An Integrated Framework for Classifying Intervention Mechanisms and Measurement Methodologies of Lower-Limb Proprioception in Patients with Stroke: A Systematic Review and Exploratory Meta-Analysis

Background/Objectives: This systematic review and exploratory meta-analysis evaluated interventions targeting lower-limb proprioception after stroke, examined their underlying therapeutic mechanisms and assessment methods, and proposed a mechanism-oriented classification framework for organizing the existing evidence. Methods: Randomized parallel-group and randomized crossover studies involving adults with stroke and reporting lower-limb proprioception outcomes were eligible. PubMed, Embase, Scopus, Web of Science Core Collection, Cochrane Library, and MEDLINE via Ovid were searched through 31 March 2026. Risk of bias was assessed using the Cochrane Risk of Bias 2 (RoB 2) tool. Interventions were classified as peripheral, central, or combined peripheral and central mechanisms according to predefined operational criteria. Assessment methods and testing procedures were summarized, and comparable joint position reproduction (JPR) angle-error outcomes were synthesized using random-effects meta-analysis. Results: Nineteen studies involving 657 participants were included, of which 10 contributed to the exploratory meta-analysis. Considerable heterogeneity was observed across intervention mechanisms, proprioception measures, and testing procedures. Using a random-effects model, the overall pooled effect showed a statistically significant improvement in proprioception outcomes in favor of the intervention groups (SMD = −0.78, 95% CI −1.30 to −0.25, p = 0.004). However, substantial between-study heterogeneity remained (I2 = 82%), and the 95% prediction interval crossed the null value, indicating uncertainty in the range of effects that might be observed in future comparable studies. No statistically significant differences were found between the mechanism subgroups. The certainty of evidence for the primary proprioceptive outcome was rated as very low. Conclusions: The proposed mechanism-oriented classification provides a structured approach for organizing heterogeneous lower-limb proprioceptive interventions after stroke, but it requires further validation. The quantitative findings should be interpreted as exploratory and hypothesis-generating because of substantial heterogeneity, methodological variability, and very low certainty of evidence. Standardized assessment protocols and higher-quality randomized studies are needed to further evaluate proprioceptive rehabilitation after stroke.