 2007 [36]	Retrospective cohort Swedish Birth Registry	SSRIs SNRIs NRIs n = 6481 Controls n = 860 215	Increased risk of cardiac congenital malformations with paroxetine use, OR 1.63 (95%CI 1.05–2.53)
Negative studies
Ericson et al., 1999 [37]	Retrospective cohort Swedish Birth Registry	SSRIs, TCAs, tertracyclics, MAOI n = 969 Controls n = 281,728	‘No increased risk’ - statistical data not reported.
Malm et al., 2005 [38]	Retrospective cohort Finnish Birth Registry	SSRIs n = 1782 Controls n = 1782	No increased risk p = 0.4
Lennestal et al., 2007 [39]	Retrospective cohort Swedish Birth Registry	SNRIs NRIs n = 732 Controls 860,215	No increased rates of major congenital malformations (3.8%) compared with general population rate (4.7%) or congenital cardiac malformations RR 1.2 (95%CI 0.6–2.2)
Positive studies	Retrospective cohort Database
Wolgelius et al., 2006 [40]	Retrospective cohort Database	SSRIs n = 1051 Controls n = 150,780	Increased rates of malformations with SSRI exposure anytime during pregnancy (n =51) 4.9% RR 1.34 (95%CI 1.0–1.79) and in 2nd/3rd trimester (n = 31), RR 1.84 (95%CI 1.25–2.71)
Cole et al., 2007 [41]	Retrospective cohort Database	Paroxetine n = 815 Controls (other antidepressant)s n = 4936	Increased risk of major malformations with paroxetine mono-therapy OR 1.89 (95%CI 1.2–2.89) and paroxetine mono or poly-therapy OR 1.76 (95%CI 1.18–2.64) No increased risk of cardiovascular malformations with paroxetine exposure
Davis et al., 2007 [42]	Retrospective cohort Database	SSRIs n = 1047 TCAs n = 221 Other antidepressants n = 173 Controls n = 49663	Increased rates of congenital malformations for TCAs exposure limb defects n = 7, RR 2.55 (95%CI 1.23–5.29), and Spina bifida RR 12.43 (95%CI 1.7–90.66) No increased rates for SSRIs over all increasead rates of congenital eye anomalies with paroxetine exposure, RR 2.36 (95%CI 1.2–4.66) No increased rates of cardiovascular defects with paroxetine exposure
Negative studies
Simon et al., 2002 [43]	Retrospective cohort	TCAs n = 209, Controls n = 209 SSRIs n = 185, Controls n = 184	No increased rates of congenital malformations
Wen et al., 2006 [44]	Retrospective cohort	SSRIs n = 972 Controls n = 3878	No increased rates of major malformations OR 0.98 (95%CI 0.59–1.64) No increased rates of minor malformations OR 1.02 (95%CI 0.69–1.51)
Wichman et al., 2009 [45]	Retrospective cohort	SSRIs, venlafaxine n = 808 Controls n = 24406	No increased rates of congenital malformations above population baseline p = 0.23
Positive studies	Case–control
Rosa, 1994 [46]	Case–control	Cases = 6454 Completed pregnancies = 104,339 TCAs, tetracyclic antidepressants and lithium	Increased risk limb reduction defects with amitriptyline n = 3 RR 4.7 (95%CI 1.5–15), Other medications no statistically significant increase in risk
Berard et al., 2006 [47]	2 Nested case–control	Total n = 1403 Paroxetine n = 542 Other SSRIs n = 443 (sertraline, citalopram, fluoxetine, fluvoxamine) Other antidepressant n = 418 (TCAs, MAOIs, SNRIs, trazodone, nefazodone) Cases n = 101 Controls n = 1302	Paroxetine exposure >25 mg/day increased rates of major congenital malformations OR 2.23 (95%CI 1.19–4.17) and congenital cardiac malformations OR 3.07 (95%CI 1.0–9.42) Other paroxetine exposure no increased rates of congenital cardiac malformations OR 1.38 (95%CI 0.49–3.92) Other SSRI exposure no increased rates of congenital cardiac malformations OR 0.89 (95%CI 0.28–2.84)
Alwan et al., 2007 [48]	Case–control Birth Registry	Cases n = 9622 Controls 4092 SSRIs	Increased rates of major congenital malformations with SSRI use Anencephaly n = 9 OR 2.4 (95%CI 1.1–5.1) Craniosynostosis n = 24 OR 2.5 (95%CI 1.5–4.0) Omphalocele n = 11 OR 2.8 (95%CI 1.3–5.7)
Louik et al., 2007 [49]	Case–control Slone Birth Defects Study	SSRIs n = 9849 Controls n = 5860	All SSRI exposure no increased rate of congenital malformations (Craniosynostosis 15 cases, Omphalocele 127 cases and heart defects 3724 cases) Sertraline exposure increased rates of Omphalocele OR 5.7 (95%CI 1.6–20.3) 3 cases, Septal defects 13 cases OR 2.0 (95%CI 1.2–4.0) Paroxetine increased rates of right ventricular outflow obstructions 6 cases, OR 3.3 (95%CI 1.3–8.8)
Negative studies
Ramos et al., 2008 [50]	Case–control	TCADs SSRIs, SNRIs, bupropion, moclobamide, trazodone, nefazodone n = 2329	No increased rates of major congenital malformations with antidepressant exposure or duration of antidepressant use compared with unexposed infants

Meta-analyses

Of four meta-analyses that collectively examined 102 835 cases, three meta-analyses did not find that first trimester paroxetine, SSRI or other antidepressant exposure increased the risk of major or cardiac malformations [16–18]. In contrast, one meta-analysis involving 2621 cases found that first trimester paroxetine use was associated with a significantly increased risk of major congenital malformations, OR 1.31 (95%CI 1.03–1.67) and cardiac congenital malformations, OR 1.72 (95%CI 1.22–2.42) [9]. In this study, however, matching to controls was incomplete with exposed women and exposed infants having statistically significantly more obstetric diagnostic tests.

Cohort studies

There were twenty-six cohort studies with matched controls. The studies were divided into the following types:

Prospective cohort studies including teratology information services and observational studies, and

Retrospective cohort studies which include birth registries, database studies and ‘other’ retrospective cohort studies

Prospective cohort studies

Fifteen prospective cohort studies were identified. Eleven studies [20–30] were conducted by Teratology Information Services (TIS), nine of which were conducted by the Motherisk Program in Toronto, Canada and [22–30] and four from unrelated centres [31–34].

Only two of the 15 studies (N = 952 out of 6663 patients) found a significant association between fluoxetine or paroxetine exposure and congenital malformations [20,21]. Diav-Citrin identified 410 women who used paroxetine and 314 who used fluoxetine during the first trimester and matched them to 1467 controls [21]. After exclusion of genetic and cytogenetic anomalies and controlling for cigarette smoking, they found fluoxetine, but not paroxetine was significantly associated with an increased risk of cardiac malformations. The authors state that the strength of the findings is open to question because of the broad definition of caseness and the wide range of the 95% confidence intervals. In the other positive study, Chambers found a significant association between fluoxetine exposure and at least three minor congenital malformations [20]. Neither of these findings has been replicated in other studies.

The best study design would take into account the effect of maternal depression on the foetus and include a control group of depressed but not medicated women in their design. Of the two studies that did this, both had negative results but were underpowered and had multiple confounders [32,34].

Retrospective cohort studies

Eleven retrospective cohort studies are included in this review with five studies yielding positive results (N = 14 803 out of 20 460 patients) Five studies utilized data from birth registries, and of these, there were two positive studies both by Kallen utilizing data from the Swedish Birth Registry [35,36]. In 2003, Kallen found TCA or tetracyclic antidepressant use in early pregnancy was significantly associated with cardiac malformations. The association was strongest with clomipramine, OR 2.03 (95%CI 1.22–3.4) [35]. In 2007, Kallen reported a significant and positive association between the use of paroxetine in early pregnancy and cardiac malformations, OR 1.63 (95%CI 1.05–2.53) but not between SSRIs in general and congenital malformations [36]. In contrast, two other registry studies also using the Swedish Birth Registry and one using data from the Finnish Birth Registry did not find an increased risk of major or cardiac malformations in infants exposed to antidepressants [37–39].

