A Novel Approach to Characterizing Patterns of Real-World Upper-Extremity Use During Walking and Activities of Daily Living in People with Subacute Stroke.

In accelerometer-based measurement, the inclusion of upper extremity (UE) activity during walking can overestimate real-world functional UE use in people post-stroke. This study introduces a comprehensive, accelerometer-based approach to examining real-world UE use, including the contribution of UE activity during walking, and compares UE use during walking versus ADL-based activities. People with subacute stroke wore bilateral wrist and ankle inertial measurement units (IMU) for one week (24 h/day). Agreement between UE-Total, UE-Continuous Walking and UE-ADL measures were assessed using Bland-Altman analyses. Variables were modeled using linear regression to explore step count, mobility aid use and UE motor impairment as predictors of IMU measures of real-world UE use. Agreement between UE-Total and UE-ADL measures showed minimal bias. Differences between UE-Continuous Walking and UE-ADL varied by mobility aid. UE-Continuous Walking was associated with increased UE activity and relatively greater paretic arm use compared with UE-ADL. Across UE-Continuous Walking and UE-ADL, step count was the strongest predictor of real-world UE use. UE activity during continuous walking exhibits different use patterns as compared to ADL-based activity. Accordingly, separating continuous walking-related UE activity from ADL-based UE use may improve clinical interpretation of IMU measures of real-world UE use in individuals with subacute stroke.