Abstract / Summary
Summary: Background: Children and adolescents with neurodevelopmental disorders (NDDs) frequently experience executive dysfunction. Although non-pharmacological interventions are commonly recommended to address these challenges, their relative effectiveness across intervention types and executive-function domains remain unclear. We aimed to systematically compare the effectiveness of non-pharmacological interventions for improving overall and domain-specific task-based executive function outcomes in children and adolescents with NDDs. Methods: We conducted a systematic review, pairwise meta-analysis, and network meta-analysis of randomized controlled trials. Comprehensive searches were performed across APA PsycINFO, CINAHL Complete, MEDLINE, Web of Science, ERIC, and Embase from database inception through July 2024, with search updates subsequently completed on September 24, 2025, and August 24, 2026. Studies identified in the 24 August 2026 update were not incorporated into the existing analyses. The quantitative synthesis therefore remains based on studies identified through 24 September 2025. We included trials of participants aged 5–18 years with formally identified neurodevelopmental disorders that compared an eligible non-pharmacological intervention with another eligible intervention or control and reported performance-based executive-function outcomes; non-randomized studies and studies without an eligible task-based outcome were excluded. Eligibility was not restricted by participant sex or gender. The primary outcome was overall task-based executive-function performance at the first post-intervention assessment. Heterogeneity was assessed using Cochran's Q and I 2 . Risk of bias and methodological quality were assessed using the Cochrane Risk of Bias 2.0 tool and the Physiotherapy Evidence Database (PEDro) scale. Confidence in the network estimates was assessed using Confidence in Network Meta-Analysis (CINeMA), informed by Grading of Recommendations Assessment, Development and Evaluation (GRADE) principles. This systematic review is registered with the International Prospective Register of Systematic Reviews (PROSPERO), CRD42024540385. Findings: Collectively, 2087 records were identified, and 118 studies involving 5863 participants (76.10% [4462 of 5863] male and 23.90% [1401 of 5863] female) and 141 independent intervention arms were included; 83 of the 118 studies involved populations with attention-deficit/hyperactivity disorder (ADHD). The 141 intervention arms were organized into 5 broad categories for pairwise meta-analysis: computerized cognitive training, non-computerized executive function training, mindfulness training, neurofeedback training, and physical activity training. Twelve specific intervention types were investigated using frequentist random-effects network meta-analysis. Risk of bias was predominantly rated as having some concerns, the mean PEDro score was 6.93, and CINeMA confidence, informed by GRADE principles, was low overall and low or very low for most domain-specific estimates; heterogeneity for the overall pooled effect was moderate (I 2 = 49.21%). Non-pharmacological interventions were associated with a moderate improvement in overall task-based executive function (EF) performance (standardized mean difference [SMD] = 0.521; 95% confidence interval [CI], 0.447–0.595), with moderate heterogeneity (I 2 = 49.21%). Egger's test indicated possible small-study effects/publication bias (P < .001). In network meta-analysis relative to treatment as usual, computerized serious games were associated with improvements in cognitive flexibility (SMD = 1.30; 95% CI, 0.60–2.00), working memory (SMD = 0.79; 95% CI, 0.47–1.11), and planning (SMD = 1.34; 95% CI, 0.44–2.24), while cognitively engaging exercise was associated with improved inhibitory control (SMD = 0.65; 95% CI, 0.43–0.86). Several other interventions yielded mixed or non-significant estimates, and the certainty of evidence was low overall and low or very low for most domain-specific estimates. Interpretation: Non-pharmacological interventions may improve task-based EF performance in children and adolescents with NDDs. Computerized serious games and cognitively engaging exercise appear promising for selected EF domains, but findings should be interpreted cautiously given low certainty, heterogeneity, publication bias, and the ADHD-dominant evidence base. Future adequately powered, diagnosis-specific randomized controlled trials (RCTs) using standardized executive-function outcomes are needed to strengthen confidence in these findings and determine whether the observed intervention rankings generalize across NDDs. Funding: No funding was received for this work.