Four retrospective cohort studies utilised databases, three with positive [40–42] and one with negative results [43]. Cole and colleagues found an increased risk of major congenital malformations with both first trimester paroxetine monotherapy OR 1.89 (95%CI 1.2–2.89) and paroxetine polytherapy OR 1.76 (95%CI 1.18–2.64) but not cardiac malformations [41]. Davis found that first trimester tricyclic antidepressant exposure was associated with an increased rate of limb abnormalities RR 2.55 (95%CI 1.23–5.29) and spina bifida RR 12.43 (95%CI 1.7–90.66) [42]. The validity of this last finding is questionable given the extremely wide range of the 95% confidence interval. Davis also found that paroxetine exposure was associated with an increased risk of congenital eye abnormalities, RR 2.36 (95%CI 1.2–4.66) [42]. The final positive study found that first trimester SSRI exposure was associated with an increased risk of major malformations, RR 1.34 (95%CI 1.0–1.79) as were second and third trimester exposure, RR 1.84 (95%CI 1.25–2.71) [40].

The two final non-database retrospective cohort studies by Wen-Shi Wu [44] and Wichman [45] failed to find a significant association between SSRIs or other antidepressant exposure and major, cardiac or minor congenital malformations.

Case–control studies

Five case–control studies were included in this review. Four studies found an association between antidepressant use and congenital malformations [46–49] and one did not [50]. Alwan and colleagues using data from the National Birth Defects Prevention Study identified 9622 cases of congenital malformations and 4092 controls. They found maternal SSRI use in early pregnancy was associated with anencephaly, OR 2.4 (95%CI 1.1–5.1), craniosynostosis, OR 2.5 (95%CI 1.5–4.0) and omphalocele, OR 2.8 (95%CI 1.3–5.7) [48]. It is important to note that the authors made 265 comparisons without adjustment to the level of significance hence increasing the likelihood that some of the positive outcomes could be due to chance. Louik and others conducted a case–control study using data from Sloane Epidemiology Centre Birth Defects Study [49]. They identified 9849 infants with congenital malformations and matched them to 5860 control infants. They found that SSRI exposure during pregnancy was not associated with congenital malformations but was associated with right ventricular outflow tract obstruction defects, OR 2.0 (95%CI 1.1–3.6). Despite the large study size, the authors warned against viewing their findings as strong evidence given the small number of exposed infants with rare outcomes. Berard and colleagues studied the association between exposure to a range of antidepressant types in a nested case–control study of 1403 cases and found a statistically significant association between paroxetine and major congenital, OR 2.23 (95%CI 1.19–5.17) and cardiac malformations, OR 3.07 (95%CI 1.0–9.42), but only with doses of paroxetine greater than 25 mg/day [47]. The final case–controlled study was reported in a letter by Rosa in 1994, it is included in this review because of the large sample size [46]. The author reported on an analysis performed by the Food and Drug Administration using the Medicaid database of 6454 infants identified with congenital malformations linked to 104 339 completed pregnancies. They found that amitriptyline use was associated with an increased risk of limb reduction malformations, RR 4.4 (95%CI 1.5–5.1).

Summary

It is reassuring that of the 35 studies reviewed, only twelve demonstrated statistically significant associations between antidepressant use in early pregnancy and congenital malformations. The methodological limitations of the study designs are notable and only some findings have been replicated. Bar-Oz's meta-analysis [19] finding of an association between paroxetine and cardiac malformations is replicated by Kallen [36], and Berard [47] but is refuted by others [21,41,42,50]. Bar-Oz's other finding of an association between paroxetine exposure and a major congenital malformation is replicated by Cole and Berard [40,41]. Though Berard found an increased risk with paroxetine only for doses greater than 25 mg [47]. Kallen's findings of an association between cardiac malformations and TCAs and tetracyclics, in particular clomipramine and Diav-Citrin's finding with fluoxetine failed to be replicated by many others [22,23,34,37,40,42,46,50]. Rosa's report of an association between amitriptyline and limb reduction malformations [46] is replicated by Davis in 2007 [42], while Alwan's finding of an association between SSRI use and omphalocele [48] was partially replicated by Louik's finding of an association between sertraline and omphalocele [49].

It is clear that the majority of studies examining the full range of antidepressants (SSRIs, TCAs, SNRIs, MAOIs, tetracyclics, moclobemide) yield negative results, but interpreting the various results from studies of different methodology is difficult. On balance, with the studies included in this review, it would appear that antidepressants as a group and individual antidepressants apart from paroxetine, are unlikely to be teratogenic.

Pregnancy outcomes

Thirty-five articles were analysed that reported pregnancy outcomes in infants exposed to antidepressants (Table 2). The outcomes measured were rates of spontaneous abortion (two out of six yielding positive results N = 3318 out of 3422), preterm birth defined as birth earlier than 37 weeks of gestation (15 out of 19 with positive results N = 8286 out of 8735), abnormal birth weight (8 out of 23 with positive results N = 6907 out of 10 890), gestational age (4 out of 17 with positive results N = 2271 out of 6853) and stillbirth (one study).

Table 2.

