The effect of lifestyle interventions on maternal body composition during pregnancy in developing countries: a systematic review

Abstract Optimal maternal body composition during pregnancy is a public health priority due to its implications on maternal health and infant development. We therefore aimed to conduct a systematic review of randomised, controlled trials, and case–control and cohort studies using lifestyle interventions to improve body composition in developing countries. Of the 1 708 articles that were searched, seven studies, representing three countries (Brazil, Iran and Argentina), were included in the review. Two articles suggested that intervention with physical activity during pregnancy may significantly reduce maternal weight gain, and five studies were scored as being of poor quality. This systematic review highlights the lack of research within developing countries on lifestyle interventions for the management of excessive weight gain during pregnancy. Similar reviews from developed countries demonstrate the efficacy of such interventions, which should be confirmed using well-designed studies with appropriate intervention methods in resource-limited environments.

Both developed and developing countries are experiencing a rapid increase in the prevalence of obesity, which places affected individuals at an increased risk for a number of different diseases, including hypertension, diabetes, heart disease, asthma and cancer. 1,2 The World Health Organisation (WHO) estimated that in 2005 there were approximately 1.6 billion adults (aged 15 years and over) globally who were overweight and at least 400 million adults who were obese. 3 Especially alarming is the high prevalence of overweight and obesity among women of childbearing age in both developed and developing countries. Around 12 to 38% of pregnant women in developed countries, 5 and 8 to 26% of pregnant women in developing countries 4,6 are reported to be overweight or obese.
Obesity during pregnancy is associated with an increased risk for maternal and neonatal complications. The associated adverse maternal effects of obesity during pregnancy include miscarriage, pre-eclampsia, gestational diabetes mellitus, infection, venous thromboembolism and haemorrhage. 7 The foetal risks associated with maternal obesity include stillbirths and neonatal deaths, preterm births, congenital abnormalities and macrosomia. 8 Long-term effects of maternal obesity on the offspring have also been observed and include increased risks of childhood and adolescent obesity, and diabetes and cardiovascular disease in adult life. 9 AFRICA Both under-and overweight pose a risk to the mother and child during and after birth. 10 It is therefore important to carefully manage weight gain during pregnancy with dietary intake of a sufficient level to ensure proper foetal nutrition, 11 but avoiding excessive maternal weight gain. The use of lifestyle interventions to attenuate such weight gain during pregnancy has been the focus of many studies in the developed world, with a recent systematic review of 88 studies, involving 182 139 women, showing that maternal weight control during pregnancy via diet, exercise or a mix of these methods is safe and improves both maternal and foetal outcomes. 12 However, no similar analysis of such data from the developing world is currently available.
The health risks of maternal obesity and excessive gestational weight gain to the mother and baby pose significant demands on the healthcare system, with an increased need for additional resources in both primary and secondary care settings. 11,13 This is particularly true in developing countries where insufficient resources exist to meet these extra demands on the public health system, and where obesity is already prevalent. It is therefore important to develop cost-effective interventions to reduce maternal obesity in such environments. In an attempt to determine the effectiveness of maternal lifestyle interventions in resource-limited environments, we conducted a systematic review of the literature on weight-management protocols for pregnant females, undertaken in developing countries.

Methods
Five electronic databases were searched; these included Public/ Publisher MEDLINE (PubMed), SCOPUS, a bibliographic database containing abstracts and citations for academic journal articles, Biomed Central, the Cochrane Library and the Cumulative Index to Nursing and Allied Health (CINHAL). Twenty-four search terms, with varying combinations, encompassing pregnancy, obesity/overweight, diet/nutrition, physical activity and developing countries (Table 1), were used. The search included all articles published up to 4 December 2013. No filters were set, in order to obtain articles in all languages and all types of documents.
Randomised, controlled trials (RCTs), case-control studies and cohort studies that were performed in developing countries and which investigated overweight/obesity in pregnant women and/or lifestyle interventions during pregnancy were considered eligible. Investigations performed in developing countries and the following types of studies were not eligible for inclusion: reviews, position statements/guidelines, reports, epidemiological studies, observational studies and prevalence studies.
The literature search was performed independently by two authors, Shelley Macaulay (SM) and Estelle Watson (EW). The search results obtained from each of the five electronic databases were pooled and duplicates were removed. At the first step, titles were screened for eligibility. Following this, the abstracts of those that were considered eligible were then obtained and read. Fulltext articles of the abstracts that fulfilled the inclusion criteria were then obtained and read. In addition, the reference lists of appropriate full-text articles were hand-searched for further relevant articles.
Data were extracted from the full-text articles by four of the reviewers: Phillipe Gradidge (PG), Elena Libhaber (EL), EW and SM. For each included article, data were extracted for country, region, sample size, gestational age, BMI, intervention details and outcomes.
The quality of each included article was assessed by three authors (EW, PG and EL) using the Cochrane Risk of Bias Tool (Cochrane, 2011). In accordance with the risk-assessment checklist, each study was assessed on: sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting and other threats to validity.
The studies were classified as being good, average or poor quality based on how many of the above criteria were met. Good-quality articles met five or more of the above criteria, average-quality articles met three to four of the above criteria, and poor-quality ones met less than three of the above criteria.

