Effectiveness of Non-Pharmacological Interventions for Lower Limb Motor Impairment in Post-Stroke Hemiplegia: A Systematic Review and Bayesian Network Meta-Analysis

Effectiveness of Non-Pharmacological Interventions for Lower Limb Motor Impairment in Post-Stroke Hemiplegia: A Systematic Review and Bayesian Network Meta-Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Effectiveness of Non-Pharmacological Interventions for Lower Limb Motor Impairment in Post-Stroke Hemiplegia: A Systematic Review and Bayesian Network Meta-Analysis Di Zhang, Yating Yang, Guixing Xu, Lu Wang, Guilin Zhang, Wanyi Song, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8591907/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract Background Hemiplegia resulting from a stroke often causes significant dysfunction in lower limb movement, greatly impacting an individual's ability to walk and maintain balance. While non-pharmacological interventions show therapeutic potential, the optimal rehabilitation approach remains unclear. This study employs a network meta-analysis (NMA) to assess these treatments and offer evidence-based recommendations for clinical application. Objective To evaluate the efficacy and comparative ranking of non-pharmacological interventions in improving lower limb motor function, balance, walking ability, and activities of daily living in individuals with post-stroke hemiplegia. Methods We conducted a search of PubMed, Embase, Cochrane Library, and Web of Science databases for randomized controlled trials (RCTs) published from January 2010 to August 2025. The Cochrane Risk of Bias Tool and Review Manager 5.4 were used to assess study quality, and evidence was graded with GRADEPro. Using R Studio software, a NMA was carried out to evaluate the clinical efficacy of various treatments in improving lower limb motor function in patients with post-stroke hemiplegia, ranked by the surface under the cumulative ranking curve (SUCRA). The study was officially registered in PROSPERO under the number CRD420251169037. Results This study employed a NMA incorporating 82 randomized controlled trials involving 3,514 patients and covering 16 non-pharmacological interventions. Results indicated that repetitive transcranial magnetic stimulation (rTMS) (SMD = − 3.68; 95% CI: −5.93 to − 0.93) demonstrated the most significant effect in improving lower limb motor function (FMA-LE score) in patients with post-stroke hemiplegia. Additionally, rTMS (SMD = − 12.32; 95% CI: −15.07 to − 9.57) (SMD = − 13.51; 95% CI: −16.32 to − 10.73) showed optimal efficacy in enhancing patients' balance ability and activities of daily living. Transcranial direct current stimulation (tDCS) (SMD = − 1.47; 95% CI: −2.54 to − 0.42) demonstrated the best efficacy in improving Functional Ambulation Classification. The combination intervention of virtual reality and robotic rehabilitation (SMD = 6.39; 95% CI: 4.56 to 7.99) yielded the most favorable results in reducing the Timed Up and Go test time. Conclusion This NMA suggests that rTMS might be the preferred non-pharmacological approach for enhancing lower limb motor function in patients with post-stroke hemiplegia. For individuals experiencing difficulties with walking and balance, a combined approach using virtual reality and robot-assisted rehabilitation is recommended. However, due to the limited number of studies and small sample sizes, the evidence remains preliminary. Therefore, future large-scale, multicenter, double-blind randomized controlled trials are needed to confirm and extend these results. Network Meta-Analysis stroke hemiplegia non-pharmacological interventions randomized controlled trials Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Introduction Stroke is a sudden neurological disorder caused by a disruption in cerebral circulation, resulting in either localized or widespread neurological impairments[ 1 ]. According to the most recent Global Burden of Disease (GBD) report, stroke is now the second leading cause of death worldwide, trailing only ischemic heart disease, and it ranks third in terms of global disability[ 2 ]. The period from 1990 to 2019 witnessed a significant surge in the annual rate of stroke cases and deaths, with stroke incidence climbing by 70% and stroke-related mortality rising by 43%[ 3 ]. Projections indicate that stroke-related deaths are expected to increase by 47% by 2050[ 4 ]. Hemiplegia, commonly observed after a stroke, is a significant neurological disorder that typically results in motor impairment or loss of function on one side of the body[ 5 ]. The pathophysiological mechanism primarily arises from neurological damage due to ischemia or hemorrhage in brain tissue[ 6 ]. The loss of limb motor function is mainly attributed to damage in the corticospinal tract, particularly affecting critical areas such as the motor cortex of the precentral gyrus, the posterior segment of the internal capsule, and the brainstem region[ 7 ]. This damage ultimately results in contralateral limb motor neuron paralysis. Patients with hemiplegia following a stroke frequently experience lower limb motor dysfunction, characterized by increased muscle tone, abnormal gait control, and diminished coordination and balance. Not only do these consequences deeply impact the patients, but they also have significant repercussions for their families, caregivers, and the wider society. Stroke survivors frequently face long-term disability and ongoing care requirements, imposing significant emotional strain, financial burdens, and practical caregiving challenges upon families[ 8 ]. Concurrently, the societal costs of stroke are substantial, encompassing increased healthcare expenditure, diminished productivity, and a growing demand for extended rehabilitation and care services[ 9 ]. Research shows that 50%–65% of stroke patients struggle with walking independently shortly after onset, and despite rehabilitation, 20%–30% still cannot walk independently by the end of treatment[ 10 ]. As a result, improving the restoration of lower limb motor ability in individuals with post-stroke hemiplegia, as well as boosting their autonomy in walking and performing daily tasks, has become a primary focus in contemporary stroke rehabilitation research and clinical practice. Rehabilitating stroke patients with hemiplegia is a challenging yet crucial process. Sole reliance on pharmacological interventions frequently proves inadequate in substantially enhancing limb function and activity levels, as medications primarily address the underlying causes and mitigate symptoms, with limited impact on neural remodeling and functional recovery. Restoring motor function and improving self-care abilities rely primarily on structured and ongoing rehabilitation programs. Non-pharmacological rehabilitation interventions can optimize the activation of residual functions, facilitate cerebral reorganization, and assist patients in regaining independent living abilities. Consequently, early intervention and consistent rehabilitation treatment are imperative for hemiplegic patients to achieve functional recovery. As research into the mechanisms of lower limb dysfunction and rehabilitation strategies after a stroke deepens, the use of various non-pharmacological therapies for functional recovery has expanded, offering a crucial supplement to the limitations of drug-based treatments. The commonly used interventions include neurostimulation techniques, Constraint-induced movement therapy, virtual reality training, robotic rehabilitation, acupuncture therapy, mirror therapy, vibration therapy, and hydrotherapy. These approaches are becoming progressively more significant in enhancing lower limb motor abilities and fostering neural regeneration and functional recovery. Nonetheless, earlier research has mainly concentrated on contrasting the impacts of individual non-pharmacological interventions with standard care or placebo controls, without direct or indirect comparisons between various interventions. This creates significant uncertainty for clinicians when selecting therapies and optimizing treatment plans. To establish a clearer and more systematic evidence base for efficacy, it is necessary to employ Bayesian network meta-analysis (NMA). Bayesian NMA, an enhanced version of conventional meta-analysis, integrates indirect comparisons across interventions, facilitating the assessment of the comparative effectiveness of different treatments. This method allows for the ranking and comparative evaluation of treatments, aiding in the identification of the most effective intervention[ 11 ]. Consequently, this research employs a NMA to comprehensively evaluate the effectiveness of different non-pharmacological treatments for improving lower limb motor function in stroke survivors with hemiplegia. The objective is to rank the various interventions, providing evidence-driven guidance for choosing suitable non-pharmacological treatment approaches in clinical settings Methods Study protocol and registration The study complies with the PRISMA extension guidelines for systematic reviews and meta-analyses and has been registered in PROSPERO with the registration number CRD420251169037[12]. Inclusion criteria The inclusion criteria were defined based on the PICOS framework, encompassing Population, Intervention, Comparison, Outcomes, and Study design[13]. (1) P: Participants must be at least 18 years old, confirmed to have had a stroke via CT or MRI scans, and present with hemiplegia on one side of the body. All participants should have stable vital signs and be fully conscious. There are no restrictions regarding stroke type, side of hemiplegia, illness duration, or participants' gender, race, or geographical location. (2) I: Non-pharmacological interventions encompass a range of rehabilitation methods, including acupuncture (ACPU), vibration therapy (VT), mirror therapy (MT), virtual reality therapy (VR), constraint-induced movement therapy (CIMT), functional electrical stimulation (FES), repetitive transcranial magnetic stimulation (rTMS), transcranial direct current stimulation (tDCS), intermittent theta burst stimulation (iTBS), robot-assisted rehabilitation (RAR), water therapy (WT), and various combined approaches (e.g., VR + RAR, RAR + FES, MT + FES). The supplementary materials provide a detailed account of the specific intervention methods. (3) C: The control group may receive one of the following standard treatments: conventional rehabilitation methods (including exercise therapy, neuromuscular therapy, training, and standard stroke care), sham treatment, or no treatment. (4) O: Outcome Measures: Primary Outcome: ① The Fugl-Meyer Assessment for Lower Extremities (FMA-LE) is used to evaluate motor function in the lower limbs, where higher scores indicate improved motor abilities (maximum score: 34 points). Secondary Outcomes: ① The Berg Balance Scale (BBS) is used to assess an individual's balance capabilities, with a total score of 56 points. Higher scores reflect improved balance function. ② The Modified Barthel Index (MBI) evaluates an individual's ability to perform activities of daily living, with a higher score signifying greater independence and proficiency in completing everyday tasks (maximum score: 100 points). ③ The Functional Ambulation Classification (FAC) is used to assess walking ability, with higher scores indicating better walking function. ④ The Timed Up and Go (TUG) test measures mobility and dynamic stability, with quicker times reflecting improved function. (5) S: Study design: Randomized Controlled Trial (RCT). Exclusion criteria The criteria for exclusion were outlined as follows: (1) Hemiplegia not caused by stroke. (2) reviews, letters, study protocols, conference abstracts, case reports, animal studies, duplicate publications. (3) literature lacking outcome measures, or studies with incomplete outcome data. (4) Research involving a sample size of less than 10. Search strategy Two researchers (DZ and TYY) independently conducted literature searches across four English-language databases: PubMed, Web of Science, Cochrane Library, and Embase. The search encompassed publications from January 2010 to August 2025, with no restrictions on language. Key search terms included stroke, hemiplegia, lower limbs, non-pharmacological interventions, randomized controlled trial, among others. For each database, detailed search strategies are available in the supplementary materials. To ensure comprehensive coverage, references from eligible studies and relevant prior research were manually reviewed. For studies with incomplete data, we reached out to the corresponding authors to obtain the required information and conducted a thorough online search for all relevant studies. Data extraction Two researchers (DZ, TYY) imported the retrieved studies into EndNote reference management software, and after removing duplicates, conducted an initial screening by reviewing the titles and abstracts. The studies were subsequently chosen independently according to pre-established eligibility criteria. Full-text articles from potentially eligible studies were obtained for additional evaluation. The two researchers (DZ, TYY) then independently extracted the relevant study details using a standardized data collection form and cross-verified the data. Conflicts were addressed through consultation with a third researcher (DHL). For every included study, the following information was extracted: lead author, year of publication, country of origin; participant characteristics (sample size, gender distribution, average age, type of stroke, affected side, disease stage); details of interventions and their duration for both experimental and control groups; and outcome measures. Risk-of-bias assessment Two researchers (DZ, TYY) evaluated the risk of bias in the selected studies using the Cochrane Risk of Bias Tool, and the analysis was conducted using Review Manager 5.4 software[14]. The assessment included seven areas: method of randomization, concealment of allocation, blinding of both participants and intervention providers, blinding of those assessing outcomes, completeness of data, selective reporting, and other possible biases. Each domain was classified as “low risk,” “unclear risk,” or “high risk.” Any discrepancies were addressed by a third researcher (DHL) who made the final inclusion decision. The evidence quality was evaluated with GRADEPro, and in the event of substantial differences or concerns regarding study inclusion, a third-party expert was consulted for resolution. Statistical analysis In the included studies, pairwise comparisons were used for two-arm trials. For multi-arm trials with three or more interventions, all possible pairwise comparisons were made for analysis. Bayesian models for NMA were constructed using the gemtc package in R Studio, which also generated treatment probability plots and ranked the interventions. For continuous outcomes, the standardized mean difference (SMD) along with its 95% confidence interval (CI) was calculated. When a closed loop of interventions was established, inconsistency tests were conducted to evaluate the agreement between direct and indirect comparisons, utilizing node-splitting for local analysis. A P value exceeding 0.05 indicates the absence of significant inconsistency. To assess the overall effectiveness of the interventions, the surface under the cumulative ranking curve (SUCRA) was determined for each treatment, and the interventions were ranked accordingly. For FMA-LE, BBS, FAC, and MBI, SUCRA values span from 0% to 100%, with 100% representing the optimal treatment effect and 0% reflecting the least effective intervention. For TUG, lower values indicate better intervention effects; therefore, a higher SUCRA value for TUG corresponds to poorer outcomes. Funnel plots were utilized to examine the possibility of publication bias or small sample size effects, with a visual check for symmetry to identify any systematic bias. Additionally, Egger's test was performed, with a P value below 0.05 suggesting the presence of publication bias. Results Literature Screening Four electronic databases provided a collection of 3,071 articles. After excluding 1,002 studies published before 2010 and 886 duplicate studies, 1,183 articles remained. The titles and abstracts of the studies were assessed, resulting in the exclusion of 970 articles. A total of 213 studies were read in full. After further examination, 132 additional studies were excluded. Ultimately, 82 studies[15–96] with 3,514 participants were included in the NMA. Figure 1 depicts the process of study selection through a flowchart, and the search queries and search results can be found in the appendix. Basic characteristics of the studies This study included 82 RCTs[15–96] published between 2010 and 2025, covering 14 countries and involving 3,514 participants. These trials evaluated 16 distinct interventions. The corresponding codes for the interventions are presented in Table 1. Specific interventions comprised: acupuncture (ACPU) (n=4)[31–34], vibration therapy (VT) (n=7)[91–96], mirror therapy (MT) (n=5)[15,16,18,19,21], constraint-induced movement therapy (CIMT) (n=3)[35–37], virtual reality therapy (VR) (n=12)[79–90], repetitive transcranial magnetic stimulation (rTMS) (n=5)[71–75], transcranial direct current stimulation (tDCS) (n=4)[41,76–78], functional electrical stimulation (FES) (n=10)[38–45], intermittent theta burst stimulation (iTBS) (n=3)[28–30], robotic assistive rehabilitation (RAR) (n=20)[47–54,56–65,68,69], water therapy (WT) (n=6)[22–27], MT+FES (n=3)[17,19,20], VR+RAR (n=4)[46,66,67,70], FES+RAR (n=2)[46,53], sham (n=17)[17,28–30,42,44,71–78,88], and conventional treatment (CT) (n=72)[15,16,18–27,31–41,43,45–70,79–87,89–96].Seven studies[19,42,45,46,71,73,94] employed a three-group design, whilst the remaining studies utilized a two-group design. The appendix contains detailed descriptions of the studies that were included. Table. 1. Codes linked to intervention actions. number code Intervention measures 1 A Conventional Therapy (CT) 2 B Acupuncture (ACPU) 3 C Vibration therapy (VT) 4 D Mirror therapy (MT) 5 E Constraint-induced movement therapy (CIMT) 6 F Virtual reality therapy (VR) 7 G intermittent theta burst stimulation (iTBS) 8 H Functional electrical stimulation (FES) 9 I Repetitive Transcranial Magnetic Stimulation (rTMS) 10 J Transcranial Direct Current Stimulation (tDCS) 11 K Robot-assisted rehabilitation (RAR) 12 L Water therapy (WT) 13 M sham 14 FK VR+RAR 15 KH RAR+FES 16 DH MT+FES Quality assessment of the included studies The Cochrane Risk of Bias Assessment Tool (Review Manager 5.4) was used to evaluate the risk of bias in 82 studies. Of these, 69 studies specified the methods used for random sequence generation and were categorized as ‘low risk,’ while the remaining 13 studies referenced randomization without providing specific details, and were labeled as ‘unclear risk of bias’. Allocation concealment was achieved using opaque sealed envelopes in 28 studies, which were classified as ‘low risk’; the other 54 studies did not specify their methods for allocation concealment and were categorized as having an ‘unclear risk of bias’. Owing to the inherent characteristics of certain non-pharmacological interventions, researchers encountered challenges in implementing blinding during the intervention process. Only 10 studies reported blinding of both participants and researchers, whereas 61 studies reported blinding of outcome assessments. 