Effectiveness of the interventions

S Thangaratinam; E Rogozińska; K Jolly; S Glinkowski; W Duda; E Borowiack; T Roseboom; J Tomlinson; J Walczak; R Kunz; BW Mol; A Coomarasamy; KS Khan

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Thangaratinam S, Rogozińska E, Jolly K, et al. Interventions to Reduce or Prevent Obesity in Pregnant Women: A Systematic Review. Southampton (UK): NIHR Journals Library; 2012 Jul. (Health Technology Assessment, No. 16.31.)

Interventions to Reduce or Prevent Obesity in Pregnant Women: A Systematic Review.

Show details

Contents

< Prev Next >

3Effectiveness of the interventions

Study selection

At the final update on 31 March 2011, 19,563 potentially relevant citations were identified from the major electronic databases to evaluate the effectiveness of weight management interventions in pregnancy for maternal and fetal outcomes. A further 23 studies were identified from the reference lists of the identified studies. In total, 88 articles were included in the review. Figure 2 shows the flow diagram of study identification, selection and exclusion.

FIGURE 2

Flow chart of study identification and selection in the effectiveness review.

A total of 56 experimental studies (40 randomised and 16 non-randomised controlled studies;⁴⁴^–⁵⁹ involving 8842 women) and 32 observational studies (26 cohort⁶⁰^–⁸⁵ and six case–control studies;⁸⁶^–⁹¹ involving 173,297 women) evaluated the effectiveness of dietary, physical activity and other lifestyle interventions in pregnancy for maternal and fetal outcomes. The 40 RCTs included 12 trials on dietary interventions,⁹²^–¹⁰³ 20 on physical activity¹⁰⁴^–¹²³ and eight on mixed approach¹²⁴^–¹³⁰ in pregnancy for the prevention or reduction of obesity. Appendix 3 provides details of the included RCTs.

Quality of included studies

Randomised controlled trials

Figure 3 demonstrates the risk of bias of the included RCTs in the seven domains. Two-thirds of studies scored a low risk of bias for selective reporting of outcomes and blinding for objective outcomes. Although there was no obvious evidence of a high risk of bias for sequence generation, allocation concealment and blinding for subjective outcomes, a large proportion of the studies were unclear in their reporting in these domains. Appendix 4 provides a detailed quality assessment of the individual RCTs.

FIGURE 3

Quality assessment of the included RCTs.

Non-randomised studies and observational studies

The internal validity of NRSs has been assessed in line with the NICE checklist.³⁴ Figure 4 presents the quality of the included NRSs. Further details of the individual study quality for non-randomised and observational studies are provided in Appendices 5 and 6. The observational studies were evaluated using the NOS and could score a maximum of nine stars, with four stars for selection, two for comparison and three for outcome assessment. In total, 7/26 (26.9%) cohort studies had a low risk of bias and scored seven or more stars, 18/26 (69.2%) had a medium risk of bias and scored between four and six stars and one study (3.8%) had a high risk of bias (see Appendix 6). All six case–control studies had a medium risk of bias.

FIGURE 4

Quality assessment of the included NRSs.

Effect of the interventions on weight-related outcomes

Maternal weight-related outcomes

Maternal weight gain in pregnancy

A total of 30 RCTs¹⁷^,⁹³^–⁹⁶^,⁹⁹^–¹⁰⁵^,¹⁰⁷^–¹⁰⁹^,¹¹¹^–¹¹⁴^,¹¹⁶^–¹²⁰^,¹²³^,¹²⁴^,¹²⁶^–¹²⁸^,¹³⁰ including 4503 women evaluated the effect of interventions on maternal weight gain in pregnancy. This included nine⁹³^–⁹⁶^,⁹⁹^–¹⁰³ trials on dietary interventions, six¹⁷^,¹²⁴^,¹²⁶^–¹²⁸^,¹³⁰ on mixed approach and 15¹⁰⁴^,¹⁰⁵^,¹⁰⁷^–¹⁰⁹^,¹¹¹^–¹¹⁴^,¹¹⁶^–¹²⁰^,¹²³ on physical activity interventions. There was a significant decrease in weight gain in pregnancy with interventions of 0.97 kg (95% CI −1.60 kg to −0.34 kg; p = 0.003; I² = 87%). The largest reduction in weight gain was observed in the dietary intervention studies, with a MD of −3.36 kg (95% CI −4.73 kg to −1.99 kg; p < 0.00001; I² = 91%), followed by mixed approach, with a MD of −0.57 kg (95% CI −1.60 kg to 0.65 kg; p = 0.27; I² = 35%). The studies were heterogeneous with an I² of 87%. There was a statistically significant difference between the intervention groups (p = 0.0005) (Figure 5).

