你在这里

Maternal and neonatal outcomes in women undergoing bariatric surgery: a systematic review and meta-analysis

European Journal of Obstetrics & Gynecology and Reproductive Biology, pages 45 - 53

Abstract

Background

Obese women are at increased risk for many pregnancy complications, and bariatric surgery (BS) before pregnancy has shown to improve some of these.

Objectives

To review the current literature and quantitatively assess the obstetric and neonatal outcomes in pregnant women who have undergone BS.

Search strategy

MEDLINE, EMBASE and Cochrane databases were searched using relevant keywords to identify studies that reported on pregnancy outcomes after BS.

Selection criteria

Pregnancy outcome in firstly, women after BS compared to obese or BMI-matched women with no BS and secondly, women after BS compared to the same or different women before BS. Only observational studies were included.

Data collection and analysis

Two investigators independently collected data on study characteristics and outcome measures of interest. These were analysed using the random effects model. Heterogeneity was assessed and sensitivity analysis was performed to account for publication bias.

Main results

The entry criteria were fulfilled by 17 non-randomised cohort or case-control studies, including seven with high methodological quality scores. In the BS group, compared to controls, there was a lower incidence of preeclampsia (OR 0.45, 95% CI 0.25–0.80;P = 0.007), GDM (OR 0.47, 95% CI 0.40–0.56;P < 0.001) and large neonates (OR 0.46, 95% CI 0.34–0.62;P < 0.001) and a higher incidence of small neonates (OR 1.93, 95% CI 1.52–2.44;P < 0.001), preterm birth (OR 1.31, 95% CI 1.08–1.58;P = 0.006), admission for neonatal intensive care (OR 1.33, 95% CI 1.02–1.72;P = 0.03) and maternal anaemia (OR 3.41, 95% CI 1.56–7.44,P = 0.002).

Conclusions

BS as a whole improves some pregnancy outcomes. Laparoscopic adjustable gastric banding does not appear to increase the rate of small neonates that was seen with other BS procedures. Obese women of childbearing age undergoing BS need to be aware of these outcomes.

Keywords: Bariatric surgery, Pregnancy, Preeclampsia, Gestational diabetes, Birth weight.

Introduction

Obesity in many developed countries is acquiring epidemic proportions. In the UK, one in three pregnant women are overweight or obese[1] and [2]. In such women, compared to those with normal body mass index (BMI), there is increased risk for many pregnancy complications, including gestational diabetes mellitus (GDM), preeclampsia, preterm birth, anaemia, caesarean section, delivery of small and large neonates and perinatal death[3], [4], [5], [6], [7], [8], [9], [10], and [11]. Bariatric surgery (BS), including malabsorptive procedures like gastric bypass and restrictive procedures like laparoscopic adjustable gastric band (LAGB), is increasingly being used as an effective method of treating obesity. Several studies including systematic reviews have reported pregnancy outcome following such therapy is improved[12], [13], [14], [15], [16], [17], and [18]. However, these studies have not been able to quantify the effect by means of meta-analysis because of the insufficient number of suitable primary studies and the heterogeneity among them.

The aim of this study was to review the current literature and quantitatively assess the obstetric and neonatal outcomes in pregnant women who have undergone BS and to examine whether the effects of surgery are the mere consequence of reduction in BMI at the onset of pregnancy.

Methods

Institutional review board (IRB) approval was not necessary as the meta-analysis did not require any patient identifying information.

Selection criteria

MEDLINE (1966 to June 2014), EMBASE (1980 to June 2014) and Cochrane (1993 to June 2014) databases were searched using the terms “obstetrics AND bariatric surgery”, “pregnancy outcomes AND bariatric surgery”, “neonatal outcomes AND bariatric surgery” and “Pregnancy outcomes AND weight-reducing surgery”.

The inclusion criteria were observational studies reporting on pregnancy outcome in firstly, women after BS compared to obese or BMI-matched women with no BS and secondly, women after BS compared to the same or different women before BS. No language restrictions were made.

The outcome measures of interest were: preeclampsia, GDM, maternal anaemia, preterm birth, caesarean section, large and small neonates, neonatal intensive care admission (NICU) and perinatal mortality.

Two reviewers (NG and ND) independently screened the titles and abstracts from the search to identify all potentially useful studies for which the full manuscripts were evaluated to assess eligibility for the meta-analysis. The reference lists in the selected manuscripts were examined for any additional relevant studies. Any inconsistency regarding the selection of papers was resolved independently by another co-author (CS).

Quality assessment

The quality of the selected studies was evaluated by two assessors (NG and ND) using the Newcastle–Ottawa scale for patient selection, comparability of the two study groups and assessment of outcome [19] . Studies that achieved 9 or more points, from a maximum of 10, were considered to be of high methodological quality.

Statistical analysis

The meta-analysis was performed in line with recommendations from the Cochrane Collaboration and the Quality of Reporting of Meta-analyses guidelines[20] and [21]. Each outcome measure in the BS group was compared to the control group using the odds ratio (OR) as the summary statistic. Risk ratios were also calculated to determine the percentage increase or decrease of an outcome measure after BS.

In the calculation of ORs the random-effect model was used. This assumes that a variation exists between studies, and therefore the calculated OR has a more conservative value[22] and [23]. Yate's correction was used for those studies that contained a zero in one cell for the number of events of interest[24] and [25]. However, such zero values created problems with the computation of ratio and this was overcome by adding 0.5 to each cell of the study in question.

Heterogeneity was evaluated by graphic exploration with funnel plots to assess publication bias and by subgroup analysis using the following: (a) high methodological quality studies comparing pregnancies in women after BS with those in women before or no BS (b) pregnancies in the same women after and before BS, (c) pregnancies in different women after and before BS, (d) pregnancies in women after BS compared to those in women with no BS, but matched to cases for prepregnancy BMI, (e) pregnancies in women after BS compared to those in obese women without BS, but matched to cases for presurgery BMI and (f) pregnancies in women after LAGB with those in women before or no BS.

Analysis was conducted by using the statistical software Review Manager Version 5.2 (The Cochrane Collaboration, Software Update, Oxford, UK).