62 studies were rated as ‘low risk’ due to having complete data. All 82 studies reported pre-specified outcome measures, rated as ‘low risk’. None detailed other sources of bias, thus rated as ‘risk of bias unclear’. Figure 2 displays the findings from the quality assessment. Certainty of evidence Using the GRADEpro tool, the certainty of the evidence was assessed. The results, presented in attachment, demonstrated a high certainty for changes in FMA-LE, a moderate certainty for MBI, FAC and TUG, and a low certainty for BBS. The lower certainty levels were largely attributed to a high risk of bias, insufficient methodological rigor, and considerable heterogeneity among the studies. Primary outcome: FMA-LE Network plot of interventions Figure 3a shows a network diagram of 47 RCTs[15,16,18,28,29,31–35,38,39,41–43,46,48–50,54,57,59–61,63–66,68–80,83,86] evaluating the efficacy of FMA-LE under different individual interventions, involving 2,134 participants and 13 non-pharmacological treatments. The treatments are as follows: A: CT, B: ACPU, D: MT, E: CIMT, F: VR, G: iTBS, H: FES, I: rTMS, J: tDCS, K: RAR, FK: VR+RAR, KH: RAR + FES, M: sham. The network primarily focuses on conventional treatment. Interventions are represented by points, with lines connecting them to show direct comparisons. The width of these lines represents the number of studies included. Figure 4a presents the forest plot that compares conventional treatment A with each intervention. NMA The NMA results revealed 156 pairwise comparisons. The final network effect indicated that compared to conventional therapy that ACPU (SMD = −2.54; 95% CI: −3.8 to −1.3), MT (SMD = −3.43; 95% CI: −5.93 to −0.93), VR combined with RAR (SMD = −2.98; 95% CI: −3.98 to −2), rTMS (SMD = −3.68; 95% CI: −5.93 to −0.93), tDCS (SMD = −3.36; 95% CI: −4.86 to −2.49), RAR (SMD = −2.52; 95% CI: −3.63 to −1.4) and RAR combined with FES (SMD = −3; 95% CI: −4.48 to −1.52) were effective in improving lower limb motor function in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions. The detailed results of the NMA on FMA-LE can be found in the appendix. Ranking The SUCRA probability ranking showed that the comparative efficacy of treatments in improving FMA-LE was as follows: rTMS (88.45%) > tDCS (81.58%) > MT (78.80%) > RAR + FES (72.41%) > VR + RAR (72.00%) > ACPU (61.81%) > RAR (60.77%) > iTBS (36.53%) > CIMT (36.20%) > FES (24.40%) > VR (18.32%) > CT (14.93%) > sham (3.81%). The results suggest that rTMS may be the most effective method for improving FMA-LE scores in patients with lower limb hemiplegia following a stroke. The rankogram appears in figure 3b, with the SUCRA graph depicted in figure 5a. Secondary outcome: BBS Network evidence graph The figure 6a presents a network diagram of 59 RCTs[15,17,19–28,30,35,42,43,45,46,49,51–53,55,55–62,64,66,67,69–71,76,78,79,81–86,88–96] evaluating the efficacy of BBS under various individual interventions, involving a total of 2,075 participants and covering 15 non-pharmacological treatments (A: CT, B: ACPU, C: VT, D: MT, F: VR, E: CIMT, H: FES, I: rTMS, J: tDCS, K: RAR, L: WT, FK: VR + RAR, KH: RAR + FES, DH: MT + FES, M: sham). Figure 4b presents the forest plot that compares conventional treatment A with each intervention. NMA The NMA results revealed that 210 pairwise comparisons were produced. Compared with conventional therapy that VT (SMD = −2.79; 95% CI: −4.19 to −1.4), MT (SMD = −11.21; 95% CI: −16.73 to −5.71), CIMT (SMD = −2.87; 95% CI: −4.68 to −1.04), iTBS (SMD = −3.81; 95% CI: −7.28 to −0.32), FES (SMD = −5.33; 95% CI: −7.25 to −3.42), VR combined with RAR (SMD = −4.41; 95% CI: −5.33 to −3.49), rTMS (SMD = −12.32; 95% CI: −15.07 to −9.57), RAR (SMD = −1.82; 95% CI: −3.25 to −0.4), RAR combined with FES (SMD = −4.07; 95% CI: −5.26 to −2.89) and WT (SMD = −1.04; 95% CI: −1.48 to −0.6) were effective in improving balance function in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions. The detailed results of the NMA on BBS can be found in the appendix. Ranking The SUCRA probability ranking showed that the comparative efficacy of treatments in improving BBS was as follows: rTMS (96.2%) > MT (95.77%) >FES (79.65%) > VR + RAR (71.07%) > iTBS (67.47%) > RAR + FES (66.33%) > tDCS (53.62%) > CIMT (47.54%) > VT (46.86%) > WT (31.58%) > RAR (26.31%) > MT+FES (23.75%) > VR (21.47%) > sham (20.45%) > CT (1.61%). The results indicate that rTMS is the most effective method for improving balance function in post-stroke hemiplegic patients. The rankogram can be seen in figure 6b, and the SUCRA graph is illustrated in figure 5b. Secondary outcome: MBI Network evidence graph The figure 7a displays a network diagram from 26 RCTs[15,28,31,32,34,38,41,42,46,49,52,57,60,61,66,71–75,93] evaluating the efficacy of MBI under various individual interventions, involving a total of 1,218 participants and covering 12 non-pharmacological treatments (A: CT, B: ACPU, C: VT, D: MT G: iTBS, H: FES, I: rTMS, J: tDCS, K:RAR, FK: VR+ RAR, KH: RAR+ FES, M: sham). Figure 4c presents the forest plot that compares conventional treatment A with each intervention. NMA The NMA results showed that 132 pairwise comparisons were performed. Compared with conventional therapy that ACPU (SMD = −7.57; 95% CI: −12.61 to −2.49), MT (SMD = −9.83; 95% CI: −17.98 to −1.64), VR combined with RAR (SMD = −7.34; 95% CI: −8.94 to −5.74), rTMS (SMD = −13.51; 95% CI: −16.32 to −10.73), tDCS (SMD = −11.76; 95% CI: −17.72 to −5.78), RAR (SMD = −6.28; 95% CI: −11.41 to −1.1) and RAR combined with FES (SMD = −4.63; 95% CI: −7.26 to −2.06) were effective in enhancing daily living abilities in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions. The detailed results of the NMA on MBI can be found in the appendix. Ranking The SUCRA probability ranking shows that the comparative efficacy of treatments in improving MBI is as follows: rTMS (94.72%) > tDCS (86.22%) > MT (75.58%) > ACPU (64.43%) > VR + RAR (64.15%) > RAR (56.38%) > RAR + FES (44.52%) > iTBS (38.38%) > VT (29.66%) > FES (21.92%) > CT (20.53%) > sham (3.69%). The results indicate that rTMS is the most effective method for improving activities of daily living in post-stroke hemiplegic patients. In figure 7b, the rankogram is depicted, and Figure 5c shows the SUCRA graph. Secondary outcome: FAC Network evidence graph The figure 10a displays a network diagram from 22 RCTs[21,24,40–42,45,47–49,51,52,54,56,57,59,62,63,68,70,84,91] evaluating the efficacy of FAC under various individual interventions, involving a total of 920 participants and covering 8 non-pharmacological treatments (A: CT, C: VT, D: MT, H: FES, J: tDCS, K: RAR, L: WT, M: sham). Figure 4d presents the forest plot that compares conventional treatment A with each intervention. NMA The results of the NMA revealed that 56 pairwise comparisons were performed. Compared with conventional therapy that VT (SMD = −0.77; 95% CI: −1.3 to −0.25), MT (SMD = −1.45; 95% CI: −2.2 to −0.7), FES (SMD = −1.28; 95% CI: −2.31 to −0.24), tDCS (SMD = −1.47; 95% CI: −2.54 to −0.42) and RAR (SMD = −0.65; 95% CI: −0.81 to −0.5) were effective in enhancing daily living abilities in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions. The detailed results of the NMA on FAC can be found in the appendix. Ranking The SUCRA probability ranking shows that the comparative efficacy of treatments in improving MBI is as follows: tDCS (87.54%) > MT (82.25%) > FES (69.74%) > VT (45.58%) > WT (43.11%) > RAR (36.28%) > sham (32.09%) > CT (3.41%). The results indicate that tDCS is the most effective method for improving functional ambulation in post-stroke hemiplegic patients. The rankogram can be found in figure 8b, and the SUCRA graph is illustrated in figure 5d. Secondary outcome: TUG Network plot of interventions The figure 9a displays a network diagram from 34 RCTs[20,22,23,25–29,36,37,39,43,46,47,53,57,65,77,81–85,87–96] evaluating the efficacy of FAC under various individual interventions, involving a total of 1,422 participants and covering 10 non-pharmacological treatments (A: CT, C: VT, E: CIMT, F: VR, H: FES, K: RAR, L: WT, DH: MT+FES, FK: VR+RAR, KH: RAR+FES). Figure 4e presents the forest plot that compares conventional treatment A with each intervention. NMA The results of the NMA revealed that 90 pairwise comparisons were performed. Compared with conventional therapy that VR combined with RAR (SMD = 6.39; 95% CI: 4.56 to 7.99), MT combined with FES (SMD = 5.26; 95% CI: 2.97 to 7.6), VT (SMD = 2.93; 95% CI: 1.96 to 3.91), FES (SMD = 2.98; 95% CI: 0.33 to 5.62) and WT (SMD = 2.04; 95% CI: 1.32 to 2.76) were effective in shortening the TUG. in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions. The detailed results of the NMA on TUG can be found in the appendix. Ranking The SUCRA probability ranking shows that the comparative efficacy of treatments in improving TUG is as follows: VR + RAR (5.05%) > RAR + FES (13.43%) > VT (35.87%) > FES (39.40%) > WT (53.67%) > RAR+FES (61.20%)> VR (61.48%) > CIMT (67.45%) > RAR (74.47%) > CT (88.05%). The results indicate that VR + RAR is the most effective method for improving TUG in post-stroke hemiplegic patients. In figure 9b, the rankogram is depicted, and figure 5e shows the SUCRA graph. Consistency analysis The assessment of network inconsistency was conducted rigorously through node-splitting analysis when applicable. For TUG, the network structure lacked closed loops, making inconsistency assessment irrelevant. The results of FMA-LE, BBS, MBI, and FAC underwent node-splitting analysis to evaluate the direct and indirect treatment comparison estimates from two types of evidence. There were no significant inconsistencies found in any comparisons, with all P-values being much higher than 0.05, which indicates a high level of concordance between the direct and indirect evidence, supporting the validity of our NMA assumptions. The specific details can be found in the appendix. Assessment of Publication Bias Utilizing a random-effects model, a corrected contrast funnel plot was developed. Overall, the distribution of study points appears relatively symmetric, with no significant signs of small-study effects or publication bias observed. It should be noted that for certain interventions within the network, the number of available studies is limited, which may affect the interpretability of the funnel plot. Hence, the associated results should be viewed with caution. Discussion Lower limb motor impairment is a common and severe outcome for individuals who have suffered a stroke. Following stroke, cortical spinal excitability in the damaged motor cortex typically diminishes markedly, whilst excitability in the contralateral region frequently increases, accompanied by heightened interhemispheric inhibition[ 97 ]. This asymmetrical alteration in excitability significantly impairs patients' gait and balance maintenance capabilities. Neuroimaging studies indicate that walking and balance depend not only on cortical activity but also require coordinated regulation involving subcortical structures and the spinal cord[ 98 ]. However, rehabilitation treatment targeting these complex sensorimotor functions continues to present substantial clinical challenges. Therefore, identifying safe, effective, and targeted rehabilitation methods to improve lower limb mobility and overall recovery post-stroke has become a central research area in stroke rehabilitation. To the best of our knowledge, this is the initial network meta-analysis that examines various non-drug interventions and assesses five outcome measures (FMA-LE, BBS, MBI, FAC, and TUG) related to lower limb motor function in hemiplegic patients. This NMA incorporated 82 RCTs involving 3,541 participants, comparing 16 distinct interventions to evaluate non-pharmacological approaches for enhancing motor abilities in the lower limbs of individuals with post-stroke hemiplegia. All included studies underwent strict quality and risk assessments. Our study demonstrates that rTMS has a significant advantage in improving FMA-LE in post-stroke hemiplegic patients, achieving a notable SUCRA ranking of 88.45%. Following rTMS, tDCS and mirror therapy also show promising results. Consistent with previous meta-analyses, rTMS demonstrates consistent efficacy across all stages of stroke rehabilitation, significantly enhancing lower limb motor function. rTMS is a non-invasive technique for brain stimulation that modulates neuronal membrane potential through pulsed magnetic fields, enabling targeted regulation of neural excitability in particular brain regions[ 99 ]. rTMS has been widely used in managing neurological and psychiatric conditions, such as cognitive deficits, depression, psychosis, and Alzheimer's disease[ 100 – 102 ]. Due to its painless and straightforward administration, it is a valuable tool for accurately modulating brain regions and predicting stroke rehabilitation success[ 103 ]. Its effects extend beyond the stimulation site, influencing cortical excitability in distant areas via synaptic transmission. The findings of this study suggest that rTMS exhibits greater effectiveness in enhancing lower limb motor function (FMA-LE) and improving balance (BBS). rTMS enhances excitability in the ipsilateral primary motor cortex through high-frequency stimulation (≥ 3 Hz) or suppresses excessive activation in the contralateral hemisphere via low-frequency stimulation (≤ 1 Hz) [ 104 , 105 ]. This restores the excitatory-inhibitory balance between both hemispheres, serving as the core mechanism for promoting recovery of motor control and coordination functions. Earlier research has demonstrated that rTMS can modulate the central nervous system neurotransmitter system, improving the balance between excitatory neurotransmitters such as glutamate and inhibitory neurotransmitters like γ-aminobutyric acid (GABA), enhancing synaptic plasticity[ 106 , 107 ]. Moreover, rTMS influences not only neurons but also plays a significant role in modulating glial cell activity, facilitating the shift of astrocytes from the pro-inflammatory A1 subtype to the neuroprotective A2 subtype, and promoting the polarization of microglia from the M1 to the M2 phenotype[ 108 ]. This process alleviates neuroinflammation, inhibits neuronal apoptosis, and preserves the integrity of the blood-brain barrier. In addition, rTMS can downregulate the levels of pro-inflammatory cytokines such as tumor necrosis factor-α, interleukin-1β, and interleukin-6, while upregulating the expression of anti-inflammatory cytokines such as interleukin-10 and transforming growth factor-β[ 109 ]. The aforementioned mechanisms collectively form a synergistic network: the restoration of hemispheric balance and enhanced synaptic plasticity directly drive improvements in motor function, while optimization of the neuroimmune microenvironment provides sustained biological support for functional remodeling. This may represent the intrinsic reason for rTMS' s superior performance on both the FMA-LE and BBS outcomes in this study. However, This study relies on only five RCTs with small sample sizes and does not include subgroup analyses of rTMS stimulation frequencies or examine the specific effects at different frequencies, which limits the accuracy and generalizability of the findings. Thus, the results should be viewed with caution. Future research urgently requires larger-scale, high-quality RCTs to explore optimal rTMS stimulation parameters and protocols, and to validate its long-term efficacy in improving lower limb function in patients with post-stroke hemiplegia. tDCS has demonstrated favorable outcomes in advancing motor skills in the lower extremities of patients who have hemiplegia due to a stroke. tDCS is a non-invasive method that shows potential for stimulating neuroplasticity within the brain's cortex and supporting recovery after a stroke[ 110 ]. tDCS modulates neuronal membrane potential and alters ion distribution across the membrane through the application of a mild direct current to the brain's cortex. This stimulates neuroplastic processes, thus facilitating the restoration of neural function in the brain tissue of stroke survivors[ 111 ]. tDCS effectively re-establishes the balance of inhibitory competition between the two hemispheres by increasing excitability in the affected hemisphere through the anode or decreasing excitability in the opposite hemisphere using the cathode[ 112 ]. Two meta-analyses indicate that tDCS demonstrates good safety and tolerability[ 113 , 114 ]. Mirror therapy ranked third in the primary outcome measure FMA-LE, demonstrating a degree of efficacy. Mirror therapy is an innovative intervention that uses mirror visual feedback (MVF) to provide new visual input, stimulating the mirror neuron system and helping to reduce lower limb motor impairments after a stroke[ 115 ]. In conventional mirror therapy, the unaffected limb is placed in the reflected area of the mirror, while the impaired limb is positioned behind the mirror, out of sight. This visual illusion, or MVF, creates the perception of both limbs functioning normally, effectively “deceiving” the brain and stimulating the sensory-motor areas, thereby promoting the recovery of motor ability[ 116 ]. Mirror therapy's benefits stem from its simple implementation, needing no advanced equipment or specialized oversight, while being non-invasive throughout the entire treatment process. Even patients with severe movement disorders can independently administer therapy at home, offering flexibility and convenience[ 117 ]. By activating the brain's motor regions through visual feedback, this approach not only enhances patients' capacity for self-directed rehabilitation but also permits uninterrupted practice free from external distractions. Consequently, it effectively improves rehabilitation outcomes and ensures treatment sustainability. This study included three combined therapies, with the regimens combining virtual reality therapy with robotic therapy and robotic therapy with FES ranking highly. Robots for stroke rehabilitation are primarily categorized into therapeutic robots and assistive robots[ 118 ]. Therapeutic robots are primarily used to promote the recovery of patients' motor functions through task-specific training, with common types including end-effector and exoskeleton robots[ 119 ]. Another category is assistive robots, which provide external assistance to help patients perform daily activities, achieving functional compensation. Robotic rehabilitation offers high-intensity, high-dose, and highly repetitive task-oriented training, creating the necessary physical stimuli for neuroplasticity, thereby promoting the reorganization and activation of damaged motor neural networks[ 120 ]. However, robotic rehabilitation typically exhibits a high degree of passivity. As training duration increases, this passive training model may lead to a decline in muscle activation levels and reduce patients' willingness to actively participate[ 121 ]. In contrast, virtual reality technology, with its immersive experience, not only effectively enhances patients' training adherence but also further optimizes the rehabilitation process by promoting active engagement[ 122 ]. The integration of virtual reality and robotic technology offers a complementary rehabilitation approach with multiple advantages. Virtual reality therapy promotes neuroplasticity by providing real-time feedback, adjustable task difficulty, and diverse contextual training, which helps activate brain regions associated with motor control and cognition. Meanwhile, robotic therapy reinforces the establishment and consolidation of neural circuits through repetitive goal-oriented motor tasks. Particularly when patients have limited voluntary movement ability, the mechanical assistance and repetitive support provided by robots facilitate relearning and functional reorganization of impaired neural networks. By combining both technologies, it is possible not only to enhance patient motivation and training adaptability through virtual reality but also to deliver precise and stable physical training support via robotic systems. This integrated approach strengthens active patient engagement while improving overall rehabilitation efficiency and functional recovery outcomes. Functional electrical stimulation (FES) induces muscle contractions by applying electrical currents to impaired nerves and muscles 117 . The integration of FES with robotic rehabilitation technology creates a synergistic therapeutic paradigm, with its principal advantages manifested in the following aspects: At the neuromuscular activation level, FES enables targeted activation of specific muscle groups, effectively compensating for potential insufficient muscle activation during robotic training. This is particularly beneficial for patients with significantly limited voluntary motor ability. In terms of motor control, robotic systems provide precise trajectory guidance and mechanical support, ensuring normative movement patterns while supplying real-time kinematic parameters for the temporal control of FES. This facilitates high-precision synchronization between electrical stimulation and motor execution. Furthermore, the combination of the two technologies can establish a closed-loop "stimulation-movement-feedback" training system. By enhancing the consistency between proprioceptive input and motor output, it promotes sensorimotor integration and neuroplasticity. Additionally, the integrated approach allows for individualized parameter adjustment, enabling dynamic modulation of robotic assistance intensity and FES stimulation parameters based on the patient's functional status. This ensures training intensity while optimizing energy efficiency and reducing the development of abnormal compensatory movement patterns. Currently, combination therapy has become a significant research direction in the field of post-stroke motor function recovery. By integrating the complementary mechanisms of different intervention approaches, this model holds the potential to overcome the limitations of single-modality therapies and achieve multi-level, multi-target synergistic rehabilitation effects. However, this area remains in the early stages of exploration, with relatively limited existing research. Future efforts should focus on conducting more high-quality clinical studies to clarify its long-term efficacy and gradually establish standardized, generalizable treatment parameters and implementation protocols. Given the limitations of traditional pharmacological treatments in rehabilitating lower limb function in stroke patients with hemiplegia—including frequent side effects, high dependency, and uncertain long-term efficacy—non-pharmacological therapies have increasingly become important complementary or even alternative strategies. Compared to drug-based interventions, non-pharmacological approaches not only alleviate symptoms but also promote sustained functional recovery by enhancing neuroplasticity and motor learning, while avoiding risks such as drug dependence. These interventions primarily focus on lifestyle modifications, systematic exercise training, physical stimulation, and neuromodulation techniques, directly targeting motor dysfunction to improve muscular coordination and promote neural network reorganization. Notably, non-pharmacological therapies offer significant advantages in improving patient treatment compliance and clinical applicability. This strategy emphasizes active patient participation and self-management, contributing to enhanced rehabilitation confidence and treatment continuity. In this context, Bayesian network meta-analysis provides a rigorous methodological framework for systematically evaluating various non-pharmacological interventions. By quantifying the efficacy of different treatment measures and enabling indirect comparisons, it offers evidence-based support for selecting optimal clinical treatment plans. In summary, non-pharmacological therapies not only demonstrate outstanding cost-effectiveness and safety but also promote comprehensive recovery of lower extremity movement through multidimensional and integrated rehabilitation pathways, showing significant clinical application