FIGURE 5

Effect of weight management interventions on maternal weight gain in pregnancy. SD, standard deviation. a, SD: from Briley 2002. b, Normal weight women. c, SD: average from Asbee 2009 and Jeffries 2009. d, Overweight women. e, SD: average from Baciuk (more...)

Maternal body mass index at delivery

Three RCTs⁹⁹^,¹⁰⁴^,¹¹³ reported on the effect of interventions on the mother's BMI at delivery. There was a significant reduction in BMI with dietary intervention, with a MD of −1.00 kg/m² (95% CI −1.67 kg/m² to −0.33 kg/m²; p = 0.003). This effect was not observed with interventions based on physical activity. The overall pooled estimate showed a MD of −0.23 kg/m² (95% CI −1.4 kg/m² to 0.94 kg/m²; p = 0.70) with a heterogeneity of I² = 58%. There was a significant difference between the subgroups (p = 0.04) (Figure 6).

FIGURE 6

Effect of weight management interventions on maternal BMI at delivery. SD, standard deviation.

Exceeding the Institute of Medicine's recommendations on weight gain in pregnancy

The IOM guidelines¹³¹ recommend the optimum weight gain in pregnancy for American women based on their BMI. The recommended gestational weight gain is 11.5–16.0 kg in women with normal BMI (BMI 18.5–24.9 kg/m²), 7.0–11.5 kg in overweight women (BMI 25–29.9 kg/m²) and 5.0–9.0 kg in obese women (BMI ≥ 30 kg/m²). Two RCTs¹²⁸^,¹³⁰ reported a reduction in the number of women exceeding IOM recommendations with a dietary and physical activity intervention, which was not statistically significant (Figure 7).

FIGURE 7

Effect of weight management interventions on IOM recommendations. a, Women with normal weight. b, Overweight women.

Fetal and neonatal weight-related outcomes

Birthweight

A total of 28 RCTs (4573 newborns) evaluated the effect of the interventions on the birthweight of the newborn. This included nine RCTs on dietary interventions,⁹⁴^–⁹⁶^,⁹⁸^–¹⁰³ five on a mixed approach intervention¹²⁵^–¹²⁸^,¹³⁰ and 14 on physical activity-based interventions.¹⁰⁴^,¹⁰⁵^,¹⁰⁷^,¹⁰⁸^,¹¹⁰^,¹¹³^–¹¹⁶^,¹¹⁸^,¹¹⁹^,¹²²^,¹³² Overall, there was a small, but statistically significant, reduction in the mean birthweight of 0.07 kg (95% CI −0.14 kg to −0.01 kg; p = 0.03). There was heterogeneity observed among the groups (I² = 68%), with no large birthweight reduction in the three intervention subgroups (Figure 8).

FIGURE 8

Effect of weight management interventions on birthweight. SD, standard deviation. a, SD(EXP): from Landon 2009. b, Women with normal weight; SD: average from Hui 2006¹⁶⁷ and Jeffries 2009. c, Overweight women; SD: from Guelinckx 2010. d, SD: average from (more...)

