Comparative outcomes following virtual reality-assisted and traditional supplementary exercise programs and usual training on balance, sprint performance, and resting-state EEG in youth academy football players

This study compared outcomes associated with an eight-week virtual reality-assisted supplementary exercise program, a traditional supplementary exercise program, and usual training in youth-academy football players. Because the two active programs were not empirically matched for exercise content or exposure dose, the study was not designed to isolate a virtual-reality-specific effect. Thirty-nine male youth-academy football players (18.38 ± 0.49 years; n = 13/group) were randomly allocated to a virtual reality-assisted exercise group, a traditional exercise group, or a usual-training control group using manual block randomization performed by an independent researcher. Balance, 30-m sprint, resting-state spectral power, and anterior–central synchronization were assessed before and after the intervention. Baseline-adjusted ANCOVA was used to estimate post-intervention between-group differences because baseline imbalances were present. Mixed-design ANOVA time × group interactions were retained as complementary sensitivity analyses and corrected using the Benjamini–Hochberg procedure across nine retained interaction tests. Baseline-adjusted ANCOVA showed between-group differences in two-leg, dominant-leg, and non-dominant-leg balance. For two-leg balance, all three pairwise comparisons were statistically detectable. For dominant-leg and non-dominant-leg balance, adjusted post-test stability-index values were lower in the virtual reality-assisted group than in both the traditional and control groups; no statistically detectable difference was found between the traditional and control groups. The 30-m sprint showed an adjusted post-test difference between the virtual reality-assisted and control groups, but no differential pre–post trajectory was detected. Resting-state EEG outcomes showed exploratory differences in anterior spectral power and anterior–central synchronization. An eight-week virtual reality-assisted supplementary exercise program was associated with greater improvements in two-leg, dominant-leg, and non-dominant-leg balance compared with traditional supplementary exercise and usual training in youth football players. Although a difference in 30-m sprint performance was observed between the virtual reality and control groups, no differential pre–post trajectory was detected. Exploratory EEG results indicated differences in anterior spectral power and anterior–central synchronization. Future studies should use dose-matched programs, prospective registration, reliability-tested balance measures, condition-specific EEG recordings, and systematic monitoring of training load. ClinicalTrials.gov NCT07572344, retrospectively registered on 22/04/2026, after data collection had been completed.