Associations between behavioral intervention dose and outcomes in children and adolescents with obesity: meta-regression of randomized clinical trials

Childhood obesity is a growing global health challenge with long-term consequences. Behavioral interventions are pivotal in the management of obesity, yet the relationship between intervention dose and effectiveness remains controversial. A study-level meta-regression analysis was conducted following PRISMA guidelines. MEDLINE, PsycINFO, and the Cochrane Library were searched from inception through December 2024, with an updated search conducted in April 2026. Randomized controlled trials (RCTs) evaluating behavioral interventions for children aged 2–18 years with study-defined overweight or obesity (i.e., with body mass index (BMI) eligibility thresholds at or above the 75th percentile for age and sex) were included. Two reviewers independently screened studies, extracted data, and assessed risk of bias. Effect sizes (Hedges’ g) were calculated from prespecified weight-related outcomes using a hierarchical approach, prioritizing BMI z score, followed by BMI, BMI percentile, percent overweight, and body weight. Random-effects meta-analyses were used to pool effect sizes. Meta-regressions evaluated the relationship between intervention dose parameters and weight-related outcomes. Seventy-eight RCTs (91 comparisons) comprising 16,514 participants were included. The pooled effect size was − 0.28 (95% CI, − 0.35 to − 0.21), indicating a small benefit of behavioral interventions. However, heterogeneity was substantial across studies (I² = 73.4%). In the adjusted meta-regression model, total contact hours were associated with larger intervention effects (β = − 0.016; 95% CI, − 0.027 to − 0.004), while intervention duration and session number were not consistently associated with effect sizes. Interventions involving both children and caregivers were associated with larger intervention effects than child-only interventions (β = -0.271; 95% CI, -0.532 to -0.010). Sensitivity analyses yielded directionally consistent findings. In this study-level meta-regression, greater total contact hours were most consistently associated with larger intervention effects on weight-related outcomes in children and adolescents with overweight or obesity. Interventions involving both children and caregivers may also be associated with larger effects. However, these study-level associations should be interpreted cautiously given the potential for ecological bias, residual study-level confounding, and substantial between-study heterogeneity.