Large for gestational age at birth

We defined LGA infants as those above the 90th centile or with a birthweight > 4 kg. Twelve RCTs⁹⁶^,⁹⁷^,⁹⁹^,¹⁰¹^,¹⁰²^,¹⁰⁵^,¹¹⁸^,¹²⁵^–¹²⁸^,¹³⁰ evaluated this outcome in 3021 newborns. There was a 27% reduction (RR 0.73, 95% CI 0.54 to 0.99; p = 0.05) in the risk of having a LGA newborn. The results were not heterogeneous, with an I² of 33% (p = 0.13). This reduction in the incidence of LGA infants was observed with all interventions in pregnancy (Figure 9). Five RCTs reported the effects of the interventions on obese and overweight women. There was no significant difference in the incidence of LGA infants between the experimental and control groups of obese and overweight women (RR 1.32, 95% CI 0.55 to 3.16; p = 0.54; I² = 78%).

FIGURE 9

Effect of weight management interventions on the incidence of LGA infants. a, Women with normal weight.

Small for gestational age at birth

Small-for-gestational-age newborns were defined as those with a birthweight below the 10th centile or < 2.5 kg. This outcome served the dual purpose of assessment of the beneficial effect of the intervention and assessment of any adverse effect of the intervention on fetal weight. Eight RCTs⁹⁶^,⁹⁸^,⁹⁹^,¹⁰⁴^,¹⁰⁵^,¹¹⁹^,¹²⁸^,¹³⁰ (2901 newborns) evaluated the effectiveness of the weight management interventions for this outcome. The summary estimate of the RCTs showed no difference in the incidence of SGA infants with a RR of 0.99 (95% CI 0.76 to 1.29). The studies were homogeneous. The effect was consistently observed with all three interventions (Figure 10).

FIGURE 10

Effect of weight management interventions on the incidence of SGA infants. a, Overweight women. b, Women with normal weight.

Ponderal index

The ponderal index for newborns assesses the relationship between the weight of the newborn and its length (kg/m³). Four RCTs¹⁰⁵^,¹⁰⁷^,¹⁰⁸^,¹¹³ (333 newborns) evaluated the effect of the weight management interventions on the ponderal index. The summary estimate of the trials showed no significant difference in ponderal index of the newborns between the intervention and the control groups, with a MD of −0.09 kg/m³ (95% CI −0.18 to 0.00 kg/m³, I² = 72%) (Figure 11).

FIGURE 11

Effect of weight management interventions on ponderal index. SD, standard deviation.

Fetal fat mass

Fetal fat mass in kilograms was reported in four trials.⁹⁵^,⁹⁹^,¹⁰⁷^,¹⁰⁸ Dietary interventions resulted in a significant reduction in fetal fat mass in the intervention group, with a MD of −0.04 kg (95% CI −0.06 kg to −0.01 kg; p = 0.005; I² = 0%) (Figure 12).

FIGURE 12

Effect of weight management interventions on fetal fat mass. SD, standard deviation.

Effect of the interventions on obstetric maternal outcomes

Gestational diabetes mellitus

Five RCTs (involving 675 women) reported on the effect of weight management interventions on gestational diabetes mellitus (GDM). Three studies included only obese or overweight pregnant women for the evaluation of a dietary intervention (two RCTs¹⁰²^,¹⁰³) and a mixed approach-based intervention (one RCT¹³⁰). There was an overall reduction in the incidence of GDM of 29% (RR 0.71, 95% CI 0.44 to 1.13; p = 0.15), which was not statistically significant (Figure 13). Weight management interventions in obese and overweight women showed a reduction of 42% (RR 0.58, 95% CI 0.30 to 1.09; p = 0.09). The findings were homogeneous (I² =0) across studies and did not reach statistical significance.

FIGURE 13

Effect of weight management interventions on GDM. a, Women with normal weight. b, Overweight women.

Pre-eclampsia

Ten studies⁹⁶^,⁹⁸^,⁹⁹^,¹⁰¹^–¹⁰³^,¹²²^,¹²⁶^,¹²⁸^,¹³⁰ (involving 3072 women) reported the effect of weight management interventions on the incidence of pre-eclampsia. There was an overall statistically significant reduction in pre-eclampsia of 26% (RR 0.74, 95% CI 0.59 to 0.92; p = 0.008; I² = 22%). The largest reduction in pre-eclampsia (33%) was observed with dietary intervention (RR 0.67, 95% CI 0.53 to 0.85; p = 0.0009) with no heterogeneity (I² = 0). A similar effect was not observed with physical activity-based intervention or a mixed approach (Figure 14). Six studies included only obese and overweight women and showed a significant reduction in pre-eclampsia with the interventions (RR 0.65, 95% CI 0.44 to 0.97; p = 0.04; I² = 0).