potential. Additional high-quality studies are required to better understand the combined effects of various therapies and to develop standardized clinical treatment protocols, thus promoting the consistent use and broader adoption of this approach in rehabilitation practice. Advantages and Limitations The primary strength of this study lies in its integration of multiple non-pharmacological interventions to provide a more comprehensive analysis, representing a first in the existing literature within this field. Previous studies have typically focused on specific non-pharmacological interventions, whereas this research fills a gap by comprehensively evaluating the effects of different interventions. However, despite efforts to ensure comprehensiveness in the literature search and the absence of significant publication bias indicated by funnel plots, the possibility of overestimation of intervention effects in unpublished studies or published reports cannot be entirely ruled out. Furthermore, the number of studies for some interventions was relatively limited, with smaller sample sizes, necessitating caution in interpreting related findings. Third, given the nature of non-pharmacological interventions, difficulties in implementing blinding for both participants and researchers undoubtedly increase the risk of potential bias, potentially affecting the accuracy of results and efficacy assessments. Finally, data limitations in certain studies resulted in inadequate evaluation of some interventions, thereby restricting our comprehensive understanding of their therapeutic effects. Consequently, future research should prioritize expanding sample sizes, adopting more rigorous randomized designs, and implementing blinding protocols to generate more reliable evidence that provides stronger support for clinical application. Conclusion This research carried out a comprehensive assessment of the effectiveness of non-pharmacological treatments targeting lower limb motor impairments in individuals with post-stroke hemiplegia. The results indicate that rTMS may be the most effective intervention for improving lower limb motor function, followed by tDCS and mirror therapy. For patients exhibiting functional limitations such as reduced walking speed and poor dynamic balance control, a comprehensive training program combining virtual reality technology with robot-assisted rehabilitation is recommended. "Nevertheless, because of constraints in both the amount and quality of existing evidence—including limited sample sizes and possible bias risks—the interpretation and broader applicability of these findings should be treated with caution. Future studies are urgently needed to conduct larger, methodologically sound, high-quality randomized controlled trials to further confirm and refine the clinical implementation of various intervention strategies. Abbreviations Network Meta-Analysis: NMA Randomized Controlled Trial: RCT Fugl-Meyer Assessment for Lower Extremities: FMA-LE Berg Balance Scale: BBS Modified Barthel Index: MBI Functional Ambulation Classification: FAC Timed Up and Go test: TUG Surface Under the Cumulative Ranking curve Analysis: SUCRA Standardized Mean Difference: SMD Conventional Therapy: CT Acupuncture: ACPU Vibration therapy: VT Mirror therapy: MT Constraint-induced movement therapy: CIMT Virtual reality: VR intermittent theta burst stimulation: iTBS Functional electrical stimulation: FES Repetitive Transcranial Magnetic Stimulation: rTMS Transcranial Direct Current Stimulation: tDCS Robot-assisted rehabilitation: RAR Water therapy: WT mirror visual feedback: MVF Declarations Ethics approval and consent to participate Not applicable Consent for publication All authors have approved the submission of this manuscript. This work is original, has not been published previously, and is not under consideration for publication elsewhere. Availability of data and materials The primary contributions of this study are detailed in the article or supplementary material. For further inquiries, please reach out to the corresponding author. Competing interests The authors state that there are no competing interests associated with this manuscript. Funding This study is funded by the National Natural Science Foundation of China (82575221) and the National Natural Science Foundation of China (82374602). Authors' contributions Study objective: DZ, DHL. Literature search: DZ, TYY. Data extraction: DZ, TYY. Methodological quality assessment: DZ, GXX, LW, GLZ. Data Analysis: GXX, LW, WYS, MYY. Manuscript writing: DZ. Critical review and manuscript approval: JL, DHL. All authors have reviewed and approved the final manuscript. Acknowledgement We gratefully acknowledge the support from the National Natural Science Foundation of China. References Marsh JD, Keyrouz SG. Stroke prevention and treatment. J Am Coll Cardiol. 2010; 56(9):683–691. Feigin VL, Brainin M, Norrving B, Martins SO, Pandian J, Lindsay P, et al. World Stroke Organization: Global Stroke Fact Sheet 2025. Int J Stroke Off J Int Stroke Soc. 2025; 20(2):132–144. GBD 2019 Stroke Collaborators. Global, regional, and national burden of stroke and its risk factors, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Neurol. 2021; 20(10):795–820. Feigin VL, Owolabi MO, World Stroke Organization–Lancet Neurology Commission Stroke Collaboration Group. Pragmatic solutions to reduce the global burden of stroke: a World Stroke Organization-Lancet Neurology Commission. Lancet Neurol. 2023; 22(12):1160–1206. Xiong J, Wang J-T, Lin S, Xie B-Y. Advances in hemiplegia rehabilitation: modern therapeutic interventions to enhance activities of daily living. Front Neurol. 2025; 16:1555990. Liu Z, Xin H, Chopp M. Axonal remodeling of the corticospinal tract during neurological recovery after stroke. Neural Regen Res. 2021; 16(5):939–943. Joy MT, Carmichael ST. Encouraging an excitable brain state: mechanisms of brain repair in stroke. Nat Rev Neurosci. 2021; 22(1):38–53. Li S. Stroke Recovery Is a Journey: Prediction and Potentials of Motor Recovery after a Stroke from a Practical Perspective. Life Basel Switz. 2023; 13:2061. Strilciuc S, Grad DA, Radu C, Chira D, Stan A, Ungureanu M, et al. The economic burden of stroke: a systematic review of cost of illness studies. J Med Life. 2021; 14(5):606–619. Jørgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil. 1995; 76(1):27–32. Salanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods. 2012; 3(2):80–97. Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med. 2015; 162(11):777–784. Amir-Behghadami M, Janati A. Population, Intervention, Comparison, Outcomes and Study (PICOS) design as a framework to formulate eligibility criteria in systematic reviews. Emerg Med J EMJ. 2020; 37(6):387. Whiting P, Savović J, Higgins JPT, Caldwell DM, Reeves BC, Shea B, et al. ROBIS: A new tool to assess risk of bias in systematic reviews was developed. J Clin Epidemiol. 2016; 69:225–234. Cui W, Huang L, Tian Y, Luo H, Chen S, Yang Y, et al. Effect and mechanism of mirror therapy on lower limb rehabilitation after ischemic stroke: A fMRI study. NeuroRehabilitation. 2022; 51(1):65–77. Arya KN, Pandian S, Kumar V. Effect of activity-based mirror therapy on lower limb motor-recovery and gait in stroke: A randomized controlled trial. Neuropsychol Rehabil. 2019; 29(8):1193–1210. Lee D, Lee G. Effect of afferent electrical stimulation with mirror therapy on motor function, balance, and gait in chronic stroke survivors: a randomized controlled trial. Eur J Phys Rehabil Med. 2019; 55(4):442–449. Mohan U, Babu SK, Kumar KV, Suresh BV, Misri ZK, Chakrapani M. Effectiveness of mirror therapy on lower extremity motor recovery, balance and mobility in patients with acute stroke: A randomized sham-controlled pilot trial. Ann Indian Acad Neurol. 2013; 16(4):634–639. Kim D-H, Jang S-H. Effects of Mirror Therapy Combined with EMG-Triggered Functional Electrical Stimulation to Improve on Standing Balance and Gait Ability in Patient with Chronic Stroke. Int J Environ Res Public Health. 2021; 18(7):3721. Lee D, Lee G, Jeong J. Mirror Therapy with Neuromuscular Electrical Stimulation for improving motor function of stroke survivors: A pilot randomized clinical study. Technol Health Care Off J Eur Soc Eng Med. 2016; 24(4):503–511. İkizler May H, Özdolap Ş, Mengi A, Sarıkaya S. The effect of mirror therapy on lower extremity motor function and ambulation in post-stroke patients: A prospective, randomized-controlled study. Turk J Phys Med Rehabil. 2020; 66(2):154–160. Kim K, Lee D-K, Kim E-K. Effect of aquatic dual-task training on balance and gait in stroke patients. J Phys Ther Sci. 2016; 28(7):2044–2047. Kim K, Lee D-K, Jung S-I. Effect of coordination movement using the PNF pattern underwater on the balance and gait of stroke patients. J Phys Ther Sci. 2015; 27(12):3699–3701. Tripp F, Krakow K. Effects of an aquatic therapy approach (Halliwick-Therapy) on functional mobility in subacute stroke patients: a randomized controlled trial. Clin Rehabil. 2014; 28(5):432–439. Eyvaz N, Dundar U, Yesil H. Effects of water-based and land-based exercises on walking and balance functions of patients with hemiplegia. NeuroRehabilitation. 2018; 43(2):237–246. Zhu Z, Cui L, Yin M, Yu Y, Zhou X, Wang H, et al. Hydrotherapy vs. conventional land-based exercise for improving walking and balance after stroke: a randomized controlled trial. Clin Rehabil. 2016; 30(6):587–593. Park J, Chung Y. The effects of an additional weight aquatic exercise program on balance and lower extremity strength in persons with stroke: randomized controlled study. Phys Ther Rehabil Sci. Korean Academy of Physical Therapy Rehabilitation Science; 2018; 7(1):6–12. ZHU P-A, LI Z-L, LU Q-Q, NIE Y-Y, LIU H, KIERNAN E, et al. Can cerebellar theta-burst stimulation improve balance function and gait in stroke patients? A randomized controlled trial. Eur J Phys Rehabil Med. 2024; 60(3):391–399. Xie Y-J, Wei Q-C, Chen Y, Liao L-Y, Li B-J, Tan H-X, et al. Cerebellar Theta Burst Stimulation on Walking Function in Stroke Patients: A Randomized Clinical Trial. Front Neurosci. 2021; 15:688569. Koch G, Bonnì S, Casula EP, Iosa M, Paolucci S, Pellicciari MC, et al. Effect of Cerebellar Stimulation on Gait and Balance Recovery in Patients With Hemiparetic Stroke. JAMA Neurol. 2019; 76(2):170–178. Wang H-Q, Hou M, Bao C-L, Min L, Li H. Effects of Acupuncture Treatment on Lower Limb Spasticity in Patients Following Hemorrhagic Stroke: A Pilot Study. Eur Neurol. 2019; 81(1-2):5–12. Wang H-Q, Hou M, Li H, Bao C-L, Min L, Dong G-R, et al. Effects of acupuncture treatment on motor function in patients with subacute hemorrhagic stroke: A randomized controlled study. Complement Ther Med. 2020; 49:102296. Wang N, Li Z. Life quality improvement of spastic hemiplegia of stroke treated with fire-needle: a randomized controlled trial. Zhongguo Zhen Jiu Chin Acupunct Moxibustion. 2015; 35(11):1105–1109. Bai Y-L, Li L, Hu Y-S, Wu Y, Xie P-J, Wang S-W, et al. Prospective, randomized controlled trial of physiotherapy and acupuncture on motor function and daily activities in patients with ischemic stroke. J Altern Complement Med N Y N. 2013; 19(8):684–689. Aloraini SM. Effects of constraint-induced movement therapy for the lower extremity among individuals post-stroke: A randomized controlled clinical trial. NeuroRehabilitation. 2022; 51(3):421–431. Choi H-S, Shin W-S, Bang D-H, Choi S-J. Effects of Game-Based Constraint-Induced Movement Therapy on Balance in Patients with Stroke: A Single-Blind Randomized Controlled Trial. Am J Phys Med Rehabil. 2017; 96(3):184–190. Menezes-Oliveira E, da Silva Matuti G, de Oliveira CB, de Freitas SF, Miyuki Kawamura C, Fernandes Lopes JA, et al. Improvement of gait and balance function in chronic post-stroke patients induced by Lower Extremity - Constraint Induced Movement Therapy: a randomized controlled clinical trial. Brain Inj. 2024; 38(7):559–568. Zhang X-H, Liu J-Y, Han P, Wang Y-L, Xiao P. Clinical Efficacy of Functional Electrical Stimulation-assisted RehabilitationCycling on the Function of Lower Limbs in Patients with Stroke. Curr Neurovasc Res. 2021; 18(3):318–323. Matsumoto S, Shimodozono M, Noma T, Miyara K, Onoda T, Ijichi R, et al. Effect of Functional Electrical Stimulation in Convalescent Stroke Patients: A Multicenter, Randomized Controlled Trial. J Clin Med. 2023; 12(7):2638. Bauer P, Krewer C, Golaszewski S, Koenig E, Müller F. Functional electrical stimulation-assisted active cycling--therapeutic effects in patients with hemiparesis from 7 days to 6 months after stroke: a randomized controlled pilot study. Arch Phys Med Rehabil. 2015; 96(2):188–196. Zhang X-H, Gu T, Liu X-W, Han P, Lv H-L, Wang Y-L, et al. The Effect of Transcranial Direct Current Stimulation and Functional Electrical Stimulation on the Lower Limb Function of Stroke Patients. Front Neurosci. 2021; 15:685931. Tan Z, Liu H, Yan T, Jin D, He X, Zheng X, et al. The effectiveness of functional electrical stimulation based on a normal gait pattern on subjects with early stroke: a randomized controlled trial. BioMed Res Int. 2014; 2014: 545408. Lee H-J, Cho K-H, Lee W-H. The effects of body weight support treadmill training with power-assisted functional electrical stimulation on functional movement and gait in stroke patients. Am J Phys Med Rehabil. 2013; 92(12):1051–1059. Lee Y, Kim K. The influence of Gait Training Combined with Portable Functional Electrical Stimulation on motor function, balance and gait ability in stroke patients. J Back Musculoskelet Rehabil. 2022; 35(6):1171–1178. Cho M-K, Kim J-H, Chung Y, Hwang S. Treadmill gait training combined with functional electrical stimulation on hip abductor and ankle dorsiflexor muscles for chronic hemiparesis. Gait Posture. 2015; 42(1):73–78. Park J. Comparison of the effects of robot-assisted gait training using FES versus virtual reality for stroke patients. A randomized clinical trial. Top Stroke Rehabil. 2025; 13:1–10. Meng G, Ma X, Chen P, Xu S, Li M, Zhao Y, et al. Effect of early integrated robot-assisted gait training on motor and balance in patients with acute ischemic stroke: a single-blinded randomized controlled trial. Ther Adv Neurol Disord. 2022; 15:17562864221123195. Li D-X, Zha F-B, Long J-J, Liu F, Cao J, Wang Y-L. Effect of Robot Assisted Gait Training on Motor and Walking Function in Patients with Subacute Stroke: A Random Controlled Study. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc. 2021; 30(7):105807. Huo C, Shao G, Chen T, Li W, Wang J, Xie H, et al. Effectiveness of unilateral lower-limb exoskeleton robot on balance and gait recovery and neuroplasticity in patients with subacute stroke: a randomized controlled trial. J Neuroengineering Rehabil. 2024; 21(1):213. Fan T, Zheng P, Zhang X, Gong Z, Shi Y, Wei M, et al. Effects of exoskeleton rehabilitation robot training on neuroplasticity and lower limb motor function in patients with stroke. BMC Neurol. 2025; 25(1):193. Kim S-Y, Yang L, Park IJ, Kim EJ, JoshuaPark MS, You SH, et al. Effects of Innovative WALKBOT Robotic-Assisted Locomotor Training on Balance and Gait Recovery in Hemiparetic Stroke: A Prospective, Randomized, Experimenter Blinded Case Control Study With a Four-Week Follow-Up. IEEE Trans Neural Syst Rehabil Eng Publ IEEE Eng Med Biol Soc. 2015; 23(4):636–642. Kim J, Kim DY, Chun MH, Kim SW, Jeon HR, Hwang CH, et al. Effects of robot-(Morning Walk®) assisted gait training for patients after stroke: a randomized controlled trial. Clin Rehabil. 2019; 33(3):516–523. Bae Y-H, Ko YJ, Chang WH, Lee JH, Lee KB, Park YJ, et al. Effects of Robot-assisted Gait Training Combined with Functional Electrical Stimulation on Recovery of Locomotor Mobility in Chronic Stroke Patients: A Randomized Controlled Trial. J Phys Ther Sci. 2014; 26(12):1949–1953. Chang WH, Kim MS, Huh JP, Lee PKW, Kim Y-H. Effects of robot-assisted gait training on cardiopulmonary fitness in subacute stroke patients: a randomized controlled study. Neurorehabil Neural Repair. 2012; 26(4):318–324. Kim D-H, In T-S, Jung K-S. Effects of robot-assisted trunk control training on trunk control ability and balance in patients with stroke: A randomized controlled trial. Technol Health Care Off J Eur Soc Eng Med. 2022; 30(2):413–422. Yoo M, Chun MH, Hong GR, Lee C, Lee JK, Lee A. Effects of Training with a Powered Exoskeleton on Cortical Activity Modulation in Hemiparetic Chronic Stroke Patients: A Randomized Controlled Pilot Trial. Arch Phys Med Rehabil. 2023; 104(10):1620–1629. Min JH, Seong HY, Ko SH, Jo W-R, Sohn H-J, Ahn YH, et al. Effects of trunk stabilization training robot on postural control and gait in patients with chronic stroke: a randomized controlled trial. Int J Rehabil Res Int Z Rehabil Rev Int Rech Readaptation. 2020; 43(2):159–166. Jin Y, Xiong B, Chen L, Zhao W, Li Z, Zhang C, et al. Efficacy and safety of using a unilateral lower limb exoskeleton combined with conventional treatment in post-stroke rehabilitation: a randomized controlled trial. Front Bioeng Biotechnol. 2024; 12:1441986. Louie DR, Mortenson WB, Durocher M, Schneeberg A, Teasell R, Yao J, et al. Efficacy of an exoskeleton-based physical therapy program for non-ambulatory patients during subacute stroke rehabilitation: a randomized controlled trial. J Neuroengineering Rehabil. 2021; 18(1):149. Haolin T, Yuanbin Y, Hu Z, Wenjing Z, Jing Z, Jingfeng T, et al. Efficacy of Daoyin combined with lower limb robot as a comprehensive rehabilitation intervention for stroke patients: a randomized controlled trial. J Tradit Chin Med Chung Tsa Chih Ying Wen Pan. 2024; 44(3):530–536. Zhang Y, Zhao W, Wan C, Wu X, Huang J, Wang X, et al. Exoskeleton rehabilitation robot training for balance and lower limb function in sub-acute stroke patients: a pilot, randomized controlled trial. J Neuroengineering Rehabil. 2024; 21(1):98. Nam Y-G, Park JW, Lee HJ, Nam KY, Choi MR, Yu CS, et al. Further effects of electromechanically assisted gait trainer (Exowalk®) in patients with chronic stroke: A randomized controlled trial. J Rehabil Med. 2020; 52(9): jrm00097. Ochi M, Wada F, Saeki S, Hachisuka K. Gait training in subacute non-ambulatory stroke patients using a full weight-bearing gait-assistance robot: A prospective, randomized, open, blinded-endpoint trial. J Neurol Sci. 2015; 353(1-2):130–136. Palmcrantz S, Wall A, Vreede KS, Lindberg P, Danielsson A, Sunnerhagen KS, et al. Impact of Intensive Gait Training With and Without Electromechanical Assistance in the Chronic Phase After Stroke–A Multi-Arm Randomized Controlled Trial With a 6 and 12 Months Follow Up. Front Neurosci. 2021; 15: 660726. Bustamante Valles K, Montes S, Madrigal M de J, Burciaga A, Martínez ME, Johnson MJ. Technology-assisted stroke rehabilitation in Mexico: a pilot randomized trial comparing traditional therapy to circuit training in a Robot/technology-assisted therapy gym. J Neuroengineering Rehabil. 2016; 13(1):83. Park J, Chung Y. The effects of robot-assisted gait training using virtual reality and auditory stimulation on balance and gait abilities in persons with stroke. NeuroRehabilitation. 2018; 43(2):227–235. Kayabinar B, Alemdaroğlu-Gürbüz İ, Yilmaz Ö. The effects of virtual reality augmented robot-assisted gait training on dual-task performance and functional measures in chronic stroke: a randomized controlled single-blind trial. Eur J Phys Rehabil Med. 2021; 57(2):227–237. Zhang H, Li X, Gong Y, Wu J, Chen J, Chen W, et al. Three-Dimensional Gait Analysis and sEMG Measures for Robotic-Assisted Gait Training in Subacute Stroke: A Randomized Controlled Trial. BioMed Res Int. 2023; 2023:7563802. Lee H-J, Lee S-H, Seo K, Lee M, Chang WH, Choi B-O, et al. Training for Walking Efficiency With a Wearable Hip-Assist Robot in Patients With Stroke: A Pilot Randomized Controlled Trial. Stroke. 2019; 50(12):3545–3552. Chen L, Zhu H, Wang J, Lu R, Tian J, Wu B, et al. Virtual reality-based robotic training for lower limb rehabilitation in stroke patients with Hemiplegia: A pilot study. Aging Health Res. 2025; 5:100233. Wang L, Wang L, Wang Z, Gao F, Wu J, Tang H. Clinical Effect Analysis of Wearable Sensor Technology-Based Gait Function Analysis in Post-Transcranial Magnetic Stimulation Stroke Patients. Sensors. 2024; 24(10):3051. Guan Y-Z, Li J, Zhang X-W, Wu S, Du H, Cui L-Y, et al. Effectiveness of repetitive transcranial magnetic stimulation (rTMS) after acute stroke: A one-year longitudinal randomized trial. CNS Neurosci Ther. 2017; 23(12):940–946. Lin Y-N, Hu C-J, Chi J, Lin L-F, Yen T-H, Lin Y-K, et al. Effects of repetitive transcranial magnetic stimulation of the unaffected hemisphere leg motor area in patients with subacute stroke and substantial leg impairment: A pilot study. J Rehabil Med. 2015; 47(4):305–310. Lu S, Wang H, Yuan Y, Ma T, Liu Z, Li Q, et al. Effects of repetitive transcranial magnetic stimulation on contralesional dorsal premotor cortex in subcortical stroke patients with moderate to severe upper limb motor impairment: a pilot randomized controlled trial. medRxiv. 2025; 2025.03.05.25323209. Huang Y-Z, Lin L-F, Chang K-H, Hu C-J, Liou T-H, Lin Y-N. Priming With 1-Hz Repetitive Transcranial Magnetic Stimulation Over Contralesional Leg Motor Cortex Does Not Increase the Rate of Regaining Ambulation Within 3 Months of Stroke: A Randomized Controlled Trial. Am J Phys Med Rehabil. 2018; 97(5):339–345. Youssef H, Mohamed NAE-H, Hamdy M. Comparison of bihemispheric and unihemispheric M1 transcranial direct current stimulations during physical therapy in subacute stroke patients: A randomized controlled trial. Neurophysiol Clin Clin Neurophysiol. 2023; 53(3):102895. Duan Q, Liu W, Yang J, Huang B, Shen J. Effect of Cathodal Transcranial Direct Current Stimulation for Lower Limb Subacute Stroke Rehabilitation. Neural Plast. 2023; 2023:1863686. Chang MC, Kim DY, Park DH. Enhancement of Cortical Excitability and Lower Limb Motor Function in Patients With Stroke by Transcranial Direct Current Stimulation. Brain Stimulat. 