Effects of Intrinsic Foot Muscle-Targeted Exercise Programs on Functional, Neuromuscular, and Biomechanical Outcomes Relevant to Lower-Limb Injury: A Systematic Review with Limited Meta-Analysis

This review evaluated the effects of intrinsic foot muscle (IFM)-targeted exercise programs on predefined ankle- and knee-related clinical endpoints and on functional, neuromuscular, and biomechanical surrogate outcomes potentially relevant to lower-limb injury. This systematic review, incorporating a limited meta-analysis where quantitative pooling was feasible, followed PRISMA guidelines. Six electronic databases were searched for randomized controlled trials investigating IFM-targeted exercise programs in adults aged 18–59 years compared with no intervention, sham, or usual care, including parallel-group, cluster-randomized, and randomized crossover or counterbalanced designs when an adequate washout period was reported. Meta-analysis used Hedges’ g under a random-effects model with restricted maximum likelihood (REML) estimation of between-study variance and Hartung–Knapp adjustment for the pooled confidence interval; outcomes with fewer than three poolable studies were synthesized narratively per SWiM guidance. Thirteen randomized controlled trial reports (462 participants in the analyzed samples; 2012–2026) were included. The included samples were predominantly young and physically active, limiting generalizability across the broader adult age range. None of the included trials reported the predefined primary clinical endpoint categories: incident ankle injuries, prospectively recorded ankle-instability events, incident acute knee injuries, or incident knee-related symptoms/clinical syndromes. Consequently, none of these primary endpoint categories could be evaluated quantitatively or through comparative narrative synthesis, resulting in an empty synthesis for the primary clinical outcomes. Therefore, no direct estimate of the effect of IFM-targeted exercise programs on incident ankle injury, ankle-instability events, acute knee injury, or the onset of knee-related symptoms/clinical syndromes could be generated. The available evidence was therefore limited to secondary surrogate outcomes. Ten secondary outcome clusters were identified; however, quantitative pooling was feasible for only one dynamic postural control (k = 3; n = 70), while the remaining outcomes were synthesized narratively. For dynamic postural control, the pooled point estimate favored IFM training in direction (Hedges’ g = 0.39), but the effect was highly uncertain under the REML/Hartung–Knapp model (95% CI −0.62 to 1.41; p = 0.238), and the certainty of evidence was rated Very Low. Individual trials reported improvements in foot muscle strength, morphology, foot posture, ankle instability, and inter-articular coordination, but these findings were generally derived from single studies and cannot establish injury-preventive efficacy. The evidence base is currently empty for the predefined primary outcomes, and the current evidence does not support clinical recommendations for isolated IFM training as an injury-prevention strategy or as an intervention with established broader rehabilitation effectiveness. PROSPERO registration: CRD420261405154.