FIGURE 14

Effect of weight management interventions on the incidence of pre-eclampsia. a, Overweight women.

Gestational hypertension

Gestational hypertension was evaluated as an outcome in six RCTs.¹⁰²^,¹⁰³^,¹²²^,¹²⁶^,¹²⁸^,¹³⁰ There was a reduction in gestational hypertension with interventions, which was not statistically significant (RR 0.77, 95% CI 0.54 to 1.1; I² = 37%) (Figure 15). Dietary intervention (two RCTs)¹⁰²^,¹⁰³ in pregnancy showed the greatest benefit by reducing gestational hypertension by 70% (RR 0.30, 95% CI 0.10 to 0.88; p = 0.03), with homogeneity between the studies (I² = 0). Both of the studies on dietary intervention were undertaken in obese and overweight women. The four studies on obese and overweight women¹⁰²^,¹⁰³^,¹²⁶^,¹³⁰ showed a reduction in gestational hypertension incidence that was not significant (RR 0.70, 95% 0.30 to 1.16; p = 0.4).

FIGURE 15

Effect of weight management interventions on the incidence of gestational hypertension. a, Women with normal weight. b, Overweight women.

Preterm delivery

Eleven RCTs (involving 2198 women)⁹⁴^,⁹⁸^,⁹⁹^,¹⁰²^,¹⁰⁴^,¹⁰⁵^,¹¹⁸^,¹¹⁹^,¹²⁵^,¹²⁸^,¹³⁰ evaluated the effectiveness of weight management interventions in pregnancy on preterm delivery before 37 weeks of gestation. There was no overall difference in the rates of preterm births between the two groups, with a RR of 0.76 (95% CI 0.56 to 1.02) (Figure 16). The studies were homogeneous (I² = 0%). The four RCTs⁹⁴^,⁹⁸^,⁹⁹^,¹⁰² that evaluated a dietary intervention (n = 1474) showed a significant reduction in preterm births of 32% (RR 0.68, 95% CI 0.48 to 0.96; p = 0.03; I² = 35%). Four RCTs⁹⁹^,¹⁰²^,¹¹⁹^,¹³⁰ (involving 1305 women) including obese and overweight women showed a reduction in preterm births that was not statistically significant (RR 0.80, 95% CI 0.53 to 1.13; p = 0.21, I² = 0%).

FIGURE 16

Effect of weight management interventions on preterm delivery before 37 weeks of gestation. a, Overweight women. b, Women with normal weight.

Gestational age at delivery

A total of 20 RCTs⁹⁶^,⁹⁸^–¹⁰⁵^,¹⁰⁷^,¹⁰⁸^,¹¹⁰^,¹¹¹^,¹¹³^–¹¹⁶^,¹²⁰^,¹²⁵^–¹²⁷ (4028 women) evaluated the effect of the interventions on the gestational age at delivery. There were no significant differences in the gestational age at delivery between the intervention and control groups, with a MD of 0.03 weeks (95% CI −0.13 weeks to 0.07 weeks; I² = 33%) (Figure 17). There was low heterogeneity between studies (I² = 33%). Dietary intervention (six RCTs, involving 2625 women) resulted in a MD in the gestational age at delivery of 0.05 weeks (95% CI −0.18 weeks to 0.08 weeks; p = 0.42; I² = 71%).

FIGURE 17

Effect of weight management interventions on gestational age at delivery. SD, standard deviation. a, Women with normal weight. b, Overweight women. c, Data from Kramer 2006 review.

