Objective: The aim of this meta-analysis was to provide an updated comprehensive analysis of the extant literature on physical activity (PA) training and cognitive performance post-stroke. We also sought to explore study and sample characteristics that may moderate the effects of PA training on cognition.
Methods: We synthesized data from 639 participants included in 13 intervention trials that involved PA training and assessments of neurocognitive function. Intervention effects were represented by Hedges’ g, calculated separately for intervention and control conditions within each trial. Effect size data were subjected to moderation analyses using the between-group heterogeneity (QB) test. Data were analyzed using random and mixed-effects approaches.
Results: A positive, moderate effect size was observed for PA training on a domain general estimate of global cognition (Hedges’ g [95% Confidence Interval] = 0.56 [0.41; 0.70], p < .001). Furthermore, a mixed-effects analysis indicated that PA training led to significantly greater cognitive gains relative to controls (QB = 9.26, p = 0.002). Among those that received PA training, performance improvements were observed in all neurocognitive domains assessed, including working memory (Hedges’ g = 0.65 [0.19; 1.12]; p = .006), executive function (Hedges’ g = 0.24 [0.10; 0.38]; p = .001), and attention and processing speed (Hedges’ g = 0.31 [0.10; 0.51]; p = .004). The effects of PA training on cognition were not moderated by intervention length or time from stroke to initiation of the intervention.
Conclusions: The effects of PA training on cognition post-stroke are moderate in magnitude, and are apparent across numerous neurocognitive domains. These findings support the application of physical activity training to reduce the burden of cognitive impairment in stroke survivors.