- Research article
- Open Access
- Open Peer Review
Bipolar disorder in pregnancy and childbirth: a systematic review of outcomes
BMC Pregnancy and Childbirthvolume 16, Article number: 331 (2016)
Bipolar Disorder (BD) is a mental disorder usually diagnosed between 18 and 30 years of age; this coincides with the period when many women experience pregnancy and childbirth. As specific problems have been reported in pregnancy and childbirth when the mother has BD, a systematic review was carried out to summarise the outcomes of pregnancy and childbirth, in mother and child, when the mother has BD diagnosed before pregnancy.
An a priori protocol was designed and a systematic search conducted in PubMed, CINAHL, Scopus, PsycINFO and Cochrane databases in March 2015. Studies of all designs were included if they involved women with a diagnosis of bipolar disorder prior to pregnancy, who were pregnant and/or followed up to one year postpartum. All stages of inclusion, quality assessment and data extraction were done by two people. All maternal or infant outcomes were examined, and narrative synthesis was used for most outcomes. Meta-analysis was used to achieve a combined prevalence for some outcomes and, where possible, case and control groups were combined and compared.
The search identified 2809 papers. After screening and quality assessement (using the EPHPP and AMSTAR tools), nine papers were included. Adverse pregnancy outcomes such as gestational hypertension and antepartum haemorrhage occur more frequently in women with BD. They also have increased rates of induction of labour and caesarean section, and have an increased risk of mood disorders in the postnatal period. Women with BD are more likely to have babies that are severely small for gestational age (<2nd-3rd percentile), and it appears that those women not being treated with mood stabilisers in pregnancy might not have an increased risk of having a baby with congenital abnormalities.
Due to heterogeneity of data, particularly the use of differing definitions of bipolar disorder, narrative synthesis was used for most outcomes, rather than a meta-analysis.
It is evident that adverse outcomes are more common in women with BD and their babies. Large cohort studies examining fetal abnormality outcomes for women with BD who are not on mood stabilisers in pregnancy are required, as are studies on maternal-infant interaction.
Bipolar Disorder (BD) is a severe affective mood disorder characterized by a wide range of lifelong mood changes varying between depressive, hypomanic, manic or mixed episodes [1, 2]. The lifetime prevalence for the most common variations within the BD spectrum, BD-I and BD-II, is reported as 1–2.4 % [3, 4]. A recent study from Sweden reports that incidence and prevalence of BD has increased during the last 20 years . BD entails a significant burden for those that are affected as well as high societal costs related to health care, sick leave and early retirement . The Swedish study has found that in persons with BD, despite similar education levels, employment levels are lower and disposable income is less compared to the general population. Also, persons with BD have more sick leave compared to the general population . BD is a severe condition, and an example of this is that the suicide risk in people with this disorder is about 20–30 times greater than that for the general population .
The time when many women experience pregnancy, childbirth and early postpartum overlaps the period between 18 and 30 years when most women are diagnosed with BD . Being pregnant and giving birth to a child is a key life transition event comprising engagement, growth and transformation . Women with mental illness, however, often experience problems and increased risk of complications related to childbearing . Female patients with psychiatric disorders seem to have more negative attributes with regard to sexuality and reproduction . The complex picture of problems that women with BD face include: sexually risky behaviour during episodes of mania, health related risks for the mother and/or the baby, decisions regarding medication during pregnancy, as well as decisions related to unplanned/undesired pregnancy, and sexually transmitted infections [12–16].
As a basis for developing high quality healthcare with best health outcomes for mother and child when the pregnant woman has BD, there is need for a scientific overview of the risks. The aim of this paper is, through a systematic literature review, to summarise outcomes of pregnancy and childbirth, in mother and child, when the mother has bipolar disorder diagnosed before pregnancy.
The research question guiding this systematic review was: What are the outcomes of pregnancy and childbirth (childbirth defined as labour and birth, and the first year postpartum) for women with bipolar disorder and their fetus/infant, when compared with outcomes for women with bipolar disorder, who are not pregnant, experiencing labour and birth, or postnatal. We were open to all scientific findings in the field.
The use of the accepted definition of bipolar disorder [1, 2] was planned. An a priori protocol was designed, outlining the aim and procedure for the review and is available in the Additional file 1. An inclusion/exclusion criteria list was developed from identifying relevant PICOS (Table 1), and then tested with the PRESS criteria , to guide inclusion of all relevant studies. A comprehensive and systematic search was done in the following databases: PubMed, CINAHL, Scopus, PsycINFO and Cochrane, from inception until 10th of March 2015, to identify studies or existing systematic reviews. The final search string for PubMed is available in the Additional file 1, Table 2. A repeat search in PubMed and Cochrane Library was conducted in March 2016 to check for any new papers that should be included in the discussion section, for comparison.
We limited our search to peer-reviewed papers. We planned to include randomised trials if the control groups could provide outcomes due to BD, but none were found.
A total of 2809 papers were found in the five databases (PubMed n = 877, CINAHL n = 240, PsychINFO n = 512, Scopus n = 1156 and Cochrane n = 24). After deletion of duplicates by two reviewers (MR, MB), 1700 papers remained. Exclusion of papers by title and abstract was made by two reviewers (MR, MB) based on assessment of the inclusion and exclusion criteria. In case of disagreement the third reviewer (CB) was involved. This process ended in exclusion of 1652 papers. The full text of the remaining papers (n = 48) was read by three teams of two researchers each (MR/MB, MR/CB, MB/CB). Five further studies were added from their reference lists. Of these 53, 35 papers were excluded (Fig. 1). Reasons for exclusion were: drug related issues (21 papers, [14, 18–37]), no focus on the chosen study group (six papers, [38–43]), outcomes in women not diagnosed before pregnancy (two papers, [44, 45]), mixed groups bipolar disorder and schizophrenia (two papers, [46, 47]), discussion paper (two papers, [48, 49]), gender mixed groups also including men (one paper, ), duplicates (one paper, ). Occasionally it was difficult to separate the different diagnoses that were included in the sample; for example, in Doyle et al.  it was not possible to find separate data for the three included groups: BD I, BD II and schizoaffective disorders bipolar type. In Mac Cabe et al.  it was difficult to determine whether women with additional diagnoses were included in the sample or not, so both these papers were excluded. Eighteen papers (15 studies and 3 reviews) were thus quality assessed by the three pairs.
