Remember me
Background:
Postural stability is vital for athletic performance and injury prevention in young athletes. While neuromuscular training (NMT) is common, its specific effects on different stability components and optimal training parameters remain unclear. This systematic review and meta-analysis evaluated the efficacy of NMT on dynamic and static postural stability in young athletes.
Methods:
Five databases (PubMed, Web of Science, Embase, Cochrane Library, and Scopus) were searched for randomized controlled trials (RCTs) examining NMT effects on postural stability in young athletes. Methodological quality and risk of bias were assessed using the PEDro scale and RoB 1.0. Evidence certainty was evaluated via the GRADE approach. Data were pooled using a random-effects model, reporting standardized mean differences (SMDs) and 95% confidence intervals (CIs).
Results:
Eighteen articles (19 independent trials, N = 605) were included. NMT significantly improved both dynamic [SMD = 0.96, 95% CI (0.70, 1.22), p < 0.00001] and static postural stability [SMD = 0.96, 95% CI (0.60, 1.32), p < 0.00001]. Subgroup analyses identified participant age as a significant source of heterogeneity for static stability outcomes.
Conclusions:
NMT effectively enhances dynamic and static postural stability in young athletes. Given the comparable efficacy across different NMT modalities, practitioners can flexibly design training programs to suit specific athletic contexts and practical constraints.
Systematic Review Registration:
https://www.crd.york.ac.uk/PROSPERO/view/, identifier CRD420261299111.
1 IntroductionPostural stability is essential for executing complex motor tasks and preventing musculoskeletal injuries (Hewett et al., 2005; Horak, 2006; Paillard, 2019). However, young athletes face unique challenges during the physiological process of maturation. The adolescent growth spurt is characterized by rapid, asynchronous increases in limb length and body mass (Malina et al., 2004), often precipitating a “neuromuscular lag.” During this phase, the maturation of motor control and proprioceptive systems fails to keep pace with accelerated skeletal growth (Quatman et al., 2006; Parsons, 2014; Parry et al., 2024). Biomechanically, limb elongation increases the segmental moment of inertia, while an elevated center of mass compromises stability—a phenomenon termed “adolescent awkwardness” (McKay et al., 2016; Borato et al., 2025). During this window, athletes may experience a transient regression in sensorimotor function and diminished joint position acuity (Williams et al., 2021). Consequently, altered biomechanics during high-risk maneuvers, such as jumping or cutting, can amplify knee valgus moments and heighten the risk of anterior cruciate ligament (ACL) tears and ankle sprains (Hewett et al., 2010; Wordeman, 2014; Gu et al., 2025).
Neuromuscular training (NMT) optimizes motor command output by stimulating sensory pathways and inducing central nervous system adaptations (Faigenbaum et al., 2011; Yang et al., 2025). Contemporary Integrative Neuromuscular Training (INT) models have moved beyond single-modality exercises to combine core stability, plyometrics, balance, and agility drills (Myer et al., 2011; Faude et al., 2017). These programs enhance postural stability through “sensory reweighting” (Peterka, 2002). By introducing destabilizing stimuli, such as unstable surfaces or visual occlusion, NMT challenges the nervous system, promoting an increased reliance on proprioceptive and vestibular inputs. This structured motor training promotes neuroplasticity, optimizing functional connectivity within the frontoparietal network that governs motor learning (Vacchini et al., 2025). Given the heightened neural plasticity of adolescence, NMT provides a critical stimulus for developing robust postural control circuitry.
While previous meta-analyses establish NMT’s efficacy in reducing lower extremity injuries by approximately 36% and ACL injuries by nearly 50% (Emery et al., 2015), evidence regarding its specific impact on postural stability remains heterogeneous. This inconsistency largely stems from variations in NMT modalities, intervention dosages, and outcome measures (Behm et al., 2015; Gebel et al., 2018). Furthermore, dynamic postural stability (e.g., Star Excursion Balance Test, Y-Balance Test) and static postural stability (e.g., Balance Error Scoring System, stabilometry) may respond differently to specific NMT components (Imai et al., 2014; Pinzón-Romero et al., 2019; Zhang et al., 2021; Daneshjoo et al., 2022; Gasim et al., 2022). Therefore, this systematic review and meta-analysis aimed to quantify the effects of NMT on dynamic and static postural stability in young athletes compared to conventional training. A secondary objective was to explore the moderating effects of NMT modality, intervention duration, outcome measures, and participant age, ultimately providing a robust empirical rationale for optimizing youth athletic development programs.
