| Design | Sixty-eight studies (66 randomized controlled trials, 2 single-group pre–post designs [included for harms outcomes only]) |
| Populations | All studies included overweight (body mass index [BMI] 25–29.9) or obese (BMI 30–39.9) adults Two interventions targeted seniors (≥ 65 yr); all other studies included adults aged 18 years or older Sixty-four studies included both sexes (1 reported data only for men); 3 included only women, 1 included only men Twenty-six studies (38%) were directed at populations with a high risk for cardiovascular disease (i.e., screened or identified as high risk or diagnosed with type 2 diabetes, hypertension or dyslipidemia) |
| Interventions | Forty-one behavioural intervention arms (8 diet, 4 exercise, 10 diet plus exercise and 19 lifestyle) in 39 studies Twenty-nine pharmacologic (25 studies involved orlistat [dosages: 23 studies, 120 mg three times daily; in 2 studies, 60 mg three times daily, included only for harms]; and 4 studies involved metformin (dosages: 500 mg once daily, 850 mg once daily, 850 mg twice daily, 1500 mg once daily]) plus behavioural (hypocaloric diet and encouragement to increase physical activity level) intervention arms in 27 studies Median intervention duration was 12 mo; 49 interventions (72%) were ≤ 12 mo; 19 interventions (28%) lasted 13−60 mo: most ran for ≤ 2 yr |
| Comparator | In behavioural intervention trials, control participants received usual care from their physicians or no intervention; in 7 studies they received a minimal component (e.g., printed materials on weight loss and healthy lifestyles) In trials using orlistat or metformin, control participants followed the same diet and exercise instructions as the intervention participants but received placebo instead of active medications |
| Outcomes | Primary weight outcomes (weight change in kilograms, loss of ≥ 5% baseline bodyweight, loss of ≥ 10% baseline bodyweight, change in BMI and change in waist circumference) Secondary health outcomes (total cholesterol, low-density lipoprotein cholesterol level, fasting blood glucose level, incidence of type 2 diabetes, systolic blood pressure and diastolic blood pressure) Treatment harms (any adverse events, serious adverse events, gastrointestinal events and withdrawal from the study because of adverse events) |
| Quality assessment | Sixty-two of the randomized controlled trials (94%) were rated as having unclear or high risk of bias primarily because of a lack of information or a lack of procedures to ensure random sequence generation, allocation concealment and blinding of participants, personnel and outcome assessment Most outcomes received moderate quality Grading of Recommendations Assessment, Development and Evaluation ratings (downgraded for risk of bias); occasional low-quality ratings were applied because of added concerns primarily regarding reporting bias |
| Study locations | Two studies were conducted in Canada, 26 in the US, 31 in European countries, 1 co-located in the US and Europe, 6 in Australia or New Zealand, 1 in Japan and 1 in China |
| Publication dates | Thirty-five studies (51%) were published between 2009 and 2013; 33 studies were published between 1985 and 2008 |