Virtual reality-based interventions targeting sports injury-related risk factors: a systematic review of biomechanical, neuromuscular, and functional outcomes

Background Virtual reality (VR)-based interventions are increasingly applied in sports training and musculoskeletal rehabilitation. However, their potential role in modifying lower-extremity injury-related risk factors in athletic populations remains incompletely understood. Methods Following PRISMA guidelines, PubMed, Scopus, Web of Science, and SportDiscus were searched up to September 25th, 2025. Eligible studies included randomized, controlled, quasi-experimental, or within-subject experimental designs evaluating immersive, semi-immersive, or non-immersive VR interventions in junior to young-adult athletes. Comparator conditions included conventional training, alternative exercise interventions, no-intervention controls, placebo conditions, or non-VR comparisons. Outcomes addressed biomechanical, neuromuscular, functional, perceptual-cognitive, or psychological factors potentially relevant to sports injury risk. Methodological quality was assessed using the Downs and Black checklist. Results Thirty studies met the inclusion criteria. Interventions ranged from single-session biofeedback exposure to 4–16-week neuromuscular, balance, perceptual-cognitive, or sensorimotor programs across multiple sports. Most studies reported improvements in at least one biomechanical, neuromuscular, functional, perceptual-cognitive, or psychological outcome; particularly, in lower-extremity movement control, balance, coordination, and reaction efficiency. However, several acute studies demonstrated transiently less favorable movement mechanics during highly immersive or cognitively demanding tasks. Importantly, no included study evaluated injury incidence as a primary preventive outcome. Downs and Black scores ranged from 14 to 26 (mean = 19.6 ± 3.1), indicating overall fair-to-good methodological quality. Discussion The evidence suggests that VR-based training can effectively modify several surrogate lower-extremity injury-related risk factors; particularly, those associated with biomechanics and neuromuscular control. Nevertheless, the lack of longitudinal data and the scarcity of injury-incidence outcomes limit conclusions regarding real-world preventive efficacy. Conclusions VR-based interventions may represent a promising adjunct to conventional neuromuscular and perceptual-cognitive training approaches for modifying surrogate lower-extremity injury-related factors. Nevertheless, substantial heterogeneity, limited longitudinal follow-up, and the absence of injury-incidence outcomes restrict conclusions regarding definitive preventive efficacy. Future adequately powered randomized trials with standardized protocols and verified injury outcomes are required. Systematic Review Registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420251163302 , identifier CRD420251163302.