Results

Primary studies included in the meta-analysis

Search of MEDLINE yielded 312 articles and no additional articles were identified from the searches of the EMBASE and Cochrane databases ( Fig. 1 ). After reading the title and abstract of the studies, 283 were excluded because they were not original cohort studies or did not provide data on the desired pregnancy outcome measures. Full manuscripts were evaluated in 29 cases but only 17 fulfilled the entry criteria[26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], and [42]while the remaining 12 papers were excluded[43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53], and [54]. There were three main reasons for excluding these 12 papers: (a) there was no comparison (control) group[44], [45], [46], [48], [50], [51], [52], and [53], (b) pregnancy outcomes were compared with different timing from BS but not against no surgery[47] and [54]and (c) different BS techniques were compared against each other but not against no surgery[43] and [49].

gr1

Fig. 1 Flowchart showing selection of studies included in the meta-analysis.

Main study characteristics and methodological quality assessment

The 17 articles included in the meta-analysis consisted of 16 retrospective and one prospective [28] non-randomised cohort or case-control studies. High methodological quality scores were obtained for seven of the 17 studies[29], [30], [33], [37], [38], [41], and [42]. A total of 166,134 participants were evaluated which included 5361 women after BS and 160,773 controls ( Table 1 )[26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], and [42]. All studies matched the comparison groups or used logistic regression models to control for confounders.

Table 1 The main characteristics of each study with the methodological quality score.

Study Study design Case group Control group Newcastle–Ottawa
Wittgrove et al. [26] Case-control, retrospective 18 23 Same women before BS 8/10
Skull et al. [27] Case-control, retrospective 49 31 Same women before BS 8/10
Dixon et al. [28] Case-control, prospective 79 A: 40 Same women before BS

B: 79 Different obese women with no BS, matched for pre surgery BMI
8/10
Ducarme et al. [29] Case-control, retrospective 13 414 Different obese women with no BS 9/10
Weintraub et al. [30] Cohort, retrospective 507 301 Different obese women before BS 9/10
Bennett et al. [31] Cohort, retrospective 316 269 Different obese women before BS 7/10
Burke et al. [32] Case-control, retrospective 354 346 Different obese women before BS 7/10
Lapolla et al. [33]
 
Cohort, retrospective

Case-control, retrospective
A: 83

B: 27
A: 120 Different obese women with no BS

B: 27 Different obese women before BS
10/10
Santulli et al. [34] Cohort, retrospective 24 120 Different obese women with no BS, matched for prepregnancy BMI 8/10
Josefsson et al. [35] Cohort, retrospective 126 241 Different obese women before BS 8/10
Aricha-Tamir et al. [36] Case-control, retrospective 144 144 Same women before BS 8/10
Belogolovkin et al. [37] Cohort, retrospective 293 131,023 Different obese women with no BS 10/10
Lesko et al. [38] Case-control, retrospective 70 A: 140 Different obese women with no BS, matched for presurgery BMI

B: 140 Different obese women with no BS, matched for prepregnancy BMI
10/10
Amsalem et al. [39] Case-control, retrospective 109 109 Same women before BS 8/10
Kjaer et al. [40] Case-control, retrospective 339 1277 Different obese women with no BS, matched for prepregnancy BMI 8/10
Roos et al. [41] Case-control, retrospective A: 2474

B: 2511
A: 11,979 Different obese women with no BS, matched for presurgery BMI

B: 12,379 Different obese women with no BS, matched for prepregnancy BMI
9/10
Shai et al. [42] Case-control, retrospective 326 1612 Different obese women with no BS 9/10

A = deliveries in women after bariatric surgery (BS); B = deliveries in women before or no bariatric surgery (BS).

Definitions of outcome measures

Preeclampsia

This was reported in 11 of the 17 studies[27], [28], [29], [30], [31], [33], [36], [37], [38], [40], and [42], but only two provided a definition for the condition[33] and [38]. In both studies, preeclampsia was defined as systolic blood pressure >140 mmHg or diastolic blood pressure >90 mmHg after 20 weeks’ gestation in previously normotensive women together with proteinuria (>100 mg/dL on urinalysis or 300 mg in a 24-h urine collection).

Gestational diabetes

This was reported in 15 of the 17 studies[26], [27], [28], [29], [30], [31], [32], [33], [34], [36], [37], [38], [39], [40], and [42], but only two provided a definition for the condition[32] and [33]. One study defined GDM as the new incidence of abnormal glucose tolerance complicating pregnancy, childbirth or puerperium [32] and the other [33] defined GDM as per the Carpenter–Courstan criteria and the recommendations after a 100 g oral glucose tolerance test[55] and [56].

Maternal anaemia

This was reported in four of the 17 studies and all defined it as haemoglobin concentration less than 10 g/dL[30], [34], [36], and [42]. None of the studies specified the gestational age at which blood samples were taken.

Caesarean section

This was reported in 13 of the 17[26], [27], [29], [30], [32], [33], [34], [35], [36], [37], [38], [39], and [40]studies and in all but one [27] data were provided for total rather than emergency or elective caesarean section.

Preterm birth

This was reported in 12 of the 17 studies[26], [28], [29], [33], [34], [35], [36], [37], [38], [39], [40], and [41]. In eight studies, preterm birth was defined as birth before the 37th week of gestation, but in four studies no definition was provided[26], [34], [36], and [38]. Only one study provided separate data on spontaneous and iatrogenic preterm birth [41] .

Large neonates

This was reported in 14 of the 17 studies[26], [28], [29], [30], [32], [33], [35], [36], [37], [38], [39], [40], [41], and [42]and in 11 of these it was defined as macrosomia in which the birth weight was over 4 kg[26], [28], [29], [30], [33], [36], [37], [38], [39], [40], and [42]. Two studies reported incidence of large for gestational age, defined as birth weight above the 95th percentile[35] and [41], another provided data separately for macrosomia and large for gestational age above the 90th percentile [33] and another reported macrosomia and LGA together but no definition was provided for either [32] . For the purpose of this study all the above outcomes were classified as large neonates.

