Effects of hand motor interventions on cognitive outcomes post-stroke: A systematic review and Bayesian Network Meta-Analysis

OBJECTIVE:To synthetize the evidence on the effects of hand rehabilitation interventions on cognition post-stroke and compare their efficacy. DATA SOURCES:PubMed, Embase, Cochrane, Scopus, Web of Science and CINAHL were searched from inception to November 2022. DATA SELECTION:Randomized controlled trials conducted in adults with stroke where the effects of hand motor interventions on any cognitive domains were assessed. DATA EXTRACTION:Data were extracted by 2 independent reviewers. A Bayesian Network Meta-analysis (NMA) was applied for measures with enough studies and comparisons. Risk of bias was assessed with the Cochrane Risk of Bias tool. DATA SYNTHESIS:Fifteen studies were included in qualitative synthesis, and 11 in NMA. Virtual reality (n=7), robot-assisted (n=5) or handgrip strength (n=3) training were the experimental interventions and conventional rehabilitation (n=14) the control intervention. Two separate NMA were performed with MoCA (n=480 participants) and MMSE (n=350 participants) as outcome measures. Both coincided that the most probable best interventions were robot-assisted and strength training, according to SUCRA and rankogram, followed by conventional rehabilitation and virtual reality training. No significant differences between any of the treatments were found in the MoCA network, but in the MMSE, robot-assisted and strength training were significantly better than conventional rehabilitation and virtual reality. No significant differences between robot-assisted and strength training were found nor between conventional rehabilitation and virtual reality. CONCLUSIONS:Motor interventions can improve MoCA/MMSE scores post-stroke. Most probable best interventions were robot-assisted and strength training. Limited literature assessing domain-specific cognitive effects was found.