2015; 8(3):561–566. Bian M, Shen Y, Huang Y, Wu L, Wang Y, He S, et al. A non-immersive virtual reality-based intervention to enhance lower-extremity motor function and gait in patients with subacute cerebral infarction: A pilot randomized controlled trial with 1-year follow-up. Front Neurol. 2022; 13:985700. Junior VADS, Santos M de S, Ribeiro NM da S, Maldonado IL. Combining Proprioceptive Neuromuscular Facilitation and Virtual Reality for Improving Sensorimotor Function in Stroke Survivors: A Randomized Clinical Trial. J Cent Nerv Syst Dis. 2019; 11:1179573519863826. Xu Y, Yao J, Ni J, Yang Y, Fu L, Xu C. Comparison of Combined Virtual Reality Combined With Standing Balance Training Versus Standard Practice in Patients With Hemiplegia: A Single-Blinded, Randomized Controlled Trial. Am J Phys Med Rehabil. 2025; 104(4):312–317. Sultan N, Khushnood K, Qureshi S, Altaf S, Khan MK, Malik AN, et al. Effects of Virtual Reality Training Using Xbox Kinect on Balance, Postural Control, and Functional Independence in Subjects with Stroke. Games Health J. 2023; 12(6):440–444. Park D-S, Lee D-G, Lee K, Lee G. Effects of Virtual Reality Training using Xbox Kinect on Motor Function in Stroke Survivors: A Preliminary Study. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc. 2017; 26(10):2313–2319. Chen S-C, Lin C-H, Su S-W, Chang Y-T, Lai C-H. Feasibility and effect of interactive telerehabilitation on balance in individuals with chronic stroke: a pilot study. J Neuroengineering Rehabil. 2021; 18(1):71. Xu Y, Ni J, Yang Y, Yao J, Fu L, Xu C. Game-based visual feedback-guided dynamic balance training versus conventional training in patients with hemiplegia: a pilot randomized controlled trial. J Rehabil Med. 2025; 57: jrm41277. Yaman F, Akdeniz Leblebicier M, Okur İ, İmal Kızılkaya M, Kavuncu V. Is virtual reality training superior to conventional treatment in improving lower extremity motor function in chronic hemiplegic patients? Turk J Phys Med Rehabil. 2022; 68(3):391–398. McEwen D, Taillon-Hobson A, Bilodeau M, Sveistrup H, Finestone H. Virtual reality exercise improves mobility after stroke: an inpatient randomized controlled trial. Stroke. 2014; 45(6):1853–1855. In T, Lee K, Song C. Virtual Reality Reflection Therapy Improves Balance and Gait in Patients with Chronic Stroke: Randomized Controlled Trials. Med Sci Monit Int Med J Exp Clin Res. 2016; 22:4046–4053. Barcala L, Grecco LAC, Colella F, Lucareli PRG, Salgado ASI, Oliveira CS. Visual biofeedback balance training using wii fit after stroke: a randomized controlled trial. J Phys Ther Sci. 2013; 25(8):1027–1032. Yatar GI, Yildirim SA. Wii Fit balance training or progressive balance training in patients with chronic stroke: a randomised controlled trial. J Phys Ther Sci. 2015; 27(4):1145–1151. Kim JW, Lee JH. Effect of whole-body vibration therapy on lower extremity function in subacute stroke patients. J Exerc Rehabil. 2021; 17(3):158–163. Lee D-K, Han J-W. Effects of active vibration exercise using a Flexi-Bar on balance and gait in patients with chronic stroke. J Phys Ther Sci. 2018; 30(6):832–834. Annino G, Alashram A, Romagnoli C, Iovane A, Youssef TM, Tancredi V, et al. Efficacy of localized muscle vibration on lower extremity functional ability in patients with stroke: A randomized controlled trial. J Bodyw Mov Ther. 2025; 42:769–776. Wei N, Cai M. Optimal frequency of whole body vibration training for improving balance and physical performance in the older people with chronic stroke: A randomized controlled trial. Clin Rehabil. 2022; 36(3):342–349. Sade I, Çekmece Ç, İnanir M, SelÇuk B, Dursun N, Dursun E. The Effect of Whole Body Vibration Treatment on Balance and Gait in Patients with Stroke. Noro Psikiyatri Arsivi. 2020; 57(4):308–311. Choi E-T, Kim Y-N, Cho W-S, Lee D-K. The effects of visual control whole body vibration exercise on balance and gait function of stroke patients. J Phys Ther Sci. 2016; 28(11):3149–3152. Shimizu T, Hosaki A, Hino T, Sato M, Komori T, Hirai S, et al. Motor cortical disinhibition in the unaffected hemisphere after unilateral cortical stroke. Brain J Neurol. 2002; 125:1896–1907. Swinnen SP. Intermanual coordination: from behavioural principles to neural-network interactions. Nat Rev Neurosci. 2002; 3(5):348–359. Dimyan MA, Cohen LG. Contribution of transcranial magnetic stimulation to the understanding of functional recovery mechanisms after stroke. Neurorehabil Neural Repair. 2010; 24(2):125–135. Chou Y-H, Ton That V, Sundman M. A systematic review and meta-analysis of rTMS effects on cognitive enhancement in mild cognitive impairment and Alzheimer’s disease. Neurobiol Aging. 2020; 86:1–10. Slotema CW, Blom JD, van Lutterveld R, Hoek HW, Sommer IEC. Review of the efficacy of transcranial magnetic stimulation for auditory verbal hallucinations. Biol Psychiatry. 2014; 76(2):101–110. George MS, Taylor JJ, Short EB. The expanding evidence base for rTMS treatment of depression. Curr Opin Psychiatry. 2013; 26(1):13–18. Sheng R, Chen C, Chen H, Yu P. Repetitive transcranial magnetic stimulation for stroke rehabilitation: insights into the molecular and cellular mechanisms of neuroinflammation. Front Immunol. 2023; 14:1197422. Hummel FC, Cohen LG. Non-invasive brain stimulation: a new strategy to improve neurorehabilitation after stroke? Lancet Neurol. 2006; 5(8):708–712. Khedr EM, Abdel-Fadeil MR, Farghali A, Qaid M. Role of 1 and 3 Hz repetitive transcranial magnetic stimulation on motor function recovery after acute ischaemic stroke. Eur J Neurol. 2009; 16(12):1323–1330. Chen Q-M, Yao F-R, Sun H-W, Chen Z-G, Ke J, Liao J, et al. Combining inhibitory and facilitatory repetitive transcranial magnetic stimulation (rTMS) treatment improves motor function by modulating GABA in acute ischemic stroke patients. Restor Neurol Neurosci. 2021; 39(6):419–434. Ikeda T, Kobayashi S, Morimoto C. Effects of repetitive transcranial magnetic stimulation on ER stress-related genes and glutamate, γ-aminobutyric acid and glycine transporter genes in mouse brain. Biochem Biophys Rep. 2019; 17:10–16. Xing Y, Zhang Y, Li C, Luo L, Hua Y, Hu J, et al. Repetitive Transcranial Magnetic Stimulation of the Brain After Ischemic Stroke: Mechanisms from Animal Models. Cell Mol Neurobiol. 2023; 43(4):1487–1497. Sheng R, Chen C, Chen H, Yu P. Repetitive transcranial magnetic stimulation for stroke rehabilitation: insights into the molecular and cellular mechanisms of neuroinflammation. Front Immunol. 2023; 14:1197422. Gowan S, Hordacre B. Transcranial Direct Current Stimulation to Facilitate Lower Limb Recovery Following Stroke: Current Evidence and Future Directions. Brain Sci. 2020; 10(5):310. Kenney-Jung DL, Blacker CJ, Camsari DD, Lee JC, Lewis CP. Transcranial Direct Current Stimulation: Mechanisms and Psychiatric Applications. Child Adolesc Psychiatr Clin N Am. 2019; 28(1):53–60. Bai X, Guo Z, He L, Ren L, McClure MA, Mu Q. Different Therapeutic Effects of Transcranial Direct Current Stimulation on Upper and Lower Limb Recovery of Stroke Patients with Motor Dysfunction: A Meta-Analysis. Neural Plast. 2019; 2019:1372138. Brunoni AR, Amadera J, Berbel B, Volz MS, Rizzerio BG, Fregni F. A systematic review on reporting and assessment of adverse effects associated with transcranial direct current stimulation. Int J Neuropsychopharmacol. 2011; 14(8):1133–1145. Aparício LVM, Guarienti F, Razza LB, Carvalho AF, Fregni F, Brunoni AR. A Systematic Review on the Acceptability and Tolerability of Transcranial Direct Current Stimulation Treatment in Neuropsychiatry Trials. Brain Stimulat. 2016; 9(5):671–681. Thieme H, Morkisch N, Mehrholz J, Pohl M, Behrens J, Borgetto B, et al. Mirror therapy for improving motor function after stroke. Cochrane Database Syst Rev. 2018; 7:CD008449. Broderick P, Horgan F, Blake C, Ehrensberger M, Simpson D, Monaghan K. Mirror therapy for improving lower limb motor function and mobility after stroke: A systematic review and meta-analysis. Gait Posture. 2018; 63:208–220. Michielsen ME, Selles RW, van der Geest JN, Eckhardt M, Yavuzer G, Stam HJ, et al. Motor recovery and cortical reorganization after mirror therapy in chronic stroke patients: a phase II randomized controlled trial. Neurorehabil Neural Repair. 2011; 25(3):223–233. Chang WH, Kim Y-H. Robot-assisted Therapy in Stroke Rehabilitation. J Stroke. 2013; 15:174–181. Mehrholz J, Pohl M. Electromechanical-assisted gait training after stroke: a systematic review comparing end-effector and exoskeleton devices. J Rehabil Med. 2012; 44(3):193–199. Bonanno L, Cannuli A, Pignolo L, Marino S, Quartarone A, Calabrò RS, et al. Neural Plasticity Changes Induced by Motor Robotic Rehabilitation in Stroke Patients: The Contribution of Functional Neuroimaging. Bioeng Basel Switz. 2023; 10(8):990. Wang J, Zhang H, Ma J, Gu L, Li X. Efficacy of combined non-invasive brain stimulation and robot-assisted gait training on lower extremity recovery post-stroke: a systematic review and meta-analysis of randomized controlled trials. Front Neurol. 2025; 16:1500020. Wankhede NL, Koppula S, Ballal S, Doshi H, Kumawat R, Raju Ss, et al. Virtual reality modulating dynamics of neuroplasticity: Innovations in neuro-motor rehabilitation. Neuroscience. 2025; 566:97–111. Additional Declarations No competing interests reported. Supplementary Files Supplementarymaterial.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 17 Mar, 2026 Reviews received at journal 17 Mar, 2026 Reviews received at journal 16 Mar, 2026 Reviews received at journal 06 Mar, 2026 Reviewers agreed at journal 18 Feb, 2026 Reviewers agreed at journal 17 Feb, 2026 Reviewers agreed at journal 16 Feb, 2026 Reviewers invited by journal 16 Feb, 2026 Editor assigned by journal 14 Jan, 2026 Submission checks completed at journal 14 Jan, 2026 First submitted to journal 13 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8591907","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":595014137,"identity":"cb3041fe-7c1f-4858-9ba9-5a4d9563438b","order_by":0,"name":"Di Zhang","email":"","orcid":"","institution":"Chengdu university of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Di","middleName":"","lastName":"Zhang","suffix":""},{"id":595014139,"identity":"cfa25e34-5cb3-4541-ba73-04ae84563d85","order_by":1,"name":"Yating Yang","email":"","orcid":"","institution":"Chengdu university of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yating","middleName":"","lastName":"Yang","suffix":""},{"id":595014143,"identity":"07ad0c44-6c7c-4ae9-bf50-3a78cf214214","order_by":2,"name":"Guixing Xu","email":"","orcid":"","institution":"Sichuan Provincial Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Guixing","middleName":"","lastName":"Xu","suffix":""},{"id":595014145,"identity":"149ac355-3087-4cbb-9594-df3813b9f567","order_by":3,"name":"Lu Wang","email":"","orcid":"","institution":"Sichuan Provincial Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Lu","middleName":"","lastName":"Wang","suffix":""},{"id":595014147,"identity":"2b9dbb59-b076-4f9e-a32e-983c7b6f4e18","order_by":4,"name":"Guilin Zhang","email":"","orcid":"","institution":"Sichuan Provincial Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Guilin","middleName":"","lastName":"Zhang","suffix":""},{"id":595014148,"identity":"f4361d8b-914d-4f94-8c20-255e65c04798","order_by":5,"name":"Wanyi Song","email":"","orcid":"","institution":"Chengdu university of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Wanyi","middleName":"","lastName":"Song","suffix":""},{"id":595014151,"identity":"1e1c1da2-b5ed-4130-b221-b8123b5aef57","order_by":6,"name":"Miaoyu Ye","email":"","orcid":"","institution":"Chengdu university of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Miaoyu","middleName":"","lastName":"Ye","suffix":""},{"id":595014155,"identity":"9fd08291-0ade-4212-870e-e43890869662","order_by":7,"name":"Jian Luo","email":"","orcid":"","institution":"Sichuan Provincial Hospital of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Jian","middleName":"","lastName":"Luo","suffix":""},{"id":595014156,"identity":"34107234-0460-4150-85cb-dbfc20f76d11","order_by":8,"name":"Dehua Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAoUlEQVRIiWNgGAWjYDACZgaGDxUVEnLypGhhnHHmjIWxYQMJ9jDOONtWkchwgFj18u0MjA0H50kkMDYwP3x0gxgtBodBWrZJ5LEzsBkb5xClhZmB/fHHbRLFjA08bNJEaZFvBtkyRyKx4QCxWhjADmsgRQvYLweOSRgbNhPrF/n+A0AtNXVy8uzNDx8T5zAG/g8Qmpk45aNgFIyCUTAKiAEA7rQuAMnwmv0AAAAASUVORK5CYII=","orcid":"","institution":"Sichuan Provincial Hospital of Traditional Chinese Medicine","correspondingAuthor":true,"prefix":"","firstName":"Dehua","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2026-01-13 12:23:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8591907/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8591907/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103505783,"identity":"bc2463a1-0398-4826-9f1c-f3610f8ee332","added_by":"auto","created_at":"2026-02-26 13:32:59","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":123473,"visible":true,"origin":"","legend":"\u003cp\u003eFlow diagram of the search and selection process for literature.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/890f5ec368805dcbd038236b.png"},{"id":103211982,"identity":"9ece70d6-4416-4d9f-ad5c-d188aad7810e","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":8151,"visible":true,"origin":"","legend":"\u003cp\u003eBias risk assessment for inclusion in randomized controlled trials.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/8f20ce9e406a14fb3283db7e.png"},{"id":103504871,"identity":"d2dd8741-b8b6-412f-bb46-9a6cfb1d152f","added_by":"auto","created_at":"2026-02-26 13:21:52","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":144655,"visible":true,"origin":"","legend":"\u003cp\u003eFMA-LE (a) Network layout of the primary outcome: FMA-LE; (b) rankogram diagram.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/ee8d1792e72a93fc4293955b.png"},{"id":103211983,"identity":"36cdcb3f-7913-4a0b-b8ab-5f8bba06ca28","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":103651,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of each intervention's effects with standard therapy shown in a forest plot. (a) FMA-LE; (b) BBS; (c) MBI; (d) FAC; (e) TUG.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/48132ed0be7018ea14f5a80a.png"},{"id":103211990,"identity":"c8e56bbf-ca71-46a6-b332-ccd135e7bb34","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":192467,"visible":true,"origin":"","legend":"\u003cp\u003eHeat map displaying NMA for each outcome index. (a) FMA-LE; (b) BBS; (c) MBI; (d) FAC; (e) TUG.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/3758128898ffc4edc9a7e32b.png"},{"id":103211988,"identity":"c2c4a288-b1c4-4158-8bed-b553b2fccbc7","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":127808,"visible":true,"origin":"","legend":"\u003cp\u003eBBS (a) Network layout of the secondary outcome: BBS; (b) rankogram diagram.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/29469be48ac328dbf9d1aa26.png"},{"id":103211986,"identity":"d1f4a806-0681-486d-980c-b073a8c1c145","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":136615,"visible":true,"origin":"","legend":"\u003cp\u003eMBI (a) Network layout of the secondary outcome: MBI; (b) rankogram diagram.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/937602befa7c0184b36c3b7b.png"},{"id":103211987,"identity":"2d7466ed-fe1a-4e19-8f9f-2e758730a9e3","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":84209,"visible":true,"origin":"","legend":"\u003cp\u003eFAC (a) Network layout of the secondary outcome: FAC; (b) rankogram diagram.\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/ab28dfbda0dc8a3da3f7c335.png"},{"id":103211991,"identity":"f9fa3353-47b1-4cdb-be91-7fe1ff0a2bda","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":112250,"visible":true,"origin":"","legend":"\u003cp\u003eTUG (a) Network layout of the secondary outcome: TUG; (b) rankogram diagram.\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/99752f406f2fcb75ff55c6de.png"},{"id":104397552,"identity":"b206a1f2-e648-4a80-8d5e-b55bf4fa73e1","added_by":"auto","created_at":"2026-03-11 11:51:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1801979,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/b44a935d-3220-4af0-bf81-707b5ac0645c.pdf"},{"id":103211989,"identity":"d3b15f68-4b71-4334-9b09-d316ccf3ca55","added_by":"auto","created_at":"2026-02-23 08:48:28","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":6912448,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-8591907/v1/6c14e39a9a1d9282c79000c5.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Effectiveness of Non-Pharmacological Interventions for Lower Limb Motor Impairment in Post-Stroke Hemiplegia: A Systematic Review and Bayesian Network Meta-Analysis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eStroke is a sudden neurological disorder caused by a disruption in cerebral circulation, resulting in either localized or widespread neurological impairments[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. According to the most recent Global Burden of Disease (GBD) report, stroke is now the second leading cause of death worldwide, trailing only ischemic heart disease, and it ranks third in terms of global disability[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. The period from 1990 to 2019 witnessed a significant surge in the annual rate of stroke cases and deaths, with stroke incidence climbing by 70% and stroke-related mortality rising by 43%[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Projections indicate that stroke-related deaths are expected to increase by 47% by 2050[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Hemiplegia, commonly observed after a stroke, is a significant neurological disorder that typically results in motor impairment or loss of function on one side of the body[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. The pathophysiological mechanism primarily arises from neurological damage due to ischemia or hemorrhage in brain tissue[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. The loss of limb motor function is mainly attributed to damage in the corticospinal tract, particularly affecting critical areas such as the motor cortex of the precentral gyrus, the posterior segment of the internal capsule, and the brainstem region[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. This damage ultimately results in contralateral limb motor neuron paralysis. Patients with hemiplegia following a stroke frequently experience lower limb motor dysfunction, characterized by increased muscle tone, abnormal gait control, and diminished coordination and balance. Not only do these consequences deeply impact the patients, but they also have significant repercussions for their families, caregivers, and the wider society. Stroke survivors frequently face long-term disability and ongoing care requirements, imposing significant emotional strain, financial burdens, and practical caregiving challenges upon families[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Concurrently, the societal costs of stroke are substantial, encompassing increased healthcare expenditure, diminished productivity, and a growing demand for extended rehabilitation and care services[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Research shows that 50%\u0026ndash;65% of stroke patients struggle with walking independently shortly after onset, and despite rehabilitation, 20%\u0026ndash;30% still cannot walk independently by the end of treatment[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. As a result, improving the restoration of lower limb motor ability in individuals with post-stroke hemiplegia, as well as boosting their autonomy in walking and performing daily tasks, has become a primary focus in contemporary stroke rehabilitation research and clinical practice.\u003c/p\u003e \u003cp\u003eRehabilitating stroke patients with hemiplegia is a challenging yet crucial process. Sole reliance on pharmacological interventions frequently proves inadequate in substantially enhancing limb function and activity levels, as medications primarily address the underlying causes and mitigate symptoms, with limited impact on neural remodeling and functional recovery. Restoring motor function and improving self-care abilities rely primarily on structured and ongoing rehabilitation programs. Non-pharmacological rehabilitation interventions can optimize the activation of residual functions, facilitate cerebral reorganization, and assist patients in regaining independent living abilities. Consequently, early intervention and consistent rehabilitation treatment are imperative for hemiplegic patients to achieve functional recovery. As research into the mechanisms of lower limb dysfunction and rehabilitation strategies after a stroke deepens, the use of various non-pharmacological therapies for functional recovery has expanded, offering a crucial supplement to the limitations of drug-based treatments. The commonly used interventions include neurostimulation techniques, Constraint-induced movement therapy, virtual reality training, robotic rehabilitation, acupuncture therapy, mirror therapy, vibration therapy, and hydrotherapy. These approaches are becoming progressively more significant in enhancing lower limb motor abilities and fostering neural regeneration and functional recovery. Nonetheless, earlier research has mainly concentrated on contrasting the impacts of individual non-pharmacological interventions with standard care or placebo controls, without direct or indirect comparisons between various interventions. This creates significant uncertainty for clinicians when selecting therapies and optimizing treatment plans. To establish a clearer and more systematic evidence base for efficacy, it is necessary to employ Bayesian network meta-analysis (NMA).