Small neonates

This was reported in 11 of the 17 studies[28], [29], [30], [33], [34], [35], [36], [37], [38], [40], and [41]and in three of these it was defined as intrauterine growth restriction (IUGR)[30], [34], and [36]. However, only one provided a definition of IUGR as a decrease in fetal growth velocity in serial ultrasound scans [34] . Seven studies reported on small for gestational age neonates, which was defined as birth weight below the 10th percentile[29], [33], [37], and [38]or below the 5th percentile[35], [40], and [41]. Low birth weight (LBW) was reported in three studies; this was defined as weight less than 2.5 kg in two[28] and [29], whereas one study did not provide a definition for LBW [30] . In these three studies, LBW was not associated with prematurity and was thought to be the result of the micronutrient restriction seen after bariatric surgery[28], [29], and [30]. Since different terms were used which were not even consistently defined, the authors decided to collectively group the above outcomes to the “small neonates” as the ultimate end point, albeit this could be the result of different mechanism.

Perinatal mortality

This was reported in five of the 17 studies[30], [32], [38], [40], and [41]. One did not give a definition [38] , three presented data only on fetal death between 22 and 40 weeks’ gestation[30], [32], and [41]while Kjaer et al. defined perinatal mortality as death from 22 weeks’ gestation to the 7th day postpartum [40] .

Neonatal intensive care admissions

This was reported in four of the 17 studies[33], [34], [38], and [40]. None of the studies provided information on the indication of admission or the length of stay.

Results for the overall meta-analysis

Table 2 and Fig. 2 summarise the obstetric and neonatal outcomes in women who underwent BS and the controls. In the BS group, compared to controls, there was a lower incidence of preeclampsia (OR 0.45, 95% CI 0.25–0.80;P = 0.007), GDM (OR 0.47, 95% CI 0.40–0.56;P < 0.001) and large neonates (OR 0.46, 95% CI 0.34–0.62;P < 0.001) and a higher incidence of small neonates (OR 1.93, 95% CI 1.52–2.44;P < 0.001), preterm birth (OR 1.31, 95% CI 1.08–1.58;P = 0.006), admission for neonatal intensive care (OR 1.33, 95% CI 1.02–1.72;P = 0.03) and maternal anaemia (OR 3.41, 95% CI 1.56–7.44,P = 0.002). There were no significant differences between the groups in incidence of caesarean section or perinatal mortality.

Table 2 Results of overall meta-analysis.

Outcome of interest Studies Participants OR (95% CI) P-value Heterogeneity

P-value
RR (95% CI)
    BS Control        
Preeclampsia 11 2075 135,473 0.45 (0.25–0.80) 0.007 <0.001 0.48 (0.29–0.82)
Gestational diabetes mellitus 15 2724 136,075 0.47 (0.40–0.56) <0.001 <0.001 0.52 (0.45–0.59)
Preterm birth 12 3809 158,263 1.31 (1.08–1.58) 0.006 0.022 1.28 (1.08–1.51)
Large neonates 14 4968 148,334 0.46 (0.34–0.62) <0.001 0.004 0.51 (0.38–0.67)
Small neonates 11 4185 158,343 1.93 (1.52–2.44) <0.001 0.012 1.83 (1.48–2.27)
Maternal anaemia 4 1270 133,170 3.41 (1.56–7.44) 0.002 <0.001 2.63 (1.25–5.53)
Caesarean section 13 2126 134,456 0.99 (0.75–1.31) 0.95 <0.001 1.00 (0.84–1.20)
Perinatal mortality 5 3804 14,532 1.05 (0.48–2.31) 0.90 0.15 1.05 (0.48–2.28)
Neonatal intensive care 4 516 1824 1.38 (1.02–1.86) 0.03 0.51 1.33 (1.02–1.72)

BS = bariatric surgery, OR = odds ratio, RR = risk ratio, CI = confidence interval.

gr2

Fig. 2 Summary of odds ratio with 95% confidence interval for preeclampsia, gestational diabetes mellitus and birth of large neonates in pregnancies of women who underwent bariatric surgery (BS) compared to those in controls. Total analysis (black), high methodological quality studies (red), after and before BS in the same women (blue), after and before BS in different women (green), after BS compared to women with no BS, but matched to cases for prepregnancy BMI (pink) and after BS compared to women without BS, but matched to cases for presurgery BMI (brown).

Subgroup analysis

The findings of subgroup analysis for the following: (a) high methodological quality studies comparing pregnancies in women after BS with those in women before or no BS, (b) pregnancies in the same women after and before BS, (c) pregnancies in different women after and before BS, (d) pregnancies in women after BS compared to those in women with no BS, but matched to cases for prepregnancy BMI, (e) pregnancies in women after BS compared to those in obese women without BS, but matched to cases for presurgery BMI and (f) pregnancies in women after LAGB with those in women before or no BS are summarised inTable 3, Table 4, Table 5, Table 6, Table 7, and Table 8. The results of the total and subgroup analysis for preeclampsia, GDM, large and small neonates and caesarean section are illustrated inFig 2 and Fig 3.

Table 3 Results of meta-analysis in the subgroup of high methodological quality studies comparing outcome in pregnancies of women after bariatric surgery to those of obese women before or no surgery.

Outcome of interest References of studies Participants OR (95% CI) P-value Heterogeneity

P-value
RR (95% CI)
    BS Control        
Preeclampsia [29], [30], [33], [37], [38], and [42] 1292 133,777 0.76 (0.52–1.11) 0.16 0.16 0.78 (0.55–1.11)
Gestational diabetes mellitus [29], [30], [33], [37], [38], and [42] 1292 133,777 0.34 (0.18–0.67) <0.001 <0.001 0.41 (0.23–0.72)
Preterm birth [33], [37], [38], and [41] 2,957 155,856 1.51 (1.33–1.72) <0.001 0.59 1.46 (1.30–1.63)
Large neonates [29], [30], [33], [37], [38], [41], and [42] 3799 146,115 0.38 (0.24–0.58) <0.001 0.009 0.42 (0.28–0.63)
Small neonates [29], [30], [33], [37], [38], and [41] 3473 156,482 1.88 (1.42–2.50) <0.001 0.09 1.80 (1.41–2.30)
Maternal anaemia [37] and [42] 619 132,635 2.40 (0.97–5.96) 0.06 <0.001 2.00 (0.82–4.89)
Caesarean section [29], [30], [33], [37], and [38] 966 132,165 1.01 (0.63–1.63) 0.96 <0.001 1.04 (0.77–1.40)
Perinatal mortality [30] and [38] 577 581 2.35 (0.06–89.67) 0.65 0.003 2.31 (0.06–82.26)
Neonatal intensive care [33] and [38] 153 427 1.56 (0.73–3.36) 0.25 0.17 1.48 (0.77–2.83)

BS = bariatric surgery, OR = odds ratio, RR = risk ratio, CI = confidence interval.