Pregnancy outcomes

Study	Design	Medication and n	Outcomes: Spontaneous abortion
Positive studies
Hemel et al., 2005 [51]	Meta-analysis	Exposed n = 1534 Non-exposed n = 2033	RR 1.45 (95%CI 1.19–1.77)
Rahimi et al., 2006 [17]	Meta-analysis	9 studies	OR 1.7 (95%CI 1.19–1.77)
Negative studies
Pastuszak et al., 1993 [22]	Prospective cohort	Fluoxetine n = 128 TCAs n = 74 Controls n = 128	RR 1.9 (95%CI 0.92–3.92). Risk with fluoxetine similar to that of TCAs
Kulin et al., 1998 [24]	Prospective cohort	SSRI n = 267 Controls n = 267	20 cases in exposed group versus 21 in non-exposed p = 0.24
Einarson et al., 2001 [25]	Prospective case–control	Venlafaxine n = 150 SSRI n = 150 Controls n = 150	12% exposed versus 7% non-exposed p = 0.24
Djulus et al., 2006 [29]	Prospective case–control	Mirtazapine n = 104 Other antidepressant n = 104 Controls n = 104	19% exposed versus 11% non-exposed p = 0.12
Study	Design	Medication and n	Outcomes-risk of preterm birth
Positive studies
Chambers et al., 1996 [20]	Prospective cohort	Fluoxetine n = 228 Controls n = 254	RR 4.8 (95%CI 1.1–20.8)
Ericson et al. 1999 [37]	Prospective cohort	SSRIs, TCAs, tetracyclics, MAOIs n = 969 Controls n = 281,728	OR 1.43 (95%CI 1.14–1.8)
Costei et al., 2002 [54]	Prospective cohort	Paroxetine n = 55 Controls n = 27	20% exposed preterm versus 4% non-exposed controls p = 0.02
Simon et al. 2002 [43]	Retrospective cohort	SSRIs n = 185 Controls n = 184 TCAs n = 209 Controls n = 209	SSRI OR 4.38 (95%CI 1.57–12.22) No difference with TCA
Kallen et al. 2004 [53]	Prospective cohort	Citalopram n = 285 Paroxetine n = 106 Fluoxetine n = 91 Sertraline n = 77 Venlafaxine n = 24 Controls: all infants born in Sweden 1995–2001	All antidepressants OR 1.96 (95%CI 1.55–2.22) >25weeks Gestation use SSRI OR 2.06 (95%CI 1.58–2.69) TCA OR 2.50 (95%CI 1.87–3.4)
Zeskind et al., 2004 [55]	Retrospective cohort	SSRIs n = 17 Controls n = 17	Increased prematurity p = 0.019
Lattimore et al., 2005 [52]	Meta-analysis	SSRIs n = 511 Controls n = 438	OR 2.03 (95%CI 1.61–2.6)
Djulus et al., 2006 [29]	Prospective cohort	Mirtazapine n = 104 Other antidepressants n = 104 Controls n = 104	10% exposed versus 7% controls p = 0.04
Oberlander et al., 2006 [56]	Prospective cohort	SSRIs n = 1451 Depressed n = 14234 Non-exposed non-depressed controls n = 92192	Difference between depressed and exposed groups p < 0.001
Wen et al. 2006 [44]	Retrospective cohort	SSRIs n = 972 Controls n = 3878	OR 1.57 (95%CI 1.28–1.92)
Davis et al. 2007 [42]	Retrospective cohort	SSRIs n = 1047 TCAs n = 221 Other antidepressants n = 173 Controls n = 49663	Exposed 9.4% Non-exposed 6.6% RR 1.45 (95%CI 1.25–1.68) Exposed 11%, non-exposed 6.6% RR 1.67 (95%CI 1.25–2.22)
Lennestal et al., 2007 [39]	Retrospective cohort	SNRIs NRIs n = 732 Controls n = 860,215	OR 1.6 (95%CI 1.19–2.15)
Suri et al. 2007 [57]	Prospective cohort	SSRIs, SNRI, TCAs n = 49 Depressed untreated control n = 22	14% Preterm in exposed versus 0% depressed 5% controls P = 0.05
Maschi et al. 2008 [58]	Prospective cohort	SSRIs, TCAs n = 200 Controls n = 1200	Significantly higher prematurity OR 2.31 (95%CI 1.14–4.63) Chronic exposure OR 4.3 (95%CI 1.31–14.07)
Wisner et al. 2009 [34]	Prospective cohort with untreated depressed control arm	SSRIs venlafaxine, bupropion, TCADs (confirmed with maternal serum levels) n = 238 categorized into 5 groups. SSRI exposure continuous n = 48, partial n = 23 Major depressive disorder continuous n = 14, partial n = 22 Controls non-depressed non-exposed SSRI n = 131	Continuous exposure 20% more likely to have preterm birth than partial or control Continuous depression RR 3.71 (95%CI 0.98–14.13) Continuous exposure RR 5.43 (95%CI 1.98–14.13)
Negative studies
Pastuszak et al., 1993 [22]	Prospective cohort	Fluoxetine n = 128 TCA n = 74 Controls n = 128	No difference noted p = 0.22
Casper et al., 2003 [32]	Prospective cohort	SSRIs n = 31 Depressed control n = 13	Exposed 3%, non-exposed depressed group 8% p = 0.53
Sivojelezova et al., 2005 [28]	Prospective case–control	Citalopram n = 132 Control n = 132	Exposed 11 preterm Controls 5 preterm P = 0.25
Pearson et al., 2007 [59]	Retrospective cohort	SSRIs n = 84 Controls n = 168	Exposed 10.7 % preterm Non-exposed 10.1% preterm p = 0.88
Study	Design	Medication and n	Outcomes: gestational age
Positive studies
Simon et al., 2002 [43]	Retrospective cohort	SSRIs 185 Controls n = 184 TCAs 209 Controls n = 209	Mean gestational age 38.5 weeks in SSRI exposed versus 39.4 unexposed. Adjusted difference −0.9 weeks (95%CI-1.3–0.5). No difference between TCAs and unexposed controls
Oberlander et al., 2004 [10]	Prospective cohort	SSRIs n = 28 SSRIs plus clonazepam n = 18 Controls n = 23	Exposed 38.8 weeks. Depressed non-exposed 39.1 p < 0.001
Malm et al., 2005 [38]	Prospective cohort	SSRI n = 1782 Controls 1782	Mean gestational age lower exposed 39.1 non exposed 39.4 p < 0.001
Suri et al., 2007 [57]	Prospective cohort	SSRIs, SNRI, TCAs n = 49 Depressed untreated controls n = 22	Mean gestational age lower for exposed group. Difference increased with higher doses.
Negative studies
Nulman et al., 1997 [60]	Prospective cohort	Fluoxetine n = 80 TCAs n = 50 Controls n = 84	No difference in gestational age p = 0.10
Kulin et al., 1998 [24]	Prospective cohort	SSRIs n = 267 Controls n = 267	No difference in gestational age detected
Cohen et al., 2000 [61]	Retrospective file review	SSRIs p = 64 Controls n = 64	No difference in gestational age between early, late and non exposed groups
Einarson et al., 2001 [25]	Prospective cohort	Venlafaxine n = 150 SSRIs n = 150 Controls n = 150	No difference in gestational age p = 0.72
Einarson et al., 2003 [26]	Prospective cohort	Trazodone n = 58 Nefazodone n = 89 Other antidepressants n = 147 Control n = 147	No difference in gestational age p = 0.57
Heikkenin et al., 2002 [12]	Prospective cohort	Citalopram n = 11 Controls n = 10	No difference in gestational age p = 0.26
Nulman et al., 2002 [62]	Prospective cohort	Fluoxetine n = 40 TCAs n = = 46 Controls n = 36	No difference in gestational age
Casper et al., 2003 [32]	Prospective cohort	SSRIs n = 31 Controls n = 13	No difference in gestational age p = 0.56
Heikkenin et al., 2003 [13]	Prospective case–control	Fluoxetine n = 11 Controls n = 10	No difference in gestational age p = 0.26
Laine et al., 2003 [63]	Prospective cohort	Fluoxetine citalopram n = 20 Controls n = 20	No difference in gestational age p = 0.06
Suri et al., 2004 [64]	Prospective case–control	Depressed fluoxetine exposed n = 28 Depressed fluoxetine non-exposedn = 18 Controls n = 16	No difference in gestational age p = 0.23
Sivojelezova et al., 2005 [28]	Prospective case–control	Citalopram n = 132 Controls n = 132	No difference in gestational age
Oberlander et al., 2008 [65]	Prospective cohort	SSRIs n = 3500	No difference between early and late exposure
Study	Design	Medication and n	Outcomes: birth weight
Positive studies
Chambers et al., 1996 [20]	Prospective cohort	Fluoxetine n = 228 Controls n = 254	Birth weight lower p = 0.04
Simon et al., 2002 [43]	Retrospective cohort	SSRIs n = 185 Controls n = 184 TCAs n = 209 Controls n = 209	Birth weight <2500 g in SSRIs exposed 6.5% versus non-exposed 3.3% OR 4.3 (95%CI 1.5–12.2). No difference with TCAs compared with non-exposed
Kallen et al., 2004 [53]	Retrospective cohort	citalopram n = 285, paroxetine n = 106, fluoxetine n = 91, sertraline n = 77 venlafaxine n = 24 Controls: all infants born in Sweden 1995–2001.	Birth weight <2500 g in TCAs OR 1.88 (95%CI 1.28–2.26), SSRIs OR 1.98 (95%CI 1.42–2.76)
Malm et al., 2005 [38]	Prospective cohort	SSRIs n = 1782 Controls n = 1782	Exposed 3476 g, non-exposed 3574 g p < 0.001
Oberlander et al., 2006 [56]	Prospective cohort	SSRIs n = 1451 Depressed n = 14234 Non-exposed non-depressed control n = 92 192	Difference between exposed and depressed p = 0.05 Difference between depressed and control p < 0.001
Wen et al., 2006 [44]	Retrospective cohort	SSRIs n = 972 Controls n = 3878	Birth weight <2500 g OR 1.58 (95%CI 1.19–2.1)
Oberlander et al., 2008 [65]	Prospective case–control	SSRIs n = 3500	Low birth weight related to length of exposure p < 0.05
Lattimore et al., 2005 [52]	Meta-analysis	SSRIs n = 511 Controls n = 438	Low birth weight OR 3.64 (95%CI 1.01–13.08)
Negative studies
Nulman et al., 1997 [60]	Prospective cohort	TCAs n = 80 Fluoxetine n = 55 Controls n = 84	No difference, n = 0.18
Cohen et al., 2000 [61]	Retrospective cohort	Fluoxetine n = 64 Controls n = 64	No difference in exposed infants
Einarson et al., 2001 [25]	Prospective cohort	Venlafaxine n = 150 SSRIs n = 150 Controls n = 150	Venlafaxine exposed 3332 g Controls 3452 g, p = 0.34
Costei et al., 2002 [54]	Prospective cohort	Paroxetine n = 55 Controls n = 27	Paroxetine exposed 3394 g, Controls 3578 g p = 0.09
Einarson et al., 2003 [26]	Prospective cohort	Nefazodone/trazedone n = 147 Other antidepressants n = 147 Controls n = 147	SSRIs mean weight 3652 g Controls 3418 g, p = 0.24
Casper et al., 2003 [32]	Prospective cohort	SSRIs n = 31 Depressed control n = 13	No difference between exposed and depressed group, p = 0.84
Suri et al., 2004 [64]	Prospective case–control	Depressed fluoxetine exposed n = 28, depressed fluoxetine non-exposed n = 18 Controls n = 16	No difference in weight, p = 0.06
Zeskind et al., 2004 [55]	Prospective case–control	SSRIs n = 17 Controls n = 17	Exposed group 3452 g, non-exposed group 3298 g, p = 0.25
Pastuszak et al., 1993 [22]	Prospective cohort	Fluoxetine n = 128 TCAs n = 74 Controls n = 128	Fluoxetine weight 3459 g, TCA weight 3519 g, p = 0.62
Djulus et al., 2006 [29]	Prospective cohort	Mirtazapine n = 104 Other antidepressant n = 104 Controls n = 104	Mirtazapine mean weight 3335 g Control mean weight 3502 g, p = 0.08
Sivojelezova et al. 2005 [28]	Prospective case–control	Citalopram n = 132 Controls n = 132	No significant difference between Citalopram mean weight 3540 g and Non-exposed control 3436 g
Suri et al. 2007 [57]	Prospective case–control	SSRIs n = 49 Depressed control n = 22	No difference in mean birth weight n = 0.85
Heikkinen et al., 2002 [12]	Prospective case–control	Citalopram n = 11 Controls n = 10	Exposed weight 3460 g Control 3560 g, p = 0.62
Laine et al., 2003 [63]	Prospective cohort	Fluoxetine or citalopram n = 20 Controls n = 20	No difference
Wisner et al., 2009 [34]	Prospective cohort	SSRIs venlafaxine, bupropion, TCADs (confirmed with maternal serum levels), n = 238 categorized into 5 groups SSRI exposure continuous n = 48 partial n = 23 Major depressive disorder continuous n = 14, partial n = 22 Controls non-depressed non-exposed SSRI n = 131	No difference

Spontaneous abortion

No specific studies focused on the impact of antidepressants on spontaneous abortion. However, it was a secondary outcome in two meta-analyses. Hemel reported that in 1534 treated patients there was a statistically significant increase in spontaneous abortions of 3.9%, RR of 1.45 (95%CI 1.19–1.77) [51]. Rahimi reviewed nine studies and also found an increase risk OR 1.7 (95%CI 1.28–2.24) [17]. Unfortunately, none of these studies adequately controlled for previous miscarriages, age, smoking, other drug use or depression.