A repeat search in PubMed and Cochrane Library was conducted in March 2016, using the same methodology, to check for any new papers that should be included in the discussion section, for comparison. Fifty-eight papers were found in PubMed in this second search and no papers in the Cochrane Library. Four new papers [51–54] were found to be topical for this review. Logsdon et al. 2015  described maternal-infant interaction at 12 months postpartum in women with BD compared to women with unipolar depression and a control group without a major mood disorder. Marengo et al. 2015  looked at how women with BD made reproductive decisions. Both of these papers are included in the discussion section only, for noting, as they were published too late to be included in the review. Taylor et al. 2015  was excluded, as their description of the characteristics of pregnant women and their use of psychotropic medication was not relevant for this review. The fourth paper, a review and meta-analysis by Wesseloo et al. 2016  was also not relevant for inclusion as it included women who had a psychotic or manic episode following childbirth, even if they had never had a diagnosis of BD in pregnancy.
To assess the studies for quality the validated Effective Public Health Practice Project (EPHPP) tool  was used for original studies (n = 15), based on six categories including study design, selection bias and data collection methods. Each item was given a rating of strong, moderate or weak, based on how well the criteria were met. An overall strong, moderate or weak rating was awarded for studies with no, one, or two or more “weak” categories. We excluded six studies rated as “weak” [56–61]. The quality of the reviews (n = 3) was assessed using the AMSTAR tool [62, 63]. Each of the eleven issues is rated with Yes, No, Can’t answer or Not Applicable. Every yes is given one mark; 8 to 11 marks is “high quality”, 4 to 7 is “medium”, and 0 to 3 is “low quality”. We excluded all three reviews due to “low quality” [64–66].
The 9 remaining papers were included, 4 of which had a “strong” rating, and 5 a “moderate” (Table 3).
Data extraction and analysis
Three teams of two reviewers extracted key findings from each study, as well as potential explanations given by the authors for results, and their recommendations, using a pre-designed data collection form. Any disagreements were resolved by consensus. Meta-analysis was not always possible because of heterogeneity in the majority of studies (due to differing definitions of BD, or non-similar cohorts), so a descriptive narrative synthesis was used for most outcomes. When appropriate (similar definitions of BD), simple meta-analysis was used to achieve a combined prevalence. In a few cases, we were able to combine and compare case and control groups, when similar methods had also been used in two or more studies. Odds ratios were computed using MedCalc online statistical software (https://www.medcalc.org/download.php) and heterogeneity between studies was calculated using RevMan, the Cochrane Collaboration software (Review Manager, 2014). The study by Jablensky et al.  had a mixed cohort that included only 55 % of women with pre-existing BD, so has been left out of all meta-analyses performed.
In the following pages we describe the identified outcomes of pregnancy and childbirth in mothers, and their children, when the mother has BD diagnosed before pregnancy.
Description of included studies
Characteristics of the nine included studies are presented in Table 3. Outcome data measured were risks or complications for the pregnancy, mother, fetus and/or baby. Seven of the nine papers had a retrospective design. Of these, three comprised data on outcomes in pregnancy and birth, one had postpartum data only (from 27 days to one year), and five comprised data on all these three episodes (Table 3). There were no restrictions for age, parity, ethnicity or other variables. We included papers written in the languages English, German, French and Swedish, and had no restrictions for age, parity, ethnicity or other variables.
The studies were published from year 2000 to 2015; one each performed in Turkey , Sweden , Australia , Taiwan , Canada  and Denmark ; two in United Kingdom [73, 74], and one study using data from an international data base of Lithium treated persons including Austria, Canada, the Czech Republic, Denmark, Germany and Sweden . Four of the publications were cohort studies, and five case control studies where outcomes of pregnancy were compared for women with BD and those without. The objectives of the papers varied.
Definitions of bipolar disorder used in the papers varied from BD (DSM-IV not specified), BD-I and BD-II (DSM-IV), BD (ICD-10 codes F30-31, F34.0 and F38), RDC , BD (ICD-9 codes 296.0 and 296.2-5), BD (ICD-8 codes 296.19, 296.39 and 298.19), BD (ICD-9, ICD-10CA not specified) and BD (ICD-9-CM codes 295.XX, 296.0X, 296.1X, 296.4X, 296.5X, 296.6X, 296.7X, 296.80 or 296.89). (Table 3) Authors’ definitions have been used in this review.
Mood episodes (Table 4)
Mood episodes during pregnancy and post-partum in the women with BD
Three studies reported results both on mood episodes during pregnancy or the post-partum period. Akdeniz et al.  showed that 32 % of the women with BD had at least one mood episode during pregnancy or within the first month after childbirth. Di Florio et al.  reported that 49.8 % of the women with BD-I and 42.2 % of women with BD-II had a mood episode in pregnancy or the post-partum period. Women who had a mood episode during their 1st pregnancy were, according to Akdeniz et al. , more likely to have another episode postpartum (OR 9.6 (95 % CI, 1.00–91.96). Also, women who had a mood episode in the 1st post-partum period were more likely to have a mood episode in the 2nd post-partum period (OR 3.6, 95 % CI, 0.25–50.33). Di Florio et al.  reported that women with BD-II had a higher incidence of any perinatal mood episode compared with women with BD-I (BD-II 18.4 % and BD-I 8.6 %, (chi-square = 10.38, d.f. = 1, p < 0.002 (calculated by review authors)) and that the mood episodes were significantly more common during the first month post-partum than during pregnancy (BD-I OR 44.5 95 % CI 26.90–76.00 and BD II OR 4.7 95 % CI 2.40–9.80). Grof et al.  show that episodes occurred significantly more often in the post-partum period than before and during pregnancy.