2 Protocol and registrationThis systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Page et al., 2021). The study protocol was prospectively registered in PROSPERO (CRD420261299111). All methodological procedures were established a priori to maintain transparency and minimize bias.
3 Data sources and search strategyTwo independent reviewers (JFZ and SCL) searched PubMed, Web of Science, Embase, the Cochrane Library, and Scopus from inception to November 1, 2025. The search strategy combined Medical Subject Headings (MeSH) and free-text terms using Boolean operators (AND/OR) (the full search strategy is provided in the Supplementary Files). Keywords targeted the population (e.g., “adolescent”, “young athletes”), interventions (e.g., “neuromuscular training”, “integrative neuromuscular training”), and outcomes (e.g., “postural stability”, “Y-balance test”). Reference lists of the retrieved articles were manually screened to identify additional eligible trials.
The initial search yielded 3256 records. After removing 1460 duplicates via NoteExpress 3.2.0, 1765 titles and abstracts were screened, leaving 31 articles for further evaluation. Following a full-text assessment, 13 articles were excluded, resulting in 18 randomized controlled trials (RCTs) for the quantitative synthesis (Figure 1).
PRISMA flow diagram of the study selection process.
4 Inclusion and exclusion criteriaEligibility criteria were defined using the PICOS framework (Chandler et al., 2019):
Population (P): Young athletes (10–24 years; UN/WHO definition), encompassing the developmental continuum from early adolescence to late-stage neuromuscular consolidation in collegiate participants.
Intervention (I): Multi-component NMT programs emphasizing neuromuscular coordination, irrespective of specific training duration, frequency, or modality.
Comparison (C): Conventional sport-specific training or other active/passive regimens lacking NMT components (e.g., traditional strength training). The NMT intervention could either supplement or substitute a segment of the routine training.
Outcomes (O): Quantitative measures of dynamic postural stability (e.g., Star Excursion Balance Test, Y-Balance Test, instrumented platforms) or static postural stability (e.g., Balance Error Scoring System, stabilometry, single-leg stance tests).
Study Design (S): Randomized controlled trials (RCTs) only.
Exclusion criteria comprised (1): non-RCT designs or review articles (2); animal models (3); unavailable full texts; and (4) insufficient or non-extractable data for quantitative synthesis.
5 Data extraction and processingTwo independent reviewers (JFZ and SCL) extracted data into a standardized spreadsheet, capturing participant characteristics, intervention details, and pre- and post-intervention outcomes (means, standard deviations [SDs], and mean change scores). Study characteristics are summarized in Tables 1, 2. Discrepancies were resolved through discussion or consultation with a third senior researcher.
StudyNAge (EG/CG), yearsSexSportOutcomesEsmailnezhad et al. (2024)2415.5 ± 0.9/15.91 ± 0.79FWrestlingYBT, BESSGong et al. (2024)3015.8 ± 0.79/15.6 ± 0.7MBasketballSEBT, SLST-ECKim et al. (2024)3016-19MBaseballYBTMitrousis et al. (2023)4212.71 ± 0.41/12.73 ± 0.46MSoccerLafayette Platform, Johnson & Nelson TestShi et al. (2023)1613.7 ± 0.39/13.6 ± 0.33MTennisLOSSikora and Linek (2022)9012.5 ± 2.2/12.4 ± 2.1MSoccerYBT, ALFA PlatformAloui et al. (2022)3414.6 ± 0.5/14.6 ± 0.4MSoccerYBTDaneshjoo et al. (2022)2414.75 ± 1.1/14.58 ± 0.51MHandballYBT, BBS, SLSTGasim et al. (2022)1817.2 ± 0.4/17.3 ± 0.5/17.7 ± 0.5MSoccerYBTGidu et al. (2022)1615.3 ± 3.0/13.6 ± 4.9MSoccerBESSDogan and Savaş (2021)3012-14MBasketballYBT, StabilometerPuzi and Choo (2021)3014.13 ± 0.83/13.6 ± 0.91M/FHandballSEBTZhang et al. (2021)5819.81 ± 1.72/19.02 ± 1.97M/FDanceYBTZacharakis et al. (2020)2513-14MBasketballLafayette Platform, Narrow Beam SLSTPinzón-Romero et al. (2019)5812.93 ± 1.4/13.21 ± 1.3M/FRoller SkatingSEBT, BESSOndra et al. (2017)2117.3 ± 1.3/16.5 ± 1.8MBasketballCOP VelocityRamírez-Campillo et al. (2015)4011.2 ± 2.3/11.4 ± 2.4男MSoccerCOP TrajectoryImai et al. (2014)1916.5 ± 0.5/16.1 ± 0.6MSoccerSEBT, COP TrajectoryParticipant characteristics and outcome measures of the included studies.
