Thirty-eight RCTs (mean number of participants 72, range 16 to 258) were included in the review. Eight trials met eight of the 10 quality criteria; results were fully reported in the review. Some trials did not follow participants after the intervention ended; other studies had follow-up in the range one month to four years.
Meta-analysis indicated a statistically significantly larger beneficial effect on weight and body composition for lifestyle compared to no treatment or a wait-list control over two years (SMD -0.97, 95% CI -1.39 to -0.55; Ι²=90%; 19 trials, 23 comparisons, 1,234 participants). There was no evidence of publication bias.
Change in BMI at the end of active treatment showed a statistically significant reduction in the lifestyle intervention group compared with usual care or minimal intervention (SMD -1.30kg/m², 95% CI -1.58 to -1.03; Ι²=75%; seven trials, 586 participants). Four trials that assessed longer follow-up (seven months to one year after programme completion) indicated that weight loss was sustained.
Lifestyle interventions showed a statistically significantly greater reduction in BMI z-score over one year when compared with written educational materials (SMD -0.06, 95% CI -0.10 to -0.02; Ι²=99%; three trials, 354 participants) but no differences for change in BMI (two trials).
Results from subgroup analyses and other findings were reported in the review.