Effects of repetitive peripheral magnetic stimulation on upper extremity motor function recovery after stroke: a meta-analysis and dose-response study

BackgroundRepetitive peripheral magnetic stimulation (rPMS), a representative non-invasive neuromodulation technique, is widely utilized for the recovery of motor dysfunction following stroke. Although its clinical efficacy has been confirmed, discrepancies among studies and the optimal rPMS stimulation parameters remain unclear. This study aims to systematically analyze and quantitatively evaluate the optimal stimulation parameters using rPMS parameter subgroups extracted from existing studies.MethodsConducted in accordance with PRISMA guidelines, this study searched for research related to rPMS and the Fugl-Meyer Assessment for Upper Extremity (FMA-UE) in stroke patients. A systematic review and meta-analysis of the aggregated studies were then performed. Furthermore, a robust error meta-regression (REMR) model was employed to explore the non-linear dose-response relationship between rPMS stimulation parameters (frequency, intensity, duration, and treatment days) and FMA-UE scores.ResultsA total of 14 trials (n = 580) were included. The results indicated that rPMS yielded a significant therapeutic effect on FMA-UE score improvement (SMD = 0.91, 95% CI 0.31–1.51; p = 0.003) and spasticity reduction (SMD = −1.15, 95% CI − 1.80 to −0.49; p = 0.0006). Dose-response analysis revealed an inverted U-shaped curve for both frequency and duration: the greatest clinical benefits were achieved with optimal stimulation at 10 Hz (peak gain: 13.82 points, 95% CI 9.65–18.00), 10–20 min per session, a plateau effect at 20–55% maximum stimulator output (MSO), and a treatment course of ≥21 days. During the subacute stroke window (14 days to 6 months), neural-targeted stimulation (e.g., brachial plexus, radial nerve) demonstrated superiority over muscle-targeted approaches (SMD = 0.81 vs. 0.47; p = 0.006).ConclusionUnder optimal parameter windows, the therapeutic mechanism of rPMS may be associated with triggering homeostatic plasticity and beta-band corticomuscular coherence. The greatest benefits are obtained particularly with neural-targeted protocols during the subacute phase of stroke, utilizing low frequency (≤20 Hz), moderate-to-low intensity (20–55% MSO), and an extended treatment course (≥21 days). In conclusion, current evidence provides a novel scientific basis and clinical reference for the application of rPMS in stroke rehabilitation.