Abstract
BACKGROUND: Integrative neuromuscular training (INT) is proposed to enhance static and dynamic balance in team athletes by improving sensorimotor control. This systematic review and meta-analysis evaluated INT effects across these two domains. METHODS: We searched PubMed/MEDLINE, Scopus, SPORTDiscus (EBSCOhost), and the Web of Science Core Collection from database inception through 3 March 2025. Random-effects and multilevel meta-analyses were conducted in R (meta, metafor). Standardized mean differences (Hedges' g) were pooled using inverse-variance weighting. For datasets with multiple effects per study, we fitted multilevel random-effects models (metafor::rma.mv) with random intercepts for study and within-study effects, estimating variance components (τ²) via REML. We report 95% confidence intervals, I², τ², and 95% prediction intervals. Risk of bias was assessed with RoB 2 and certainty of evidence with GRADE; small-study effects were examined using Egger's regression (k ≥ 10) and bubble plots. RESULTS: Fifteen randomized controlled trials (n = 609 athletes) were included. Overall risk of bias (RoB 2) was judged as some concerns across studies. INT improved dynamic balance (SMD = 1.48, 95% CI 0.97-1.99; I² = 86.8) and static balance (SMD = 0.89, 95% CI 0.32-1.47; I² = 51.2). Dynamic balance improved more with ≥ 3 sessions/week (SMD = 2.01, 95% CI 1.23-2.80; I² = 91.2%), whereas static balance benefited more from programs ≤ 8 weeks (SMD = 1.16, 95% CI 0.77-1.54). CONCLUSIONS: INT improves balance in team athletes, with larger gains for dynamic tasks. Dynamic balance was preferentially enhanced by ≥ 3 sessions/week, while static balance gains were optimal at ≤ 8 weeks. Evidence certainty was moderate for dynamic balance and low for static balance.