EG, experimental group; CG, control group; M, male; F, female; YBT, Y-Balance Test; SEBT, Star Excursion Balance Test; BESS, Balance Error Scoring System; LOS, Limits of Stability; SLST, Single-Leg Stance Test; SLST-EC, Single-Leg Stance Test with eyes closed; BBS, Berg Balance Scale; COP, Center of Pressure.
StudyDuration & FreqInterventions (EG vs. CG)Training contentTraining protocolEsmailnezhad et al. (2024)8 wks, 3x/wkEG (NMT): Core Stability [Primary], Balance & Proprioception [Primary], Dynamic Strength & Functional, Agility & Reactive ControlCharacteristics of the neuromuscular training interventions.
EG, experimental group; CG, control group; NMT, neuromuscular training; TST, traditional strength training; PT, placebo training; KB, kettlebell; W/R, work/rest; Prog, progression; Iso, isometric.
Data transformations, including the aggregation of multiple correlated testing conditions (e.g., various stances, surfaces, or test variations) into composite means and SDs, and the imputation of missing change score SDs (assuming a conservative intra-individual correlation coefficient of r = 0.5), were conducted in accordance with the Cochrane Handbook for Systematic Reviews of Interventions (Higgins et al., 2019).
Specifically, to prevent disproportionate weighting from studies reporting multiple testing conditions, the following aggregation formulas were applied:
When pre-to-post intervention change scores were not reported, they were estimated as follows:
6 Methodological quality assessmentTwo independent reviewers (JFZ and SCL) evaluated study quality and risk of bias using the Physiotherapy Evidence Database (PEDro) scale (Maher et al., 2003) and the Cochrane Risk of Bias tool (RoB 1.0) (Higgins et al., 2011). Based on the 10-point PEDro scale, study quality was classified as excellent (9–10), high (6–8), moderate (4–5), or low (< 4). Concurrently, RoB 1.0 classified trials into Grade A (≥ 4 low-risk domains), Grade B 2–3), or Grade C (≤ 1).
PEDro scores ranged from 4 to 9 (two excellent, eleven high, and five moderate) (Figures 2, 3), while RoB 1.0 identified eight Grade A and ten Grade B studies. No low-quality or Grade C studies were included (Figures 4, 5). The trials generally demonstrated strong methodological rigor in random sequence generation, baseline comparability, and incomplete outcome data management. Notably, blinding of participants and personnel universally presented a high risk of bias—an inherent constraint of physical exercise interventions rather than a methodological flaw. Overall, the risk of bias was low to moderate, supporting the validity of the quantitative synthesis.
Methodological quality assessment of the included studies based on the PEDro scale.
Summary of methodological quality criteria across the included studies (PEDro scale).
Risk of bias summary: review authors’ judgements about each risk of bias item for each included study.
Risk of bias graph: review authors’ judgements about each risk of bias item presented as percentages across all included studies.
7 Statistical analysisAll statistical analyses were performed using Review Manager (RevMan, version 5.4) and Stata/SE (version 15.0). Effect sizes were pooled as standardized mean differences (SMDs) with 95% confidence intervals (CIs) due to the variation in postural stability assessment tools. The magnitude of SMDs was classified as small (< 0.5), moderate (0.5–0.79), or large (≥ 0.8). Statistical heterogeneity was assessed via the I2 statistic, with values of < 25%, 25–50%, and > 50% representing low, moderate, and high heterogeneity, respectively.
A random-effects model was applied for all syntheses to account for anticipated clinical and methodological heterogeneity. Subgroup analyses explored potential sources of variance across predefined components. Additionally, a leave-one-out sensitivity analysis was conducted to evaluate the robustness of the pooled estimates. Potential publication bias was evaluated using Egger’s regression test, with statistical significance set at p < 0.05.
8 Assessment of publication biasEgger’s regression test (Egger et al., 1997) indicated no significant publication bias for either dynamic [t = 1.36, p = 0.195, 95% CI (-0.942, 4.225)] or static postural stability outcomes [t = 1.33, p = 0.212, 95% CI (-1.386, 5.510)]. The regression intercepts for both domains firmly encompassed zero, confirming the absence of small-study effects and substantiating the robustness of the pooled estimates (Table 3).
Outcome measureParameterCoefficientSEt
Comments (0)