Preterm birth

Nineteen studies examine the incidence of preterm birth in women exposed to antidepressants in late pregnancy (see Table 2). These include one meta-analysis, 12 prospective cohort studies, and six retrospective cohort studies. In contrast, 17 studies looked at mean gestational age (rather than preterm birth) and of these only four showed a lowering of gestational age in exposed babies (see Table 2).

Lattimore's meta-analysis of nine studies showed a significant increase in preterm births with an OR 2.23 (95%CI 1.61–2.6) [52]. Of the 12 prospective cohort studies, nine (involving antidepressants (SSRIs, Mirtazapine, TCA and SNRI)) demonstrated a statistically significant increase in preterm delivery rates. Two of these studies showed a relationship between an increase in preterm delivery rates and late exposure [20,53]. Maschi and colleagues revealed a correlation between preterm birth and chronic exposure to antidepressants but not with short term exposure [58]. Five out of six retrospective cohort studies also demonstrated a positive finding of exposure and preterm birth (see Table 2). All of these studies have similar weaknesses, with no measures of actual drug exposure or controls for confounders such as smoking and in particular the effects of underlying depression.

Some of the more rigorous studies have tried to tease out the role of untreated maternal depression. Oberlander [56] examined 119 547 prescription records matched with hospital separation records and found that SSRI-exposed babies had a higher rate of preterm birth than babies exposed to depression alone (p < 0.01). A further study, by the same group, of 3500 cases found that increased length of drug exposure was related to higher risk of low gestational age and low birth weight [65]. These studies are weakened by the fact that hospital records may be unreliable and that the authors did not correct for multiple comparisons. A small prospective study compared women with depression alone and treatment with SSRIs and found a statistically significant higher preterm birth rate in the exposed infants (14% exposed group, 0% depressed group) [62]. In contrast, another small study did not detect any difference between treated and depressed groups [32].

Finally, a recent carefully designed study looked at five overlapping groups using a prospective design [34]. Patients were classified into those with no SSRI or depression exposure, those with continuous or partial SSRI exposure and those with continuous or partial depression exposure. Unlike most other studies, exposure was confirmed with serum SSRI levels. The group with continuous SSRI exposure had a significantly higher preterm delivery rate with a RR of 5.43 (95%CI 1.98–14.84). The group with continuous depression also had a higher rate but this was no longer statistically significant when controlled for maternal age and race, RR 3.7 (95%CI 0.98–14.13). The partial SSRI and partial depression groups were no different from controls.

Birth weight

Twenty-three studies were identified that examined birth weight, with eight yielding positive findings. Lattimore's meta-analysis revealed a statistically significant risk of low birth weight for gestational age [52]. Seven other studies showed similar findings [20,38,43,44,53,56,65), but fifteen did not [12,22,25,26,28,29,32,34,54,57,60,61,64,66]. Oberlander attempted to increase the sensitivity of the study by using propensity scoring to control for confounders and still found an effect on birth weight related to exposure [56]. However, only two studies controlled for confounders such as depression, smoking, maternal age or maternal weight, and these showed no differences in birth weight in babies born to exposed or non-exposed mothers [34,64].

An increased possible risk of large birth weights following TCA treatment was reported in one study [37]. Although partially related to higher body mass index in mothers taking antidepressants, the effect did not completely disappear when this was controlled for.

Stillbirth

No studies specifically designed to look at the risk of stillbirth (i.e. late trimester obstetric loss) upon exposure to antidepressants were identified. Very large numbers would be needed to show any statistically significant differences. Of note, none of the large birth registry studies has shown any increased incidence of stillbirth in exposed infants.

Pregnancy complications

One study has shown an increased risk of hypertension and possibly pre-eclampsia in women exposed to SSRIs beyond the first trimester with a RR 4.9 (95%CI 2.7–8.8) [66]. However, this retrospective study was flawed, as it relied on patient recall of the diagnosis rather than any objective medical measures of blood pressure or other criteria for pre-eclampsia.

Summary

It is evidence that incidence of preterm birth is elevated in the majority of the studies that specifically examined this outcome, including some that attempt to control for the impact of the underlying depressive illness. The relationship is strengthened by one study finding that longer drug exposure is more likely to decrease gestational age. Other pregnancy outcomes such as reduced birth weight are weaker in their association. The data on rates of spontaneous abortions is limited and contradictory.

Antidepressants and neonatal outcomes

Neonatal adaptation

Seventeen controlled studies, including one meta-analysis, were analysed for information on neonatal outcomes after antidepressant exposure (Table 3). Fifteen of the studies show some association between gestational exposure to antidepressants and neonatal adaptation difficulties (N = 7869 out of 7900). Neonatal adaptation is measured in various ways, from gross markers such as NICU admission to more subtle evaluations such as behavioural observations, in different studies making it difficult to draw general conclusions.

Table 3.

