Optimal exercise prescription for depression and anxiety in children and adolescents: a Bayesian dose–response network meta-analysis protocol

Abstract Background Depression and anxiety are among the most common mental health problems affecting children and adolescents worldwide. Exercise is a widely used and potentially cost-effective non-pharmacological approach that may improve mood and mental health. However, the optimal exercise modalities and doses for alleviating depressive and anxiety symptoms in children and adolescents remain uncertain. Previous evidence has primarily relied on pairwise meta-analyses or conventional network meta-analyses: the former are unable to compare multiple exercise formats simultaneously, while the latter, although capable of integrating different interventions, have not quantified dose characteristics such as intensity, frequency, and duration. Consequently, systematic dose–response evidence regarding depressive and anxiety symptoms in children and adolescents is lacking. This study aims to examine the quantitative relationship between exercise dose and changes in depressive and anxiety symptoms. Methods This protocol outlines a systematic review and Bayesian model-based dose–response network meta-analysis. A systematic search will be conducted of PubMed, Embase, Web of Science, the Cochrane Library, Scopus, PsycINFO, SPORTDiscus, and the China National Knowledge Infrastructure databases through May 2026. Randomized controlled trials enrolling children and adolescents aged 6–18 years with depressive or anxiety symptoms and comparing different types and doses of exercise training will be eligible for inclusion. Study quality will be appraised using the Cochrane Risk of Bias 2.0 tool. Exercise interventions will be categorized by type (e.g., aerobic, resistance, mind–body, and combined exercise-only) prior to dose–response modeling. A Bayesian model-based dose–response network meta-analysis will be performed, with exercise dose quantified as weekly metabolic equivalent of task (MET) minutes (MET-min/week) by integrating intensity, session duration, and frequency. Nonlinear dose–response curves will be fitted for distinct exercise modalities. Meta-classification and regression tree (meta-CART) analysis will be employed to identify potential effect modifiers. Discussion This study will systematically evaluate the nonlinear dose–response relationships between exercise dose and changes in depressive and anxiety symptoms in children and adolescents, and estimate dose ranges associated with symptom change across exercise modalities. The findings may help inform future evidence-based recommendations and provide methodological guidance for dose–response research in child and adolescent mental health. Systematic review registration This protocol has been registered with the International Prospective Register of Systematic Reviews (PROSPERO), registration number CRD420251174947.