Table 4 Results of meta-analysis in the subgroup of studies comparing outcome in pregnancies of women after bariatric surgery and those in the same women before surgery.

Outcome of interest References of studies Participants OR (95% CI) P-value Heterogeneity

P-value
RR (95% CI)
    BS Control        
Preeclampsia [27], [28], and [33] 155 98 0.20 (0.08–0.51) <0.001 0.56 0.24 (0.11–0.56)
Gestational diabetes mellitus [26], [28], [33], [36], and [39] 377 343 0.71 (0.45–0.1.11) 0.14 <0.001 0.73 (0.48–1.10)
Preterm birth [26], [33], [36], and [39] 298 303 0.75 (0.45–1.26) 0.28 84 0.78 (0.49–1.23)
Large neonates [26], [33], [36], and [39] 298 303 0.75 (0.35–1.63) 0.47 0.19 0.81 (0.44–1.50)
Small neonates [33] and [36] 171 171 0.81 (0.16–4.00) 0.79 0.26 0.83 (0.18–3.77)
Maternal anaemia            
Caesarean section [26], [27], [33], [36], and [39] 344 334 1.26 (0.87–1.83) 0.22 0.35 1.18 (0.92–1.52)
Perinatal mortality            
Neonatal intensive care            

BS = bariatric surgery, OR = odds ratio, RR = risk ratio, CI = confidence interval.

Table 5 Results of meta-analysis in the subgroup of studies comparing outcome in pregnancies of women after bariatric surgery and those in a different group of women before surgery.

Outcome of interest References of studies Participants OR (95% CI) P-value Heterogeneity

P-value
RR (95% CI)
    BS Control        
    BS Control        
Preeclampsia [30] and [31] 823 570 0.14 (0.06–0.31) <0.001 0.18 0.16 (0.10–0.27)
Gestational diabetes mellitus [30], [31], and [32] 1171 916 0.42 (0.22–0.79) 0.007 0.002 0.48 (0.29–0.81)
Preterm birth            
Large neonates [30] and [32] 987 888 0.52 (0.31–0.86) 0.01 0.16 0.54 (0.32–0.92)
Small neonates [30] and [35] 633 542 1.58 (1.02–2.45) 0.04 0.31 1.54 (0.95–2.50)
Maternal anaemia            
Caesarean section [30], [32], and [35] 987 888 0.92 (0.38–2.22) 0.82 <0.001 0.96 (0.51–1.81)
Perinatal mortality [30] and [32] 861 647 0.52 (0.19–1.41) 0.20 0.47 0.52 (0.20–1.40)
Neonatal intensive care            

BS = bariatric surgery, OR = odds ratio, RR = risk ratio, CI = confidence interval.

Table 6 Results of meta-analysis in the subgroup of studies comparing outcome in pregnancies of women after bariatric surgery and those in obese women without surgery, but matched for prepregnancy body mass index.

Outcome of interest References of studies Participants OR (95% CI) P-value Heterogeneity

P-value
RR (95% CI)
    BS Control        
Preeclampsia [38] and [40] 407 1417 0.63 (0.38–1.06) 0.08 0.53 0.67 (0.42–1.07)
Gestational diabetes mellitus [34], [38], and [40] 433 1537 0.77 (0.22–2.65) 0.68 0.10 0.81 (0.26–2.51)
Preterm birth [34], [38], [40], and [41] 2944 13,916 1.67 (1.27–2.18) <0.001 0.26 1.60 (1.27–2.02)
Large neonates [40] and [41] 2916 13,755 0.44 (0.27–0.70) <0.001 0.17 0.46 (0.30–0.72)
Small neonates [34], [38], [40], and [41] 2940 13,875 2.30 (1.53–3.44) <0.001 0.16 2.18 (1.51–3.13)
Maternal anaemia            
Caesarean section [34], [38], and [40] 433 1537 1.16 (0.92–1.46) 0.20 0.78 1.11 (0.95–1.30)
Perinatal mortality [38] and [40] 409 1417 2.08 (0.14–30.87) 0.60 0.08 2.04 (0.14–28.75)
Neonatal intensive care [34], [38], and [40] 433 1537 2.00 (0.80–4.99) 0.14 0.08 1.87 (0.81–4.34)

BS = bariatric surgery, OR = odds ratio, RR = risk ratio, CI = confidence interval.

Table 7 Results of meta-analysis in the subgroup of studies comparing outcome in pregnancies of women after bariatric surgery and those in obese women without surgery, but matched for presurgery body mass index.

Outcome of interest References of studies Participants OR (95% CI) P-value Heterogeneity

P-value
RR (95% CI)
    BS Control        
Preeclampsia [28], [29], [33], [37], [38], and [42] 864 133,388 0.69 (0.43–1.12) 0.13 0.05 0.73 (0.48–1.11)
Gestational diabetes mellitus [28], [29], [33], [37], [38], and [42] 864 133,388 0.24 (0.10–0.54) <0.001 <0.001 0.30 (0.15–0.61)
Preterm birth [28], [33], [37], [38], and [41] 2999 142,489 1.35 (1.18–1.54) <0.001 0.43 1.32 (1.17–1.48)
Large neonates [28], [29], [33], [37], [38], and [42] 864 133,388 0.35 (0.18–0.66) 0.001 0.02 0.39 (0.20–0.76)
Small neonates [28], [29], [33], [37], [38], and [41] 3008 143,755 1.90 (1.39–2.77) <0.001 0.14 1.84 (1.41–2.40)
Maternal anaemia [37] and [42] 619 132,635 2.40 (0.97–5.96) 0.06 <0.001 2.00 (0.82–4.89)
Caesarean section [29], [33], [37], and [38] 459 131,697 0.69 (0.34–1.42) 0.32 <0.001 0.84 (0.57–1.23)
Perinatal mortality            
Neonatal intensive care [33] and [38] 153 260 2.27 (1.26–4.07) 0.006 0.93 2.03 (1.22–3.36)

BS = bariatric surgery, OR = odds ratio, RR = risk ratio, CI = confidence interval.