Results
A total of 6 988 records were identified from the five databases, after which 5 280 duplicates were removed. The title screen therefore involved 1 708 articles, of which 73 were considered appropriate, and their abstracts were obtained and reviewed. After reviewing the abstracts, 23 full-text articles were obtained and read. In addition, the bibliography of the full-text articles were hand-searched for further appropriate articles. Six additional articles were obtained through hand-searching. Together with the hand-searched articles, a final total of seven articles were considered eligible for this systematic review. Articles that were excluded at this stage were those conducted in high-income countries and articles involving women post delivery of their babies (Fig. 1).
The results of the Cochrane Risk of Bias Tool are displayed in Table 2, highlighting the criteria for assessing quality and risk of bias. Two of the included six articles displayed adequate sequence generation, allocation concealment, and addressed incomplete outcome data. 14,15 However, only one of these reported a low risk of bias for 'blinding' of participants. 14 The final outcome of the quality assessment showed that five out of the seven articles (71%) were of poor quality.

AFRICA
The articles comprised five RCTs from Brazil and Iran and two non-RCTs from Argentina and Iran, as shown in Table 3. 11,[14][15][16][17][18][19] Characteristics of the interventions across the seven studies are detailed in Table 4. Six of the trials studied the impact of exercise alone on maternal and birth outcomes, and one study investigated using fortified food to enhance micronutrient nutritional status.
A study by Prevedel et al. 17 was one of two that used aquatic physical activity as an intervention. The relative body fat percentage of the experimental group remained at 29%, however, the control group increased by 1.9%.
A study by Cavalcante et al. 18 also used an intervention of aquatic exercise during pregnancy to determine its effectiveness on maternal outcomes. No differences were noted between control and intervention groups for weight gain during pregnancy, body fat percentage, fat-free mass or body mass index (BMI).
The effects of supervised aerobic exercise on the maternal outcomes of overweight pregnant women were evaluated by Santos et al. 14 Although oxygen consumption of the exercise group at anaerobic threshold was higher post intervention, neither groups showed any differences in weight change after the intervention.
Two Iranian interventions 16,19 evaluated the effect of landbased exercise on low-back pain during pregnancy. The typical exercise programme for these studies included a combination of midwife-supervised anaerobic and aerobic exercise performed three days per week at a moderate intensity. In the study by Garshabi et al., 19 lordosis was reduced in the exercise group after the intervention, but weight gain was similar between the study groups. In addition, spinal flexibility was significantly lower in the exercise group post intervention, and this was correlated with BMI. Weight gain was lower in the control group, and body weight of the neonate was higher than in the exercise group. Although Sedaghati et al. 16 showed intensity of low-back pain was higher in the control group, weight gain during pregnancy was higher in the exercise group.
The intervention that aimed to determine the possibility of improving maternal outcomes using fortified foods 11 found that the prevalence of folic acid and serum retinol deficiency decreased, while vitamin A deficiency remained the same post intervention. No differences were noted for body composition, and the proportions of overweight and obesity in the groups were at a moderate level of 20 and 26.3%, respectively, post intervention.