Mode of delivery

The rate of caesarean section was evaluated as an outcome in 14 RCTs⁹⁶^,⁹⁷^,⁹⁹^,¹⁰²^–¹⁰⁴^,¹¹⁴^–¹¹⁶^,¹²⁴^–¹²⁶^,¹²⁸^,¹³⁰ involving 3312 women. This included five trials⁹⁶^,⁹⁷^,⁹⁹^,¹⁰²^,¹⁰³ on dietary interventions, four¹⁰⁴^,¹¹⁴^–¹¹⁶ on physical activity-based interventions and five¹²⁴^–¹²⁶^,¹²⁸^,¹³⁰ on a mixed approach. There were no differences between the experimental and the control groups with any intervention. The summary estimate for caesarean section was a RR of 0.93 (95% CI 0.85 to 1.03; p = 0.15) (Figure 18). There was no significant heterogeneity between the groups (p = 0.22, I² = 21%). A total of 6 of the 14 RCTs involved obese and overweight women and showed no change in the rate of caesarean section (RR 0.97, 95% CI 0.73 to 1.28; I² = 61%).

FIGURE 18

Effect of weight management interventions on rate of caesarean section. a, Overweight women. b, Women with normal weight.

The rate of vaginal delivery was evaluated in five RCTs.⁹⁹^,¹⁰¹^,¹⁰⁴^,¹¹⁵^,¹²⁵ There was no difference in the rate of vaginal delivery with any intervention. The pooled estimate showed a RR of 1.00 (95% CI 0.94 to 1.07; p = 1). The studies were homogeneous (Figure 19). The effect of dietary intervention on vaginal delivery in obese and overweight mothers was studied in two RCTs.⁹⁹^,¹⁰¹ The rate of vaginal delivery did not change with the intervention, with a RR of 0.97 (95% CI 0.89 to 1.07; I² = 0).

FIGURE 19

Effect of weight management interventions on rate of vaginal delivery.

Induction of labour

The effect of weight management interventions in pregnancy on induction of labour was studied in five RCTs (involving 2362 women).⁹⁶^,⁹⁹^,¹⁰¹^,¹⁰²^,¹²⁶ There was a slight increase in induction of labour in the intervention arm that was not significantly different from that of the control arm (RR 1.12, 95% CI 1.00 to 1.26; p = 0.05; I² = 47%) (Figure 20). Obese and overweight women only were included in four RCTs⁹⁹^,¹⁰¹^,¹⁰²^,¹²⁶ (involving 1362 women); in these studies there was no difference in the rate of induction of labour between the intervention and control groups (RR 0.99, 95% CI 0.84 to 1.16; I² = 0%).

FIGURE 20

Effect of weight management interventions on induction of labour.

Post-partum haemorrhage

Two RCTs⁹⁶^,¹⁰² (n = 1232) compared the rates of post-partum haemorrhage between the weight management intervention group and the control group. The pooled estimate of the studies did not show any significant differences between the groups (RR 0.90, 95% CI 0.57 to 1.42; I² = 0%) (Figure 21).

FIGURE 21

Effect of weight management interventions on post-partum haemorrhage.

Two observational case–control studies⁷⁷^,⁷⁸ studied the effect of physical activity-based interventions on post-partum haemorrhage and found no difference between the intervention and control groups.

Low back pain

Low back pain was reported as an outcome in two RCTs¹¹¹^,¹²⁶ (involving 302 women) evaluating physical activity-based interventions. The severity of low back pain was increased in one study¹¹¹ and decreased in the other study.¹²⁰ The pooled estimate did not show any differences in back pain between the two groups (MD 0.16, 95% CI −10.16 to 10.48; I² = 97%) (Figure 22).

FIGURE 22

Effect of weight management interventions on low back pain in pregnancy. SD, standard deviation.