Table 8 Results of meta-analysis in the subgroup of studies comparing outcome in pregnancies of women after LAGB and those in obese women with no BS.

Outcome of interest References of studies Participants OR (95% CI) P-value Heterogeneity

P-value
RR (95% CI)
    BS Control        
Preeclampsia [27], [28], [29], and [33] 224 711 0.34 (0.14–0.81) 0.01 0.20 0.39 (0.18–0.85)
Gestational diabetes mellitus [27–29, 33] 224 711 0.16 (0.09–0.32) <0.001 0.35 0.22 (0.13–0.38)
Preterm birth [28], [29], and [33] 175 640 0.99 (0.43–2.30) 0.98 0.21 0.99 (0.47–2.09)
Large neonates [28], [29], and [33] 175 640 0.62 (0.38–1.00) 0.05 0.97 0.71 (0.49–1.01)
Small neonates [28], [29], and [33] 175 640 0.57 (0.22–1.44) 0.23 0.68 0.60 (0.25–1.42)
Maternal anaemia            
Caesarean section [27], [29], and [33] 145 592 0.74 (0.29–1.90) 0.54 0.08 0.87 (0.47–1.58)
Perinatal mortality            
Neonatal intensive care            

LAGB = laparoscopic adjustable gastric band, BS = bariatric surgery, OR = odds ratio, RR = risk ratio, CI = confidence interval.

gr3

Fig. 3 Summary of odds ratio with 95% confidence interval for birth of small neonates and caesarean section in pregnancies of women who underwent bariatric surgery (BS) compared to those in controls. Total analysis (black), high methodological quality studies (red), after and before BS in the same women (blue), after and before BS in different women (green), after BS compared to women with no BS, but matched to cases for prepregnancy BMI (pink) and after BS compared to women without BS, but matched to cases for presurgery BMI (brown).

Publication bias

Evaluation of funnel plots demonstrated publication bias was present for preeclampsia and large neonates, even after sensitivity analysis of high methodological quality studies. A funnel plot of all the studies used in the meta-analysis reporting on GDM is shown in Fig. 4 (top). All studies (except from two[28] and [30]) lie inside the 95% CIs and are distributed evenly about the vertical, showing no evidence of publication bias. Heterogeneity was observed among the studies (P = 0.001). When only high methodological quality studies were considered ( Fig. 4 , bottom) all studies lied within the 95% CIs and were distributed evenly about the vertical, showing no evidence of publication bias (P = 0.55).

gr4

Fig. 4 Funnel plot illustrating heterogeneity in regards to the incidence of gestational diabetes mellitus in the overall meta-analysis (top) and in the sensitivity analysis of high methodological quality studies (bottom).

Discussion

Main findings

The findings of this meta-analysis demonstrate that in women who underwent BS, compared to controls without BS, there is a decrease in the incidence of preeclampsia, GDM and large neonates, an increase in the incidence of small neonates, preterm birth, admission for neonatal intensive care, maternal anaemia and no significant difference in incidence of caesarean section or perinatal mortality.

Interpretation

Preeclampsia

There is a well-reported association between obesity and the development of preeclampsia[57] and [58]. There is also good evidence that the incidence of preeclampsia is decreased by BS[13], [14], [16], and [18]. This meta-analysis has confirmed that BS is associated with reduction in preeclampsia and has quantified this effect as being about half.

Gestational diabetes mellitus

Previous systematic reviews have concluded that the incidence of GDM is reduced in women who had BS before pregnancy compared to obese women[13], [14], [16], and [18]. Our meta-analysis quantifies this effect as being about half.

It has been previously speculated that this reduction in the incidence of GDM might be due to the metabolic or absorptive changes that occur in response to BS [38] . In addition, studies investigating post-BS pregnancies with those of the general population, found that the incidence of GDM was higher in the post-BS group[12] and [59]. Our subgroup analysis revealed no significant difference in GDM between post-BS pregnancies and pregnancies with matched prepregnancy BMI (P = 0.68). This suggests that BMI and the hormonal and metabolic environment associated with it are responsible for the development or worsening of diabetes and not the postoperative status per se.

Small and large neonates

Previous systematic reviews have concluded that in women who had BS before pregnancy, compared to obese pregnant women, the incidence of large neonates is reduced and that of small neonates is increased[13], [14], [16], [17], and [18]. The present meta-analysis found that the risk of large neonates is halved and that of small neonates is increased by about 80%.

Our subgroup analysis revealed that unlike the case of GDM, the incidence of small neonates in the BS group was significantly higher than in pregnancies with matched prepregnancy BMI. It was previously suggested that the BS-related increase in incidence of small neonates is the consequence of malnutrition and microenvironment deficiencies secondary to the therapeutic postoperative state [18] . Interestingly, a subgroup analysis of women who had undergone LAGB compared to women with no surgery, found no difference in the rates of small neonates (P = 0.23). This suggests that restrictive rather than malabsorptive BS (like gastric bypass) should be preferred in young women planning to have children in order to minimise this complication.

This issue requires further investigation because the beneficial long-term effects in the reduction of large neonates in regards to obesity and type 2 diabetes [15] may be counterbalanced by the detrimental effects of fetal growth restriction[60] and [61].

Preterm birth

This meta-analysis found that BS is associated with a 28% increase in preterm birth. Similarly, subgroup analysis on prepregnancy BMI-matched obese women and on high methodological quality studies also found that in the BS group there is an increase in the incidence of preterm birth. These findings contradict those of a previous review which reported that BS is associated with a decrease in preterm birth [13] . The studies did not report on spontaneous or iatrogenic preterm birth separately.

