This systematic review and meta-analysis pooled randomized controlled trials comparing contralesional low-frequency repetitive transcranial magnetic stimulation (LF-rTMS) (≤1 Hz) versus sham in people after stroke, with comparable rehabilitation interventions in both arms. Thirty trials (33 comparisons) enrolling a total of 1,668 participants were included. The prespecified primary outcome was change in the Fugl-Meyer Assessment for the upper extremity (FMA-UE).
Across trials, LF-rTMS produced greater improvement in FMA-UE than sham, with a mean difference of 4.11 points (95% CI 2.83–5.39). Standardized effect size was Hedges g 0.64 (95% CI 0.45–0.84). The pooled estimates indicate a statistically significant but modest clinical benefit in upper-limb motor recovery. Substantial heterogeneity was present across studies, which the authors accounted for using random-effects meta-analysis with restricted maximum likelihood estimation and Knapp–Hartung adjustment.
The investigators explored whether baseline clinical severity of motor impairment modified LF-rTMS effects. Continuous meta-regression using baseline severity did not demonstrate a significant effect modification. In other words, baseline motor impairment alone, measured continuously across trials, did not reliably predict who would benefit more from contralesional LF-rTMS in the included studies.
The authors emphasize that these results do not support using baseline clinical severity by itself as a selection criterion for LF-rTMS candidacy based on the available aggregate trial data.
Within-trial, biomarker-stratified analyses reported in some included trials were synthesized exploratorily. These analyses suggested that participants with preserved corticospinal tract (CST) integrity, or those with positive motor-evoked potential (MEP) status, showed larger treatment effects favoring LF-rTMS compared with sham.
The review frames these observations as hypothesis-generating rather than definitive. The authors state that CST integrity and MEP status are candidate biomarkers that require confirmation in adequately powered, prospectively biomarker-stratified randomized trials before they can be used to guide clinical decision-making.
Beyond clinical scales, the meta-analysis synthesized available neurophysiological outcomes reported in trials. Contralesional LF-rTMS was associated with shortening of MEP latency and central motor conduction time. However, despite these physiological changes, the authors caution that these measures could not be validated as surrogate endpoints for clinical motor recovery within the evidence base reviewed.
Therefore, while LF-rTMS exerts measurable neurophysiological effects on corticospinal conduction metrics, their translation to meaningful functional improvement or their suitability as trial surrogate markers remains unproven in the current literature.
The systematic search covered seven databases from inception to July 2026: PubMed, Embase, Web of Science, Cochrane CENTRAL, CNKI, Wanfang Data, and VIP. Eligible studies were randomized controlled trials of contralesional LF-rTMS (≤1 Hz) versus sham after stroke with matched rehabilitation in both arms. The analysis used aggregate data extracted from published reports; no new individual participant data were collected.
The review was prospectively registered in PROSPERO (CRD420261441561). Extracted study-level data supporting the meta-analysis are provided in the manuscript and supporting files, and the underlying dataset is available at the cited Zenodo DOI. The authors declared no competing interests and reported adherence to relevant ethical and reporting guidelines.
The pooled randomized evidence indicates that contralesional LF-rTMS produces a modest but statistically significant improvement in upper-limb motor function after stroke as measured by FMA-UE. Baseline clinical motor severity alone did not modify response in continuous meta-regression analyses, limiting its utility as a standalone selection criterion.
Exploratory within-trial stratified analyses point to CST integrity and MEP status as potential biomarkers associated with larger treatment effects. These findings are explicitly described as exploratory and hypothesis-generating; the authors call for adequately powered biomarker-stratified randomized trials to test whether CST integrity or MEP status can reliably identify responders and guide clinical use of LF-rTMS.
Neurophysiological changes (shorter MEP latency and central motor conduction time) were observed but are not validated surrogate endpoints for clinical recovery based on current trial evidence.
In practice, contralesional LF-rTMS may be considered as an adjunct to rehabilitation for post-stroke upper-limb recovery, recognizing that the overall benefit size is modest and that biomarker-based patient selection remains investigational. Future work should focus on prospective validation of CST-related biomarkers, standardization of outcome measurement, and adequately powered trials that test biomarker-stratified treatment strategies.