Predictors of gait speed post-stroke: A systematic review and meta-analysis

Abstract

Background: While gait speed serves as a clinical marker and health indicator, there is a paucity of a consolidated review of the factors that are most predictive of gait speed across the clinical stages of stroke recovery.

Research question: What are predictors of gait speed in the acute, subacute and chronic phase of stroke, categorized according to International Classification of Functioning, Health, and Disability (ICF)?

Methods: A systematic search was conducted using four electronic databases following the PRISMA guideline. Included studies were cross-sectional, cohort and case-control reporting the predictors of gait speed, published from January 2000 to July 2024, and involved participants 18 years or older with diagnosis of stroke. Four meta-analyses were performed.

Results: The initial search yielded 311 articles. After screening, 32 articles were included in the final analysis. In all clinical stages of stroke, age was the most common predictor of gait speed, followed by admission walking speed, Berg Balance Scale (BBS) score and time since onset. Knee extensor strength emerged as a predictor in three studies, all in the chronic stage. The meta-analyses indicated that older age was associated with slower discharge gait speed (SMD: -0.004 [-0.01, -0.001]; p < 0.0001) while higher BBS scores at admission were associated with a larger change in gait speed between admission and discharge (SMD: 0.17 [0.06, 0.28]; p = 0.002).

Significance: Understanding the modifiable factors can help clinicians target interventions and seek additional care while non-modifiable factors can guide the prognosis of walking function in people post stroke.

Keywords: Gait speed; Meta-analysis; Post stroke; Systematic review.