Twenty-three studies (1,869 participants) were included: 12 randomised controlled trials, 10 pre-post single group studies and one non-randomised patient preference study. Four studies also evaluated a weight-loss maintenance intervention and followed up participants for between three and 12 months. Most studies were considered to be at moderate to high risk of bias (median quality score of 3). Only three studies were considered at low risk of bias. Seven studies used intention-to-treat analysis, five accounted for confounders in their analysis and five were adequately powered.
Meta-analysis of seven trials found a statistically significant benefit for weight-loss interventions with a greater reduction in weight compared with no intervention (weighted mean difference -5.66kg, 95% CI -6.35 to -4.97; Ι²=39%). The 11 non-randomised studies varied in interventions, duration and intensity of contact and most were at high risk of bias. Most reported a statistically significant reduction in mean weight after the intervention.
Of the four studies that assessed a weight-loss maintenance intervention, two were randomised controlled trials considered to be at moderate risk of bias and two were pre-post studies considered to be at high risk of bias. Both randomised trials reported statistically significant but comparable weight regain for intervention and control groups.
Study characteristics found to have a possible relationship with intervention effectiveness were a prescribed energy restriction, interventions that included group face-to-face contact, higher frequency of contact (>2.7 per month) and a younger population (≤42.8 years).