Discussion
Pregnancy appears to be a pivotal time for both maternal and foetal health. Emerging research has highlighted the profound effects of the in utero environment on the lifelong health of the baby. More specifically, both underweight and overweight babies are at risk of obesity later on in life. 20 The perinatal period has been cited by Lawlor and Chaturvedi 21 as one of the three critical periods in life for the prevention of obesity.
Maternal obesity is perhaps one of the major causes of intrauterine over-nutrition during pregnancy, and can lead to largefor-gestational-age deliveries. In addition, excessive gestational  AFRICA weight gain in both overweight and normal-weight women has been shown to increase obesity in the offspring in both childhood 22 and adolescence. 23 For the mother, obesity-related complications and gestational diabetes mellitus may predispose her to the risk of metabolic and vascular diseases later on in life. 24 Therefore, with the current epidemic of obesity, maternal obesity has serious implications on the health of both current and future generations.
Due to the potential health consequences of maternal obesity, pregnancy is a pivotal period to implement health interventions, 25 however, little research exists on health promotion during this period. 24 A previous systematic review by Thangaratinam et al. 12 found 44 randomised, controlled trials, conducted in developed countries, which implemented dietary or physical activity interventions to influence maternal weight during pregnancy. In their review, 14 studies implemented physical activity interventions, while 10 looked at dietary interventions and 10 addressed a mixed approach. Similarly, in our study, the majority of interventions focused on physical activity over dietary programmes. It is interesting to note that Thangaratinam et al. 12 found that diet was more cost effective than physical activity in the management of weight in this population.
The type of interventions used to limit weight gain during pregnancy appears to vary widely between studies. In their systematic review of lifestyle interventions in pregnancy, Oteng-Ntim et al. 26 found a variety of individual, group and seminar interventions, while in the review by Thangaratinam et al., 12 interventions varied from a balanced diet and exercise prescription to counselling and educational sessions. In our study, six out of the seven interventions focused on physical activity, while only one used a nutritional intervention.
Although previous systematic reviews have analysed the literature dealing with interventions during pregnancy for limiting excessive weight gain, 12,26,27 ours is the first review of such studies performed solely in developing countries. Changes in diet and activity levels resulting from globalisation and movement of populations from rural to urban environments have led to a rapid rise in the prevalence of obesity in developing countries, 28 and have caused this disease to move to the top of the public health agenda in many of these countries. 29 Although there have been calls to focus interventions on maternal nutrition in order to reduce the risk of obesity later on in life, 30 our review found only seven articles covering maternal obesity interventional studies, with only one specifically addressing nutrition. In addition, although the rate of obesity is high and affects many developing countries, our study showed that only three countries (Brazil, Iran and Argentina) have reported on interventions to curb obesity during pregnancy.
Despite the growing prevalence of obesity in developing countries and the well-recognised detrimental effects of maternal obesity on both maternal and foetal outcomes, this review demonstrates the lack of pertinent research in this area within developing countries. Two of the studies in our review found an increase in weight 16 and fat percentage 17 in their control groups, and weight or weight gain was often a secondary outcome measure within the studies reviewed, the majority of which (six of seven studies) were not targeted to overweight women. Other reviews have demonstrated the effectiveness and safety of lifestyle interventions for reducing gestational weight gain, 12 but this was not strongly demonstrated in the current study.
Very few studies exist to address the issues of intervention for obesity and weight gain during pregnancy in developing countries. This review summarises the existing literature, of which 71% were of poor quality. Although our review focused on lifestyle interventions for overweight and obesity during pregnancy, the search yielded only one study that aimed the intervention at overweight women, and 50% of the studies were not primarily measuring weight gain as an outcome. In addition, the studies varied significantly from type of intervention to outcome measure, and additionally, the methodology was often poorly described, making comparative and accumulative analysis difficult.

Conclusion
Dietary and lifestyle interventions during pregnancy may well be the key to addressing the prevention of obesity in future generations. 28 Physical activity 29 and dietary 12 interventions have been shown to play an effective role in maternal weight management in the developed world. To our knowledge, this review is the first to address interventions for weight gain and obesity in developing countries, and few articles appear to have addressed this important issue. Lifestyle interventions may be a cost-effective and useful way to curb the growing epidemic of nutrition-related non-communicable diseases. Despite maternal health and obesity being a public health priority, few robust studies have addressed this critical area.
This review has highlighted the need for further research, and in particular, carefully designed randomised, controlled trials, addressing primarily the issues of weight gain and obesity in pregnancy. Such studies are essential to determine the effectiveness and safety of appropriate lifestyle interventions during pregnancy in resource-limited settings.

Key messages:
• Lifestyle interventions may be a cost-effective and useful way to manage maternal overweight and obesity as well as gestational weight gain • Few good-quality studies assessing the efficacy of lifestyle interventions on maternal body composition have been conducted in developing countries • In this systematic review of seven studies, two suggested that a physical activity intervention during pregnancy may significantly reduce maternal weight gain; and five were scored as being of poor quality. • Future, well-designed lifestyle-intervention studies aimed at managing maternal body composition are much needed in developing countries