Neonatal outcomes

Authors	Study design	Medications and n	Findings
Positive studies
Chambers et al., 1996 [20]	Prospective cohort	Fluoxetine n = 228 Controls n = 224	Increased risk of SCN admission RR = 2.6 (95%CI 1.1–6.9) Increased risk of poor neonatal adaptation RR = 8.7 (95%CI 2.9–26.6)
Boucher et al., 2008 [76]	Case–control	SSRIs n = 73 Controls n = 73	Exposed neonates more likely than non-exposed to exhibit symptoms of: altered alertness RR 17 (95%CI 4–68), OR 37 (95%CI 8–174); muscular tone RR11 (95%CI 3–35) OR 20 (95%CI 5–71); feeding/ gastrointestinal difficulties RR 2.27 (95%CI 1.36–3.79) OR 3.8 (95%CI 1.7–8.1); tachypnoea RR 1.63 (95%CI 1.10–2.61); and globally affected RR 2.04 (95%CI 1.49–2.79) OR 7.0 (95%CI 3.2–15.3)
Costei et al., 2002 [54]	Prospective cohort	Paroxetine: neonates exposed in 3rd trimester n = 55 Neonates exposed in 1st or second trimester n = 27 Neonates of women using non-teratogenic drugs n = 27	12/55 infants exposed in third trimester had prolonged hospitalization (p = 0.03) Respiratory distress occurred in 9 of these. All recovered. In the paroxetine exposure group, 8/36 breastfeeding reported problems (i.e. alertness, constipation, sleepiness, or irritability) compared to 0/44 in the control group (p = 0.001)
Laine et al., 2003 [66]	Prospective cohort	SSRIs n = 20 (citalopram and fluoxetine) Controls n = 20	SSRIs group lower Apgar score at 15 min p < 0.02, increased HR p < 0.049 Lower cord blood 5HIAA concentrations p = 0.02 Risk of at least one serotonergic symptom OR 6.9 (95%CI 1.6–29.2) p = 0.008 At 2 weeks and 2 months no difference in symptom scores p = 0.22
Oberlander et al., 2004 [10]	Prospective cohort	SSRIs n = 28 SSRIs plus clonazepam n = 18 Controls n = 23	Increased risk of SCN admission (p = 0.018)
Zeskind and Stephens, 2004 [55]	Prospective cohort	SSRIs n = 17 Controls 17	Increased rates of tremulousness p = 0.038, number of different behavioural states p = 0.009, number of changes in behavioural states p = 0.05, number of bouts of active sleep p = 0.001
Lattimore et al., 2005 [52]	Meta-analysis of cohort studies	SSRIs n = 511 Controls n = 438	Late term SSRI use associated three times increased risk of SCN/NICU admission
Malm et al., 2005 [38]	Prospective cohort	SSRIs n = 1782 Controls n = 1782	Increased risk of admission to SCN/NICU in third trimester exposure compared with first trimester exposure OR 1.6 (95%CI 1.1–2.2) p = 0.009
Sivojelezova et al., 2005 [28]	Prospective cohort	Citalopram n = 132 Depressed controls n = 132 Non-depressed controls n = 132	Increased risk of NICU admission RR 4.2 (95%CI 1.71–10.26)
Kallen et al., 2004 [53]	Retrospective cohort	Citalopram n = 285, Paroxetine n = 106 Fluoxetine n = 91 Sertraline n = 77 Venlafaxine n = 24 Controls: all infants born in Sweden 1995–2001	Increased respiratory distress, OR = 2.21 for all antidepressants. No significant effect on BSL. Low Apgar score OR = 2.33 for all antidepressants. Convulsions OR = 4.7 Paroxetine not significantly worse than others. SSRIs and TCAs not different; >24 wk exposure not significantly different from any exposure
Oberlander et al., 2006 [56]	Retrospective cohort	SSRIs n = 1451 Depressed group n = 14234 Non-exposed non-depressed control n = 92192	Longer hospital stays p < 0.07, OR 0.43 (95%CI 0.12 to 0.74) than depressed group or non-exposed non-depressed control.
Wen et al., 2006 [44]	Retrospective cohort	SSRIs n = 972 Control n = 3878	Increased rate of seizures, OR 3.87 (95%CI 1.0–14.99).
Levinson-Castiel et al., 2006 [75]	Prospective cohort	SSRIs n = 60 Control n = 60	Symptoms of ‘neonatal abstinence syndrome’ present in 18 (30%) exposed infants versus no controls (p < 0.001). Maximum mean daily symptoms occurred within 48h of birth. Brief seizures occurred in 2. No infant required treatment. Relationship to SSRI dosage only measured in paroxetine group. Mean dose increased with severity of symptoms (p = 0.01)
Davis et al., 2007 [42]	Case–control from managed care organization data	SSRIs n = 2201 Controls n = not reported	Increased respiratory distress syndrome, endocrine, metabolic disturbance, hypoglycaemia, temperature dysregulation, convulsions
Ferreira et al., 2007 [68]	Case–control	SSRIs or venlafaxine n = 76 60.5% taking paroxetine Controls n = 90	Increased abnormal movements, tonus abnormality and respiratory problems (p < 0.001). All signs appeared in day 1, median duration 3 days
Negative studies
Casper et al., 2003 [32]	Case–control	SSRIs n = 31 Controls (children of non-medicated depressed mothers) n = 13	Drug-exposed had lower Apgars than non-exposed; otherwise no statistically significant differences
Maschi et al., 2008 [58]	Prospective cohort	Paroxetine n = 58 Fluoxetine n = 32	Higher rate of prematurity. No difference in rates of poor neonatal adaptation, prolonged hospitalization or NICU admission

Large database or registry studies have variously suggested an increased risk of neonatal seizures [41–44,53,56], 1.5 times increased risk of NICU admission with third trimester exposure compared to first trimester exposure [38], and increased risks of respiratory distress and low APGAR scores [53].

The 2005 meta-analysis [52] of prospective controlled trials included consideration of 1066 mother–infant pairs and found a three-fold increased risk of SCN/NICU admission. None of the trials in the meta-analysis included a depressed, non-drug-treated group. Two subsequent studies do include such a group. Of these, Ferreira [68] found a significant (p < 0.001) increase in abnormal movements, tone and respiratory symptoms in 76 infants exposed to SSRIs or venlafaxine in the third trimester. Sivojelezova [28] found that 132 infants exposed to citalopram in pregnancy had a four-fold increased incidence of NICU admission compared to matched infants exposed to untreated maternal depression or controls.

Of the two controlled studies that did not find an association between antidepressant exposure and neonatal adaptation difficulties Casper [32] was possibly underpowered, whereas Maschi [58] was a larger trial including 200 women treated with antidepressants and 1200 controls.

Studies examining the neonatal effects of antenatal exposure to TCAs are limited to case reports and case series. Several case reports describe an association between gestational clomipramine [69–71] and imipramine [71–73] exposure with signs of neonatal adaptation difficulties. In a prospective case series [74] of 18 pregnant women on TCAs (predominantly imipramine), all nine with third trimester exposure had infants with adaptational difficulties.

With regard to the newer antidepressants, evidence is scant. Some of the controlled trials include venlafaxine exposure [53,68,75,76] and suggest similar neonatal adaptation difficulties to SSRIs. Mirtazapine features only in case reports [77,78] and in one controlled study [21,29], with a suggestion of both respiratory and thermoregulatory problems. One case report exists for duloxetine [79] documenting neonatal adaptation symptoms. No data were found for desvenlafaxine or reboxetine.

Summay

Taken together, the evidence strongly suggests that antenatal antidepressant exposure is associated with a higher risk of neonatal adaptation difficulties, though it is not possible at present to quantify that risk or to distinguish between individual medications on the basis of risk. Less credible evidence implicates TCAs, SNRIs and mirtazapine. While some of the syndromes have included such serious problems as seizures, or warranted NICU admission, the vast majority of cases appear to have been mild-to-moderate manifestations that resolved in a matter of days.

Persistent pulmonary hypertension of the newborn

Chambers et al., in a retrospective case–control design nested within the larger Slone Epidemiology Centre Birth Defects Study, noted an association between SSRI exposure beyond 20 weeks and persistent pulmonary hypertension of the newborn (PPHN) [80]. PPHN is a condition occurring in infants, leading to respiratory distress, and hypoxaemia and is a rare but potentially fatal condition with the possibility of adverse longer term developmental outcomes. There were 377 women whose infants had PPHN and 836 matched controls. The risk of PPHN associated with SSRI use at any time during pregnancy was not elevated. However, SSRI use after the 20^th week of gestation was found to be significantly associated with PPHN OR 5.1 (95%CI 1.9 to 13.3). There was no increased risk with TCAs and venlafaxine. Extrapolating from population prevalence data for PPHN of 1–2 infants per 1000 live births, Chambers suggests that the absolute risk with SSRI use in late pregnancy would be 6–12 per 1000.

Subsequently, Kallen and Olausson [81], using the Swedish Medical Birth Register, conducted a prospective study that supports Chamber's findings. There were 504 infants with PPHN of the 831 324 born, with a mortality rate of 8.9%. Particularly for gestational ages greater than 34 weeks, PPHN was associated with early exposure RR 2.40 (95%CI 1.2–4.3) and with early and late SSRI exposure RR 3.6 (95%CI 1.2–8.3). In contrast, Andrade [82] conducted a retrospective study of SSRIs in the third trimester with 1104 exposed infants and 1104 matched unexposed infants. There was no increased risk of PPHN in the exposed group.

In all three studies to date addressing PPHN, there have been significant methodological weaknesses. Chambers’ study was retrospective and used self-report data, with the potential for recall bias [80]. Furthermore, it is unclear as to why 20 weeks gestation was chosen as a demarcation point. Kallen's group used prospective data, however, most of the information was collected at interview in the first trimester, thus missing women who ceased or started medication later [81]. Furthermore, multiple comparisons were made to demonstrate a statistically significant finding and there were large confidence intervals. One negative study [82], had relatively small numbers and it was only adequately powered to find a 6-fold increase in risk and may in fact have missed a smaller, increase.

Summay

The evidence concerning the association between late pregnancy SSRI exposure and this rare but serious condition could be summarized as suggestive of a real effect that is nonetheless small in absolute terms. It cannot be construed as sufficient to contraindicate the use of these agents in pregnancy. Complicating the risk–benefit analysis is the list of other putative maternal risk factors for PPHN, these include higher pre-pregnancy BMI, diabetes, Caesarean section delivery, and the use of non-steroid anti-inflammatory drugs.

Longer term developmental outcomes

Seven studies examined longer-term development as a primary outcome (Table 4). Six of these studies reported no significant difference in developmental outcomes in children exposed to antidepressant medication compared with the variously used control groups (N = 375 out of 402) [23,60,62,83–85]. The measures of developmental outcomes have ranged from the use of formal scales (Bayley Scales of Infant Development, Carey Temperament Scales, Achenbach Behaviour Checklist) to standard hospital physical examination.

Table 4.