In the subgroup analysis of women who had undergone LAGB compared to women with no surgery, there was no difference in the rate of preterm birth (P = 0.98). Therefore, our findings may be related to the increase in small neonates after BS as a whole, but not after LAGB, which is likely to cause iatrogenic rather than spontaneous preterm birth. This, however, requires further investigation.

Caesarean section

There was considerable inconsistency in the literature regarding the incidence of caesarean section after BS. One systematic review concluded that the incidence of caesarean section after BS is increased compared to that of obese women with no BS [14] . Moreover, one primary study investigating pregnancy outcomes after BS compared to the normal population concluded that BS is an independent risk factor for caesarean section [59] . The authors of the same study acknowledged the absence of any possible physiological cause for such effect and concluded that this may be a result of care-giver bias. On the other hand, a review by Vrebosch et al. found that the incidence of caesarean sections was lower after BS [16] . Two other reviews concluded that the incidence of caesarean section after BS is unclear and will need to be investigated further[13] and [18].

Our meta-analysis did not demonstrate any significant difference in caesarean section rates in women after BS, compared to obese women with no BS, in either the total group analysis or any of the subanalyses. We conclude that BS does not appear to have an effect on the rate of caesarean section.

Maternal anaemia

The present study found significant increase in the incidence of anaemia in pregnant women after BS, compared to obese women with no BS. However, only four studies contributed to this meta-analysis[30], [36], [37], and [42]. Obesity is associated with poor iron status and higher iron requirements [62] . Moreover, Salgado et al. reported an increased iron loss and decreased iron absorption after BS [63] . The same study also found that the incidence of anaemia post BS was higher in younger and in female patients.

We suggest that these women should have more frequent haematological investigations during pregnancy in order to receive prompt and effective supplementation. This may avoid the need of blood transfusion after delivery which was reported to be higher after BS in one study [38] .

Strengths and limitations

This is the first meta-analysis exploring the obstetric and neonatal outcomes after BS. A total number of 166,134 participants were taken into account and six different subgroups were created to provide more information and minimise potential bias.

Meta-analyses are thought to provide the highest level of evidence [64] , but they have been criticised in that they reinforce the inherent systematic biases of the included studies, produce spurious statistical stability and discourage further research [65] . On the other hand, it is argued that pooling results from many primary studies and statistical quantification provide an excellent tool for identifying reasons for variability and inconsistency and that the finding of heterogeneity sets the stage for further research on a given topic [66] .

Given that randomised control trials are not feasible for assessing pregnancy outcomes, observational studies represent the best available evidence. However, such studies are subject to confounding and bias [67] . The results of this meta-analysis should be interpreted with caution because of the retrospective nature of the included studies, the varied definitions of the outcome measures and heterogeneity between the individual studies that occasionally persisted during subgroup analysis. Furthermore, most studies did not provide separate data on each BS procedure, but rather presented pregnancy outcome after BS as a whole. Four studies compared pregnancy outcome in women who underwent LABG only and these were presented as a subgroup[27], [28], [29], and [33]. Future studies will need to compare pregnancy outcome between different BS procedures. Finally, the studies included did not provide separate data on the outcomes in regard to the time from BS in order to identify the optimum time for conception after BS. However, this question has been specifically addressed elsewhere [54] .

Conclusion

The present study concludes that BS improves pregnancy outcomes but potentially increases the risk of small neonates and preterm birth. However, these complications disappear after LAGB which should be the preferred procedure in women planning to become pregnant. This meta-analysis should be regarded as a valuable working document for the growing number of healthcare professionals who need to be aware of the pregnancy outcomes after BS in order to inform when caring for obese women of childbearing age.

Contribution to authorship

N Galazis: Performed the literature search, data analysis and composed first draft.

N Docheva: Performed literature search and extraction of relevant study characteristics.

C Simillis: Performed the statistical analysis.

KH Nicolaides: Supervised work throughout and composed second, third and fourth drafts.

Acknowledgement

Only the listed authors are responsible for this study.