\u003c/p\u003e \u003cp\u003eBayesian NMA, an enhanced version of conventional meta-analysis, integrates indirect comparisons across interventions, facilitating the assessment of the comparative effectiveness of different treatments. This method allows for the ranking and comparative evaluation of treatments, aiding in the identification of the most effective intervention[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Consequently, this research employs a NMA to comprehensively evaluate the effectiveness of different non-pharmacological treatments for improving lower limb motor function in stroke survivors with hemiplegia. The objective is to rank the various interventions, providing evidence-driven guidance for choosing suitable non-pharmacological treatment approaches in clinical settings\u003c/p\u003e"},{"header":"Methods","content":"Study protocol and registration\n\nThe study complies with the PRISMA extension guidelines for systematic reviews and meta-analyses and has been registered in PROSPERO with the registration number CRD420251169037[12].\n\nInclusion criteria\n\nThe inclusion criteria were defined based on the PICOS framework, encompassing Population, Intervention, Comparison, Outcomes, and Study design[13].\n\n(1) P: Participants must be at least 18 years old, confirmed to have had a stroke via CT or MRI scans, and present with hemiplegia on one side of the body. All participants should have stable vital signs and be fully conscious. There are no restrictions regarding stroke type, side of hemiplegia, illness duration, or participants' gender, race, or geographical location.\n\n(2) I: Non-pharmacological interventions encompass a range of rehabilitation methods, including acupuncture (ACPU), vibration therapy (VT), mirror therapy (MT), virtual reality therapy (VR), constraint-induced movement therapy (CIMT), functional electrical stimulation (FES), repetitive transcranial magnetic stimulation (rTMS), transcranial direct current stimulation (tDCS), intermittent theta burst stimulation (iTBS), robot-assisted rehabilitation (RAR), water therapy (WT), and various combined approaches (e.g., VR + RAR, RAR + FES, MT + FES). The supplementary materials provide a detailed account of the specific intervention methods.\n\n(3) C: The control group may receive one of the following standard treatments: conventional rehabilitation methods (including exercise therapy, neuromuscular therapy, training, and standard stroke care), sham treatment, or no treatment.\n\n(4) O: Outcome Measures:\n\nPrimary Outcome: \n\n① The Fugl-Meyer Assessment for Lower Extremities (FMA-LE) is used to evaluate motor function in the lower limbs, where higher scores indicate improved motor abilities (maximum score: 34 points).\n\nSecondary Outcomes:\n\n① The Berg Balance Scale (BBS) is used to assess an individual's balance capabilities, with a total score of 56 points. Higher scores reflect improved balance function. \n\n② The Modified Barthel Index (MBI) evaluates an individual's ability to perform activities of daily living, with a higher score signifying greater independence and proficiency in completing everyday tasks (maximum score: 100 points). \n\n③ The Functional Ambulation Classification (FAC) is used to assess walking ability, with higher scores indicating better walking function. \n\n④ The Timed Up and Go (TUG) test measures mobility and dynamic stability, with quicker times reflecting improved function. \n\n(5) S: Study design: Randomized Controlled Trial (RCT).\n\nExclusion criteria\n\nThe criteria for exclusion were outlined as follows: \n\n(1) Hemiplegia not caused by stroke.\n\n(2) reviews, letters, study protocols, conference abstracts, case reports, animal studies, duplicate publications.\n\n(3) literature lacking outcome measures, or studies with incomplete outcome data.\n\n(4) Research involving a sample size of less than 10.\n\nSearch strategy\n\nTwo researchers (DZ and TYY) independently conducted literature searches across four English-language databases: PubMed, Web of Science, Cochrane Library, and Embase. The search encompassed publications from January 2010 to August 2025, with no restrictions on language. Key search terms included stroke, hemiplegia, lower limbs, non-pharmacological interventions, randomized controlled trial, among others. For each database, detailed search strategies are available in the supplementary materials. To ensure comprehensive coverage, references from eligible studies and relevant prior research were manually reviewed. For studies with incomplete data, we reached out to the corresponding authors to obtain the required information and conducted a thorough online search for all relevant studies.\n\nData extraction\n\nTwo researchers (DZ, TYY) imported the retrieved studies into EndNote reference management software, and after removing duplicates, conducted an initial screening by reviewing the titles and abstracts. The studies were subsequently chosen independently according to pre-established eligibility criteria. Full-text articles from potentially eligible studies were obtained for additional evaluation. The two researchers (DZ, TYY) then independently extracted the relevant study details using a standardized data collection form and cross-verified the data. Conflicts were addressed through consultation with a third researcher (DHL). For every included study, the following information was extracted: lead author, year of publication, country of origin; participant characteristics (sample size, gender distribution, average age, type of stroke, affected side, disease stage); details of interventions and their duration for both experimental and control groups; and outcome measures.\n\nRisk-of-bias assessment\n\nTwo researchers (DZ, TYY) evaluated the risk of bias in the selected studies using the Cochrane Risk of Bias Tool, and the analysis was conducted using Review Manager 5.4 software[14]. The assessment included seven areas: method of randomization, concealment of allocation, blinding of both participants and intervention providers, blinding of those assessing outcomes, completeness of data, selective reporting, and other possible biases. Each domain was classified as “low risk,” “unclear risk,” or “high risk.” Any discrepancies were addressed by a third researcher (DHL) who made the final inclusion decision. The evidence quality was evaluated with GRADEPro, and in the event of substantial differences or concerns regarding study inclusion, a third-party expert was consulted for resolution.\n\nStatistical analysis\n\nIn the included studies, pairwise comparisons were used for two-arm trials. For multi-arm trials with three or more interventions, all possible pairwise comparisons were made for analysis. Bayesian models for NMA were constructed using the gemtc package in R Studio, which also generated treatment probability plots and ranked the interventions. For continuous outcomes, the standardized mean difference (SMD) along with its 95% confidence interval (CI) was calculated. When a closed loop of interventions was established, inconsistency tests were conducted to evaluate the agreement between direct and indirect comparisons, utilizing node-splitting for local analysis. A P value exceeding 0.05 indicates the absence of significant inconsistency. To assess the overall effectiveness of the interventions, the surface under the cumulative ranking curve (SUCRA) was determined for each treatment, and the interventions were ranked accordingly. For FMA-LE, BBS, FAC, and MBI, SUCRA values span from 0% to 100%, with 100% representing the optimal treatment effect and 0% reflecting the least effective intervention. For TUG, lower values indicate better intervention effects; therefore, a higher SUCRA value for TUG corresponds to poorer outcomes. Funnel plots were utilized to examine the possibility of publication bias or small sample size effects, with a visual check for symmetry to identify any systematic bias. Additionally, Egger's test was performed, with a P value below 0.05 suggesting the presence of publication bias."},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eLiterature Screening\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFour electronic databases provided a collection of 3,071 articles. After excluding 1,002 studies published before 2010 and 886 duplicate studies, 1,183 articles remained. The titles and abstracts of the studies were assessed, resulting in the exclusion of 970 articles. A total of 213 studies were read in full. After further examination, 132 additional studies were excluded. Ultimately, 82 studies[15\u0026ndash;96]\u0026nbsp;with 3,514 participants were included in the NMA. Figure 1 depicts the process of study selection through a flowchart, and the search queries and search results can be found in the appendix.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBasic characteristics of the studies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study included 82 RCTs[15\u0026ndash;96] published between 2010 and 2025, covering 14 countries and involving 3,514 participants. These trials evaluated 16 distinct interventions.\u0026nbsp;The corresponding codes for the interventions are presented in Table 1.\u0026nbsp;Specific interventions comprised: acupuncture (ACPU) (n=4)[31\u0026ndash;34], vibration therapy (VT) (n=7)[91\u0026ndash;96], mirror therapy (MT) (n=5)[15,16,18,19,21], constraint-induced movement therapy (CIMT) (n=3)[35\u0026ndash;37], virtual reality therapy (VR) (n=12)[79\u0026ndash;90], repetitive transcranial magnetic stimulation (rTMS) (n=5)[71\u0026ndash;75], transcranial direct current stimulation (tDCS) (n=4)[41,76\u0026ndash;78], functional electrical stimulation (FES) (n=10)[38\u0026ndash;45],\u0026nbsp;intermittent theta burst stimulation (iTBS)\u0026nbsp;(n=3)[28\u0026ndash;30], robotic assistive rehabilitation (RAR) (n=20)[47\u0026ndash;54,56\u0026ndash;65,68,69], water therapy (WT) (n=6)[22\u0026ndash;27], MT+FES (n=3)[17,19,20], VR+RAR (n=4)[46,66,67,70], FES+RAR (n=2)[46,53], sham (n=17)[17,28\u0026ndash;30,42,44,71\u0026ndash;78,88], and conventional treatment (CT) (n=72)[15,16,18\u0026ndash;27,31\u0026ndash;41,43,45\u0026ndash;70,79\u0026ndash;87,89\u0026ndash;96].Seven studies[19,42,45,46,71,73,94]\u0026nbsp;employed a three-group design, whilst the remaining studies utilized a two-group design. The appendix contains detailed descriptions of the studies that were included.\u003c/p\u003e\n\u003cp\u003eTable. 1. Codes linked to intervention actions.\u003c/p\u003e\n\u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e\u003cstrong\u003enumber\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003e\u003cstrong\u003ecode\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eIntervention measures\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eConventional Therapy (CT)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eAcupuncture (ACPU)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eVibration therapy (VT)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eMirror therapy (MT)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eConstraint-induced movement therapy (CIMT)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eVirtual reality therapy (VR)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eintermittent theta burst stimulation (iTBS)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eFunctional electrical stimulation (FES)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eRepetitive Transcranial Magnetic Stimulation (rTMS)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eJ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eTranscranial Direct Current Stimulation (tDCS)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eRobot-assisted rehabilitation (RAR)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eWater therapy (WT)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003esham\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eFK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eVR+RAR\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eKH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eRAR+FES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 15.5668%;\"\u003e\n \u003cp\u003e16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.2284%;\"\u003e\n \u003cp\u003eDH\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69.2047%;\"\u003e\n \u003cp\u003eMT+FES\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQuality assessment of the included studies\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Cochrane Risk of Bias Assessment Tool (Review Manager 5.4) was used to evaluate the risk of bias in 82 studies. Of these, 69 studies specified the methods used for random sequence generation and were categorized as \u0026lsquo;low risk,\u0026rsquo; while the remaining 13 studies referenced randomization without providing specific details, and were labeled as \u0026lsquo;unclear risk of bias\u0026rsquo;.\u0026nbsp;Allocation concealment was achieved using opaque sealed envelopes in 28 studies, which were classified as \u0026lsquo;low risk\u0026rsquo;; the other 54 studies did not specify their methods for allocation concealment and were categorized as having an \u0026lsquo;unclear risk of bias\u0026rsquo;. Owing to the inherent characteristics of certain non-pharmacological interventions, researchers encountered challenges in implementing blinding during the intervention process. Only 10 studies reported blinding of both participants and researchers, whereas 61 studies reported blinding of outcome assessments. 62 studies were rated as \u0026lsquo;low risk\u0026rsquo; due to having complete data. All 82 studies reported pre-specified outcome measures, rated as \u0026lsquo;low risk\u0026rsquo;. None detailed other sources of bias, thus rated as \u0026lsquo;risk of bias unclear\u0026rsquo;. Figure 2 displays the findings from the quality assessment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCertainty of evidence\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUsing the GRADEpro tool, the certainty of the evidence was assessed. The results, presented in\u0026nbsp;attachment, demonstrated a high certainty for changes in FMA-LE, a moderate certainty for MBI, FAC and TUG, and a low certainty for BBS. The lower certainty levels were largely attributed to a high risk of bias, insufficient methodological rigor, and considerable heterogeneity among the studies.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePrimary outcome: FMA-LE\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNetwork plot of interventions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFigure 3a shows a network diagram of 47 RCTs[15,16,18,28,29,31\u0026ndash;35,38,39,41\u0026ndash;43,46,48\u0026ndash;50,54,57,59\u0026ndash;61,63\u0026ndash;66,68\u0026ndash;80,83,86]\u0026nbsp;evaluating the efficacy of FMA-LE under different individual interventions, involving 2,134 participants and 13 non-pharmacological treatments. The treatments are as follows: A: CT, B: ACPU, D: MT, E: CIMT, F: VR, G: iTBS, H: FES, I: rTMS, J: tDCS, K: RAR, FK: VR+RAR, KH: RAR + FES, M: sham. The network primarily focuses on conventional treatment. Interventions are represented by points, with lines connecting them to show direct comparisons. The width of these lines represents the number of studies included. Figure 4a presents the forest plot that compares conventional treatment A with each intervention.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNMA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe NMA results revealed 156 pairwise comparisons. The final network effect indicated that compared to conventional therapy that ACPU\u0026nbsp;(SMD = \u0026minus;2.54; 95% CI: \u0026minus;3.8 to \u0026minus;1.3), MT (SMD = \u0026minus;3.43; 95% CI: \u0026minus;5.93 to \u0026minus;0.93), VR combined with RAR (SMD = \u0026minus;2.98; 95% CI: \u0026minus;3.98 to \u0026minus;2), rTMS (SMD = \u0026minus;3.68; 95% CI: \u0026minus;5.93 to \u0026minus;0.93), tDCS (SMD = \u0026minus;3.36; 95% CI: \u0026minus;4.86 to \u0026minus;2.49), RAR (SMD = \u0026minus;2.52; 95% CI: \u0026minus;3.63 to \u0026minus;1.4) and RAR combined with FES (SMD = \u0026minus;3; 95% CI: \u0026minus;4.48 to \u0026minus;1.52) were effective in\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eimproving lower limb motor function in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions. The detailed results of the NMA on FMA-LE can be found in the appendix.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRanking\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe SUCRA probability ranking showed that the comparative efficacy of treatments in improving FMA-LE was as follows: rTMS (88.45%) \u0026gt; tDCS (81.58%) \u0026gt; MT (78.80%) \u0026gt; RAR + FES (72.41%) \u0026gt; VR + RAR (72.00%) \u0026gt; ACPU (61.81%) \u0026gt; RAR (60.77%) \u0026gt; iTBS (36.53%) \u0026gt; CIMT (36.20%) \u0026gt; FES (24.40%) \u0026gt; VR (18.32%) \u0026gt; CT (14.93%) \u0026gt; sham (3.81%). The results suggest that rTMS may be the most effective method for improving FMA-LE scores in patients with lower limb hemiplegia following a stroke.\u0026nbsp;The rankogram appears in figure 3b, with the SUCRA graph depicted in figure 5a.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSecondary outcome: BBS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNetwork evidence graph\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe figure 6a presents a network diagram of 59 RCTs[15,17,19\u0026ndash;28,30,35,42,43,45,46,49,51\u0026ndash;53,55,55\u0026ndash;62,64,66,67,69\u0026ndash;71,76,78,79,81\u0026ndash;86,88\u0026ndash;96]\u0026nbsp;evaluating the efficacy of BBS under various individual interventions, involving a\u0026nbsp;total of 2,075 participants and covering 15 non-pharmacological treatments (A: CT, B: ACPU, C: VT, D: MT, F: VR, E: CIMT, H: FES, I: rTMS, J: tDCS, K: RAR, L: WT, FK: VR + RAR, KH: RAR + FES, DH: MT + FES, M: sham). Figure 4b presents the forest plot that compares conventional treatment A with each intervention.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNMA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe NMA results revealed that 210 pairwise comparisons were produced. Compared with conventional therapy that VT\u0026nbsp;(SMD = \u0026minus;2.79; 95% CI: \u0026minus;4.19 to \u0026minus;1.4), MT (SMD = \u0026minus;11.21; 95% CI: \u0026minus;16.73 to \u0026minus;5.71), CIMT (SMD = \u0026minus;2.87; 95% CI: \u0026minus;4.68 to \u0026minus;1.04), iTBS (SMD = \u0026minus;3.81; 95% CI: \u0026minus;7.28 to \u0026minus;0.32), FES (SMD = \u0026minus;5.33; 95% CI: \u0026minus;7.25 to \u0026minus;3.42), VR combined with RAR (SMD = \u0026minus;4.41; 95% CI: \u0026minus;5.33 to \u0026minus;3.49), rTMS (SMD = \u0026minus;12.32; 95% CI: \u0026minus;15.07 to \u0026minus;9.57), RAR (SMD = \u0026minus;1.82; 95% CI: \u0026minus;3.25 to \u0026minus;0.4), RAR combined with FES (SMD = \u0026minus;4.07; 95% CI: \u0026minus;5.26 to \u0026minus;2.89) and WT (SMD = \u0026minus;1.04; 95% CI: \u0026minus;1.48 to \u0026minus;0.6) were effective in\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eimproving balance function in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions.\u0026nbsp;The detailed results of the NMA on BBS can be found in the appendix.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRanking\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe SUCRA probability ranking showed that the comparative efficacy of treatments in improving BBS was as follows: rTMS (96.2%) \u0026gt; MT (95.77%) \u0026gt;FES (79.65%) \u0026gt; VR + RAR (71.07%) \u0026gt; iTBS (67.47%) \u0026gt; RAR + FES (66.33%) \u0026gt; tDCS (53.62%) \u0026gt; CIMT (47.54%) \u0026gt; VT (46.86%) \u0026gt; WT (31.58%) \u0026gt; RAR (26.31%) \u0026gt; MT+FES (23.75%) \u0026gt; VR (21.47%) \u0026gt; sham (20.45%) \u0026gt; CT (1.61%). The results indicate that rTMS is the most effective method for improving balance function in post-stroke hemiplegic patients. The rankogram can be seen in figure 6b, and the SUCRA graph is illustrated in figure 5b.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSecondary outcome: MBI\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNetwork evidence graph\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe figure 7a displays a network diagram from 26 RCTs[15,28,31,32,34,38,41,42,46,49,52,57,60,61,66,71\u0026ndash;75,93]\u0026nbsp;evaluating the efficacy of MBI under various individual interventions, involving a total of 1,218 participants and covering 12 non-pharmacological treatments (A: CT, B: ACPU, C: VT, D: MT G: iTBS, H: FES, I: rTMS, J: tDCS, K:RAR, FK: VR+ RAR, KH: RAR+ FES, M: sham). Figure 4c presents the forest plot that compares conventional treatment A with each intervention.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNMA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe NMA results showed that 132 pairwise comparisons were performed. Compared with conventional therapy that ACPU\u0026nbsp;(SMD = \u0026minus;7.57; 95% CI: \u0026minus;12.61 to \u0026minus;2.49), MT (SMD = \u0026minus;9.83; 95% CI: \u0026minus;17.98 to \u0026minus;1.64), VR combined with RAR (SMD = \u0026minus;7.34; 95% CI: \u0026minus;8.94 to \u0026minus;5.74),\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003erTMS (SMD = \u0026minus;13.51; 95% CI: \u0026minus;16.32 to \u0026minus;10.73), tDCS (SMD = \u0026minus;11.76; 95% CI: \u0026minus;17.72 to \u0026minus;5.78), RAR (SMD = \u0026minus;6.28; 95% CI: \u0026minus;11.41 to \u0026minus;1.1) and RAR combined with FES (SMD = \u0026minus;4.63; 95% CI: \u0026minus;7.26 to \u0026minus;2.06) were effective in\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eenhancing daily living abilities in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions.\u0026nbsp;The detailed results of the NMA on MBI can be found in the appendix.