Developmental outcomes

Authors	Study design	Medications and n	Findings
Casper et al., 2003 [32]	Prospective case–control	SSRIs n = 31 Controls (children of non-medicated depressed mothers) n = 13	No difference between groups on the Bayley Scales of Infant Developmental. However, exposed children scored lower on the Psychomotor Development Index (p = 0.02). There was no difference between the groups on the individual items within the Psychomotor Development Index
Nulman et al., 1996 [23]	Prospective case–control	Fluoxetine n = 44	No differences between groups on scales: Bayley Scales of Infant Development, McCarthy Scales of Children's Abilities
Nulman et al., 1997 [60]	Prospective case–control	TCAs n = 80 Fluoxetine n = 55 Controls (exposed to non-teratogens) n = 84	No difference between groups on the scales: Bayley Scales of Infant Development, McCarthy Scales of Children's Abilities; Carey Temperament Scales; Achenbach Child Behaviour Checklist; Reynell Developmental Language Scales
Mattson et al., 1999 [83]	Prospective case–control	Fluoxetine n = 66 Controls (exposed to non-teratogens) n = 30	No differences between groups on the scale: Wechsler Preschool and Primary Scale of Intelligence – Revised
Nulman et al., 2002 [62]	Prospective case–control	TCAs n = 46 Fluoxetine n = 40 Controls (not exposed) n = 36	No difference between groups on the scales: Bayley Scales of Infant Development, McCarthy Scales of Children's Abilities; Carey Temperament Scales; Achenbach Child Behaviour Checklist; Reynell Developmental Language Scales
Misri et al., 2006 [84]	Prospective case–control	SSRIs n = 22 Controls (not exposed) n = 14	No difference between groups on the internalizing behaviours subscales of the Child Behaviour Checklist and the Child-Teacher Report Form
Oberlander et al., 2007 [85]	Prospective case–control	SSRIs n = 22 Controls (not exposed) n = 14	No difference between groups on the externalizing behaviours subscales of the Child Behaviour Checklist and the Child-Teacher Report Form

The only positive study was by Casper [32], which looked at 31 children of depressed mothers treated with SSRIs and a control group of 13 children of depressed mothers, who had elected not to take medication. At follow-up, the children ranged from 6 months to 40 months of age. There was no difference between the two groups on the Bayley Scales of Infant Development – Mental Developmental Index. There was a significant difference in the overall score for the Psychomotor Development Index with exposed children scoring lower. However, there was no difference in scores on individual items.

There are a numbers of studies comparing anti-depressants in pregnancy in terms of obstetric and neonatal outcomes [10,12,13,43]. Some of these studies also report on infant development. None of these studies reported a significant difference in developmental outcomes.

Summay

Of the small number of studies available, it is reassuring that the majority of these report no developmental differences between children exposed to antidepressant medication during pregnancy and those who are unexposed. Most single agent data is available for fluoxetine. There is little to no information on the newer antidepressants: venlafaxine, mirtazapine, escitalopram, reboxetine and duloxetine. Further, methodological issues include the wide range in age at which the children are assessed, and this heterogeneity lowers the possibility of positive findings. Whilst the majority of the studies were prospective and aimed to use a structured approach to developmental assessment, each of the scales has limitations, especially in terms of the clinical significance of statistical findings.

Discussion

The results outlined above reveal the complexity of studies in this area. Although a large number of pregnant women whose infants have been exposed to antidepressants have been investigated, a clear interpretation of the results is difficult. Studies vary in quality and many of the studies had insufficient numbers for adequate power to detect rare events. The larger register studies were limited by lack of proof of drug exposure and difficulty with follow-up. Study design is a further issue of concern. All designs had multiple confounders, which in many lacked controls. Few of the studies had controls for the underlying illness – a significant weakness. Retrospective designs are weakened by recall bias and often a high rate of non-responders. Teratology information studies (which were the prominent source of malformation studies) also had difficulties such as susceptibility to selection bias due to self recruiting. Finally, many studies conducted statistical tests on multiple parameters without the necessary correction.

With these limitations in mind, the overall impression is that these medications remain relatively safe in pregnancy but some notable areas of concern exist. In particular, a higher risk of preterm birth, neonatal adaptation difficulties, congenital cardiac malformations (with paroxetine) and possibly PPHN are outcomes for which there is more credible evidence.

Recommendations

As for all therapy, the clinician faced with the decision of how to treat depression and/or anxiety in a pregnant woman has to balance overall risks and benefits to both the mother and the foetus.

The decision in cases of mild illness is straightforward as evidence is available for the efficacy of non-pharmacological treatments such as psychotherapy. Similarly in women with severe illness, the risk of not treating far outweighs the risk of treating for both mother and infant.

The more complex decision making involves women with moderate illness where the risk–benefit balance is far more subtle. In these situations, thorough clinical assessment and detailed discussion with the patient and partner, providing clear information about options, is essential. Close liaison with other clinicians involved is prudent clinical practice. It is also vital that if the decision is made to withdraw medication, the patient is closely supervised so that early intervention is available should the woman relapse. In women presenting with their first episode of major depressive or anxiety disorder, who are medication naïve, we would recommend first line treatment with a SSRI (other than paroxetine).

In women with a clear history of recurrent illness and relapse off medication a case can be made for continuation of the most effective previous medication, with the exception of paroxetine where a trial of another SSRI could be considered. In these women, a history of premature labour or recurrent spontaneous abortions may raise the threshold for consideration of antidepressant use.

There is insufficient information available at the moment to give recommendations to women taking venlafaxine, mirtazapine, and the newer antidepressants such as duloxetine. In these cases, the clinician should discuss the limitations of knowledge and guide the patient through the process of evaluating their history, and the risks of altering beneficial treatment versus the risks of continuing or changing to other options.

Given the evidence for neonatal adaptation difficulties with antidepressant use, the availability of paediatric support is recommended.

Some guidelines such as the Therapeutic Guidelines [86] and NICE [87] recommend the use of TCAs over SSRIs. Our review does not support this, as TCAs are represented in positive studies outlined above, and there have been no specific studies comparing the efficacy of TCAs and SSRIs in pregnancy. Our findings and recommendations broadly concur with those of a recent report published jointly by the American Psychiatric Association and the American College of Obstetricians and Gynaecologists [88].

Withdrawal or tapering of antidepressants before delivery is recommended by some clinicians, with the rationale of preventing neonatal adaptation difficulties. However, one study [89] that specifically addressed this found no differences in neonatal outcomes after cessation of SSRIs fourteen days prior to delivery. In view of the absence of evidence for positive outcomes and the high risk of leaving mothers without antidepressant cover at the time of peak vulnerability, this practice is not recommended.

Future directions

Further research is essential to increase the information available, to clarify areas of concern and to widen the range of medications studied. It is also important that some of the methodological difficulties in study design are addressed so that clarity can be provided. Future studies should aim to ensure adequate sample sizes to detect real differences, have clear stated prior outcomes to minimize multiple comparisons and should control for confounders including the women's underlying illness so that drug effects and illness effects can be separated.

Footnotes

Acknowledgements

The authors are grateful to the other members of the Melbourne Perinatal Psychotropic Review Group: Anne Buist, Megan Galbally, Matthew Roberts, Fiona Judd, Lia Laios, Justin Dwyer; and would also like to thank Ross Coppell and Malcolm Hopwood.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References

Johnson

. Birth outcomes in pregnant women taking fluoxetine. J Fam Pract 1997; 44:32.

Cohen

Altshuler

Harlow

Nonacs

Newport

Viguera

. Relapse of major depression during pregnancy in women who maintain or discontinue antidepressant treatment. JAMA 2006; 295:499–507.

Shadigian

Bauer

. Pregnancy-associated death: a qualitative systematic review of homicide and suicide. Obstet Gynecol Surv 2005; 60:183–90.

Oates

. Suicide: the leading cause of maternal death. Br J Psychiatry 2003; 183:279–81.

Bonari

Pinto

Ahn

Einarson

Steiner

Koren

. Perinatal risks of untreated depression during pregnancy. Can J Psychiatry 2004; 49:726–35.

Alder

Fink

Bitzer

Hosli

Holgreve

. Depression and anxiety during pregnancy: A risk factor for obstetric, fetal and neonatal outcome? A review of the literature. J Maternal Fetal Neonatal Medicine 2007; 20:189–209.

Kurki

Hiilesmaa

Raitasalo

Mattila

Ylikorkala

. Depression and anxiety in early pregnancy and risk for preeclampsia. Obstet Gynecol 2000; 95:487–90.

Austin

Leader

. Maternal stress and obstetric and infant outcomes: epidemiological findings and neuroendocrine mechanisms. Aust N Z J Obstet Gynaecol 2000; 40:331–7.

Murray

Cooper

. Postpartum depression and child development. Psychol Med 1997; 27:253–60.

10.

Oberlander

Misri

Fitzgerald

Kostaras

Rurak

Riggs

. Pharmacologic factors associated with transient neonatal symptoms following prenatal psychotropic medication exposure. J Clin Psychiatry 2004; 65:230–7.

