This review examined the effects of aerobic exercise on lipids and lipoproteins in overweight and obese adults. The authors concluded that aerobic exercise can decrease triglyceride levels, but a consistently positive effect on total cholesterol is less clear. The conclusions reflect the evidence presented, but some methodological weaknesses in the review mean that their reliability is unclear.

To examine the effects of aerobic exercise on lipids and lipoproteins in overweight and obese adults.

MEDLINE, EMBASE, SPORTDiscus, Current Contents and Dissertation Abstracts International were searched for relevant studies published in the English language from January 1955 to January 2003; the search terms were reported. Additional studies were sought from the reference lists of review articles and relevant trials, through handsearches of selected journals, and through contact with an expert in the field.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs) were eligible for inclusion in the review.

Specific interventions included in the review

Studies of interventions containing prescribed aerobic exercise of at least 8 weeks' duration compared with a non-exercise group were eligible for inclusion. Resistance training was excluded, as were studies that combined exercise with a dietary intervention. A variety of single component and combination activities were reported; these involved a range of exercise prescriptions (although the majority were reported to be consistent with American College of Sports Medicine Guidelines) and levels of supervision.

Participants included in the review

Studies of overweight and/or obese adults aged 18 years and older, with a body mass index (BMI) greater than or equal to 25 kg/m2, were eligible for inclusion. The included studies comprised males and females (pre- and postmenopausal). Some participants were receiving medication, smoked cigarettes, consumed alcohol, had experienced diet change, or had co-morbidities. The prior physical activity status of the participants was unclear. The majority of the included studies were conducted in the USA.

Outcomes assessed in the review

The primary outcomes of interest were one or more lipid and/or lipoprotein variables in the fasting state. These were total cholesterol (TC), high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol and triglycerides (TG). All were measured in milligrammes per decilitre (mg/dL). A range of lipid assessment characteristics was reported. The secondary outcomes of interest were body weight, BMI, percentage body fat, and maximum oxygen consumption (VO2 max).

How were decisions on the relevance of primary studies made?

Two reviewers independently assessed studies for inclusion in the review. Any disagreements were resolved by consensus.

Two reviewers independently assessed study quality using the Jadad scale (score 0 to 5 from lowest to highest). Any disagreements were resolved by consensus, with a third reviewer if necessary.

Two reviewers independently extracted the data for the review. Any discrepancies were resolved by consensus, with a third reviewer if necessary. Mean values were extracted using per protocol data to calculate the net changes (exercise versus control; initial versus final results) across all categories of variables. Values were reported with the corresponding standard deviation (SD) or standard error (SE) as appropriate.

How were the studies combined?

Seventeen outcomes from the included studies were pooled in a meta-analysis using a random-effects model, using the inverse of the variance as the weighting method. The pooled weighted mean difference was calculated with 95% confidence intervals (CIs). Publication bias was assessed using regression analysis and in funnel plots.

How were differences between studies investigated?

Statistical heterogeneity was examined using the Q and I-squared statistics. A sensitivity analysis considered the impact of diet, medication, smoking and menopausal status on the primary outcomes. A cumulative meta-analysis examined the impact of secular trends. A meta-regression explored the relationship between changes in the primary outcomes and selected continuous variables. A subgroup analysis was performed according to gender, diabetic status and level of exercise supervision, using random-effects and analysis of variance models.

Thirteen RCTs (n=613) were included in the review. For the meta-analysis, there were 31 study groups (17 exercise and 14 control).

The loss to follow-up ranged from 0 to 67% (21.7 +/- SD 20.4%) in the exercise groups and from 0 to 40% (19 +/- SD 13.8%) in the control groups. Programme compliance (percentage of exercise sessions attended) was reported in less than half of the studies. The median score for study quality was 2 out of a possible 5 points.

TC.

A statistically significant mean reduction in TC was reported (2%; -3.4 +/- SE 1.7 mg/dL, 95% CI: -6.7, -0.2), and this had been the trend since 2001. Changes in TC became non significant when further analyses were conducted.

HDL cholesterol.

The mean increase for HDL (3%) was statistically non significant, and this had been the trend since 1995. However, there was a statistically significant increase in HDL when studies of postmenopausal women were excluded from the analysis (1.9 +/- SE 0.8 mg/dL, 95% CI: 0.4, 3.4). Increased HDL was also associated with increased VO2 max (p=0.002) and decreased body weight (p<0.001).

LDL cholesterol.

The mean reduction in LDL (0.3%) was statistically non significant, and this had been the trend since 2001. Greater decreases in LDL were associated with decreased body weight (p=0.009).

TG.

A statistically significant reduction in TG was reported (11%; -16.1 +/- SE 7.3 mg/dL, 95% CI: -30.2, -2.1), and this had been the trend since 2002. Statistical significance remained in further sensitivity analyses.

Publication bias was not statistically significant in any of the above analyses.

Secondary outcomes.

Statistically significant reductions were found for body weight (p=0.02) and BMI (-1.1 +/- SE 0.3, 95% CI: -1.8, -0.5), while increases were reported for VO2 max ( p<0.001).

Aerobic exercise decreased TG in overweight and obese adults. The consistent effect of aerobic exercise on decreasing TC was not clear.

The review question was clear and inclusion criteria for the participants, interventions, outcomes and study design were specific. The search strategy involved some relevant sources, but failure to seek unpublished studies and those not published in English means that important studies might have been missed. The review process was systematic and transparent. It was unclear whether meta-analysis was appropriate given the heterogeneity of the included studies, and it was not possible to determine exactly which exercise prescription produced the effect. However, some appropriate subgroup and sensitivity analyses were performed. The conclusions reflect the evidence presented but (given some methodological weaknesses) their reliability is unclear.

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

Research: The authors stated that further RCTs conducted with intention-to-treat analysis are required in various overweight and/or obese populations, and in countries other than the USA. Future research should fully report the race and ethnicity of the participants, along with complete information on medications and on lifestyle behaviours that might affect lipids and lipoproteins.

National Institutes of Health-National Heart, Lung and Blood Institute, award #R01-HL069802.

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.