Moving towards precision post-stroke rehabilitation: A systematic review and meta-analysis of wearable sensor-derived walking activities in daily living

BackgroundPhysical rehabilitation interventions can enhance functional capacity after stroke, but improved ability doesnt always lead to better real-world performance. Understanding this gap is key to optimizing post-stroke recovery strategies. ObjectivesThis research aimed to evaluate the effectiveness of physical rehabilitation interventions specifically exercise, behavior change techniques (BCTs), or their combination on real-world walking measured using wearable sensors and capacity (gait speed and walking endurance) outcomes in stroke survivors. MethodsThis systematic review and meta-analysis followed PRISMA guidelines. Comprehensive searches were conducted in Medline, Embase, CINAHL, and Scopus up to January 2025. Randomized controlled trials involving stroke survivors receiving physical rehabilitation interventions--exercise, BCTs, or both--compared to exercise-only or usual care were included. Outcomes assessed were daily steps, gait speed (comfortable and fastest), and endurance (6-minute walk test). Meta-analyses using random-effects models (STATA 18) reported standardized mean differences (SMDs) and 95% confidence intervals (CIs). Heterogeneity was evaluated using I{superscript 2} statistics. ResultsOf 1,782 references screened, 28 studies met the inclusion criteria, and 23 were included in meta-analyses comprising 2,327 participants. Exercise-only interventions produced a small but significant improvement in daily steps (SMD = 0.23; 95% CI: 0.03 to 0.44; I2 = 36.6%; moderate certainty), and moderate improvements in both comfortable gait speed (SMD = 0.38; 95% CI: 0.19 to 0.57; I2 = 35.5%; moderate certainty) and endurance (SMD = 0.39; 95% CI: 0.26 to 0.52; I2 = 0%; moderate certainty). BCT-only interventions demonstrated a larger effect on daily steps (SMD = 0.41; 95% CI: 0.19 to 0.63; I2 = 0%; moderate certainty). In contrast, combined exercise and BCT interventions did not yield significant improvements in any outcomes and were supported by very low to low certainty of evidence. ConclusionExercise-only interventions improve gait speed and endurance after stroke, with small gains in daily steps. BCT-only interventions yield greater improvements in daily walking activity. Combined interventions show limited added benefit. Protocol registrationThis study has been registered in PROSPERO (No. CRD42023411679)