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRanking\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe SUCRA probability ranking shows that the comparative efficacy of treatments in improving MBI is as follows: rTMS (94.72%) \u0026gt; tDCS (86.22%) \u0026gt; MT (75.58%) \u0026gt; ACPU (64.43%) \u0026gt; VR + RAR (64.15%) \u0026gt; RAR (56.38%) \u0026gt; RAR + FES (44.52%) \u0026gt; iTBS (38.38%) \u0026gt; VT (29.66%) \u0026gt; FES (21.92%) \u0026gt; CT (20.53%) \u0026gt; sham (3.69%). The results indicate that rTMS is the most effective method for improving activities of daily living in post-stroke hemiplegic patients. In figure 7b, the rankogram is depicted, and Figure 5c shows the SUCRA graph.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSecondary outcome: FAC\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNetwork evidence graph\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe figure 10a displays a network diagram from 22 RCTs[21,24,40\u0026ndash;42,45,47\u0026ndash;49,51,52,54,56,57,59,62,63,68,70,84,91]\u0026nbsp;evaluating the efficacy of FAC under various individual interventions, involving a total of 920 participants and covering 8 non-pharmacological treatments (A: CT, C: VT, D: MT, H: FES, J: tDCS, K: RAR, L: WT, M: sham). Figure 4d presents the forest plot that compares conventional treatment A with each intervention.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNMA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe results of the NMA revealed that 56 pairwise comparisons were performed. Compared with conventional therapy that VT\u0026nbsp;(SMD = \u0026minus;0.77; 95% CI: \u0026minus;1.3 to \u0026minus;0.25), MT (SMD = \u0026minus;1.45; 95% CI: \u0026minus;2.2 to \u0026minus;0.7), FES (SMD = \u0026minus;1.28; 95% CI: \u0026minus;2.31 to \u0026minus;0.24), tDCS (SMD = \u0026minus;1.47; 95% CI: \u0026minus;2.54 to \u0026minus;0.42) and RAR (SMD = \u0026minus;0.65; 95% CI: \u0026minus;0.81 to \u0026minus;0.5) were effective in\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eenhancing daily living abilities in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions.\u0026nbsp;The detailed results of the NMA on FAC can be found in the appendix.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRanking\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe SUCRA probability ranking shows that the comparative efficacy of treatments in improving MBI is as follows: tDCS (87.54%) \u0026gt; MT (82.25%) \u0026gt; FES (69.74%) \u0026gt; VT (45.58%) \u0026gt; WT (43.11%) \u0026gt; RAR (36.28%) \u0026gt; sham (32.09%) \u0026gt; CT (3.41%). The results indicate that tDCS is the most effective method for improving functional ambulation in post-stroke hemiplegic patients. The rankogram can be found in figure 8b, and the SUCRA graph is illustrated in figure 5d.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSecondary outcome: TUG\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNetwork plot of interventions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe figure 9a displays a network diagram from 34 RCTs[20,22,23,25\u0026ndash;29,36,37,39,43,46,47,53,57,65,77,81\u0026ndash;85,87\u0026ndash;96]\u0026nbsp;evaluating the efficacy of FAC under various individual interventions, involving a total of 1,422 participants and covering 10 non-pharmacological treatments\u0026nbsp;(A: CT, C: VT, E: CIMT, F: VR, H: FES, K: RAR, L: WT, DH: MT+FES, FK: VR+RAR, KH: RAR+FES). Figure 4e presents the forest plot that compares conventional treatment A with each intervention.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNMA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe results of the NMA revealed that 90 pairwise comparisons were performed. Compared with conventional therapy that VR combined with RAR (SMD = 6.39; 95% CI: 4.56 to 7.99), MT combined with FES (SMD = 5.26; 95% CI: 2.97 to 7.6), VT (SMD = 2.93; 95% CI: 1.96 to 3.91), FES (SMD = 2.98; 95% CI: 0.33 to 5.62) and WT (SMD = 2.04; 95% CI: 1.32 to 2.76) were effective in\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eshortening the TUG. in stroke patients with hemiplegia. No significant statistical differences were observed among the other interventions.\u0026nbsp;The detailed results of the NMA on TUG can be found in the appendix.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eRanking\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe SUCRA probability ranking shows that the comparative efficacy of treatments in improving TUG is as follows: VR + RAR (5.05%) \u0026gt; RAR + FES (13.43%) \u0026gt; VT (35.87%) \u0026gt; FES (39.40%) \u0026gt; WT (53.67%) \u0026gt; RAR+FES (61.20%)\u0026gt; VR (61.48%) \u0026gt; CIMT (67.45%) \u0026gt; RAR (74.47%) \u0026gt; CT (88.05%). The results indicate that VR + RAR is the most effective method for improving TUG in post-stroke hemiplegic patients. In figure 9b, the rankogram is depicted, and figure 5e shows the SUCRA graph.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsistency analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe assessment of network inconsistency was conducted rigorously through node-splitting analysis when applicable. For TUG, the network\u0026nbsp;structure\u0026nbsp;lacked\u0026nbsp;closed\u0026nbsp;loops,\u0026nbsp;making\u0026nbsp;inconsistency assessment\u0026nbsp;irrelevant. The results of FMA-LE, BBS, MBI, and FAC underwent node-splitting analysis to evaluate the direct and indirect treatment comparison estimates from two types of evidence. There were no significant inconsistencies found in any comparisons, with all P-values being much higher than 0.05, which indicates a high level of concordance between the direct and indirect evidence, supporting the validity of our NMA assumptions. The specific details can be found in the appendix.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Publication Bias\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUtilizing a random-effects model, a corrected contrast funnel plot was developed. Overall, the distribution of study points appears relatively symmetric, with no significant signs of small-study effects or publication bias observed. It should be noted that for certain interventions within the network, the number of available studies is limited, which may affect the interpretability of the funnel plot. Hence, the associated results should be viewed with caution.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eLower limb motor impairment is a common and severe outcome for individuals who have suffered a stroke. Following stroke, cortical spinal excitability in the damaged motor cortex typically diminishes markedly, whilst excitability in the contralateral region frequently increases, accompanied by heightened interhemispheric inhibition[\u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e]. This asymmetrical alteration in excitability significantly impairs patients' gait and balance maintenance capabilities. Neuroimaging studies indicate that walking and balance depend not only on cortical activity but also require coordinated regulation involving subcortical structures and the spinal cord[\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e]. However, rehabilitation treatment targeting these complex sensorimotor functions continues to present substantial clinical challenges. Therefore, identifying safe, effective, and targeted rehabilitation methods to improve lower limb mobility and overall recovery post-stroke has become a central research area in stroke rehabilitation.\u003c/p\u003e \u003cp\u003eTo the best of our knowledge, this is the initial network meta-analysis that examines various non-drug interventions and assesses five outcome measures (FMA-LE, BBS, MBI, FAC, and TUG) related to lower limb motor function in hemiplegic patients. This NMA incorporated 82 RCTs involving 3,541 participants, comparing 16 distinct interventions to evaluate non-pharmacological approaches for enhancing motor abilities in the lower limbs of individuals with post-stroke hemiplegia. All included studies underwent strict quality and risk assessments. Our study demonstrates that rTMS has a significant advantage in improving FMA-LE in post-stroke hemiplegic patients, achieving a notable SUCRA ranking of 88.45%. Following rTMS, tDCS and mirror therapy also show promising results. Consistent with previous meta-analyses, rTMS demonstrates consistent efficacy across all stages of stroke rehabilitation, significantly enhancing lower limb motor function. rTMS is a non-invasive technique for brain stimulation that modulates neuronal membrane potential through pulsed magnetic fields, enabling targeted regulation of neural excitability in particular brain regions[\u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e]. rTMS has been widely used in managing neurological and psychiatric conditions, such as cognitive deficits, depression, psychosis, and Alzheimer's disease[\u003cspan additionalcitationids=\"CR101\" citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e102\u003c/span\u003e]. Due to its painless and straightforward administration, it is a valuable tool for accurately modulating brain regions and predicting stroke rehabilitation success[\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e]. Its effects extend beyond the stimulation site, influencing cortical excitability in distant areas via synaptic transmission. The findings of this study suggest that rTMS exhibits greater effectiveness in enhancing lower limb motor function (FMA-LE) and improving balance (BBS). rTMS enhances excitability in the ipsilateral primary motor cortex through high-frequency stimulation (\u0026ge;\u0026thinsp;3 Hz) or suppresses excessive activation in the contralateral hemisphere via low-frequency stimulation (\u0026le;\u0026thinsp;1 Hz) [\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e, \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e105\u003c/span\u003e]. This restores the excitatory-inhibitory balance between both hemispheres, serving as the core mechanism for promoting recovery of motor control and coordination functions. Earlier research has demonstrated that rTMS can modulate the central nervous system neurotransmitter system, improving the balance between excitatory neurotransmitters such as glutamate and inhibitory neurotransmitters like γ-aminobutyric acid (GABA), enhancing synaptic plasticity[\u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e106\u003c/span\u003e, \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e107\u003c/span\u003e]. Moreover, rTMS influences not only neurons but also plays a significant role in modulating glial cell activity, facilitating the shift of astrocytes from the pro-inflammatory A1 subtype to the neuroprotective A2 subtype, and promoting the polarization of microglia from the M1 to the M2 phenotype[\u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e108\u003c/span\u003e]. This process alleviates neuroinflammation, inhibits neuronal apoptosis, and preserves the integrity of the blood-brain barrier. In addition, rTMS can downregulate the levels of pro-inflammatory cytokines such as tumor necrosis factor-α, interleukin-1β, and interleukin-6, while upregulating the expression of anti-inflammatory cytokines such as interleukin-10 and transforming growth factor-β[\u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e109\u003c/span\u003e]. The aforementioned mechanisms collectively form a synergistic network: the restoration of hemispheric balance and enhanced synaptic plasticity directly drive improvements in motor function, while optimization of the neuroimmune microenvironment provides sustained biological support for functional remodeling. This may represent the intrinsic reason for rTMS' s superior performance on both the FMA-LE and BBS outcomes in this study. However, This study relies on only five RCTs with small sample sizes and does not include subgroup analyses of rTMS stimulation frequencies or examine the specific effects at different frequencies, which limits the accuracy and generalizability of the findings. Thus, the results should be viewed with caution. Future research urgently requires larger-scale, high-quality RCTs to explore optimal rTMS stimulation parameters and protocols, and to validate its long-term efficacy in improving lower limb function in patients with post-stroke hemiplegia. tDCS has demonstrated favorable outcomes in advancing motor skills in the lower extremities of patients who have hemiplegia due to a stroke. tDCS is a non-invasive method that shows potential for stimulating neuroplasticity within the brain's cortex and supporting recovery after a stroke[\u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e110\u003c/span\u003e]. tDCS modulates neuronal membrane potential and alters ion distribution across the membrane through the application of a mild direct current to the brain's cortex. This stimulates neuroplastic processes, thus facilitating the restoration of neural function in the brain tissue of stroke survivors[\u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e111\u003c/span\u003e]. tDCS effectively re-establishes the balance of inhibitory competition between the two hemispheres by increasing excitability in the affected hemisphere through the anode or decreasing excitability in the opposite hemisphere using the cathode[\u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e112\u003c/span\u003e]. Two meta-analyses indicate that tDCS demonstrates good safety and tolerability[\u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e113\u003c/span\u003e, \u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e114\u003c/span\u003e]. Mirror therapy ranked third in the primary outcome measure FMA-LE, demonstrating a degree of efficacy. Mirror therapy is an innovative intervention that uses mirror visual feedback (MVF) to provide new visual input, stimulating the mirror neuron system and helping to reduce lower limb motor impairments after a stroke[\u003cspan citationid=\"CR115\" class=\"CitationRef\"\u003e115\u003c/span\u003e]. In conventional mirror therapy, the unaffected limb is placed in the reflected area of the mirror, while the impaired limb is positioned behind the mirror, out of sight. This visual illusion, or MVF, creates the perception of both limbs functioning normally, effectively \u0026ldquo;deceiving\u0026rdquo; the brain and stimulating the sensory-motor areas, thereby promoting the recovery of motor ability[\u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e116\u003c/span\u003e]. Mirror therapy's benefits stem from its simple implementation, needing no advanced equipment or specialized oversight, while being non-invasive throughout the entire treatment process. Even patients with severe movement disorders can independently administer therapy at home, offering flexibility and convenience[\u003cspan citationid=\"CR117\" class=\"CitationRef\"\u003e117\u003c/span\u003e]. By activating the brain's motor regions through visual feedback, this approach not only enhances patients' capacity for self-directed rehabilitation but also permits uninterrupted practice free from external distractions. Consequently, it effectively improves rehabilitation outcomes and ensures treatment sustainability. This study included three combined therapies, with the regimens combining virtual reality therapy with robotic therapy and robotic therapy with FES ranking highly. Robots for stroke rehabilitation are primarily categorized into therapeutic robots and assistive robots[\u003cspan citationid=\"CR118\" class=\"CitationRef\"\u003e118\u003c/span\u003e]. Therapeutic robots are primarily used to promote the recovery of patients' motor functions through task-specific training, with common types including end-effector and exoskeleton robots[\u003cspan citationid=\"CR119\" class=\"CitationRef\"\u003e119\u003c/span\u003e]. Another category is assistive robots, which provide external assistance to help patients perform daily activities, achieving functional compensation. Robotic rehabilitation offers high-intensity, high-dose, and highly repetitive task-oriented training, creating the necessary physical stimuli for neuroplasticity, thereby promoting the reorganization and activation of damaged motor neural networks[\u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e120\u003c/span\u003e]. However, robotic rehabilitation typically exhibits a high degree of passivity. As training duration increases, this passive training model may lead to a decline in muscle activation levels and reduce patients' willingness to actively participate[\u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e121\u003c/span\u003e]. In contrast, virtual reality technology, with its immersive experience, not only effectively enhances patients' training adherence but also further optimizes the rehabilitation process by promoting active engagement[\u003cspan citationid=\"CR122\" class=\"CitationRef\"\u003e122\u003c/span\u003e]. The integration of virtual reality and robotic technology offers a complementary rehabilitation approach with multiple advantages. Virtual reality therapy promotes neuroplasticity by providing real-time feedback, adjustable task difficulty, and diverse contextual training, which helps activate brain regions associated with motor control and cognition. Meanwhile, robotic therapy reinforces the establishment and consolidation of neural circuits through repetitive goal-oriented motor tasks. Particularly when patients have limited voluntary movement ability, the mechanical assistance and repetitive support provided by robots facilitate relearning and functional reorganization of impaired neural networks. By combining both technologies, it is possible not only to enhance patient motivation and training adaptability through virtual reality but also to deliver precise and stable physical training support via robotic systems. This integrated approach strengthens active patient engagement while improving overall rehabilitation efficiency and functional recovery outcomes. Functional electrical stimulation (FES) induces muscle contractions by applying electrical currents to impaired nerves and muscles\u003csup\u003e117\u003c/sup\u003e. The integration of FES with robotic rehabilitation technology creates a synergistic therapeutic paradigm, with its principal advantages manifested in the following aspects: At the neuromuscular activation level, FES enables targeted activation of specific muscle groups, effectively compensating for potential insufficient muscle activation during robotic training. This is particularly beneficial for patients with significantly limited voluntary motor ability. In terms of motor control, robotic systems provide precise trajectory guidance and mechanical support, ensuring normative movement patterns while supplying real-time kinematic parameters for the temporal control of FES. This facilitates high-precision synchronization between electrical stimulation and motor execution. Furthermore, the combination of the two technologies can establish a closed-loop \"stimulation-movement-feedback\" training system. By enhancing the consistency between proprioceptive input and motor output, it promotes sensorimotor integration and neuroplasticity. Additionally, the integrated approach allows for individualized parameter adjustment, enabling dynamic modulation of robotic assistance intensity and FES stimulation parameters based on the patient's functional status. This ensures training intensity while optimizing energy efficiency and reducing the development of abnormal compensatory movement patterns. Currently, combination therapy has become a significant research direction in the field of post-stroke motor function recovery. By integrating the complementary mechanisms of different intervention approaches, this model holds the potential to overcome the limitations of single-modality therapies and achieve multi-level, multi-target synergistic rehabilitation effects. However, this area remains in the early stages of exploration, with relatively limited existing research. Future efforts should focus on conducting more high-quality clinical studies to clarify its long-term efficacy and gradually establish standardized, generalizable treatment parameters and implementation protocols. Given the limitations of traditional pharmacological treatments in rehabilitating lower limb function in stroke patients with hemiplegia\u0026mdash;including frequent side effects, high dependency, and uncertain long-term efficacy\u0026mdash;non-pharmacological therapies have increasingly become important complementary or even alternative strategies. Compared to drug-based interventions, non-pharmacological approaches not only alleviate symptoms but also promote sustained functional recovery by enhancing neuroplasticity and motor learning, while avoiding risks such as drug dependence. These interventions primarily focus on lifestyle modifications, systematic exercise training, physical stimulation, and neuromodulation techniques, directly targeting motor dysfunction to improve muscular coordination and promote neural network reorganization. Notably, non-pharmacological therapies offer significant advantages in improving patient treatment compliance and clinical applicability. This strategy emphasizes active patient participation and self-management, contributing to enhanced rehabilitation confidence and treatment continuity. In this context, Bayesian network meta-analysis provides a rigorous methodological framework for systematically evaluating various non-pharmacological interventions. By quantifying the efficacy of different treatment measures and enabling indirect comparisons, it offers evidence-based support for selecting optimal clinical treatment plans. In summary, non-pharmacological therapies not only demonstrate outstanding cost-effectiveness and safety but also promote comprehensive recovery of lower extremity movement through multidimensional and integrated rehabilitation pathways, showing significant clinical application potential. Additional high-quality studies are required to better understand the combined effects of various therapies and to develop standardized clinical treatment protocols, thus promoting the consistent use and broader adoption of this approach in rehabilitation practice.