Mood episodes during pregnancy
Occurrence of mood episode during pregnancy was reported in four studies. In a study by Grof et al.  18 % (five out of 28 women) had a relapse during pregnancy, all in the last five weeks of pregnancy. Combining the results of Akdeniz et al.  and Di Florio et al.  (as both studies used the DSM-IV definition) gives an overall prevalence for a mood episode during pregnancy of 9.3 % (104 women out of 1116).
Di Florio et al.  found that 35 % of the women with BD-I reported an episode (mania/psychotic depression) in the first pregnancy, 20.5 % in second pregnancy and 14.6 % in subsequent pregnancies. For women with BD-II the rates of depression in first pregnancy were 46 % and second pregnancy 33 %. Rates of depression were similar across all pregnancies and postpartum periods.
Mood episodes post-partum
Occurrence of mood episodes in the post-partum period was reported in four studies. In the study by Grof et al. , 25 % (seven out of 28) of the women had a postpartum relapse from 1 to 5 times, lasting 3.5 months on average; 42 % of these were manic. Combining the results of Akdeniz et al.  and Di Florio et al.  for mood episodes post-partum shows a prevalence of 79.2 % (812 out of 1024 births).
There was no significant association between parity and mood episodes in pregnancy or later post-partum according to Di Florio . However, a significant association between parity and mood episodes within 6 weeks postpartum was found, with primiparous women being more likely to experience an episode.
Admission during pregnancy
Mei-Dan et al.  reported that 3.6 % (n = 66) of the women with BD were hospitalised for psychiatric reasons during pregnancy. This is more than for women with major depressive disorder, where 1.9 % (n = 69) were admitted during pregnancy. Risk calculations for admission during pregnancy were not reported.
Munk-Olsen et al.  showed in a register-based cohort study (study period 1973–2005) that 26.9 % of all Danish women with BD with previous psychiatric admission(s) before birth of their first child were readmitted during the first year post-partum. That is almost twice as high compared to women with schizophrenia-like disorders (15.7 %). The highest risk of readmission was 10 to 19 days postpartum (RR, 37.22; 95 % CI, 13.58–102.04) compared with mothers with the same diagnoses who gave birth 6 to 11 months earlier. The cumulative incidence of admission 0 to 3 months postpartum was 22 %.
Pregnancy and childbirth related issues and complications (Table 5)
Non spontaneous (artificial) start of childbirth (induction or planned caesarean section)
One retrospective cohort study  found that women without BD had a rate of induction of labour of 20.7 %; in women with BD without treatment it was 30.9 %, Adjusted Odds Ratio (AOR) 1.57 (95 % CI 1.30 to 1.90); in women with BD with treatment it was 37.5 %, AOR 2.12 (95 % CI 1.68 to 2.67). Women without BD had a rate of caesarean birth of 16.8 %; in women with BD without treatment it was 23.5 %, AOR 1.45 (95 % CI 1.18 to 1.78); in women with BD with treatment the rate of caesarean birth was 25.6 %, AOR 1.56 (95 % CI 1.20 to 2.03) .
One Swedish register study examined the relationship between BD and gestational diabetes. The analysis did not find any increased risk for gestational diabetes in either pharmaceutical-treated (OR 1.18 (95 % CI 0.55–2.56)) or untreated women with BD, compared with those without (OR 1.06 (95 % CI 0.56–2.02)) .
One population-based national study in Taiwan found that women with BD were more likely to have gestational hypertension than pregnant women with no history of mental health difficulties. The odds ratio (AOR) was 2.81 when adjusted for maternal age, education level, marital status, infant gender, parity, family months income, parental age difference and parental education level (95 % CI 2.53–3.10) .
One cohort study showed that women with BD in Australia were more likely to have antepartum haemorrhage than pregnant women with no history of mental health difficulties (unadjusted OR 1.66 (95 % CI 1.15–2.39)). When adjusting for maternal age, parity, plurality, marital status, originality, and sex, the increased risk of obstetric complications remained (AOR 1.60, 95 % CI = 1.11–2.32). However, those who developed BD after the index birth were at no more risk during pregnancy and birth than the women without mental health difficulties (OR 1.02, 95 % CI = 0.92–1.12) .
One cohort study showed that women with BD were more likely to have placenta praevia than pregnant women with no history of mental health difficulties (OR 2.04 95 % CI 1.11–3.73). After adjustment for maternal age, parity, plurality, marital status aboriginality and sex the risk remained (AOR 2.13; 95 % CI 1.15–3.94) .
Preterm birth (before 37 weeks gestation)
Three studies examined this issue and found a combined prevalence of 10.68 % for preterm birth (PTB) in women with BD (328 out of 3070), compared with a rate of 6.12 % in women with no mental health problems (79,030 out of 1,291,682). This was a statistically significant difference (Z = 10.37, p < 0.0001, OR 1.83, 95 % CI 1.64 to 2.06).
Bodén et al.  found that the risk of PTB was increased for women with BD compared with women without BD (4.80 %), both for those treated (8.10 %, AOR 1.50, 95 % CI 1.01 to 2.24) and untreated (7.60 %, AOR 1.48, 95 % CI 1.08 to 2.03), when adjusted for birth order, maternal age, cohabitation, smoking, height, alcoholism and substance misuse. Lee and Lin  similarly showed that women with BD were more likely to have PTB (14.20 % vs 6.90 %) than women with no history of mental health difficulties (AOR 2.08, 95 % CI 1.53–2.83), when adjusted for maternal age, education level, marital status, infant gender, parity, family months income, parental age difference and parental education level. Mei-Dan et al.  also found a higher prevalence of PTB for women with BD (11.40 %), (AOR 1.95, 95 % CI 1.68–2.26) compared with the control group (6.20 %), when adjusted for maternal age, parity, obesity, substance abuse and medical and obstetric complications.