Effect of the interventions on fetal and neonatal morbidity and mortality

Shoulder dystocia

Four RCTs⁹⁶^,⁹⁹^,¹⁰¹^,¹²⁸ (2317 newborns) evaluated the effect of interventions (three dietary⁹⁶^,⁹⁹^,¹⁰¹ and one mixed¹²⁸ approach) on the incidence of shoulder dystocia. Overall, there was a 61% reduction in the incidence of shoulder dystocia (RR 0.39, 95% CI 0.22 to 0.70; p = 0.02). The studies were homogeneous (I² = 0%). The largest proportion of women in the analysis were in the dietary intervention group, which showed a similar effect (Figure 23). This beneficial effect was increased in the population of obese and overweight women (RR 0.33, 95% CI 0.14 to 0.74; p = 0.008).

FIGURE 23

Effect of weight management interventions on shoulder dystocia.

Intrauterine death

Two RCTs⁹⁶^,⁹⁸ (involving 1320 women) evaluated the effect of dietary intervention on stillbirths. There was a reduction in the incidence of intrauterine death, which was not statistically significant (RR 0.15, 95% CI 0.02 to 1.20; p = 0.07; I² = 0%) (Figure 24).

FIGURE 24

Effect of weight management interventions on intrauterine death.

One observational cohort study by Perichart et al.⁸² evaluated the effect of a dietary intervention compared with no intervention on intrauterine death. There were no significant differences between the groups. This effect was consistent for women with type 2 diabetes [unadjusted odds ratio (OR) 0.96, 95% CI 0.12 to 1.09] or GDM (unadjusted OR 1.00, 95% CI 0.06 to 16.57).

Respiratory distress syndrome

Two RCTs⁹⁶^,⁹⁹ (involving 1962 women) evaluated respiratory distress syndrome with the newborn in mothers undergoing a weight management intervention in pregnancy. The two studies were on dietary interventions and the pooled estimate did not show a difference between the intervention and control groups (RR 1.05, 95% CI 0.48 to 2.28; I² = 58%) (Figure 25).

FIGURE 25

Effect of weight management interventions on respiratory distress syndrome.

Admission to the neonatal intensive care unit

Admission to NICU was reported as an outcome in two RCTs⁹⁶^,⁹⁹ (involving 1962 women) evaluating dietary interventions. The studies were heterogeneous (I² = 77%) and the pooled estimate did not show any difference between the groups (RR 0.98, 95% CI 0.66 to 1.47) (Figure 26). One observational study⁸² evaluating a dietary intervention in pregnancy reported on NICU admission in two groups: women with type 2 diabetes and those with GDM. The reported unadjusted OR was significant only in the case of women with type 2 diabetes (OR 0.21, 95% CI 0.03 to 0.51).

FIGURE 26

Effect of weight management interventions on admission to NICU.

Apgar scores

Apgar scores were evaluated as an outcome in six RCTs⁹⁶^,¹⁰²^,¹⁰⁵^,¹¹⁵^,¹¹⁶^,¹²⁸ studying the effect of weight management interventions in pregnancy. Three studies⁹⁶^,¹⁰²^,¹²⁸ reported scores of < 7 at 5 minutes and three studies¹⁰⁵^,¹¹⁵^,¹¹⁶ provided the scores at 5 minutes for comparison. There were no differences in the abnormal scores (< 7 at 5 minutes) (RR 0.64, 95% CI 0.27 to 1.49; p = 0.3, I² = 0%; Figure 27) or in the mean scores (MD 0.0, 95% CI −0.05 to 0.05; p = 0.94; Figure 28) between the two groups.

FIGURE 27

Effect of weight management interventions on abnormal Apgar scores (< 7 at 5 minutes).

FIGURE 28

Effect of weight management interventions on Apgar scores at 5 minutes. SD, standard deviation.

Infant hypoglycaemia

Hypoglycaemia in the first few days after birth is defined as blood glucose < 40 mg/dl. In preterm infants, repeated blood glucose levels of < 50 mg/dl may be associated with neurodevelopmental delay. Five RCTs⁹⁶^,⁹⁹^,¹⁰¹^,¹²⁵^,¹²⁸ reported the rate of hypoglycaemia among the children of studied mothers. Neither a comprehensive approach nor dietary interventions had any significant influence on hypoglycaemia rate (Figure 29).

FIGURE 29

Effect of weight management interventions on infant hypoglycaemia.