11.

Hendrick

Stowe

Altshuler

Hwang

Lee

Haynes

. Placental passage of antidepressant medications. Am J Psychiatry 2003; 160:993–6.

12.

Heikkinen

Ekblad

Kero

Ekblad

Laine

. Citalopram in pregnancy and lactation. Clin Pharmacol Ther 2002; 72:184–91.

13.

Heikkinen

Ekblad

Palo

Laine

. Pharmacokinetics of fluoxetine and norfluoxetine in pregnancy and lactation. Clin Pharmacol Ther 2003; 73:330–7.

14.

Loughhead

Fisher

Newport

Ritchie

Owens

DeVane

. Antidepressants in amniotic fluid: another route of fetal exposure. Am J Psychiatry 2006; 163:145–7.

15.

Loughhead

Stowe

Newport

Ritchie

DeVane

Owens

. Placental passage of tricyclic antidepressants. Biol Psychiatry 2006; 59:287–90.

16.

O'Brien

Einarson

Sarkar

Einarson

Koren

. Does paroxetine cause cardiac malformations? J Obstet Gynaecol Can 2008; 30:696–701.

17.

Rahimi

Nikfar

Abdollahi

. Pregnancy outcomes following exposure to serotonin reuptake inhibitors: a meta-analysis of clinical trials. Reprod Toxicol 2006; 22:571–5.

18.

Einarson

. Newer antidepressants in pregnancy and rates of major malformations: a meta-analysis of prospective comparative studies. Pharmacoepidemiol Drug Saf 2005; 14:823–7.

19.

Bar-Oz

Einarson

Boskovic

O'Brien

Malm

. Paroxetine and congenital malformations: meta-Analysis and consideration of potential confounding factors. Clin Ther 2007; 29:918–26.

20.

Chambers

Johnson

Dick

Felix

Jones

. Birth outcomes in pregnant women taking fluoxetine. N Engl J Med 1996; 335:1010–5.

21.

Diav-Citrin

OSS

Weinbaum

Wajnberg

Avgil

Gianantonio

Clementi

Weber-Schoendorfer

Schaefer

Ornoy

. Paroxetine and Fluoxetine in pregnancy: a prospective, multicentre, controlled observational study. British journal of Clinical Pharmacology 2008; 66:695–705.

22.

Pastuszak

Schick-Boschetto

Zuber

Feldkamp

Pinelli

Sihn

. Pregnancy outcome following first-trimester exposure to fluoxetine (Prozac). JAMA 1993; 269:2246–8.

23.

Nulman

Koren

. The safety of fluoxetine during pregnancy and lactation. Teratology 1996; 53:304–8.

24.

Kulin

Pastuszak

Sage

Schick-Boschetto

Spivey

Feldkamp

. Pregnancy outcome following maternal use of the new selective serotonin reuptake inhibitors: a prospective controlled multicenter study. JAMA 1998; 279:609–10.

25.

Einarson

Fatoye

Sarkar

Lavigne

Brochu

Chambers

. Pregnancy outcome following gestational exposure to venlafaxine: a multicenter prospective controlled study. Am J Psychiatry 2001; 158:1728–30.

26.

Einarson

Bonari

Voyer-Lavigne

Addis

Matsui

Johnson

. A multicentre prospective controlled study to determine the safety of trazodone and nefazodone use during pregnancy. Can J Psychiatry 2003; 48:106–10.

27.

Chun-Fai-Chan

Koren

Fayez

Kalra

Voyer-Lavigne

Boshier

. Pregnancy outcome of women exposed to bupropion during pregnancy: a prospective comparative study. Am J Obstet Gynecol 2005; 192:932–6.

28.

Sivojelezova

Shuhaiber

Sarkissian

Einarson

Koren

. Citalopram use in pregnancy: prospective comparative evaluation of pregnancy and fetal outcome. Am J Obstet Gynecol 2005; 193:2004–9.

29.

Djulus

Koren

Einarson

Wilton

Shakir

Diav-Citrin

. Exposure to mirtazapine during pregnancy: a prospective, comparative study of birth outcomes. J Clin Psychiatry 2006; 67:1280–4.

30.

Einarson

Pistelli

DeSantis

Malm

Paulus

Panchaud

. Evaluation of the risk of congenital cardiovascular defects associated with use of paroxetine during pregnancy. Am J Psychiatry 2008; 165:749–52.

31.

Goldstein

Corbin

Sundell

. Effects of first-trimester fluoxetine exposure on the newborn. Obstet Gynecol 1997; 89 (5 Pt 1):713–8.

32.

Casper

Fleisher

Lee-Ancajas

Gilles

Gaylor

DeBattista

. Follow-up of children of depressed mothers exposed or not exposed to antidepressant drugs during pregnancy. J Pediatr 2003; 142:402–8.

33.

Hendrick

Smith

Suri

Hwang

Haynes

Altshuler

. Birth outcomes after prenatal exposure to antidepressant medication. Am J Obstet Gynecol 2003; 188:812–5.

34.

Wisner

Sit

Hanusa

Moses-Kolko

Bogen

Hunker

. Major depression and antidepressant treatment: impact on pregnancy and neonatal outcomes. Am J Psychiatry 2009; 166:557–66.

35.

Kallen

Otterblad Olausson

. Maternal drug use in early pregnancy and infant cardiovascular defect. Reprod Toxicol 2003; 17:255–61.

36.

Kallen

Otterblad Olausson

. Maternal use of selective serotonin re-uptake inhibitors in early pregnancy and infant congenital malformations. Birth Defects Res A Clin Mol Teratol 2007; 79:301–8.

37.

Ericson

Kallen

Wiholm

. Delivery outcome after the use of antidepressants in early pregnancy. Eur J Clin Pharmacol 1999; 55:503–8.

38.

Malm

Klaukka

Neuvonen

. Risks associated with selective serotonin reuptake inhibitors in pregnancy. Obstet Gynecol 2005; 106:1289–96.

39.

Lennestal

Kallen

. Delivery outcome in relation to maternal use of some recently introduced antidepressants. J Clin Psychopharmacol 2007; 27:607–13.

40.

Wogelius

Norgaard

Gislum

Pedersen

Munk

Mortensen

. Maternal use of selective serotonin reuptake inhibitors and risk of congenital malformations. Epidemiology 2006; 17:701–4.

41.

Cole

Ephross

Cosmatos

Walker

. Paroxetine in the first trimester and the prevalence of congenital malformations. Pharmacoepidemiol Drug Saf 2007; 16:1075–85.

42.

Davis

Rubanowice

McPhillips

Raebel

Andrade

Smith

. Risks of congenital malformations and perinatal events among infants exposed to antidepressant medications during pregnancy. Pharmacoepidemiol Drug Saf 2007; 16:1086–94.

43.

Simon

Cunningham

Davis

. Outcomes of prenatal antidepressant exposure. Am J Psychiatry 2002; 159:2055–61.

44.

Wen

Yang

Garner

Fraser

Olatunbosun

Nimrod

. Selective serotonin reuptake inhibitors and adverse pregnancy outcomes. Am J Obstet Gynecol 2006; 194:961–6.

45.

Wichman

Moore

Lang

St Sauver

Heise

Jr. Watson

. Congenital heart disease associated with selective serotonin reuptake inhibitor use during pregnancy. Mayo Clin Proc 2009; 84:23–7.

46.

Rosa

. Medicaid antidepressant pregnancy exposure outcomes. Reproductive Toxicology [Abstract]. 1994; 8:444–5.

47.

Berard

Ramos

Rey

Blais

St-Andre

Oraichi

. First trimester exposure to paroxetine and risk of cardiac malformations in infants: the importance of dosage. Birth Defects Res B Dev Reprod Toxicol 2007; 80:18–27.

48.

Alwan

Reefhuis

Rasmussen

Olney

Friedman

. Use of selective serotonin-reuptake inhibitors in pregnancy and the risk of birth defects. N Engl J Med 2007; 356:2684–92.

49.

Louik

Lin

Werler

Hernandez-Diaz

Mitchell

. First-trimester use of selective serotonin-reuptake inhibitors and the risk of birth defects. N Engl J Med 2007; 356:2675–83.

50.

Ramos

St-Andre

Rey

Oraichi

Berard

. Duration of antidepressant use during pregnancy and risk of major congenital malformations. Br J Psychiatry 2008; 192:344–50.

51.

Hemels

Einarson

Koren

Lanctot

Einarson

. Antidepressant use during pregnancy and the rates of spontaneous abortions: a meta-analysis. Ann Pharmacother 2005; 39:803–9.

52.