Early preterm occurrence was studied in the Canadian study. Preterm birth defined as < 32 gestational weeks was increased: 9.1 % vs 1.1 %. AOR: 1.70 (95 % CI 1.16–2.48). Preterm birth defined as < 28 gestational weeks show no significant increased risk: BD 0.9 % vs 0.6 % in the referent group (AOR 1.66; 95 % CI 0.92–3.02) .
Mode of birth
Mode of birth was reported in one study, the Swedish cohort. Compared to women without BD emergency caesarean section was about 50 % increased in women with BD. Adjustment was done for birth order, maternal age, cohabitation, smoking, height, alcoholism and substance misuse. For untreated women the AOR was 1.45 (95 % CI 1.18 to 1.78), and in treated women it was 1.56 (95 % CI 1.20–2.03). Similar results were found for vaginal instrumental birth (vacuum extraction or forceps): AOR in untreated BD was 1.49 (95 % CI 1.18–1.78), and treated BD women AOR 1.39 (95 % CI 1.20–2.03) .
Fetal and neonatal outcomes (Table 6)
Congenital abnormalities (CA) were examined in three studies. Jablensky et al.  found no difference in congenital abnormalities in women with BD (n = 62 out of 1,301, 4.80 %) compared with those with no mental health difficulties (n = 152 out of 3,129, 4.90 %), but is not included in the meta-analysis.
The combined results of the other two studies [69, 71] showed that 21,632 women without BD had a baby with congenital abnormality, out of 766,750 (2.82 %), while 175 women with BD, out of 4034, had one (4.34 %). This difference is statistically significant (chi-square = 33.59, p < 0.0001, OR 1.56, 95 % CI 1.34 to 1.82). There was no heterogeneity shown between the two studies: Chi2 = 0.02, df = 1 (P = 0.90); I2 = 0 %.
Mei-Dan et al.  found that BD presented increased risk for congenital anomalies (n = 90 out of 1859, 5.00 %) compared with the referent group (n = 14,963 out of 432,358, 3.50 %), when adjusted for maternal age and parity (AOR 1.48, 95 % CI 1.20–1.82). Bodén et al.  also found the prevalence of congenital malformations was 2 % for infants born to women without BD (i.e., the normal population). For women with BD who were not treated with mood stabilisers the rate was 1.90 %, and those women with BD who were treated with mood stabilisers had rates ranging from 0 to 3.50 %, depending on the drug used. The authors did not compare these results statistically; however, if the results for women without BD and the women with no treatment for BD are combined (6528 abnormalities out of 331,817, 2.00 %) and compared with the 12 abnormalities out of 320 women with BD who were treated (3.75 %), the difference is statistically significant (Z = 2.19, p < 0.03, OR 1.91, 95 % CI 1.07–3.39), indicating that medications for the treatment of BD in pregnancy are associated with fetal abnormalities.
Small for gestational age (SGA)
Findings were mixed in the four studies examining this outcome, and definitions differed. Boden et al.  used ≤2nd-3rd centile (severe SGA) and Mei-Dan et al.  used <3rd centile (severe SGA), while Jablensky et al. , Lee and Lin  and Mei-Dan et al’s  secondary outcomes used <10th centile. No significant results for severe SGA were reported for women with BD in Mei-Dan et al.  (4.6 % versus 3.9 %), or in Bodén et al’s study  (3.33 % versus 2.30 %), neither for the group treated with mood stabilisers nor for those not treated. However, combining results from the two studies gives prevalence rates of 4.14 % for women with BD (113 out of 2729), compared with 3.20 % for those without (24,418 out of 762,619), a statistically significant difference (Z = 2.79, p < 0.006, OR 1.31, 95 % CI 1.08 to 1.58).
Results from the three studies that used the <10th centile definition differed. Jablensky et al.  found no difference in women with BD compared with those with no mental health difficulties, but was not included in the meta-analysis.
Women with BD in Lee and Lin’s research  were more likely to have SGA (22.3 % vs.15.7 %) than those with no history of mental health difficulties (OR 1.47, 95 % CI 1.14–1.91). Mei-Dan et al. , in testing their secondary outcome of SGA using the <10th centile definition, found that women with BD had an increased risk of SGA compared to the reference group, 13.9 % compared to 12.7 %, AOR 1.17, 95 % CI 1.03–1.34, when adjusted for maternal age and parity. When results from these last two studies are combined (15.2 % versus 14.3 %), there is no significant difference using unadjusted odds ratio (Z = 1.19, p = 0.24, OR 1.07, 95 % CI 0.96 to 1.21). It would thus appear that women with BD are more likely to have babies with severe SGA (<2nd-3rd centile) but not SGA (<10th centile).
Large for gestational age fetus (LGA)
In Mei-Dan et al.’s  study, severe LGA, defined as >97th centile, was significantly more common among women with BD (3.8 %), AOR 1.29, 95 % CI 1.08–1.54, compared to women with no mental difficulties (2.7 %), when adjusted for maternal age and parity. However, when LGA was defined as >90th centile, no difference was seen between the rates for women with BD (9.10 %) compared to women without mental illness (8.2 %), AOR 1.13 95 % CI 0.96–1.32.
Two studies measured low birthweight (LBW), defined as the birth weight <10th centile  or <2,500 g . In Jablensky et al’s  study women with BD were not more likely to have LBW infants (9.9 % vs.9.3 %, OR 1.06 95 % CI 0.82–1.36) than pregnant women with no history of mental health difficulties. Women with BD in Lee and Lin’s study  were more likely to have LBW infants (9.8 % vs.5.7 %) than pregnant women with no history of mental health difficulties (OR 1.66, 95 % CI 1.16–2.38).
Fetal distress/low Apgar score
Jablensky et al.  found no difference in fetal distress or low 5-min Apgar scores in women with BD compared to those with no mental health difficulties. Boden et al.  also showed no difference in low Apgar scores between women without BD and those untreated (AOR 1.56, 95 % CI = 0.85–2.86) and treated (AOR 0.88, 95 % CI = 0.33–2.34) for BD in pregnancy.