\u003c/p\u003e \u003cdiv id=\"Sec38\" class=\"Section2\"\u003e \u003ch2\u003eAdvantages and Limitations\u003c/h2\u003e \u003cp\u003eThe primary strength of this study lies in its integration of multiple non-pharmacological interventions to provide a more comprehensive analysis, representing a first in the existing literature within this field. Previous studies have typically focused on specific non-pharmacological interventions, whereas this research fills a gap by comprehensively evaluating the effects of different interventions. However, despite efforts to ensure comprehensiveness in the literature search and the absence of significant publication bias indicated by funnel plots, the possibility of overestimation of intervention effects in unpublished studies or published reports cannot be entirely ruled out. Furthermore, the number of studies for some interventions was relatively limited, with smaller sample sizes, necessitating caution in interpreting related findings. Third, given the nature of non-pharmacological interventions, difficulties in implementing blinding for both participants and researchers undoubtedly increase the risk of potential bias, potentially affecting the accuracy of results and efficacy assessments. Finally, data limitations in certain studies resulted in inadequate evaluation of some interventions, thereby restricting our comprehensive understanding of their therapeutic effects. Consequently, future research should prioritize expanding sample sizes, adopting more rigorous randomized designs, and implementing blinding protocols to generate more reliable evidence that provides stronger support for clinical application.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis research carried out a comprehensive assessment of the effectiveness of non-pharmacological treatments targeting lower limb motor impairments in individuals with post-stroke hemiplegia. The results indicate that rTMS may be the most effective intervention for improving lower limb motor function, followed by tDCS and mirror therapy. For patients exhibiting functional limitations such as reduced walking speed and poor dynamic balance control, a comprehensive training program combining virtual reality technology with robot-assisted rehabilitation is recommended. \"Nevertheless, because of constraints in both the amount and quality of existing evidence\u0026mdash;including limited sample sizes and possible bias risks\u0026mdash;the interpretation and broader applicability of these findings should be treated with caution. Future studies are urgently needed to conduct larger, methodologically sound, high-quality randomized controlled trials to further confirm and refine the clinical implementation of various intervention strategies.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eNetwork Meta-Analysis: NMA\u003c/p\u003e\n\u003cp\u003eRandomized Controlled Trial: RCT\u003c/p\u003e\n\u003cp\u003eFugl-Meyer Assessment for Lower Extremities: FMA-LE\u003c/p\u003e\n\u003cp\u003eBerg Balance Scale: BBS\u003c/p\u003e\n\u003cp\u003eModified Barthel Index: MBI\u003c/p\u003e\n\u003cp\u003eFunctional Ambulation Classification: FAC\u003c/p\u003e\n\u003cp\u003eTimed Up and Go test: TUG\u003c/p\u003e\n\u003cp\u003eSurface\u0026nbsp;Under the\u0026nbsp;Cumulative\u0026nbsp;Ranking curve\u0026nbsp;Analysis: SUCRA\u003c/p\u003e\n\u003cp\u003eStandardized\u0026nbsp;Mean\u0026nbsp;Difference: SMD\u003c/p\u003e\n\u003cp\u003eConventional Therapy:\u0026nbsp;CT\u003c/p\u003e\n\u003cp\u003eAcupuncture: ACPU\u003c/p\u003e\n\u003cp\u003eVibration therapy: VT\u003c/p\u003e\n\u003cp\u003eMirror therapy: MT\u003c/p\u003e\n\u003cp\u003eConstraint-induced movement therapy: CIMT\u003c/p\u003e\n\u003cp\u003eVirtual reality: VR\u003c/p\u003e\n\u003cp\u003eintermittent theta burst stimulation: iTBS\u003c/p\u003e\n\u003cp\u003eFunctional electrical stimulation: FES\u003c/p\u003e\n\u003cp\u003eRepetitive Transcranial Magnetic Stimulation: rTMS\u003c/p\u003e\n\u003cp\u003eTranscranial Direct Current Stimulation: tDCS\u003c/p\u003e\n\u003cp\u003eRobot-assisted rehabilitation: RAR\u003c/p\u003e\n\u003cp\u003eWater therapy: WT\u003c/p\u003e\n\u003cp\u003emirror visual feedback: MVF\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have approved the submission of this manuscript. This work is original, has not been published previously, and is not under consideration for publication elsewhere.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe primary contributions of this study are detailed in the article or supplementary material. For further inquiries, please reach out to the corresponding author.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors state that there are no competing interests associated with this manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study is funded by the National Natural Science Foundation of China (82575221) and the National Natural Science Foundation of China (82374602).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026apos; contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eStudy objective: DZ, DHL. Literature search: DZ, TYY. Data extraction: DZ, TYY.\u003cbr\u003e\u0026nbsp;Methodological quality assessment: DZ, GXX, LW, GLZ. Data Analysis: GXX, LW, WYS, MYY. Manuscript writing: DZ. Critical review and manuscript approval: JL, DHL. All authors have reviewed and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe gratefully acknowledge the support from the National Natural Science Foundation of China.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eMarsh JD, Keyrouz SG. Stroke prevention and treatment. J Am Coll Cardiol. 2010; 56(9):683\u0026ndash;691. \u003c/li\u003e\n\u003cli\u003eFeigin VL, Brainin M, Norrving B, Martins SO, Pandian J, Lindsay P, et al. World Stroke Organization: Global Stroke Fact Sheet 2025. Int J Stroke Off J Int Stroke Soc. 2025; 20(2):132\u0026ndash;144. \u003c/li\u003e\n\u003cli\u003eGBD 2019 Stroke Collaborators. Global, regional, and national burden of stroke and its risk factors, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Neurol. 2021; 20(10):795\u0026ndash;820. \u003c/li\u003e\n\u003cli\u003eFeigin VL, Owolabi MO, World Stroke Organization\u0026ndash;Lancet Neurology Commission Stroke Collaboration Group. Pragmatic solutions to reduce the global burden of stroke: a World Stroke Organization-Lancet Neurology Commission. Lancet Neurol. 2023; 22(12):1160\u0026ndash;1206. \u003c/li\u003e\n\u003cli\u003eXiong J, Wang J-T, Lin S, Xie B-Y. Advances in hemiplegia rehabilitation: modern therapeutic interventions to enhance activities of daily living. Front Neurol. 2025; 16:1555990. \u003c/li\u003e\n\u003cli\u003eLiu Z, Xin H, Chopp M. Axonal remodeling of the corticospinal tract during neurological recovery after stroke. Neural Regen Res. 2021; 16(5):939\u0026ndash;943.\u003c/li\u003e\n\u003cli\u003eJoy MT, Carmichael ST. Encouraging an excitable brain state: mechanisms of brain repair in stroke. Nat Rev Neurosci. 2021; 22(1):38\u0026ndash;53. \u003c/li\u003e\n\u003cli\u003eLi S. Stroke Recovery Is a Journey: Prediction and Potentials of Motor Recovery after a Stroke from a Practical Perspective. Life Basel Switz. 2023; 13:2061.\u003c/li\u003e\n\u003cli\u003eStrilciuc S, Grad DA, Radu C, Chira D, Stan A, Ungureanu M, et al. The economic burden of stroke: a systematic review of cost of illness studies. J Med Life. 2021; 14(5):606\u0026ndash;619. \u003c/li\u003e\n\u003cli\u003eJ\u0026oslash;rgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil. 1995; 76(1):27\u0026ndash;32. \u003c/li\u003e\n\u003cli\u003eSalanti G. Indirect and mixed-treatment comparison, network, or multiple-treatments meta-analysis: many names, many benefits, many concerns for the next generation evidence synthesis tool. Res Synth Methods. 2012; 3(2):80\u0026ndash;97. \u003c/li\u003e\n\u003cli\u003eHutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH, Cameron C, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med. 2015; 162(11):777\u0026ndash;784. \u003c/li\u003e\n\u003cli\u003eAmir-Behghadami M, Janati A. Population, Intervention, Comparison, Outcomes and Study (PICOS) design as a framework to formulate eligibility criteria in systematic reviews. Emerg Med J EMJ. 2020; 37(6):387.\u003c/li\u003e\n\u003cli\u003eWhiting P, Savović J, Higgins JPT, Caldwell DM, Reeves BC, Shea B, et al. ROBIS: A new tool to assess risk of bias in systematic reviews was developed. J Clin Epidemiol. 2016; 69:225\u0026ndash;234. \u003c/li\u003e\n\u003cli\u003eCui W, Huang L, Tian Y, Luo H, Chen S, Yang Y, et al. Effect and mechanism of mirror therapy on lower limb rehabilitation after ischemic stroke: A fMRI study. NeuroRehabilitation. 2022; 51(1):65\u0026ndash;77. \u003c/li\u003e\n\u003cli\u003eArya KN, Pandian S, Kumar V. Effect of activity-based mirror therapy on lower limb motor-recovery and gait in stroke: A randomized controlled trial. Neuropsychol Rehabil. 2019; 29(8):1193\u0026ndash;1210. \u003c/li\u003e\n\u003cli\u003eLee D, Lee G. Effect of afferent electrical stimulation with mirror therapy on motor function, balance, and gait in chronic stroke survivors: a randomized controlled trial. Eur J Phys Rehabil Med. 2019; 55(4):442\u0026ndash;449. \u003c/li\u003e\n\u003cli\u003eMohan U, Babu SK, Kumar KV, Suresh BV, Misri ZK, Chakrapani M. Effectiveness of mirror therapy on lower extremity motor recovery, balance and mobility in patients with acute stroke: A randomized sham-controlled pilot trial. Ann Indian Acad Neurol. 2013; 16(4):634\u0026ndash;639. \u003c/li\u003e\n\u003cli\u003eKim D-H, Jang S-H. Effects of Mirror Therapy Combined with EMG-Triggered Functional Electrical Stimulation to Improve on Standing Balance and Gait Ability in Patient with Chronic Stroke. Int J Environ Res Public Health. 2021; 18(7):3721. \u003c/li\u003e\n\u003cli\u003eLee D, Lee G, Jeong J. Mirror Therapy with Neuromuscular Electrical Stimulation for improving motor function of stroke survivors: A pilot randomized clinical study. Technol Health Care Off J Eur Soc Eng Med. 2016; 24(4):503\u0026ndash;511. \u003c/li\u003e\n\u003cli\u003eİkizler May H, \u0026Ouml;zdolap Ş, Mengi A, Sarıkaya S. The effect of mirror therapy on lower extremity motor function and ambulation in post-stroke patients: A prospective, randomized-controlled study. Turk J Phys Med Rehabil. 2020; 66(2):154\u0026ndash;160. \u003c/li\u003e\n\u003cli\u003eKim K, Lee D-K, Kim E-K. Effect of aquatic dual-task training on balance and gait in stroke patients. J Phys Ther Sci. 2016; 28(7):2044\u0026ndash;2047. \u003c/li\u003e\n\u003cli\u003eKim K, Lee D-K, Jung S-I. Effect of coordination movement using the PNF pattern underwater on the balance and gait of stroke patients. J Phys Ther Sci. 2015; 27(12):3699\u0026ndash;3701. \u003c/li\u003e\n\u003cli\u003eTripp F, Krakow K. Effects of an aquatic therapy approach (Halliwick-Therapy) on functional mobility in subacute stroke patients: a randomized controlled trial. Clin Rehabil. 2014; 28(5):432\u0026ndash;439. \u003c/li\u003e\n\u003cli\u003eEyvaz N, Dundar U, Yesil H. Effects of water-based and land-based exercises on walking and balance functions of patients with hemiplegia. NeuroRehabilitation. 2018; 43(2):237\u0026ndash;246. \u003c/li\u003e\n\u003cli\u003eZhu Z, Cui L, Yin M, Yu Y, Zhou X, Wang H, et al. Hydrotherapy vs. conventional land-based exercise for improving walking and balance after stroke: a randomized controlled trial. Clin Rehabil. 2016; 30(6):587\u0026ndash;593. \u003c/li\u003e\n\u003cli\u003ePark J, Chung Y. The effects of an additional weight aquatic exercise program on balance and lower extremity strength in persons with stroke: randomized controlled study. Phys Ther Rehabil Sci. Korean Academy of Physical Therapy Rehabilitation Science; 2018; 7(1):6\u0026ndash;12. \u003c/li\u003e\n\u003cli\u003eZHU P-A, LI Z-L, LU Q-Q, NIE Y-Y, LIU H, KIERNAN E, et al. Can cerebellar theta-burst stimulation improve balance function and gait in stroke patients? A randomized controlled trial. Eur J Phys Rehabil Med. 2024; 60(3):391\u0026ndash;399. \u003c/li\u003e\n\u003cli\u003eXie Y-J, Wei Q-C, Chen Y, Liao L-Y, Li B-J, Tan H-X, et al. Cerebellar Theta Burst Stimulation on Walking Function in Stroke Patients: A Randomized Clinical Trial. Front Neurosci. 2021; 15:688569. \u003c/li\u003e\n\u003cli\u003eKoch G, Bonn\u0026igrave; S, Casula EP, Iosa M, Paolucci S, Pellicciari MC, et al. Effect of Cerebellar Stimulation on Gait and Balance Recovery in Patients With Hemiparetic Stroke. JAMA Neurol. 2019; 76(2):170\u0026ndash;178. \u003c/li\u003e\n\u003cli\u003eWang H-Q, Hou M, Bao C-L, Min L, Li H. Effects of Acupuncture Treatment on Lower Limb Spasticity in Patients Following Hemorrhagic Stroke: A Pilot Study. Eur Neurol. 2019; 81(1-2):5\u0026ndash;12. \u003c/li\u003e\n\u003cli\u003eWang H-Q, Hou M, Li H, Bao C-L, Min L, Dong G-R, et al. Effects of acupuncture treatment on motor function in patients with subacute hemorrhagic stroke: A randomized controlled study. Complement Ther Med. 2020; 49:102296. \u003c/li\u003e\n\u003cli\u003eWang N, Li Z. Life quality improvement of spastic hemiplegia of stroke treated with fire-needle: a randomized controlled trial. Zhongguo Zhen Jiu Chin Acupunct Moxibustion. 2015; 35(11):1105\u0026ndash;1109. \u003c/li\u003e\n\u003cli\u003eBai Y-L, Li L, Hu Y-S, Wu Y, Xie P-J, Wang S-W, et al. Prospective, randomized controlled trial of physiotherapy and acupuncture on motor function and daily activities in patients with ischemic stroke. J Altern Complement Med N Y N. 2013; 19(8):684\u0026ndash;689. \u003c/li\u003e\n\u003cli\u003eAloraini SM. Effects of constraint-induced movement therapy for the lower extremity among individuals post-stroke: A randomized controlled clinical trial. NeuroRehabilitation. 2022; 51(3):421\u0026ndash;431. \u003c/li\u003e\n\u003cli\u003eChoi H-S, Shin W-S, Bang D-H, Choi S-J. Effects of Game-Based Constraint-Induced Movement Therapy on Balance in Patients with Stroke: A Single-Blind Randomized Controlled Trial. Am J Phys Med Rehabil. 2017; 96(3):184\u0026ndash;190. \u003c/li\u003e\n\u003cli\u003eMenezes-Oliveira E, da Silva Matuti G, de Oliveira CB, de Freitas SF, Miyuki Kawamura C, Fernandes Lopes JA, et al. Improvement of gait and balance function in chronic post-stroke patients induced by Lower Extremity - Constraint Induced Movement Therapy: a randomized controlled clinical trial. Brain Inj. 2024; 38(7):559\u0026ndash;568. \u003c/li\u003e\n\u003cli\u003eZhang X-H, Liu J-Y, Han P, Wang Y-L, Xiao P. Clinical Efficacy of Functional Electrical Stimulation-assisted RehabilitationCycling on the Function of Lower Limbs in Patients with Stroke. Curr Neurovasc Res. 2021; 18(3):318\u0026ndash;323. \u003c/li\u003e\n\u003cli\u003eMatsumoto S, Shimodozono M, Noma T, Miyara K, Onoda T, Ijichi R, et al. Effect of Functional Electrical Stimulation in Convalescent Stroke Patients: A Multicenter, Randomized Controlled Trial. J Clin Med. 2023; 12(7):2638. \u003c/li\u003e\n\u003cli\u003eBauer P, Krewer C, Golaszewski S, Koenig E, M\u0026uuml;ller F. Functional electrical stimulation-assisted active cycling--therapeutic effects in patients with hemiparesis from 7 days to 6 months after stroke: a randomized controlled pilot study. Arch Phys Med Rehabil. 2015; 96(2):188\u0026ndash;196. \u003c/li\u003e\n\u003cli\u003eZhang X-H, Gu T, Liu X-W, Han P, Lv H-L, Wang Y-L, et al. The Effect of Transcranial Direct Current Stimulation and Functional Electrical Stimulation on the Lower Limb Function of Stroke Patients. Front Neurosci. 2021; 15:685931. \u003c/li\u003e\n\u003cli\u003eTan Z, Liu H, Yan T, Jin D, He X, Zheng X, et al. The effectiveness of functional electrical stimulation based on a normal gait pattern on subjects with early stroke: a randomized controlled trial. BioMed Res Int. 2014; 2014: 545408. \u003c/li\u003e\n\u003cli\u003eLee H-J, Cho K-H, Lee W-H. The effects of body weight support treadmill training with power-assisted functional electrical stimulation on functional movement and gait in stroke patients. Am J Phys Med Rehabil. 2013; 92(12):1051\u0026ndash;1059. \u003c/li\u003e\n\u003cli\u003eLee Y, Kim K. The influence of Gait Training Combined with Portable Functional Electrical Stimulation on motor function, balance and gait ability in stroke patients. J Back Musculoskelet Rehabil. 2022; 35(6):1171\u0026ndash;1178. \u003c/li\u003e\n\u003cli\u003eCho M-K, Kim J-H, Chung Y, Hwang S. Treadmill gait training combined with functional electrical stimulation on hip abductor and ankle dorsiflexor muscles for chronic hemiparesis. Gait Posture. 2015; 42(1):73\u0026ndash;78. \u003c/li\u003e\n\u003cli\u003ePark J. Comparison of the effects of robot-assisted gait training using FES versus virtual reality for stroke patients. A randomized clinical trial. Top Stroke Rehabil. 2025; 13:1\u0026ndash;10. \u003c/li\u003e\n\u003cli\u003eMeng G, Ma X, Chen P, Xu S, Li M, Zhao Y, et al. Effect of early integrated robot-assisted gait training on motor and balance in patients with acute ischemic stroke: a single-blinded randomized controlled trial. Ther Adv Neurol Disord. 2022; 15:17562864221123195. \u003c/li\u003e\n\u003cli\u003eLi D-X, Zha F-B, Long J-J, Liu F, Cao J, Wang Y-L. Effect of Robot Assisted Gait Training on Motor and Walking Function in Patients with Subacute Stroke: A Random Controlled Study. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc. 2021; 30(7):105807. \u003c/li\u003e\n\u003cli\u003eHuo C, Shao G, Chen T, Li W, Wang J, Xie H, et al. Effectiveness of unilateral lower-limb exoskeleton robot on balance and gait recovery and neuroplasticity in patients with subacute stroke: a randomized controlled trial. J Neuroengineering Rehabil. 2024; 21(1):213. \u003c/li\u003e\n\u003cli\u003eFan T, Zheng P, Zhang X, Gong Z, Shi Y, Wei M, et al. Effects of exoskeleton rehabilitation robot training on neuroplasticity and lower limb motor function in patients with stroke. BMC Neurol. 2025; 25(1):193. \u003c/li\u003e\n\u003cli\u003eKim S-Y, Yang L, Park IJ, Kim EJ, JoshuaPark MS, You SH, et al. Effects of Innovative WALKBOT Robotic-Assisted Locomotor Training on Balance and Gait Recovery in Hemiparetic Stroke: A Prospective, Randomized, Experimenter Blinded Case Control Study With a Four-Week Follow-Up. IEEE Trans Neural Syst Rehabil Eng Publ IEEE Eng Med Biol Soc. 2015; 23(4):636\u0026ndash;642. \u003c/li\u003e\n\u003cli\u003eKim J, Kim DY, Chun MH, Kim SW, Jeon HR, Hwang CH, et al. Effects of robot-(Morning Walk\u0026reg;) assisted gait training for patients after stroke: a randomized controlled trial. Clin Rehabil. 2019; 33(3):516\u0026ndash;523. \u003c/li\u003e\n\u003cli\u003eBae Y-H, Ko YJ, Chang WH, Lee JH, Lee KB, Park YJ, et al. Effects of Robot-assisted Gait Training Combined with Functional Electrical Stimulation on Recovery of Locomotor Mobility in Chronic Stroke Patients: A Randomized Controlled Trial. J Phys Ther Sci. 2014; 26(12):1949\u0026ndash;1953. \u003c/li\u003e\n\u003cli\u003eChang WH, Kim MS, Huh JP, Lee PKW, Kim Y-H. Effects of robot-assisted gait training on cardiopulmonary fitness in subacute stroke patients: a randomized controlled study. Neurorehabil Neural Repair. 2012; 26(4):318\u0026ndash;324. \u003c/li\u003e\n\u003cli\u003eKim D-H, In T-S, Jung K-S. Effects of robot-assisted trunk control training on trunk control ability and balance in patients with stroke: A randomized controlled trial. Technol Health Care Off J Eur Soc Eng Med. 2022; 30(2):413\u0026ndash;422. \u003c/li\u003e\n\u003cli\u003eYoo M, Chun MH, Hong GR, Lee C, Lee JK, Lee A. Effects of Training with a Powered Exoskeleton on Cortical Activity Modulation in Hemiparetic Chronic Stroke Patients: A Randomized Controlled Pilot Trial. Arch Phys Med Rehabil. 2023; 104(10):1620\u0026ndash;1629. \u003c/li\u003e\n\u003cli\u003eMin JH, Seong HY, Ko SH, Jo W-R, Sohn H-J, Ahn YH, et al. Effects of trunk stabilization training robot on postural control and gait in patients with chronic stroke: a randomized controlled trial. Int J Rehabil Res Int Z Rehabil Rev Int Rech Readaptation. 2020; 43(2):159\u0026ndash;166. \u003c/li\u003e\n\u003cli\u003eJin Y, Xiong B, Chen L, Zhao W, Li Z, Zhang C, et al. Efficacy and safety of using a unilateral lower limb exoskeleton combined with conventional treatment in post-stroke rehabilitation: a randomized controlled trial. Front Bioeng Biotechnol. 2024; 12:1441986. \u003c/li\u003e\n\u003cli\u003eLouie DR, Mortenson WB, Durocher M, Schneeberg A, Teasell R, Yao J, et al. Efficacy of an exoskeleton-based physical therapy program for non-ambulatory patients during subacute stroke rehabilitation: a randomized controlled trial. J Neuroengineering Rehabil. 2021; 18(1):149. \u003c/li\u003e\n\u003cli\u003eHaolin T, Yuanbin Y, Hu Z, Wenjing Z, Jing Z, Jingfeng T, et al. Efficacy of Daoyin combined with lower limb robot as a comprehensive rehabilitation intervention for stroke patients: a randomized controlled trial. J Tradit Chin Med Chung Tsa Chih Ying Wen Pan. 2024; 44(3):530\u0026ndash;536. \u003c/li\u003e\n\u003cli\u003eZhang Y, Zhao W, Wan C, Wu X, Huang J, Wang X, et al. Exoskeleton rehabilitation robot training for balance and lower limb function in sub-acute stroke patients: a pilot, randomized controlled trial. J Neuroengineering Rehabil. 