The review found that moderate intensity exercise programmes of six to 12 months duration were associated with modest improvements in weight, waist circumference and cardiovascular risk in overweight and obese populations. These conclusions were of questionable reliability, mainly due to the rather restricted search, the small number and limited quality of the included studies and heterogeneity between the studies.

To evaluate the effectiveness of isolated aerobic exercise for weight loss in overweight and obese populations.

The Cochrane Library and MEDLINE were searched for studies published in English or French from 1970 to January 2010. Google Scholar was used to searched for systematic reviews. Search terms were reported. Reference lists of relevant articles were checked.

Eligible studies were randomised controlled trials (RCTs) that compared aerobic exercise (without concurrent caloric restriction or weight loss drugs) versus inactivity in adults (aged over 18 years) with a body mass index of at least 25kg/m²). Control participants were instructed not to change their current exercise habits; medical attention and stretching were acceptable as control interventions. Exercise interventions were required to be of over 12 weeks duration and at least 120 minutes per week. Exercise intensity (such as percentage heart rate reserve) had to be reported and at least one measure of obesity (such as percentage fat) had to be reported in numerical form. Outcomes of interest in the review included anthropometric measures, cardiovascular risk factors, lipids and safety endpoints. Studies of participants with concurrent conditions unrelated to obesity (such as cancer, pregnancy) were excluded.

The age of participants in the included studies ranged from 19 years to over 70 years. Most studies were of all men or all women. More than half of the studies were restricted to sedentary individuals. Other studies had various inclusion criteria (such as diabetes, hypertension, employment as policeman or fireman). All participants were overweight or obese. Where reported, baseline mean body mass index ranged from 27 to 32.7kg/m² and weight ranged from 53.5 to 94.9kg. In the few studies that reported baseline blood pressure, all participants were hypertensive or pre-hypertensive. Intervention durations ranged from 12 weeks to 12 months. Exercise interventions varied and included walking, cycling, cycle ergometer and treadmill. In most studies the intervention group could choose their modality of exercise. Duration of exercise ranged from 120 to 240 minutes per week at an intensity of 40% to 80% of maximum heart rate and 40% to 70% of maximum oxygen uptake. In most cases control groups were asked to continue their current lifestyle habits. Outcomes reported in the review included change in weight, waist circumference, blood pressure and cholesterol triglyceride level.

The authors did not state how many reviewers performed study selection.

Sequence generation, allocation concealment, blinding of assessment, completeness of outcome data and selective outcome reporting were evaluated. Study quality was rated as high, unclear or low. Studies that did not use intention-to-treat analysis were excluded.

The authors did not state how many reviewers conducted the assessment.

Mean differences in change from baseline in each group were extracted or calculated with standard deviations (where reported) at the longest duration of follow-up. Studies were grouped by follow-up (12 to 16 weeks, six months and 12 months).

Two reviewers extracted data in duplicate and resolved discrepancies by consensus or with help of a third reviewer.

Data were combined to calculate pooled weighted mean differences (WMDs) and 95% confidence intervals using a random-effects DerSimonian Laird model. Data for studies with 12 to 16 week follow-up were combined in a narrative synthesis: these studies could not be pooled statistically as standard deviations for change scores were not available.

Fourteen RCTs were included in the review (1,540 participants, range 13 to 561). There was a discrepancy between text and tables for total participants. Only one study described sequence generation, three described allocation concealment, four blinded either outcome measurement or analysis, seven reported the reasons for participant loss to follow-up and 13 were free of selective outcome reporting. Only six studies reported sufficient data to enable statistical pooling.

An exercise intervention of 12 to 16 weeks was associated with a difference in weight change in the intervention group (relative to controls) that ranged from a gain of 0.8kg to a loss of 2.5kg (six RCTs) and a change in waist circumference in the intervention group that ranged from a gain of 0.9cm to a loss of 0.5cm (two RCTs).

When studies were pooled, a six-month exercise intervention was associated with significantly greater reduction in weight (WMD -1.6kg, 95% CI -1.64 to -1.56; three RCTs), waist circumference (WMD -2.12cm, 95% CI -2.81 to -1.44; two RCTs), systolic blood pressure (WMD -2.92mmHg, 95% CI -4.76 to -1.08; two RCTs) and diastolic blood pressure (WMD -1.8mmHg, 95% CI -3.43 to -0.16) than control conditions.

When studies were pooled, a 12-month exercise intervention was associated with a significantly greater reduction in weight (WMD -1.7kg, 95% CI -2.29 to -1.11; two RCTs), waist circumference (WMD -1.95cm, 95% CI -3.62 to -0.29; two RCTs) and systolic blood pressure (WMD - 1.70mmHg, 95% CI -2.29 to -1.11; one RCT) than control conditions.

Moderate intensity exercise programmes of six to 12 months duration were associated with modest improvements in weight, waist circumference and cardiovascular risk in overweight and obese populations. Such programmes were of limited value as an independent intervention for weight loss in this population but may be effective in combination with diet.

The objectives and inclusion criteria of the review were clear. Relevant sources were searched for studies. Relatively few databases were searched and the search was limited by language with one study excluded for this reason. The search was apparently limited to published studies and the risk of publication bias was not formally assessed. It was unclear whether appropriate steps were taken to limit potential for reviewer bias and error during study selection and quality assessment. The review was restricted to studies that used intention-to-treat analysis but most studies were of questionable quality with regard to other important quality criteria such as allocation concealment.

Methods used to combine studies were appropriate but it did not appear that statistical heterogeneity was formally assessed and effect estimates were not presented for all comparisons. Examination of forest plots suggested that statistical heterogeneity was likely for several analyses. There were few studies for most comparisons and it was not clear how many were included in analysis for some outcomes. As the authors noted, the primary studies may have been confounded (for example by non-adherence or increased caloric intake) or influenced by under-prescription of aerobic exercise or a predominance of non-hypertensive participants in study populations. They also noted that the review was potentially limited by publication bias, inadequate data reporting in the primary studies and heterogeneity in population characteristics and interventions.

The authors' conclusions were of questionable reliability, mainly due to the rather restricted search, the small number and limited quality of the included studies and differences between the studies.

Practice: The authors stated that isolated aerobic exercise was not an effective weight loss policy in overweight and obese individuals. It had modest benefits for weight, waist circumference, blood pressure and lipid levels and may have value in conjunction with dietary restriction. Patients and health care workers should be aware of this and should pursue alternative or combination intervention strategies for weight loss.

Research: The authors did not state any implications for research.

Canadian Institutes of Health Research.

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.