Infant hyperbilirubinaemia

Two RCTs⁹⁶^,⁹⁹ evaluated the effect of dietary interventions on the rates of hyperbilirubinaemia in 1898 newborns. The studies were homogeneous. There was a trend towards a reduction in hyperbilirubinaemia with the interventions, which was not significant (Figure 30).

FIGURE 30

Effect of weight management interventions on infant hyperbilirubinaemia.

Birth trauma

Two RCTs⁹⁶^,⁹⁹ evaluated the effect of dietary interventions on the risk of birth trauma. The studies showed a reduction in the risk of birth trauma (RR 0.36, 95% CI 0.11 to 1.23; I² = 0%), which was not statistically significant (Figure 31).

FIGURE 31

Effect of weight management interventions on birth trauma.

Effect of interventions on neonatal anthropometric measurements at birth

Child's birth length

Five RCTs⁹⁵^,¹⁰³^,¹⁰⁵^,¹²⁵^,¹²⁶ (323 newborns) evaluated the birth length of the newborn. The birth length of the newborn was reduced with the interventions, but the difference was not statistically significant (Figure 32).

FIGURE 32

Effect of weight management interventions on birth length. SD, standard deviation.

Abdominal circumference of the newborn

Two RCTs¹⁰³^,¹⁰⁷ evaluated the effect of dietary weight management interventions on abdominal circumference in 62 newborns. The studies were heterogeneous and overall there was no significant change in the intervention group in comparison with the control group (MD −1.26 cm, 95% CI −3.71 cm to 1.19 cm; p = 0.31; I² = 91%) (Figure 33).

FIGURE 33

Effect of weight management interventions on abdominal circumference. SD, standard deviation.

Crown–heel length

Three RCTs¹⁰⁷^,¹⁰⁸^,¹¹³ evaluated the effect of physical activity based weight management interventions on crown–heel length in 181 newborns. The studies were heterogeneous and overall there was no significant change in the intervention group in comparison with the control group (MD −0.18 cm, 95% CI −1.80 cm to 1.44 cm; p = 0.83; I² = 92%) (Figure 34).

FIGURE 34

Effect of weight management interventions on crown–heel length. SD, standard deviation.

Subgroup analyses

Subgroup analyses on the basis of period of publication, country of study (developed vs developing), GDM status and risk of bias from allocation concealment showed no differences in the summary estimates of gestational weight gain, birthweight and incidence of LGA and SGA infants. The type of intervention resulted in significant differences (p = 0.003) between the groups for weight gain in pregnancy, with the maximum reduction in gestational weight gain seen in the dietary intervention group (MD −3.36 kg, 95% CI −4.73 kg to −1.99 kg). Women with diabetes in pregnancy showed a significant reduction in the incidence of pre-eclampsia with weight management interventions (RR 0.65, 95% CI 0.50 to 0.84) compared with women without diabetes (RR 1.16, 95% CI 0.70 to 1.93), and the difference in the summary estimates between the groups was statistically significant (p = 0.04). There was a significant reduction in pre-eclampsia in the responders – women with significantly reduced gestational weight gain with intervention (RR 0.61, 95% CI 0.47 to 0.79) – compared with the group with no significant change in weight (RR 1.33, 95% CI 0.84 to 2.11) (p = 0.004). There was a significant difference between the responders (MD −0.29 kg, 95% CI −0.46 kg to −0.12 kg) and non-responders (MD −0.02 kg, 95% CI −0.06 kg to −0.03 kg) for birthweight (p = 0.002). Subgroup analysis of the summary estimates of birthweight and incidence of LGA and SGA infants did not show a statistically significant difference according to the type of intervention (Table 6).

TABLE 6

Subgroup analyses for trial methodology, clinical characteristics and publication for maternal and fetal outcomes in the evaluation of weight management interventions in pregnancy.

