Dose-response analysis of the effects of digital interventions on executive function after stroke in middle-aged and older adults: A systematic review and preliminary multilevel meta-analysis.

This study evaluated the effects of digital interventions on executive function in middle-aged and older adults with stroke and explored the dose-response relationship between intervention dose and executive function improvement. We systematically searched PubMed, Web of Science, Embase, and the Cochrane Library to identify relevant studies. Effect sizes were calculated using Hedges' g, and a three-level random-effects model was fitted using the metafor package in R. Potential dose-response associations and moderator variables were further examined. Nine studies from five countries were included, involving 396 people with stroke aged 56-73 years. The three-level meta-analysis showed that digital interventions had a small but statistically significant effect on overall executive function (Hedges' g = 0.25, 95% confidence interval [0.02, 0.49]). However, no statistically significant effects were observed for working memory (Hedges' g = 0.23, 95% confidence interval [-0.23, 0.49]), inhibitory control (Hedges' g = 0.44, 95% confidence interval [-0.24, 1.12]), or cognitive flexibility (Hedges' g = 0.24, 95% confidence interval [-0.50, 0.98]). Exploratory dose-response analysis suggested that, within the range of the included studies, higher predicted effects on executive function were observed around 945 min of total intervention duration and around 15 intervention sessions. Moderator analysis suggested that computer-based cognitive interventions showed a relatively larger observed effect estimate for executive function (Hedges' g = 0.65, 95% confidence interval [0.48, 0.82]). Based on low-certainty evidence, digital interventions may produce small improvements in executive function among middle-aged and older adults with stroke, with effects potentially influenced by intervention characteristics. These findings should be interpreted cautiously and require confirmation in future high-quality randomized controlled trials.