The review concluded that evidence for the effectiveness of interventions to promote healthy weight in general populations was limited, although multi-component interventions in schools and those that encouraged reduced children’s television viewing were promising strategies. Due to weaknesses and uncertainties in the review methodology and omissions in the reporting of pertinent results, these conclusions may not be reliable.

To evaluate interventions promoting healthy weight in general populations within Western countries.

Seven databases (including PubMed, EMBASE and DARE) were searched up to November 2008. Search terms were reported. References from relevant systematic reviews were also searched.

Controlled, randomised, or quasi-experimental studies that focused on the promotion of healthy weight in the general populations (not selected by weight) and provided body mass index or percentage of body fat outcome data were eligible for inclusion. Studies with a mean baseline body mass index not within normal or overweight ranges, studies that focused solely on treatment or included treatment for eating disorders, and studies that differed substantially from healthy weight promotion were excluded.

Most included studies involved only children (0 to 12 years old) or adolescents (12 to 18 years old). Most of these studies used school based interventions and did not involve the family within the intervention. Interventions used one or more of the following methods to promote healthy weight: regular exercise (physical activity); dissemination or teaching (education); change in at least one major daily meal (nutrition); and reduction in television viewing (TV). The outcome most commonly used was body mass index. Studies that assessed the percentage of body fat used one of four methods: skin fold thickness; bioelectrical impedance; dual-energy x-ray absorptiometry; or underwater weighing. The length of intervention ranged from one month to over seven years. Over half the studies were conducted in the USA.

The authors did not state how many reviewers selected studies for inclusion.

Criteria used to assess study quality were based on the Jadad scale and the Cochrane Effective Practice and Organisation of Care Group data collection checklist. Criteria assessed were: reporting of selection criteria, intervention, attrition rate, and statistical methods; prespecified power calculation; blinding; randomisation; and comparability of groups at baseline.

The authors did not state how many reviewers assessed quality.

The mean difference in outcomes between intervention and control groups and 95% confidence intervals were extracted or, if these were not available, derived through point estimates or mean outcome values. Authors were contacted if further information was required.

Data extraction was undertaken by one reviewer.

Outcomes were pooled using a random-effects meta-analysis. Pooled mean differences were calculated separately for children and adolescents (0 to 18 years) and adults (over 18 years) and grouped by outcome measured (body mass index or percentage of fat) and type of intervention. Heterogeneity was assessed using Ι².

The sensitivity of the results were analysed by changing the method of pooling, methods used to derive mean differences, adjustments used for cluster design studies, and by removing each study in turn from the results. Subgroup analyses were undertaken based on randomisation and various aspects of the study design, and characteristics of participants.

Publication bias was assessed using funnel plots.

Sixty-eight studies were included in the review (over 74,206 participants). Sample sizes ranged from 30 to 41,173 participants (details in separate online appendix). Fifty-four studies were randomised controlled trials (RCTs); most of these used a cluster design. Most studies provided a clear description of their intervention, attrition rates, and statistical analysis. A few studies reported information on an a priori power calculation, blinded outcome assessment, or the method of randomisation.

Children and adolescents

Statistically significant reductions in mean body mass index were obtained in studies that promoted: healthy weight through education only (MD -0.15kg/m², 95% CI -0.24 to -0.07; 15 studies; Ι²=65%); physical activity and education (MD -0.19kg/m², 95% CI -0.37 to -0.02; 21 studies; Ι²=94%); physical activity, education, and nutrition (MD -0.1kg/m², 95% CI -0.17 to -0.04; six studies; Ι²=5%), and TV (MD -0.27kg/m², 95% CI -0.4 to -0.13; eight studies; Ι²=20%). Results were not statistically significant for physical activity only, nutrition only, and education with nutrition.

Statistically significant reductions in the percentage of body fat was obtained in studies that promoted healthy weight through physical activity only (MD -0.7, 95% CI -1.05 to -0.31; six studies; Ι²=59%). Results were not statistically significant for education only, physical activity and education, education and nutrition, physical activity with education and nutrition, and TV only.

Adults

Statistically significant reductions in body mass index were obtained in studies that promoted healthy weight through physical activity only (MD -1.24kg/m², 95% CI -1.62 to -0.85; two studies; Ι²=0) and education only (MD -0.41kg/m², 95% CI -0.63 to -0.19; eight studies; Ι²=73%). Results were not statistically significant for nutrition only.

Statistically significant reductions in the percentage of body fat were obtained in studies that promoted healthy weight through physical activity only (MD -1.97; 95% CI -2.66 to -1.29; two studies; Ι²=0) and education only (MD -1.22, 95% CI -1.92 to -0.52; four studies; Ι²=38%).

Sensitivity analysis and subgroup analysis results were also reported.

The evidence for the effectiveness of promoting healthy weight in the general population was limited. Reductions in body composition resulting from preventive interventions of one or multiple components were at best modest. However, multi-component interventions in schools and those that encouraged reduced children’s television viewing were promising strategies.

The review question was clear. Inclusion criteria were well specified for intervention, participants, and outcomes of interest. Several databases were searched, but it was not clear whether any language restrictions were included. The search for unpublished studies was limited, so some studies may have been missed. It was unclear whether attempts were made to reduce the likelihood of reviewer error or bias in the selection of studies or assessment of study quality, but it may have been present in the extraction of data as this was performed by only one reviewer.

Appropriate methods were used to assess study quality; the results of this assessment were reported in separate online supplementary tables. Study details were also provided in a separate online supplementary table, but this was limited and details about control or comparator interventions were not reported. The pooling of the results may not have been appropriate given the statistical heterogeneity and differences in study design, interventions, and participants. However, heterogeneity was explored using appropriate methods, and the intervention and participant diversity was partly addressed by analysing the type of intervention and age group separately. The authors acknowledged that no adjustments were made for undertaking multiple statistical comparisons, which could lead to false positive results.

The authors concluded that interventions through school were promising strategies, but separate results for interventions provided through schools were not reported within the review. This, together with the weaknesses and uncertainties surrounding the methodologies used within the review means that the authors’ conclusions may not be reliable.

Practice: The authors did not state any implications for practice.

Research: The authors stated that improvements in the design of reporting of overweight and obesity prevention trials were required to ensure that changes in body composition remained comparable across studies. Long-term follow-up was needed to gain better insight into maintenance.

National Institute for Health Research (NIHR), UK; South Australian Department of Health, Strategic Health Research Grant.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.