Sensitivity analysis that excluded studies on women with diabetes in pregnancy consistently showed a overall reduction in gestational weight gain with interventions (MD −0.88 kg, 95% CI −1.85 kg to 0.09 kg; p = 0.001), including diet (MD −5.18 kg, 95% CI −9.44 kg to −0.91 kg; p < 0.00001) and physical activity (MD −0.07 kg, 95% CI −1.08 kg to 0.93 kg; p < 0.00001). The reduction in birthweight with intervention persisted (MD −0.08 kg, 95% CI −0.16 kg to 0.0 kg; p = 0.04) with no differences in the incidence of SGA and LGA infants or shoulder dystocia between the groups. The estimates of other studies for the effect of diet on the incidence of gestational hypertension, preterm birth, vaginal delivery, caesarean section and SGA infants were similar after excluding studies on women with diabetes. There was a trend towards a reduction in the incidence of pre-eclampsia with diet in these studies.

Summary

This review on the effectiveness of weight management interventions has identified a large number of RCTs, especially for the primary weight-related outcomes in the mother and the fetus. Two-thirds of the included studies showed a low risk of bias for addressing incomplete outcome data, selective reporting and blinding for objective outcomes. Fewer than one-sixth of the studies showed a high risk of bias for addressing incomplete outcome data and selective reporting. The commonly reported outcomes were maternal weight gain in pregnancy and birthweight of the newborn.

Weight management interventions in pregnancy resulted in a statistically significant reduction in weight-related outcomes such as maternal weight gain in pregnancy, and birthweight of the newborn. However, there were no differences between the intervention and control groups for incidence of SGA fetuses. Although we did not observe a beneficial effect of reduction in growth restriction in the babies with intervention, it was a reassuring finding because there have been concerns over fetal weight reduction with weight management interventions.

There was a significant decrease in the rates of key obstetric outcomes such as pre-eclampsia and shoulder dystocia in the analysis of outcomes for all interventions. It is likely that this reduction in shoulder dystocia will be of greatest benefit in women with GDM or pre-existing diabetes. There was a trend towards a reduction in the rates of obstetric complications such as GDM, gestational hypertension and preterm birth before 37 weeks with weight management interventions.

Of the three interventions, dietary intervention showed the most beneficial effect by significantly reducing rates of obstetric complications such as gestational hypertension, preterm births, pre-eclampsia and shoulder dystocia. The significant reduction in the rate of preterm births with dietary interventions is likely to be reflected in the finding of increased gestational age with dietary interventions. For fetal outcomes the evidence was limited to dietary interventions only and showed a trend towards a reduction in rates of intrauterine deaths, birth trauma and hyperbilirubinaemia.

The dietary components of the interventions evaluated a balanced diet of carbohydrates, fat and protein, moderate energy and caloric restriction based on individual requirements, low-fat and -cholesterol diets and the use of a food diary for monitoring. The physical activity-based interventions included weight-bearing sessions, walking for 30 minutes a day and low-intensity resistance training. The mixed approach group included dietary and physical activity interventions with associated in-depth behavioural risk assessments and tailored counselling.

The main strengths of the effectiveness review were the peer-reviewed protocol, the comprehensive search strategy without any language restrictions and the use of randomised data to draw inferences. Non-randomised data were included only when there was a paucity of evidence. This review has identified the largest body of evidence on this topic, for both weight-related outcomes and clinically relevant obstetric and fetal outcomes. Dietary interventions in pregnancy have consistently shown a beneficial effect on weight-related, obstetric and fetal and neonatal outcomes compared with other interventions. The review findings are limited by the lack of detail about the components of the intervention in some of the included studies, gestational age at which the intervention was commenced, its frequency and the method of delivery. Furthermore, there are very few studies for important clinical outcomes such as intrauterine death, maternal admission to the high-dependency unit (HDU) and neonatal admissions to NICU. There are no data available to assess the long-term effects of these outcomes on the mother and the fetus.

[Note: The results of this systematic review for effectiveness of weight management interventions in pregnancy includes only studies published before March 2011. The findings with the updated search (until January 2012) can be accessed at BMJ 2012;344:e2088 doi10.1136/bmj.e2088.]

Included under terms of UK Non-commercial Government License.

Bookshelf ID: NBK109459

Contents