Lattimore

Donn

Kaciroti

Kemper

Neal

Jr. Vazquez

. Selective serotonin reuptake inhibitor (SSRI) use during pregnancy and effects on the fetus and newborn: a meta-analysis. J Perinatol 2005; 25:595–604.

53.

Kallen

. Neonate characteristics after maternal use of antidepressants in late pregnancy. Arch Pediatr Adolesc Med 2004; 158:312–6.

54.

Costei

Kozer

Ito

Koren

. Perinatal outcome following third trimester exposure to paroxetine. Arch Pediatr Adolesc Med 2002; 156:1129–32.

55.

Zeskind

Stephens

. Maternal selective serotonin reuptake inhibitor use during pregnancy and newborn neurobehavior. Pediatrics 2004; 113:368–75.

56.

Oberlander

Warburton

Misri

Aghajanian

Hertzman

. Neonatal outcomes after prenatal exposure to selective serotonin reuptake inhibitor antidepressants and maternal depression using population-based linked health data. Arch Gen Psychiatry 2006; 63:898–906.

57.

Suri

Altshuler

Hellemann

Burt

Aquino

Mintz

. Effects of antenatal depression and antidepressant treatment on gestational age at birth and risk of preterm birth. Am J Psychiatry 2007; 164:1206–13.

58.

Maschi

Clavenna

Campi

Schiavetti

Bernat

Bonati

. Neonatal outcome following pregnancy exposure to antidepressants: a prospective controlled cohort study. BJOG 2008; 115:283–9.

59.

Pearson

Nonacs

Viguera

Heller

Petrillo

Brandes

. Birth outcomes following prenatal exposure to antidepressants. J Clin Psychiatry 2007; 68:1284–9.

60.

Nulman

Rovet

Stewart

Wolpin

Gardner

Theis

. Neurodevelopment of children exposed in utero to antidepressant drugs. N Engl J Med 1997; 336:258–62.

61.

Cohen

Heller

Bailey

Grush

Ablon

Bouffard

. Birth outcomes following prenatal exposure to fluoxetine. Biol Psychiatry 2000; 48:996–1000.

62.

Nulman

Rovet

Stewart

Wolpin

Pace-Asciak

Shuhaiber

. Child development following exposure to tricyclic antidepressants or fluoxetine throughout fetal life: a prospective, controlled study. Am J Psychiatry 2002; 159:1889–95.

63.

Laine

Kytola

Bertilsson

. Severe adverse effects in a newborn with two defective CYP2D6 alleles after exposure to paroxetine during late pregnancy. Ther Drug Monit 2004; 26:685–7.

64.

Suri

Altshuler

Hendrick

Rasgon

Lee

Mintz

. The impact of depression and fluoxetine treatment on obstetrical outcome. Arch Womens Ment Health 2004; 7:193–200.

65.

Oberlander

Warburton

Misri

Aghajanian

Hertzman

. Effects of timing and duration of gestational exposure to serotonin reuptake inhibitor antidepressants: population-based study. Br J Psychiatry 2008; 192:338–43.

66.

Laine

Heikkinen

Ekblad

Kero

. Effects of exposure to selective serotonin reuptake inhibitors during pregnancy on serotonergic symptoms in newborns and cord blood monoamine and prolactin concentrations. Arch Gen Psychiatry 2003; 60:720–6.

67.

Toh

Mitchell

Louik

Werler

Chambers

Hernandez-Diaz

. Antidepressant use during pregnancy and the risk of preterm delivery and fetal growth restriction. J Clin Psychopharmacol 2009; 29:555–60.

68.

Ferreira

Carceller

Agogue

Martin

St-Andre

Francoeur

. Effects of selective serotonin reuptake inhibitors and venlafaxine during pregnancy in term and preterm neonates. Pediatrics 2007; 119:52–9.

69.

Eggermont

Raveschot

Deneve

Casteels-van Daele

. The adverse influence of imipramine on the adaptation of the newborn infant to extrauterine life. Acta Paediatr Belg 1972; 26:197–204.

70.

Cowe

Lloyd

Dawling

. Neonatal convulsions caused by withdrawal from maternal clomipramine. Br Med J (Clin Res Ed) 1982; 284:1837–8.

71.

Ostergaard

Pedersen

. Neonatal effects of maternal clomipramine treatment. Pediatrics 1982; 69:233–4.

72.

Eggermont

. Withdrawal symptoms in neonates associated with maternal imipramine therapy. Lancet 1973; 2:680.

73.

Idanpaan-Heikkila

Saxen

. Possible teratogenicity of imipramine-chloropyramine. Lancet 1973 Aug 11; 2:282–4.

74.

Misri

Sivertz

. Tricyclic drugs in pregnancy and lactation: a preliminary report. Int J Psychiatry Med 1991; 21:157–71.

75.

Levinson-Castiel

Merlob

Linder

Sirota

Klinger

. Neonatal abstinence syndrome after in utero exposure to selective serotonin reuptake inhibitors in term infants. Arch Pediatr Adolesc Med 2006; 160:173–6.

76.

Boucher

Bairam

Beaulac-Baillargeon

. A new look at the neonate's clinical presentation after in utero exposure to antidepressants in late pregnancy. J Clin Psychopharmacol 2008; 28:334–9.

77.

Schwarzer

Heep

Gembruch

Rohde

. Treatment resistant hyperemesis gravidarum in a patient with type 1 diabetes mellitus: neonatal withdrawal symptoms after successful antiemetic therapy with mirtazapine. Arch Gynecol Obstet 2008; 277:67–9.

78.

Sokolover

Merlob

Klinger

. Neonatal recurrent prolonged hypothermia associated with maternal mirtazapine treatment during pregnancy. Can J Clin Pharmacol 2008; 15:e188–90.

79.

Eyal

Yaeger

. Poor neonatal adaptation after in utero exposure to duloxetine. Am J Psychiatry 2008; 165:651.

80.

Chambers

Hernandez-Diaz

Van Marter

Werler

Louik

Jones

. Selective serotonin-reuptake inhibitors and risk of persistent pulmonary hypertension of the newborn. N Engl J Med 2006; 354:579–87.

81.

Kallen

Olausson

. Maternal use of selective serotonin re-uptake inhibitors and persistent pulmonary hypertension of the newborn. Pharmacoepidemiol Drug Saf 2008; 17:801–6.

82.

Andrade

McPhillips

Loren

Raebel

Lane

Livingston

. Antidepressant medication use and risk of persistent pulmonary hypertension of the newborn. Pharmacoepidemiol Drug Saf 2009; 18:246–52.

83.

Mattson

Eastvold

Jones

Harris

Chambers

. Neurobehavioural follow-up of children prenatally exposed to fluoxetine. Teratology [Abstract]. 1999; 59:376.

84.

Misri

Reebye

Kendrick

Carter

Ryan

Grunau

. Internalizing behaviors in 4-year-old children exposed in utero to psychotropic medications. Am J Psychiatry 2006; 163:1026–32.

85.

Oberlander

Reebye

Misri

Papsdorf

Kim

Grunau

. Externalizing and attentional behaviors in children of depressed mothers treated with a selective serotonin reuptake inhibitor antidepressant during pregnancy. Arch Pediatr Adolesc Med 2007; 161:22–9.

86.

Norquay

. Therapeutic Guidelines: Psychotropics Version 6. West Melbourne, Victoria: Therapeutic Guidelines Limited; 2008.

87.

National Institute of Health and Clinical Excellence: Antenatal and Postnatal Mental Health: Clinical Management and Service Guidance Clinical Guideline CG45. 2007 [cited 18 March 2010] Available from URL: http://guidance.nice.org.uk/CG45

88.

Yonkers

Wisner

Stewart

Oberlander

Dell

Stotland

. The management of depression during pregnancy: a report from the American Psychiatric Association and the American College of Obstetricians and Gynecologists. Obstet Gynecol 2009; 114:703–13.

89.

Warburton

Hertzman

Oberlander

. A register study of the impact of stopping third trimester selective serotonin reuptake inhibitor exposure on neonatal health. Acta Psychiatr Scand 2009.

Antidepressants in Pregnancy: a Systematic Review

Abstract

Keywords

Methodology

Results

Congenital malformations

Meta-analyses

Cohort studies

Prospective cohort studies

Retrospective cohort studies

Case–control studies

Summary

Pregnancy outcomes

Spontaneous abortion

Preterm birth

Birth weight

Stillbirth

Pregnancy complications

Summary

Antidepressants and neonatal outcomes

Neonatal adaptation

Summay

Persistent pulmonary hypertension of the newborn

Summay

Longer term developmental outcomes

Summay

Discussion

Recommendations

Future directions

Footnotes

Acknowledgements

References