Predicting long-term movement behavior patterns after stroke: Development of a clinical prediction rule

Area of Science:

• Neurology
• Cardiology
• Rehabilitation Medicine
• Behavioral Science

Background:

• Post-stroke movement behavior often declines, increasing recurrent cardiovascular event risk.
• Early identification of at-risk patients is crucial for secondary prevention strategies.
• Tailored interventions can improve movement behavior and reduce cardiovascular risks after stroke.

Purpose of the Study:

• To develop and validate a clinical prediction rule for identifying stroke patients at risk of inactive movement behavior.
• To predict inactive movement patterns within two years post-stroke at hospital discharge.
• To support secondary prevention of cardiovascular events through targeted interventions.

Main Methods:

• Prospective cohort study of 200 first-ever stroke patients discharged home.
• Objective assessment of movement behavior (physical activity, sedentary time) at multiple time points post-discharge.
• Multinomial logistic regression used to identify predictors of inactive movement patterns (sedentary movers, sedentary prolongers).

Main Results:

• Female sex, older age, and increased fatigue predicted inactive movement behavior.
• Inactive behavior with prolonged sedentary bouts was associated with a 'prolonger' pattern, slower walking speed, and lower anxiety.
• The prediction model demonstrated good fit (QICC=737.02) and acceptable discrimination (AUC=0.74) with robust internal validation (shrinkage=0.96).

Conclusions:

• A validated clinical prediction rule can identify stroke patients at risk of inactive movement behavior.
• Stratification based on age, sex, and fatigue enables tailored behavior change interventions.
• Further external validation is necessary before widespread clinical implementation for secondary cardiovascular event prevention.