Abstract
BACKGROUND: Stroke is a leading cause of long-term disability, but advances for rehabilitation have lagged those for acute treatment. Large biological studies (eg, omics) may offer mechanistic insights for recovery but require collecting detailed recovery phenotypes at scale, for example, in thousands of people with minimal burden for participants and researchers. This study investigates the concurrent validity between remotely collected wearable sensor data and in-clinic assessments of motor recovery poststroke. METHODS: Utilizing a large, harmonized multisite dataset of adults at various stages of recovery poststroke, we analyzed cross-sectional (N=198; from 0 to >52 weeks) and longitudinal (N=98; from 0 to 26 weeks) changes in the use ratio, the Action Research Arm Test, and the Fugl-Meyer Assessment upper extremity subscale. The use ratio is the ratio of the time the paretic arm is active divided by the time the nonparetic arm is active. RESULTS: Our findings indicate strong concurrent validity of the use ratio, the Action Research Arm Test, and the Fugl-Meyer Assessment upper extremity subscale both cross-sectionally (differences between people) and longitudinally (changes within a person), for example, r=0.87 (95% CI, 0.80-0.91) at 0 to 6 weeks, declining to r=0.58 (95% CI, 0.39-0.72) at ≥52 weeks for correlations between use ratio and Action Research Arm Test. CONCLUSIONS: Although the use ratio strongly correlated with the Fugl-Meyer Assessment upper extremity subscale and Action Research Arm Test early after stroke, these correlations reduced with longer elapsed time poststroke. This decreasing correlation might be explained by the increasing influence that personal and environmental factors play as recovery progresses.