Concurrent and Predictive Validity of Arm Kinematics With and Without a Trunk Restraint During a Reaching Task in Individuals With Stroke

ObjectiveTo examine the concurrent and predictive validity of measurements of kinematic variables during reaching tasks with and without a trunk constraint in individuals with stroke.DesignRandomized controlled trials.SettingsHospitals and a laboratory.ParticipantsIndividuals with stroke (N=95) enrolled in previous and ongoing clinical trials.InterventionsUpper limb training protocols were 90 to 120 minutes of intervention every weekday for 3 to 4 weeks.Main Outcome MeasuresFunctional capacity was assessed using the Action Research Arm Test and motor impairment using the Fugl-Meyer Assessment for the Upper Extremity. Movement kinematics were measured during a reaching task with and without a trunk constraint. We derived 5 endpoint control variables and 3 joint recruitment variables for estimating concurrent and predictive validity.ResultsThe adjusted R2 values for the constraint tasks ranged from .24 to .38 and for the unconstraint tasks from .29 to .40. Movement time was the most prominent kinematic variable for the Fugl-Meyer Assessment for the Upper Extremity before and after the intervention (P<.05). For the Action Research Arm Test, movement time and endpoint displacement were the most significant variables before and after the intervention, respectively (P<.05).ConclusionsMeasuring kinematic performance during an unconstrained task is appropriate and possibly sufficient to represent motor impairment and functional capacity of individuals with stroke. Movement time is the dominant variable associated with motor impairment and functional capacity, and endpoint displacement is unique in reflecting functional capacity of individuals with stroke.