Mei-Dan et al.  studied neonatal admission up to 28 days of life. Babies of women with BD did have an increased risk (2.0 %) compared to those of women with no mental health difficulties (0.9 %) AOR 2.41, 95 % CI 1.76–3.31, when adjusted for maternal age and parity.
Infant mortality defined as death before the 28th day of life was studied by Mei-Dan et al. . The risk for women with BD was the same as that for women with no mental health problems (0.2 %) AOR 0.72, 95 % CI 0.23–2.23.
Neonatal morbidity, studied by Mei-Dan et al.  only, was defined as RDS (respiratory distress syndrome), seizure, sepsis, IVH (intravenous hyperalimentation), persistent fetal circulation, and neonatal abstinence syndrome. The risk of having any of these neonatal morbidities was higher in women with BD (5.4 %) compared with those without mental illness (1.9 %) AOR 2.99, 95 % CI 2.44–3.66. When analysed separately, babies of women with BD had higher rates for RDS, seizure, sepsis, IVH and neonatal abstinence syndrome. No difference was seen between the two groups of women in persistent fetal circulation (AOR 1.89, 95 % CI 0.78–4.58).
This systematic review has presented the wide-ranging health outcomes for women with bipolar disorder and their fetuses/babies. Mental health risks to women include the occurrence of mood episodes during pregnancy, rates of which varied across the studies from 9 to 18 %. Mood episodes were more common in the postpartum period [73, 75], ranging from 25 to 79 %, with differences possibly due to small sample sizes in some studies [36, 44]. It has been suggested that this may be due to placental hormones increasing throughout pregnancy, and their abrupt cessation after birth . In addition, mood episodes were found to be more common in first pregnancies . Similar results were found in a recent meta-analysis ; this review differs in methodology from ours as it included women who had a psychotic or manic episode following childbirth, even if they had never had a diagnosis of BD in pregnancy. It also included papers that compared different medication regimes, or stopping treatment at different times, which we chose to exclude as there are so many different treatments. Perhaps because of this, our postpartum relapse rate (79 %) was much higher than that quoted by Wesseloo et al. .
Single studies showed some incidence of increased risks in pregnancy; induction of labour occurred in women with BD, both treated and untreated, with rates of 31 and 38 % respectively, compared with 21 % in women without BD. Caesarean section rates were similarly increased. Women with BD were also more likely to have gestational hypertension , antepartum haemorrhage and placenta praevia . Emergency caesarean section was about 50 % increased in women with BD, and similar results were found for vaginal instrumental birth . There was no increased risk of gestational diabetes found [69, 70]. Pre-term birth showed a difference (10.68 % versus 6.12 %, two studies, total population 1,294,752). Lee and Lin  believed that smoking could be a large part of the causation, as in their (and other) studies women with BD had a higher prevalence of smoking behaviour.
Babies of women with BD have a higher prevalence of congenital abnormalities (4.34 % versus 2.82 %, two studies, total population 770,784), although the three papers examined differed in their individual findings. The difference is likely due to the smaller sample size in Jablensky’s work , at just over 6,000, and to the fact that this cohort included only 55 % of women with pre-existing BD. Other factors include the separation of women treated and un-treated for BD in Boden’s study , and the fact that Mei-Dan’s study population  were women previously hospitalised for bipolar disorder, so it is likely that they were on medication for their BD symptoms. There are many studies showing that mood stabilisers do cause congenital abnormalities, and that was not the focus of this review. It would appear from the three papers summarised here that women with BD who are not being treated with mood stabilisers in pregnancy might not be at the same level of increased risk of congenital abnormalities. Lee and Lin  recommended active monitoring of the fetus in pregnancy, and early intervention if abnormalities are noted.
A new result demonstrated by the meta-analysis in this review was that severe SGA (<2nd-3rd centile) was increased in women with BD (4.14 % versus 3.20 %, two studies, total population 765,348), in addition to previously reported increases in SGA based on the <10th centile definition [70, 71]. Again, smoking behaviour may be the main cause, as suggested by both Lee and Lin  and Jablensky et al. .
Large for gestational age babies, low birthweight, fetal distress, low Apgar score, still-birth, neonatal readmissions, infant mortality and neonatal morbidity were all examined by individual studies only, or in two studies that could not be merged for meta-analysis. Large for gestational age babies defined as >97th centile, but not as >90th centile, were more common in women with BD , as were low birthweight babies , neonatal readmissions  and any neonatal morbidity . Fetal distress , low Apgar score [67, 69], still-birth [67, 71], and infant mortality  showed no difference.
One paper published after the systematic review was finished  showed that women with BD had lower scores on maternal-infant interaction at 12 months postpartum; this is an area that would benefit from further study. Marengo et al. , also published after the systematic review was finished, highlighted results related to reproduction decisions of women with BD. They showed that women with BD had more unplanned pregnancies (37.7 % versus 9.6 %) and fewer planned pregnancies (32.8 % versus 78.1 %) than women in the control group. They also found that women with BD more often than those without had electively interrupted at least one pregnancy (42.4 % versus 13.5 %). These findings indicate that the complex picture of problems for women with BD, mentioned in the introduction, need to be further explored.
Review strengths and limitations
This is the first review, to our knowledge, that systematically searched the literature and summarised the risks associated with bipolar disorder for pregnant women and their babies. A key strength is the sourcing of literature from all languages. The result is a concise, extensive review that includes outcomes for both mother and baby, and spans both physical and mental health. Given the heterogeneity of data, particularly the use of differing definitions of bipolar disorder, and of some outcomes, we were unable to perform a meta-analysis for all outcomes.