References

  • [1] CMACE/RCOG. Joint guideline: management of women with obesity in pregnancy. (CMACE, 2010)
  • [2] N. Heslehurst, L.J. Ells, H. Simpson, A Batterham, J. Wilkinson, C.D. Summerbell. Trends in maternal obesity incidence rates, demographic predictors, and health inequalities in 36,821 women over a 15-year period. BJOG. 2007;114:187-194
  • [3] N.J. Sebire, M. Jolly, J.P. Harris, et al. Maternal obesity and pregnancy outcome: a study of 287,213 pregnancies in London. Int J Obesity. 2001;25:1175-1182
  • [4] T.E. O’Brien, J.G. Ray, W.S. Chan. Maternal body mass index and the risk of preeclampsia: a systematic overview. Epidemiology. 2003;14:368-374
  • [5] M.I. Cedergren. Maternal morbid obesity and the risk of adverse pregnancy outcome. Obstet Gynecol. 2004;103:219-224
  • [6] S.Y. Chu, S.Y. Kim, C.H. Schmid, et al. Maternal obesity and risk of cesarean delivery: a meta-analysis. Obes Rev. 2007;8:385-394
  • [7] S.Y. Chu, W.M. Callaghan, S.Y. Kim, et al. Maternal obesity and risk of gestational diabetes mellitus. Diabetes Care. 2007;30:2070-2076
  • [8] S.Y. Chu, S.Y. Kim, J. Lau, et al. Maternal obesity and risk of stillbirth: a meta-analysis. Am J Obstet Gynecol. 2007;197:223-228
  • [9] S.D. McDonald, Z. Han, S. Mulla, J. Beyene, Knowledge Synthesis Group. Overweight and obesity in mothers and risk of preterm birth and low birth weight infants: systematic review and meta-analyses. BMJ. 2010;341:c3428
  • [10] A. Syngelaki, F.E. Bredaki, E. Vaikousi, N. Maiz, K.H. Nicolaides. Body mass index at 11–13 weeks’ gestation and pregnancy complications. Fetal Diagn Ther. 2011;30:250-265
  • [11] K. Tenenbaum-Gavish, M. Hod. Impact of maternal obesity on fetal health. Fetal Diagn Ther. 2013;34:1-7
  • [12] A. Karmon, E. Sheiner. Pregnancy after bariatric surgery: a comprehensive review. Arch Gynecol Obstet. 2008;277:381-388
  • [13] M.A. Maggard, I. Yermilov, Z. Li, et al. Pregnancy and fertility following bariatric surgery: a systematic review. JAMA. 2008;300:2286-2296
  • [14] M.A. Kominiarek. Preparing for and managing a pregnancy after bariatric surgery. Semin Perinatol. 2011;35:356-361
  • [15] M.G. Dalfrà, L. Busetto, N.C. Chilelli, A. Lapolla. Pregnancy and foetal outcome after bariatric surgery: a review of recent studies. J Matern Fetal Neonatal Med. 2012;25:1537-1543
  • [16] L. Vrebosch, S. Bel, G. Vansant, I. Guelinckx, R. Devlieger. Maternal and neonatal outcome after laparoscopic adjustable gastric banding: a systematic review. Obes Surg. 2012;22:1568-1579
  • [17] M.M. Kjaer, L. Nilas. Pregnancy after bariatric surgery—A review of benefits and risks. Acta Obstet Gynecol Scan. 2013;92:264-271
  • [18] E. Sheiner, K. Willis, Y. Yogev. Bariatric surgery: impact on pregnancy outcomes. Curr Diab Rep. 2013;13:19-26
  • [19] M. Clarke, R. Horton. Bringing it all together. Lancet–Cochrane collaborate on systematic reviews. Lancet. 2001;357:1728
  • [20] D. Moher, D.J. Cook, S. Eastwood, I. Olkin, D. Rennie, D.F. Stroup. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Onkologie. 2000;23:597-602
  • [21] R. DerSimonian, N. Laird. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177-188
  • [22] N. Mantel, W. Haenszel. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst. 1959;22:719-748
  • [23] M. Egger, G. Davey Smith, M. Schneider, C. Minder. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629-634
  • [24] S. Yusuf, R. Peto, J. Lewis, R. Collins, P. Sleight. Beta blockade during and after myocardial infarction: an overview of the randomized trials. Prog Cardiovasc Dis. 1985;27:335-371
  • [25] M. Egger, G.D. Smith. Misleading meta-analysis. BMJ. 1995;311:753-754
  • [26] A.C. Wittgrove, L. Jester, P. Wittgrove, G.W. Clark. Pregnancy following gastric bypass for morbid obesity. Obes Surg. 1998;8:461-464
  • [27] A.J. Skull, G.H. Slater, J.E. Duncombe, G.A. Fielding. Laparoscopic adjustable banding in pregnancy: safety, patient tolerance and effect on obesity-related pregnancy outcomes. Obes Surg. 2004;14:230-235
  • [28] J.B. Dixon, M.E. Dixon, P.E. O’Brien. Birth outcomes in obese women after laparoscopic adjustable gastric banding. Obstet Gynecol. 2005;106:965-972
  • [29] G. Ducarme, A. Revaux, A. Rodrigues, F. Aissaoui, I. Pharisien, M. Uzan. Obstetric outcome following laparoscopic adjustable gastric banding. Int J Gynaecol Obstet. 2007;98:244-247
  • [30] A.Y. Weintraub, A. Levy, I. Levi, M. Mazor, A. Wiznitzer, E. Sheiner. Effect of bariatric surgery on pregnancy outcome. Int J Gynaecol Obstet. 2008;103:246-251
  • [31] W.L. Bennett, M.M. Gilson, R. Jamshidi, et al. Impact of bariatric surgery on hypertensive disorders in pregnancy: retrospective analysis of insurance claims data. BMJ. 2010;340:c1662
  • [32] A.E. Burke, W.L. Bennett, R.M. Jamshidi, et al. Reduced incidence of gestational diabetes with bariatric surgery. J Am Coll Surg. 2010;211:169-175
  • [33] A. Lapolla, M. Marangon, M.G. Dalfrà, et al. Pregnancy outcome in morbidly obese women before and after laparoscopic gastric banding. Obes Surg. 2010;20:1251-1257
  • [34] P. Santulli, L. Mandelbrot, E. Facchiano, et al. Obstetrical and neonatal outcomes of pregnancies following gastric bypass surgery: a retrospective cohort study in a French referral centre. Obes Surg. 2010;20:1501-1508
  • [35] A. Josefsson, M. Blomberg, M. Bladh, S.G. Frederiksen, G. Sydsjö. Bariatric surgery in a national cohort of women: sociodemographics and obstetric outcomes. Am J Obstet Gynecol. 2011;205(206):e1-e8
  • [36] B. Aricha-Tamir, A.Y. Weintraub, I. Levi, E. Sheiner. Downsizing pregnancy complications: a study of paired pregnancy outcomes before and after bariatric surgery. Surg Obes Relat Dis. 2012;8:434-439
  • [37] V. Belogolovkin, H.M. Salihu, H. Weldeselasse, et al. Impact of prior bariatric surgery on maternal and fetal outcomes among obese and non-obese mothers. Arch Gynecol Obstet. 2012;285:1211-1218
