Introduction Severe stroke frequently leads to adult disability and a loss of functional independence. Specifically, stroke-induced upper limb motor impairment significantly compromises the quality of daily life for these patients. Therefore, evaluating the efficacy of non-invasive brain stimulation combined with motor imagery for the recovery of upper limb motor function is critical for improving patient outcomes. Methods A systematic search for relevant randomized controlled trials (RCTs) was conducted across PubMed, Cochrane Library, ScienceDirect, Web of Science, EMBASE, CNKI, Wanfang Data, and VIP databases, spanning from their inception through September 4, 2025. Key outcome measures included upper extremity motor function (assessed via FMA-UE), upper extremity functional activity (ARAT and WMFT), and activities of daily living (MBI). Statistical analyses were performed using Stata 18.0. Effect sizes were calculated as weighted mean differences (WMD) or standardized mean deviations (SMD) with 95% confidence intervals (CI) using random-effects models. Results The meta-analysis, comprising 21 studies from 16 articles, demonstrated that MI combined with NIBS significantly improved upper extremity motor function (WMD = 5.75; 95% CI = 3.67, 7.82; P 0.001), upper extremity functional activity (SMD = 1.18; 95% CI = 0.78, 1.57; P 0.001), and activities of daily living (WMD = 8.94; 95% CI = 4.36, 13.53; P 0.001). Subgroup analyses identified stroke chronicity, NIBS modality, stimulation and training duration, MI paradigm, and comparator type as significant moderators of the intervention effect. Conclusion The combined intervention of MI and NIBS constitutes an effective and safe rehabilitation strategy that significantly enhances upper limb motor function and activities of daily living following stroke. Nevertheless, further high-quality, large-scale randomized controlled trials are warranted to validate these findings and establish standardized clinical protocols. Systematic review registration PROSPERO (Registration number: CRD420251167952).
Hou et al. (Wed,) studied this question.