A clear negative impact of bipolar disorder on mothers’ and their babies’ health has been shown. Adverse outcomes such as gestational hypertension and antepartum haemorrhage occur in pregnant women with bipolar disorder. They are also prone to increased rates of induction of labour and caesarean section, and have an increased risk of mood disorders in the postnatal period. The risk of fetuses developing severe growth retardation (<2nd-3rd centile) is increased in women with BD, and increased neonatal morbidity is also found. In addition, bipolar disorder is linked to a higher incidence of congenital abnormalities (if woman are treated with mood stabilisers) and neonatal readmissions. These negative outcomes lead to longer duration of hospital stay and increase the costs of care. Active monitoring of the fetus in pregnancy, early intervention if abnormalities are noted, encouragement to decrease smoking behaviour and close support and monitoring of the mother’s mood postpartum are all likely to bring benefits in terms of improved future health and decreased monetary outlay.
As only nine papers were suitable for inclusion in this review, and as some of the studies did not control for confounding variables, or included differing populations, further research is required. In particular, large cohort studies examining fetal abnormality outcomes for women with BD who are not on mood stabilisers in pregnancy are required, as are studies on maternal-infant interaction. More qualitative studies are also recommended regarding these women's perspectives of pregnancy, childbirth and early motherhood. The compilation of these outcomes should be used to guide more informed and supportive care for women with bipolar disorder who embark on pregnancy and motherhood.
APA. Diagnostic and Statistical Manual of Mental Disorders, 5th Edition. In. https://www.psychiatry.org/psychiatrists/practice/dsm. Accessed 12 Jul 2016: American Psychiatric Association; 2013
WHO. International Statistical Classification of Diseases, 10th Revision. In. http://www.who.int/classifications/icd/en/. Accessed 12 Jul 2016: World Health Organization; 2011
Merikangas KR, Akiskal HS, Angst J, et al. LIfetime and 12-month prevalence of bipolar spectrum disorder in the national comorbidity survey replication. Arch Gen Psychiatry. 2007;64(5):543–52.
Merikangas KR, Jin R, He J, et al. PRevalence and correlates of bipolar spectrum disorder in the world mental health survey initiative. Arch Gen Psychiatry. 2011;68(3):241–51.
Carlborg A, Ferntoft L, Thuresson M, Bodegard J. Population study of disease burden, management, and treatment of bipolar disorder in Sweden: a retrospective observational registry study. Bipolar Disord. 2015;17(1):76–85.
Ekman M, Granström O, Omérov S, Jacob J, Landén M. The societal cost of bipolar disorder in Sweden. Soc Psychiatry Psychiatr Epidemiol. 2013;48(10):1601–10.
Pompili M, Gonda X, Serafini G, Innamorati M, Sher L, Amore M, Rihmer Z, Girardi P. Epidemiology of suicide in bipolar disorders: a systematic review of the literature. Bipolar Disord. 2013;15(5):457–90.
Kessler RC, Berglund P, Demler O, Jin R, Merikangas KR, Walters EE. LIfetime prevalence and age-of-onset distributions of dsm-iv disorders in the national comorbidity survey replication. Arch Gen Psychiatry. 2005;62(6):593–602.
Nelson AM. Transition to motherhood. J Obstet Gynecol Neonatal Nurs. 2003;32(4):465.
Patel V, Kirkwood BR, Pednekar S, et al. Gender disadvantage and reproductive health risk factors for common mental disorders in women: A community survey in india. Arch Gen Psychiatry. 2006;63(4):404–13.
Özcan NK, Boyacıoğlu NE, Enginkaya S, Dinç H, Bilgin H. Reproductive health in women with serious mental illnesses. J Clin Nurs. 2014;23(9–10):1283–91.
McCandless F, Sladen C. Sexual health and women with bipolar disorder. J Adv Nurs. 2003;44(1):42–8.
Miller LJ, Ghadiali NY, Larusso EM, Wahlen KJ, Avni-Barron O, Mittal L, Greene JA. Bipolar Disorder in Women. Health care for women international. 2015;36(4):475–98.
Viguera AC, Whitfield T, Baldessarini RJ, Newport DJ, Stowe Z, Reminick A, Zurick A, Cohen LS. Risk of recurrence in women with bipolar disorder during pregnancy: prospective study of mood stabilizer discontinuation. Am J Psychiatr. 2007;164(12):1817.
Grof P, Robbins W, Alda M, Berghoefer A, Vojtechovsky M, Nilsson A, Robertson C. Protective effect of pregnancy in women with lithium-responsive bipolar disorder. J Affect Disord. 2000;61(1–2):31–9.
Hariri AG, Karadag F, Gokalp P, Essizoglu A. Risky Sexual Behavior among Patients in Turkey with Bipolar Disorder, Schizophrenia, and Heroin Addiction. J Sex Med. 2011;8(8):2284–91.
Sampson M, McGowan J, Cogo E, Grimshaw J, Moher D, Lefebvre C. An evidence-based practice guideline for the peer review of electronic search strategies. J Clin Epidemiol. 2009;62(9):944–52.
Aoki FY, Ruedy J. Severe lithium intoxication: management without dialysis and report of a possible teratogenic effect of lithium. Can Med Assoc J. 1971;105(8):847–8.
Austin M-PV, Mitchell PB. Use of psychotropic medications in breast-feeding women: Acute and prophylactic treatment. Aust N Z J Psychiatry. 1998;32(6):778–84.
Barbui C, Conti V, Purgato M, Cipriani A, Fortino I, Rivolta AL, Lora A. Use of antipsychotic drugs and mood stabilizers in women of childbearing age with schizophrenia and bipolar disorder: Epidemiological survey. Epidemiol psychiatr sci. 2013;22(4):355–61.
Bergink V, Bouvy PF, Vervoort JSP, Koorengevel KM, Steegers EAP, Kushner SA. Prevention of postpartum psychosis and mania in women at high risk. Am J Psychiatry. 2012;169(6):609–15.
Clark CT, Klein AM, Perel JM, Helsel J, Wisner KL. Lamotrigine dosing for pregnant patients with bipolar disorder. Am J Psychiatr. 2013;170(11):1240–7.
