Mindfulness-based interventions for children and adolescents with attention-deficit/hyperactivity disorder: a Bayesian meta-analysis of randomized controlled trials

Background: Attention-deficit/hyperactivity disorder (ADHD) in children and adolescents is characterized by inattention, hyperactivity, and impulsivity, and is frequently accompanied by impairments in executive functioning, task performance, and emotion regulation. Mindfulness-based interventions (MBIs) have been increasingly evaluated as non-pharmacological approaches for ADHD, but findings remain heterogeneous. Objective: To synthesize evidence from randomized controlled trials (RCTs) on the effects of MBIs for youths with ADHD using a pre-registered Bayesian random-effects systematic review and meta-analysis, and to examine potential moderators (age) and dose–response relationships (contact hours). Methods: We conducted a pre-registered Bayesian random-effects systematic review and meta-analysis of RCTs evaluating MBIs in children and adolescents with ADHD. Seven databases were searched from inception to April 30, 2025, prioritizing immediate post-intervention outcomes. Seventeen RCTs (total n = 2,991) were included. Pooled effects were summarized as Hedges' g with 95% credible intervals (CrIs). Symptom-domain subgroup models were performed, and heterogeneity was quantified using I² and τ². Age-stratified analyses (mean age >10 years vs. ≤ 10 years) and dose–response modeling based on contact hours were conducted. Results: Across all outcomes, MBIs showed a small-to-moderate advantage over control conditions (Hedges' g = 0.49, 95% CrI 0.37–0.62), with substantial heterogeneity (I² = 81.6%; τ² = 0.16). Domain-specific subgroup models indicated statistically credible improvements in inattention (Hedges' g = 0.30, 95% CrI 0.12–0.50), hyperactivity/impulsivity (Hedges' g = 0.54, 95% CrI 0.31–0.78), executive functions (Hedges' g = 0.23, 95% CrI 0.05–0.43), global ADHD measures (Hedges' g = 1.23, 95% CrI 0.65–1.80), and task performance (Hedges' g = 0.37, 95% CrI 0.07–0.70). The estimate for emotion regulation was imprecise and included the null (Hedges' g = 0.42, 95% CrI −0.08–0.92). Age-stratified analyses suggested larger effects in samples with mean age >10 years than in those with mean age ≤ 10 years. Dose–response modeling suggested that higher contact hours may be associated with greater improvements in selected domains (notably hyperactivity/impulsivity), although uncertainty remained in several domains. Conclusion: MBIs may be a promising complementary approach for improving ADHD-related outcomes in youths. However, substantial heterogeneity and risk-of-bias considerations warrant cautious interpretation and underscore the need for larger, methodologically rigorous RCTs. Systematic review registration: https://www.crd.york.ac.uk/PROSPERO/view, Identifier: CRD420251079766 Public.