  • [38] J. Lesko, A. Peaceman. Pregnancy outcomes in women after bariatric surgery compared with obese and morbidly obese controls. Obstet Gynecol. 2012;119:547-554
  • [39] D. Amsalem, B. Aricha-Tamir, I. Levi, D. Shai, E. Sheiner. Obstetric outcomes after restrictive bariatric surgery: what happens after 2 consecutive pregnancies?. Surg Obes Relat Dis. 2013;
  • [40] M.M. Kjær, J. Lauenborg, B.M. Breum, L. Nilas. The risk of adverse pregnancy outcome after bariatric surgery: a nationwide register-based matched cohort study. Am J Obstet Gynecol. 2013;208(6):464 e1–5
  • [41] N. Roos, M. Neovius, S. Cnattingius, et al. Perinatal outcomes after bariatric surgery: nationwide population based matched cohort study. BMJ. 2013;347:f6460
  • [42] D. Shai, I. Shoham-Vardi, D. Amsalem, D. Silverberg, I. Levi, E. Sheiner. Pregnancy outcome of patients following bariatric surgery as compared with obese women: a population-based study. J Matern Fetal Neonatal Med. 2014;27(3):275-278
  • [43] E. Sheiner, E. Balaban, J. Dreiher, I. Levi, A. Levy. Pregnancy outcome in patients following different types of bariatric surgeries. Obes Surg. 2009;19:1286-1292
  • [44] F.E. Bebber, J. Rizzolli, D.S. Casagrande, et al. Pregnancy after bariatric surgery: 39 pregnancies follow-up in a multidisciplinary team. Obes Surg. 2011;21(10):1546-1551
  • [45] A.M. Carelli, C.J. Ren, H.A. Youn, et al. Impact of laparoscopic adjustable gastric banding on pregnancy, maternal weight, and neonatal health. Obes Surg. 2011;21(10):1552-1558
  • [46] C.M. Dell’Agnolo, M.D. Carvalho, S.M. Pelloso. Pregnancy after bariatric surgery: implications for mother and newborn. Obes Surg. 2011;21(6):699-706
  • [47] E. Sheiner, A. Edri, E. Balaban, I. Levi, B. Aricha-Tamir. Pregnancy outcome of patients who conceive during or after the first year following bariatric surgery. Am J Obstet Gynecol. 2011;204(50):e1-e6
  • [48] R.A. Stone, J. Huffman, N. Istwan, et al. Pregnancy outcomes following bariatric surgery. J Womens Health (Larchmt). 2011;20(9):1363-1366
  • [49] E. Facchiano, A. Iannelli, P. Santulli, L. Mandelbrot, S. Msika. Pregnancy after laparoscopic bariatric surgery: comparative study of adjustable gastric banding and Roux-en-Y gastric bypass. Surg Obes Relat Dis. 2012;8(4):429-433
  • [50] I. Guelinckx, R. Devlieger, P. Donceel, et al. Lifestyle after bariatric surgery: a multicenter, prospective cohort study in pregnant women. Obes Surg. 2012;22(9):1456-1464
  • [51] G. Ducarme, L. Parisio, P. Santulli, L. Carbillon, L. Mandelbrot, D. Luton. Neonatal outcomes in pregnancies after bariatric surgery: a retrospective multi-centric cohort study in three French referral centers. J Matern Fetal Neonatal Med. 2013;26:275-278
  • [52] S.M. Han, W.W. Kim, R. Moon, R.J. Rosenthal. Pregnancy outcomes after laparoscopic sleeve gastrectomy in morbidly obese Korean patients. Obes Surg. 2013;23(6):756-759
  • [53] A. Josefsson, M. Bladh, A.B. Wiréhn, G. Sydsjö. Risk for congenital malformations in offspring of women who have undergone bariatric surgery. A national cohort. BJOG. 2013;120(12):1477-1482
  • [54] M.M. Kjaer, L. Nilas. Timing of pregnancy after gastric bypass—A national register-based cohort study. Obes Surg. 2013;23:1281-1285
  • [55] D.R. Coustan, M.W. Carpenter. The diagnosis of gestational diabetes. Diabetes Care. 1998;21(Suppl 2):B5-B8
  • [56] B.E. Metzger, D.R. Coustan. Summary and recommendations of the fourth international workshop—Conference on gestational diabetes mellitus. Diabetes Care. 1998;21(Suppl 2):B161-B167
  • [57] E. Burstein, A. Levy, M. Mazor, A. Wiznitzer, E. Sheiner. Pregnancy outcome among obese women: a prospective study. Am J Perinatol. 2008;25:561-566
  • [58] B.M. Sibai, M. Ewell, R.J. Levine, et al. Risk factors associated with preeclampsia in healthy nulliparous women. The calcium for preeclampsia prevention (CPEP) study group. Am J Obstet Gynecol. 1997;177:1003-1010
  • [59] E. Sheiner, A. Levy, D. Silverberg, et al. Pregnancy after bariatric surgery is not associated with adverse perinatal outcome. Am J Obstet Gynecol. 2004;190:1335-1340
  • [60] C.N. Hales. Barker DJ. Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia. 1992;35:595-601
  • [61] C. Levy-Marchal, D. Jaquet, P. Czernichow. Long-term metabolic consequences of being born small for gestational age. Semin Neonatol. 2004;9:67-74
  • [62] J. Bertinato, C. Aroche, L.J. Plouffe, et al. Diet-induced obese rats have higher iron requirements and are more vulnerable to iron deficiency. Eur J Nutr. 2013;53
  • [63] W. Salgado Jr., C. Modotti, C.B. Nonino, R. Ceneviva. Anemia and iron deficiency before and after bariatric surgery. Surg Obes Relat Dis. 2013;(June)10.1016/j.soard.2013.06.012 [Epub ahead of print]
  • [64] E. Fineout-Overholt, L. Gallagher-Ford, B. Mazurek Melnyk, S.B. Stillwell. Evidence-based practice, step by step: evaluating and disseminating the impact of an evidence-based intervention: show and tell. Am J Nurs. 2011;111:56-59
  • [65] S. Shapiro. Is there is or is there ain’t no baby? Dr Shapiro replies to Drs Petitti and Greenland. Am J Epidemiol. 1994;140:788-791
  • [66] D.B. Petitti. Of babies and bathwater. Am J Epidemiol. 1994;140:779-782
  • [67] S. Thompson, U. Ekelund, S. Jebb, et al. A proposed method of bias adjustment for meta-analyses of published observational studies. Int J Epidemiol. 2011;40:765-777

Footnotes

a Department of Obstetrics and Gynaecology, Whittington Hospital, London, UK

b School of Clinical Sciences, University of Nottingham, UK

c Department of General Surgery, Ealing Hospital, London, UK

d Harris Birthright Research Centre for Fetal Medicine, King's College, London, UK

lowast Corresponding author at: Department of Obstetrics and Gynaecology, Whittington Hospital, Magdala Avenue, N195NF, London, UK. Tel.: +44 020 7272 3070; fax: +44 020 7288 5550.