Diav-Citrin O, Einarson A. Lithium exposure during pregnancy linked with cardiovascular defects. Brown University Psychopharmacol Update. 2014;25(8):1–6.
Gentile S. Prophylactic treatment of bipolar disorder in pregnancy and breastfeeding: Focus on emerging mood stabilizers. Bipolar Disord. 2006;8(3):207–20.
Goldstein DJ, Corbin LA, Fung MC. Olanzapine-exposed pregnancies and lactation: Early experience. J Clin Psychopharmacol. 2000;20(4):399–403.
Iqbal MM, Sohhan T, Mahmud SZ. The effects of lithium, valproic acid, and carbamazepine during pregnancy and lactation. J Toxicol Clin Toxicol. 2001;39(4):381–92.
Bergink V, Burgerhout KM, Koorengevel K, Kamperman AM, Hoogendijk WJ, den Berg MP L-v, Kushner SA. Treatment of psychosis and mania in the postpartum period. Am J Psychiatry. 2015;172(2):115–23.
Diav-Citrin O, Shechtman S, Tahover E, Finkel-Pekarsky V, Arnon J, Kennedy D, Erebara A, Einarson A, Ornoy A. Pregnancy outcome following in utero exposure to lithium: a prospective, comparative, observational study. Am J Psychiatr. 2014;171(7):785–94.
Even C, Dorocant ES, Thuile J, Kalck-Stern M, Guelfi JD. Pregnancy, breast feeding and mood stabilisers: Review and recommendations for practice. L’Encéphale. 2006;32(2):224–30.
Galbally M, Roberts M, Buist A. Mood stabilizers in pregnancy: a systematic review. Aust NZ J Psychiatry. 2010;44(11):967–77.
Newport DJ, Viguera AC, Beach AJ, Ritchie JC, Cohen LS, Stowe ZN. Lithium placental passage and obstetrical outcome: implications for clinical management during late pregnancy. Am J Psychiatr. 2005;162(11):2162–70.
Troyer WA, Pereira GR, Lannon RA, Belik J, Yoder MC. Association of maternal lithium exposure and premature delivery. J Perinatol. 1993;13(2):123–7.
Yacobi S, Ornoy A. Is lithium a real teratogen? What can we conclude from the prospective versus retrospective studies? A review. Isr J Psychiatry Relat Sci. 2008;45(2):95–106.
Zalzstein E, Koren G, Einarson T, Freedom RM. A case–control study on the association between first trimester exposure to lithium and Ebstein’s anomaly. Am J Cardiol. 1990;65(11):817–8.
Kallen B, Tandberg A. Postpartum mania. Acta Psychiatr Scand. 1983;68(2):134–9.
Cohen LS, Sichel DA, Robertson LM, Heckscher E, Rosenbaum JF. Postpartum prophylaxis for women with bipolar disorder. Am J Psychiatry. 1995;152(11):1641–5.
Viguera AC, Nonacs R, Cohen LS, Tondo L, Murray A, Baldessarini RJ. Risk of recurrence of bipolar disorder in pregnant and nonpregnant women after discontinuing lithium maintenance. Am J Psychiatr. 2000;157(2):179–84.
Sharma V, Xie B, Campbell MK, Penava D, Hampson E, Mazmanian D, Pope CJ. A prospective study of diagnostic conversion of major depressive disorder to bipolar disorder in pregnancy and postpartum. Bipolar Disord. 2014;16(1):16–21.
van der Lugt NM, van de Maat JS, van Kamp IL, der Klein EAM K-v, Hovens JGFM, Walther FJ. Fetal, neonatal and developmental outcomes of lithium-exposed pregnancies. Early Hum Dev. 2012;88(6):375–8.
Heron J, Haque S, Oyebode F, Craddock N, Jones I. A longitudinal study of hypomania and depression symptoms in pregnancy and the postpartum period. Bipolar Disord. 2009;11(4):410–7.
Kinney DK, Yurgelun-Todd DA, Levy DL, Medoff D, Lajonchere CM, Radford-Paregol M. Obstetrical complications in patients with bipolar disorder and their siblings. Psychiatry Res. 1993;48(1):47–56.
Webb RT, Pickles AR, King-Hele SA, Appleby L, Mortensen PB, Abel KM. Parental mental illness and fatal birth defects in a national birth cohort. Psychol Med. 2008;38(10):1495–503.
Howard LM, Goss C, Leese M, Appleby L, Thornicroft G. The psychosocial outcome of pregnancy in women with psychotic disorders. Schizophr Res. 2004;71(1):49–60.
Browne R, Byrne M, Mulryan N, Scully A, Morris M, Kinsella A, McNeil TF, Walsh D, O’Callaghan E. Labour and delivery complications at birth and later mania: An Irish case register study. Br J Psychiatry. 2000;176:369–72.
Kadrmas A, Winokur G, Crowe RR. Postpartum mania. Br J Psychiatry. 1979;135:551–4.
Doyle K, Heron J, Berrisford G, Whitmore J, Jones L, Wainscott G, Oyebode F. The management of bipolar disorder in the perinatal period and risk factors for postpartum relapse. Eur Psychiatry. 2012;27(8):563–9.
MacCabe JH, Martinsson L, Lichtenstein P, Nilsson E, Cnattingius S, Murray RM, Hultman CM. Adverse pregnancy outcomes in mothers with affective psychosis. Bipolar Disord. 2007;9(3):305–9.
Gowan J, Roller L. Bipolar disorder: Pregnancy and breastfeeding. Aust J of Pharm. 2004;85(1008):203–7.
Weinstein MR, Goldfield MD. Cardiovascular malformations with lithium use during pregnancy. Am J Psychiatr. 1975;132(5):529–31.
El-Badri SM, Ashton HC, Ferrier IN, Moore PB. Family illness history, obstetric complications and age of onset in bipolar patients. Open Neuropsychopharmacol J. 2009;2:11–5.
