|Interventions to prevent obesity in children and adolescents: a systematic literature review
|Flodmark C E, Marcus C, Britton M
This review evaluated the effectiveness of interventions to prevent obesity in children and adolescents. The authors concluded that obesity is preventable using limited, school-based programmes that combine the promotion of physical activity and a healthy diet. However, the authors' conclusions are not supported by the data and limitations in the review methodology suggest that they may not be reliable.
To evaluate the effectiveness of interventions to prevent obesity in children and adolescents.
The literature search was an update of a previous Swedish Health Technology Assessment (see Other Publications of Related Interest, no.1). PubMed and the Cochrane Library were searched from 2001 to May 2004; the search terms were reported. Additional studies were located through searches of recently published medical journals (unspecified) and the reference lists of relevant articles and review articles. Only studies published in Swedish, Norwegian, Danish, English, German or French were included in the review.
Study designs of evaluations included in the review
Controlled studies with a follow-up of at least 12 months were eligible for inclusion. The studies included in the review were either randomised controlled trials (RCTs) or controlled clinical trials (CCTs). The duration of follow-up ranged from 1 year to 8 years/ongoing (1 study only); most were between 1 and 3 years.
Specific interventions included in the review
Studies comparing any intervention to prevent excessive weight or obesity with a control group were eligible for inclusion. The interventions included in the review used one or more of the following components: education (i.e. health, nutrition or lifestyle), counselling, changes in diet, changes in physical activity, reminders, and contact or follow-up (e.g. with a nurse, physician or other). Most of the studies in the review compared interventions with usual care or education.
Participants included in the review
Studies that included a normal group of children and/or adolescents who had not been specially selected were eligible for inclusion; eligible age ranges for children and adolescents were not defined. Studies of high-risk groups were also eligible as long as the participants were not selected because of their high risk of obesity. All but one of the included studies recruited school children; age ranges in these studies varied between 5 and 14 years. One study recruited children aged from 7 months old. One study concerned participants at increased risk of obesity and cardiovascular disease. Four studies were based in the USA, two in Great Britain, one in Germany, one in Finland and two in Russia. Only one study evaluated participants (Mexican Americans) with an elevated risk of obesity.
Outcomes assessed in the review
The studies had to assess a relevant outcome measure, primarily percentage of obese or overweight participants, body mass index (BMI) or skinfold thickness, to be eligible for inclusion. The majority of the included studies assessed BMI; three assessed skinfold thickness and three assessed the percentage of overweight or obese participants.
How were decisions on the relevance of primary studies made?
Two independent reviewers carried out an initial selection of studies based on abstracts. However, the review did not state how the papers were finally selected, or how many reviewers performed the final selection.
Assessment of study quality
Studies were evaluated according to: the type of study design used (RCT, matched control studies, or poorly defined control studies), the length of follow-up (more than 5 years, 3 to 5 years, or 1 to 2 years), the percentage drop-out (less than 20%, 20 to 30%, or 30 to 40%) and the study size (more than 1,000 participants, 500 to 1,000 participants, or 25 to 500 participants). The studies were awarded a high, medium or low rating for each criterion. Studies with mainly low or high ratings were awarded a low or high rating (respectively) overall; all other studies were assigned a medium rating. Studies lacking essential information could have their rating lowered by one level. Two independent reviewers carried out the assessments, but the authors did not report how any disagreements were resolved.
The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction.
Interventions were classified as having a positive effect if outcomes in the intervention group decreased with respect to the control group; only statistically significant differences (p<0.05) were considered. Overall study conclusions were presented as a narrative, or as a numerical summary table with an indication of statistical significance.
Methods of synthesis
How were the studies combined?
Differences in outcome measurements precluded a statistical analysis. The studies were combined using a narrative summary based on the overall positive or negative effects of the intervention.
How were differences between studies investigated?
Some differences and similarities were reported in the data tables and text of the review. The narrative synthesis did not appear to specifically consider differences between the studies.
Results of the review
Ten studies (n=10,182) were included in the updated review: 7 RCTs (n=4,966), two with CCT follow-up, and 3 CCTs (n=5,216). Twenty-four studies (n=25,896) were included when the original assessment was combined with the updated review.
Only one study was assessed as being of high quality, 5 studies were medium quality and 4 studies were low quality.
Interventions were classified as having a positive effect on preventing obesity compared with the control groups in 3 studies (one RCT, one RCT/CCT and one CCT; number of participants unclear). No significant effects were observed in the remaining 7 studies.
Results when combined with original health technology assessment (24 studies; n=25,896). Interventions were classified as having a positive effect on preventing obesity compared with the control groups in 8 studies (32% of total participants; study designs not reported); no significant differences were reported in the remaining 16 studies. No differences in the occurrence of positive effects were reported for low-quality studies in comparison with high- and medium-quality studies.
One study, based on a RCT, estimated intervention costs to be US$4,305 per quality-adjusted life-year; this equated to a net cost-saving to society.
It was possible to prevent obesity in children and adolescents using limited, school-based programmes combining the promotion of physical activity and a healthy diet.
This review was based on an update (see Other Publications of Related Interest, no.2) of a previous Swedish Health Technology Assessment (see Other Publications of Related Interest, no.1). It answered a clear research question that was defined in terms of the outcome and study design; however, the population and intervention were only broadly defined. Several relevant sources were searched and the search terms were stated. Studies published in a range of languages were eligible, thereby reducing the risk of language bias. However, it was unclear whether appropriate attempts to locate unpublished material were made, thus introducing the possibility of publication bias. The authors did not explicitly state how papers were selected for the review, or how the data were extracted, which makes it difficult to assess the potential for errors and bias. Study validity was assessed according to four factors, but it was unclear how this validity assessment was carried out and how the cut-off points were derived for categorising studies as low, medium or high quality. The studies may also have been downgraded because of poor reporting, but it was unclear how rigorously these judgments were made and enforced. Study data were listed in tables and outcomes often summarised in terms of the overall positive or negative effect, rather than reporting individual numerical data.
Overall, given the differences in outcome measures and interventions, the authors' use of a narrative synthesis would seem appropriate, though the review's findings were based on overall positive or negative effects without apparent consideration of the differences between the studies. The authors' conclusions also seemed to contradict the primary studies and their findings that few interventions showed positive effects. Given the limited reporting of the review process and study results, and the inconsistency between the results and the authors' interpretation of them, the authors' conclusions should be treated with caution.
Implications of the review for practice and research
The authors did not state any implications for practice or further research.
Flodmark C E, Marcus C, Britton M. Interventions to prevent obesity in children and adolescents: a systematic literature review. International Journal of Obesity 2006; 30(4): 579-589
Other publications of related interest
1. Swedish Council on Technology Assessment in Health Care. Obesity - problems and interventions. A systematic review. Stockholm: the Council; 2002. Report No.: SBU-report 160. 2. Swedish Council on Technology Assessment in Health Care. Forebyggande dtgarder mot fetma. Statens beredning fur medicinsk utvurdering. Stockholm: the Council; 2005. Report No.: SBU-rapport 173.
Subject indexing assigned by NLM
Adolescent; Child; Child, Preschool; Diet; Exercise; Female; Health Promotion /methods; Humans; Male; Obesity /diet therapy /prevention & control; School Health Services
Database entry date
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.