Proceedings of The Physiological Society

University of Oxford (2011) Proc Physiol Soc 23, PC184

Poster Communications

Recovery of Precise Upper Limb Force Production Enables Better Hand Position Control Early After a Stroke.

D. L. Turner1,2, X. Tang1, W. Winterbotham2, M. Kmetova2

1. Neurorehabilitation Unit, University of East London, London, United Kingdom. 2. Lewin Stroke Unit, Addenbrookes Hospital, Cambridge, United Kingdom.


The recovery of upper limb force production is highly variable across patients early after stroke and many have long-term arm impairment and disability. Arm strength is often measured by maximal voluntary isometric contractions (MVC), which engage large muscle masses that do not mimic activities of daily living (ADLs) and are accompanied by little attention to precision of force production and position control. The main purpose of this study was to demonstrate that position control using precise low-level force production and its modulation against direction-dependent force perturbations recovers in a way that is more correlated to functional clinical outcome measures than MVCs. Consecutive patients were recruited from an inpatient stroke unit (n=10). Each patient was in early recovery (<7 weeks post-lesion) from their first ever stroke. Each patient was instructed to maintain a central “hold” position task with a hand-held joystick, whilst a robot device imposed a time-varying planar force perturbation in each of 8 directions (IMT2 robot device; Boston, MA, USA). Evaluations of precise force production (Newtons; N) and position error (meters; m) during the task, as well as elbow flexion MVC and clinical outcome measures of motor impairment (Fugl-Meyer score; FM) were performed before (PRE), once a week and then after (POST) 5 weeks of recovery. Changes in mean peak force produced by the patient to counteract the robot-induced force perturbation and mean peak planar X-Y position error of the joystick (averaged across the 8 force directions), MVC (elbow flexion measured with the robot device) and FM score were compared across the recovery period using repeated measures ANOVA and post-hoc paired Student’s T-tests adjusted for multiple comparisons. Significant differences in PRE vs. POST values (mean ± SEM) are reported at the p < 0.0125 level. Mean peak force production improved during recovery after stroke (17 ± 2 vs. 26 ± 2 N; PRE vs. POST; p = 0.002), whereas elbow flexion MVC did not (25 ± 3 vs. 23 ± 5 N; p = 0.71). As a result, mean peak position error decreased during recovery (0.08 ± 0.01 vs. 0.06 ± 0.01 m; p = 0.0001). Mean peak position error (y) was inversely and linearly correlated to FM score (x) during recovery (y = 60[±8] - 409[±104] x; r = -0.60; p = 0.0005). This study demonstrated that improvement in precise force production can operate as a mechanism enabling better motor behaviour such as hand position control and suggests that low-level force production and its modulation could be a target for future neurorehabilitation strategies during early recovery from a stroke. Secondarily, robot-based kinematic measures were significantly correlated with clinical outcome and may act as surrogate indices of recovery.

Where applicable, experiments conform with Society ethical requirements