Logsdon MC, Mittelberg M, Jacob AE, Luther JF, Wisniewski SR, Confer A, Eng H, Wisner KL. Maternal-Infant interaction in women with unipoloar and bipolar depression. Appl Nurs Res. 2015;28(4):381–3.
Marengo E, Martino DJ, Igoa A, Scapola M, Fassi G, Baamonde MU, Strejilevich SA. Unplanned pregnancies and reproductive health among women with bipolar disorder. J Affect Disord. 2015;178:201–5.
Taylor CL, Stewart R, Ogden J, Broadbent M, Pasupathy D, Howard LM. The characteristics and health needs of pregnant women with schizophrenia compared with bipolar disorder and affective psychoses. BMC psychiatry. 2015;15:88.
Wesseloo R, Kamperman AM, Munk-Olsen T, Pop VJ, Kushner SA, Bergink V. Risk of Postpartum Relapse in Bipolar Disorder and Postpartum Psychosis: A Systematic Review and Meta-Analysis. Am J Psychiatry. 2016;173(2):117–27.
EPHPP. Effective Public Health Practice Project. In. http://ephpp.ca/. Accessed 12 Jul 2016: The Effective Public Health Practice Project; 1999.
Blackmore ER, Jones I, Doshi M, Haque S, Holder R, Brockington I, Craddock N. Obstetric variables associated with bipolar affective puerperal psychosis. Br J Psychiatry. 2006;188(1):32–6.
Bratfos O, Haug J. Puerperal mental disorders in manic-depressive females. Acta Psychiatr Scand. 1996;42(3):285–94.
Di Florio A, Jones L, Forty L, Gordon‐Smith K, Craddock N, Jones I. Bipolar disorder, miscarriage, and termination. Bipolar Disord. 2014;17(1):102–5.
Hunt N, Silverstone T. Does puerperal illness distinguish a subgroup of bipolar patients? J Affect Disord. 1995;34(2):101–7.
Reich T, Winokur G. Postpartum psychoses in patients with manic depressive disease. J Nerv Ment Dis. 1970;151(1):60–8.
Robertson E, Jones I, Haque S, Holder R, Craddock N. Risk of puerperal and non-puerperal recurrence of illness following bipolar affective puerperal (post-partum) psychosis. Br J Psychiatry. 2005;186(3):258–9.
Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, Porter AC, Tugwell P, Moher D, Bouter LM. Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews. BMC Med Res Methodol. 2007;7(1):10.
Sharif MO, Janjua-Sharif FN, Sharif FNJ, Ali H, Ahmed F. Systematic reviews explained: AMSTAR-how to tell the good from the bad and the ugly. Oral health dental manag. 2013;12(1):9.
Chaudron LH, Pies RW. The relationship between postpartum psychosis and bipolar disorder: A review. J Clin Psychiatry. 2003;64(11):1284–92.
Sharma V, Pope CJ. Pregnancy and bipolar disorders: A systematic review. J Clin Psychiatry. 2012;73(11):1447–55.
Özerdem A, Akdeniz F. Pregnancy and postpartum in bipolar disorder. Neuropsychiatry. 2014;4(1):95–107.
Jablensky AV, Morgan V, Zubrick SR, Bower C, Yellachich LA. Pregnancy, delivery, and neonatal complications in a population cohort of women with schizophrenia and major affective disorders. Am J Psychiatr. 2005;162(1):79–91.
Akdeniz F, Vahip S, Pirildar S, Vahip I, Doganer I, Bulut I. Risk Factors Associated with Childbearing-Related Episodes in Women with Bipolar Disorder. Psychopathology. 2003;36(5):234–8.
Bodén R, Lundgren M, Brandt L, Reutfors J, Andersen M, Kieler H. Risks of adverse pregnancy and birth outcomes in women treated or not treated with mood stabilisers for bipolar disorder: Population based cohort study. BMJ. 2012;345:e7085.
Lee H-C, Lin H-C. Maternal bipolar disorder increased low birthweight and preterm births: A nationwide population-based study. J Affect Disord. 2010;121(1–2):100–5.
Mei-Dan E, Ray JG, Vigod SN. Perinatal outcomes among women with bipolar disorder: a population-based cohort study. Am J Obstet Gynecol. 2015;212(3):367. e361-368.
Munk-Olsen T, Laursen T, Mendelson T, Pedersen CB, Mors O, Mortensen P. RIsks and predictors of readmission for a mental disorder during the postpartum period. Arch Gen Psychiatry. 2009;66(2):189–95.
Di Florio A, Forty L, Gordon-Smith K, Heron J, Jones L, Craddock N, Jones I. Perinatal episodes across the mood disorder spectrum. JAMA psychiatry. 2013;70(2):168–75.
Di Florio A, Jones L, Forty L, Gordon-Smith K, Blackmore ER, Heron J, Craddock N, Jones I. Mood disorders and parity—A clue to the aetiology of the postpartum trigger. J Affect Disord. 2014;152–154:334–9.
Grof P, Robbins W, Alda M. Protective effect of pregnancy in women with lithium-responsive bipolar disorder. Primary Care Companion J Clin Psychiatry. 2001;3(2):87.
Spitzer RL, Endicott J, Robins E. Research diagnostic criteria: Rationale and reliability. Arch Gen Psychiatry. 1978;35(6):773–82.
The authors would like to thank all the authors and publishers of the original studies.
This study was not funded by any external sources.
Availability of data and material
Data sharing is not applicable to this article as no datasets were generated or analysed during the current study. The papers that were included in the literature searches have been organised in an EndNote Library held by the first author.
All authors (MR, MB and CB) have made substantial contributions to conception and design, revision of the literature searches, assessement of the studies for quality, data extraction, analysis and interpretation of the data as well as drafting and revising the manuscript. All authors have approved the final manuscript. MR undertook the literature searches and coordinated the systematic review process.
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
Ethics approval was not needed for this systematic literature review.
Bipolar disorder in pregnancy and childbirth: a systematic review of outcomes. (DOC 214 kb)