2024; 21(1):98. \u003c/li\u003e\n\u003cli\u003eNam Y-G, Park JW, Lee HJ, Nam KY, Choi MR, Yu CS, et al. Further effects of electromechanically assisted gait trainer (Exowalk\u0026reg;) in patients with chronic stroke: A randomized controlled trial. J Rehabil Med. 2020; 52(9): jrm00097. \u003c/li\u003e\n\u003cli\u003eOchi M, Wada F, Saeki S, Hachisuka K. Gait training in subacute non-ambulatory stroke patients using a full weight-bearing gait-assistance robot: A prospective, randomized, open, blinded-endpoint trial. J Neurol Sci. 2015; 353(1-2):130\u0026ndash;136. \u003c/li\u003e\n\u003cli\u003ePalmcrantz S, Wall A, Vreede KS, Lindberg P, Danielsson A, Sunnerhagen KS, et al. Impact of Intensive Gait Training With and Without Electromechanical Assistance in the Chronic Phase After Stroke\u0026ndash;A Multi-Arm Randomized Controlled Trial With a 6 and 12 Months Follow Up. Front Neurosci. 2021; 15: 660726.\u003c/li\u003e\n\u003cli\u003eBustamante Valles K, Montes S, Madrigal M de J, Burciaga A, Mart\u0026iacute;nez ME, Johnson MJ. Technology-assisted stroke rehabilitation in Mexico: a pilot randomized trial comparing traditional therapy to circuit training in a Robot/technology-assisted therapy gym. J Neuroengineering Rehabil. 2016; 13(1):83. \u003c/li\u003e\n\u003cli\u003ePark J, Chung Y. The effects of robot-assisted gait training using virtual reality and auditory stimulation on balance and gait abilities in persons with stroke. NeuroRehabilitation. 2018; 43(2):227\u0026ndash;235. \u003c/li\u003e\n\u003cli\u003eKayabinar B, Alemdaroğlu-G\u0026uuml;rb\u0026uuml;z İ, Yilmaz \u0026Ouml;. The effects of virtual reality augmented robot-assisted gait training on dual-task performance and functional measures in chronic stroke: a randomized controlled single-blind trial. Eur J Phys Rehabil Med. 2021; 57(2):227\u0026ndash;237. \u003c/li\u003e\n\u003cli\u003eZhang H, Li X, Gong Y, Wu J, Chen J, Chen W, et al. Three-Dimensional Gait Analysis and sEMG Measures for Robotic-Assisted Gait Training in Subacute Stroke: A Randomized Controlled Trial. BioMed Res Int. 2023; 2023:7563802.\u003c/li\u003e\n\u003cli\u003eLee H-J, Lee S-H, Seo K, Lee M, Chang WH, Choi B-O, et al. Training for Walking Efficiency With a Wearable Hip-Assist Robot in Patients With Stroke: A Pilot Randomized Controlled Trial. Stroke. 2019; 50(12):3545\u0026ndash;3552. \u003c/li\u003e\n\u003cli\u003eChen L, Zhu H, Wang J, Lu R, Tian J, Wu B, et al. Virtual reality-based robotic training for lower limb rehabilitation in stroke patients with Hemiplegia: A pilot study. Aging Health Res. 2025; 5:100233. \u003c/li\u003e\n\u003cli\u003eWang L, Wang L, Wang Z, Gao F, Wu J, Tang H. Clinical Effect Analysis of Wearable Sensor Technology-Based Gait Function Analysis in Post-Transcranial Magnetic Stimulation Stroke Patients. Sensors. 2024; 24(10):3051. \u003c/li\u003e\n\u003cli\u003eGuan Y-Z, Li J, Zhang X-W, Wu S, Du H, Cui L-Y, et al. Effectiveness of repetitive transcranial magnetic stimulation (rTMS) after acute stroke: A one-year longitudinal randomized trial. CNS Neurosci Ther. 2017; 23(12):940\u0026ndash;946. \u003c/li\u003e\n\u003cli\u003eLin Y-N, Hu C-J, Chi J, Lin L-F, Yen T-H, Lin Y-K, et al. Effects of repetitive transcranial magnetic stimulation of the unaffected hemisphere leg motor area in patients with subacute stroke and substantial leg impairment: A pilot study. J Rehabil Med. 2015; 47(4):305\u0026ndash;310. \u003c/li\u003e\n\u003cli\u003eLu S, Wang H, Yuan Y, Ma T, Liu Z, Li Q, et al. Effects of repetitive transcranial magnetic stimulation on contralesional dorsal premotor cortex in subcortical stroke patients with moderate to severe upper limb motor impairment: a pilot randomized controlled trial. medRxiv. 2025; 2025.03.05.25323209.\u003c/li\u003e\n\u003cli\u003eHuang Y-Z, Lin L-F, Chang K-H, Hu C-J, Liou T-H, Lin Y-N. Priming With 1-Hz Repetitive Transcranial Magnetic Stimulation Over Contralesional Leg Motor Cortex Does Not Increase the Rate of Regaining Ambulation Within 3 Months of Stroke: A Randomized Controlled Trial. Am J Phys Med Rehabil. 2018; 97(5):339\u0026ndash;345.\u003c/li\u003e\n\u003cli\u003eYoussef H, Mohamed NAE-H, Hamdy M. Comparison of bihemispheric and unihemispheric M1 transcranial direct current stimulations during physical therapy in subacute stroke patients: A randomized controlled trial. Neurophysiol Clin Clin Neurophysiol. 2023; 53(3):102895. \u003c/li\u003e\n\u003cli\u003eDuan Q, Liu W, Yang J, Huang B, Shen J. Effect of Cathodal Transcranial Direct Current Stimulation for Lower Limb Subacute Stroke Rehabilitation. Neural Plast. 2023; 2023:1863686. \u003c/li\u003e\n\u003cli\u003eChang MC, Kim DY, Park DH. Enhancement of Cortical Excitability and Lower Limb Motor Function in Patients With Stroke by Transcranial Direct Current Stimulation. Brain Stimulat. 2015; 8(3):561\u0026ndash;566. \u003c/li\u003e\n\u003cli\u003eBian M, Shen Y, Huang Y, Wu L, Wang Y, He S, et al. A non-immersive virtual reality-based intervention to enhance lower-extremity motor function and gait in patients with subacute cerebral infarction: A pilot randomized controlled trial with 1-year follow-up. Front Neurol. 2022; 13:985700. \u003c/li\u003e\n\u003cli\u003eJunior VADS, Santos M de S, Ribeiro NM da S, Maldonado IL. Combining Proprioceptive Neuromuscular Facilitation and Virtual Reality for Improving Sensorimotor Function in Stroke Survivors: A Randomized Clinical Trial. J Cent Nerv Syst Dis. 2019; 11:1179573519863826. \u003c/li\u003e\n\u003cli\u003eXu Y, Yao J, Ni J, Yang Y, Fu L, Xu C. Comparison of Combined Virtual Reality Combined With Standing Balance Training Versus Standard Practice in Patients With Hemiplegia: A Single-Blinded, Randomized Controlled Trial. Am J Phys Med Rehabil. 2025; 104(4):312\u0026ndash;317. \u003c/li\u003e\n\u003cli\u003eSultan N, Khushnood K, Qureshi S, Altaf S, Khan MK, Malik AN, et al. Effects of Virtual Reality Training Using Xbox Kinect on Balance, Postural Control, and Functional Independence in Subjects with Stroke. Games Health J. 2023; 12(6):440\u0026ndash;444. \u003c/li\u003e\n\u003cli\u003ePark D-S, Lee D-G, Lee K, Lee G. Effects of Virtual Reality Training using Xbox Kinect on Motor Function in Stroke Survivors: A Preliminary Study. J Stroke Cerebrovasc Dis Off J Natl Stroke Assoc. 2017; 26(10):2313\u0026ndash;2319. \u003c/li\u003e\n\u003cli\u003eChen S-C, Lin C-H, Su S-W, Chang Y-T, Lai C-H. Feasibility and effect of interactive telerehabilitation on balance in individuals with chronic stroke: a pilot study. J Neuroengineering Rehabil. 2021; 18(1):71. \u003c/li\u003e\n\u003cli\u003eXu Y, Ni J, Yang Y, Yao J, Fu L, Xu C. Game-based visual feedback-guided dynamic balance training versus conventional training in patients with hemiplegia: a pilot randomized controlled trial. J Rehabil Med. 2025; 57: jrm41277.\u003c/li\u003e\n\u003cli\u003eYaman F, Akdeniz Leblebicier M, Okur İ, İmal Kızılkaya M, Kavuncu V. Is virtual reality training superior to conventional treatment in improving lower extremity motor function in chronic hemiplegic patients? Turk J Phys Med Rehabil. 2022; 68(3):391\u0026ndash;398. \u003c/li\u003e\n\u003cli\u003eMcEwen D, Taillon-Hobson A, Bilodeau M, Sveistrup H, Finestone H. Virtual reality exercise improves mobility after stroke: an inpatient randomized controlled trial. Stroke. 2014; 45(6):1853\u0026ndash;1855. \u003c/li\u003e\n\u003cli\u003eIn T, Lee K, Song C. Virtual Reality Reflection Therapy Improves Balance and Gait in Patients with Chronic Stroke: Randomized Controlled Trials. Med Sci Monit Int Med J Exp Clin Res. 2016; 22:4046\u0026ndash;4053. \u003c/li\u003e\n\u003cli\u003eBarcala L, Grecco LAC, Colella F, Lucareli PRG, Salgado ASI, Oliveira CS. Visual biofeedback balance training using wii fit after stroke: a randomized controlled trial. J Phys Ther Sci. 2013; 25(8):1027\u0026ndash;1032. \u003c/li\u003e\n\u003cli\u003eYatar GI, Yildirim SA. Wii Fit balance training or progressive balance training in patients with chronic stroke: a randomised controlled trial. J Phys Ther Sci. 2015; 27(4):1145\u0026ndash;1151. \u003c/li\u003e\n\u003cli\u003eKim JW, Lee JH. Effect of whole-body vibration therapy on lower extremity function in subacute stroke patients. J Exerc Rehabil. 2021; 17(3):158\u0026ndash;163. \u003c/li\u003e\n\u003cli\u003eLee D-K, Han J-W. Effects of active vibration exercise using a Flexi-Bar on balance and gait in patients with chronic stroke. J Phys Ther Sci. 2018; 30(6):832\u0026ndash;834. \u003c/li\u003e\n\u003cli\u003eAnnino G, Alashram A, Romagnoli C, Iovane A, Youssef TM, Tancredi V, et al. Efficacy of localized muscle vibration on lower extremity functional ability in patients with stroke: A randomized controlled trial. J Bodyw Mov Ther. 2025; 42:769\u0026ndash;776. \u003c/li\u003e\n\u003cli\u003eWei N, Cai M. Optimal frequency of whole body vibration training for improving balance and physical performance in the older people with chronic stroke: A randomized controlled trial. Clin Rehabil. 2022; 36(3):342\u0026ndash;349.\u003c/li\u003e\n\u003cli\u003eSade I, \u0026Ccedil;ekmece \u0026Ccedil;, İnanir M, Sel\u0026Ccedil;uk B, Dursun N, Dursun E. The Effect of Whole Body Vibration Treatment on Balance and Gait in Patients with Stroke. Noro Psikiyatri Arsivi. 2020; 57(4):308\u0026ndash;311. \u003c/li\u003e\n\u003cli\u003eChoi E-T, Kim Y-N, Cho W-S, Lee D-K. The effects of visual control whole body vibration exercise on balance and gait function of stroke patients. J Phys Ther Sci. 2016; 28(11):3149\u0026ndash;3152. \u003c/li\u003e\n\u003cli\u003eShimizu T, Hosaki A, Hino T, Sato M, Komori T, Hirai S, et al. Motor cortical disinhibition in the unaffected hemisphere after unilateral cortical stroke. Brain J Neurol. 2002; 125:1896\u0026ndash;1907. \u003c/li\u003e\n\u003cli\u003eSwinnen SP. Intermanual coordination: from behavioural principles to neural-network interactions. Nat Rev Neurosci. 2002; 3(5):348\u0026ndash;359. \u003c/li\u003e\n\u003cli\u003eDimyan MA, Cohen LG. Contribution of transcranial magnetic stimulation to the understanding of functional recovery mechanisms after stroke. Neurorehabil Neural Repair. 2010; 24(2):125\u0026ndash;135. \u003c/li\u003e\n\u003cli\u003eChou Y-H, Ton That V, Sundman M. A systematic review and meta-analysis of rTMS effects on cognitive enhancement in mild cognitive impairment and Alzheimer\u0026rsquo;s disease. Neurobiol Aging. 2020; 86:1\u0026ndash;10. \u003c/li\u003e\n\u003cli\u003eSlotema CW, Blom JD, van Lutterveld R, Hoek HW, Sommer IEC. Review of the efficacy of transcranial magnetic stimulation for auditory verbal hallucinations. Biol Psychiatry. 2014; 76(2):101\u0026ndash;110. \u003c/li\u003e\n\u003cli\u003eGeorge MS, Taylor JJ, Short EB. The expanding evidence base for rTMS treatment of depression. Curr Opin Psychiatry. 2013; 26(1):13\u0026ndash;18. \u003c/li\u003e\n\u003cli\u003eSheng R, Chen C, Chen H, Yu P. Repetitive transcranial magnetic stimulation for stroke rehabilitation: insights into the molecular and cellular mechanisms of neuroinflammation. Front Immunol. 2023; 14:1197422. \u003c/li\u003e\n\u003cli\u003eHummel FC, Cohen LG. Non-invasive brain stimulation: a new strategy to improve neurorehabilitation after stroke? Lancet Neurol. 2006; 5(8):708\u0026ndash;712. \u003c/li\u003e\n\u003cli\u003eKhedr EM, Abdel-Fadeil MR, Farghali A, Qaid M. Role of 1 and 3 Hz repetitive transcranial magnetic stimulation on motor function recovery after acute ischaemic stroke. Eur J Neurol. 2009; 16(12):1323\u0026ndash;1330. \u003c/li\u003e\n\u003cli\u003eChen Q-M, Yao F-R, Sun H-W, Chen Z-G, Ke J, Liao J, et al. Combining inhibitory and facilitatory repetitive transcranial magnetic stimulation (rTMS) treatment improves motor function by modulating GABA in acute ischemic stroke patients. Restor Neurol Neurosci. 2021; 39(6):419\u0026ndash;434. \u003c/li\u003e\n\u003cli\u003eIkeda T, Kobayashi S, Morimoto C. Effects of repetitive transcranial magnetic stimulation on ER stress-related genes and glutamate, \u0026gamma;-aminobutyric acid and glycine transporter genes in mouse brain. Biochem Biophys Rep. 2019; 17:10\u0026ndash;16.\u003c/li\u003e\n\u003cli\u003eXing Y, Zhang Y, Li C, Luo L, Hua Y, Hu J, et al. Repetitive Transcranial Magnetic Stimulation of the Brain After Ischemic Stroke: Mechanisms from Animal Models. Cell Mol Neurobiol. 2023; 43(4):1487\u0026ndash;1497. \u003c/li\u003e\n\u003cli\u003eSheng R, Chen C, Chen H, Yu P. Repetitive transcranial magnetic stimulation for stroke rehabilitation: insights into the molecular and cellular mechanisms of neuroinflammation. Front Immunol. 2023; 14:1197422. \u003c/li\u003e\n\u003cli\u003eGowan S, Hordacre B. Transcranial Direct Current Stimulation to Facilitate Lower Limb Recovery Following Stroke: Current Evidence and Future Directions. Brain Sci. 2020; 10(5):310. \u003c/li\u003e\n\u003cli\u003eKenney-Jung DL, Blacker CJ, Camsari DD, Lee JC, Lewis CP. Transcranial Direct Current Stimulation: Mechanisms and Psychiatric Applications. Child Adolesc Psychiatr Clin N Am. 2019; 28(1):53\u0026ndash;60. \u003c/li\u003e\n\u003cli\u003eBai X, Guo Z, He L, Ren L, McClure MA, Mu Q. Different Therapeutic Effects of Transcranial Direct Current Stimulation on Upper and Lower Limb Recovery of Stroke Patients with Motor Dysfunction: A Meta-Analysis. Neural Plast. 2019; 2019:1372138.\u003c/li\u003e\n\u003cli\u003eBrunoni AR, Amadera J, Berbel B, Volz MS, Rizzerio BG, Fregni F. A systematic review on reporting and assessment of adverse effects associated with transcranial direct current stimulation. Int J Neuropsychopharmacol. 2011; 14(8):1133\u0026ndash;1145. \u003c/li\u003e\n\u003cli\u003eApar\u0026iacute;cio LVM, Guarienti F, Razza LB, Carvalho AF, Fregni F, Brunoni AR. A Systematic Review on the Acceptability and Tolerability of Transcranial Direct Current Stimulation Treatment in Neuropsychiatry Trials. Brain Stimulat. 2016; 9(5):671\u0026ndash;681. \u003c/li\u003e\n\u003cli\u003eThieme H, Morkisch N, Mehrholz J, Pohl M, Behrens J, Borgetto B, et al. Mirror therapy for improving motor function after stroke. Cochrane Database Syst Rev. 2018; 7:CD008449. \u003c/li\u003e\n\u003cli\u003eBroderick P, Horgan F, Blake C, Ehrensberger M, Simpson D, Monaghan K. Mirror therapy for improving lower limb motor function and mobility after stroke: A systematic review and meta-analysis. Gait Posture. 2018; 63:208\u0026ndash;220. \u003c/li\u003e\n\u003cli\u003eMichielsen ME, Selles RW, van der Geest JN, Eckhardt M, Yavuzer G, Stam HJ, et al. Motor recovery and cortical reorganization after mirror therapy in chronic stroke patients: a phase II randomized controlled trial. Neurorehabil Neural Repair. 2011; 25(3):223\u0026ndash;233. \u003c/li\u003e\n\u003cli\u003eChang WH, Kim Y-H. Robot-assisted Therapy in Stroke Rehabilitation. J Stroke. 2013; 15:174\u0026ndash;181. \u003c/li\u003e\n\u003cli\u003eMehrholz J, Pohl M. Electromechanical-assisted gait training after stroke: a systematic review comparing end-effector and exoskeleton devices. J Rehabil Med. 2012; 44(3):193\u0026ndash;199. \u003c/li\u003e\n\u003cli\u003eBonanno L, Cannuli A, Pignolo L, Marino S, Quartarone A, Calabr\u0026ograve; RS, et al. Neural Plasticity Changes Induced by Motor Robotic Rehabilitation in Stroke Patients: The Contribution of Functional Neuroimaging. Bioeng Basel Switz. 2023; 10(8):990.\u003c/li\u003e\n\u003cli\u003eWang J, Zhang H, Ma J, Gu L, Li X. Efficacy of combined non-invasive brain stimulation and robot-assisted gait training on lower extremity recovery post-stroke: a systematic review and meta-analysis of randomized controlled trials. Front Neurol. 2025; 16:1500020. \u003c/li\u003e\n\u003cli\u003eWankhede NL, Koppula S, Ballal S, Doshi H, Kumawat R, Raju Ss, et al. Virtual reality modulating dynamics of neuroplasticity: Innovations in neuro-motor rehabilitation. Neuroscience. 2025; 566:97\u0026ndash;111. \u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"journal-of-neuroengineering-and-rehabilitation","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jner","sideBox":"Learn more about [Journal of NeuroEngineering and Rehabilitation](http://jneuroengrehab.biomedcentral.com/)","snPcode":"12984","submissionUrl":"https://submission.nature.com/new-submission/12984/3","title":"Journal of NeuroEngineering and Rehabilitation","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Network Meta-Analysis, stroke, hemiplegia, non-pharmacological interventions, randomized controlled trials","lastPublishedDoi":"10.21203/rs.3.rs-8591907/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8591907/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eHemiplegia resulting from a stroke often causes significant dysfunction in lower limb movement, greatly impacting an individual's ability to walk and maintain balance. While non-pharmacological interventions show therapeutic potential, the optimal rehabilitation approach remains unclear. This study employs a network meta-analysis (NMA) to assess these treatments and offer evidence-based recommendations for clinical application.\u003c/p\u003e\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eTo evaluate the efficacy and comparative ranking of non-pharmacological interventions in improving lower limb motor function, balance, walking ability, and activities of daily living in individuals with post-stroke hemiplegia.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe conducted a search of PubMed, Embase, Cochrane Library, and Web of Science databases for randomized controlled trials (RCTs) published from January 2010 to August 2025. The Cochrane Risk of Bias Tool and Review Manager 5.4 were used to assess study quality, and evidence was graded with GRADEPro. Using R Studio software, a NMA was carried out to evaluate the clinical efficacy of various treatments in improving lower limb motor function in patients with post-stroke hemiplegia, ranked by the surface under the cumulative ranking curve (SUCRA). The study was officially registered in PROSPERO under the number CRD420251169037.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThis study employed a NMA incorporating 82 randomized controlled trials involving 3,514 patients and covering 16 non-pharmacological interventions. Results indicated that repetitive transcranial magnetic stimulation (rTMS) (SMD\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;3.68; 95% CI: \u0026minus;5.93 to \u0026minus;\u0026thinsp;0.93) demonstrated the most significant effect in improving lower limb motor function (FMA-LE score) in patients with post-stroke hemiplegia. Additionally, rTMS (SMD\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;12.32; 95% CI: \u0026minus;15.07 to \u0026minus;\u0026thinsp;9.57) (SMD\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;13.51; 95% CI: \u0026minus;16.32 to \u0026minus;\u0026thinsp;10.73) showed optimal efficacy in enhancing patients' balance ability and activities of daily living. Transcranial direct current stimulation (tDCS) (SMD\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;1.47; 95% CI: \u0026minus;2.54 to \u0026minus;\u0026thinsp;0.42) demonstrated the best efficacy in improving Functional Ambulation Classification. The combination intervention of virtual reality and robotic rehabilitation (SMD\u0026thinsp;=\u0026thinsp;6.39; 95% CI: 4.56 to 7.99) yielded the most favorable results in reducing the Timed Up and Go test time.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThis NMA suggests that rTMS might be the preferred non-pharmacological approach for enhancing lower limb motor function in patients with post-stroke hemiplegia. For individuals experiencing difficulties with walking and balance, a combined approach using virtual reality and robot-assisted rehabilitation is recommended. However, due to the limited number of studies and small sample sizes, the evidence remains preliminary. Therefore, future large-scale, multicenter, double-blind randomized controlled trials are needed to confirm and extend these results.\u003c/p\u003e","manuscriptTitle":"Effectiveness of Non-Pharmacological Interventions for Lower Limb Motor Impairment in Post-Stroke Hemiplegia: A Systematic Review and Bayesian Network Meta-Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-23 08:48:23","doi":"10.21203/rs.3.rs-8591907/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-17T05:49:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-17T05:45:40+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-16T08:24:42+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-03-06T13:15:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"83977499731913502709982522000342013742","date":"2026-02-18T13:42:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"123386857567222860956238181841611481589","date":"2026-02-17T16:03:44+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"76246383102959072030504212787942183011","date":"2026-02-17T02:31:30+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-17T01:50:03+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-14T05:26:26+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-14T05:23:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"Journal of NeuroEngineering and Rehabilitation","date":"2026-01-13T12:03:10+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"journal-of-neuroengineering-and-rehabilitation","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jner","sideBox":"Learn more about [Journal of NeuroEngineering and Rehabilitation](http://jneuroengrehab.biomedcentral.com/)","snPcode":"12984","submissionUrl":"https://submission.nature.com/new-submission/12984/3","title":"Journal of NeuroEngineering and Rehabilitation","twitterHandle":"@BioMedCentral","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"19de5280-908a-4c6b-983a-d7d3f7cdd7ba","owner":[],"postedDate":"February 23rd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-20T23:38:28+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-23 08:48:23","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8591907","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8591907","identity":"rs-8591907","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}