Effect of bilateral reaching on affected arm motor control in stroke with and without loading on unaffected arm

Disability and Rehabilitation, December 2006; 28(24): 1507 1516 RESEARCH PAPER Effect of bilateral reaching on affected arm motor control in stroke with and without loading on unaffected arm J. J. CHANG 1,2, W. L. TUNG 1,W.L.WU 3 &F.C.SU 1 1 Institute of Biomedical Engineering, National Cheng Kung University, Tainan, Taiwan, 2 Faculty of Occupational Therapy, College of Health Science, Kaohsiung Medical University, Kaohsiung, Taiwan, and 3 Faculty of Sport Medicine, College of Health Science, Kaohsiung Medical University, Kaohsiung, Taiwan Accepted February 2006 Abstract Purpose. To investigate the effect of bilateral reaching, with/without inertial loading on the unaffected arm, on hemiparetic arm motor control in stroke. Methods. Twenty unilateral stroke patients were recruited. A three-dimensional optical motion capture system was used to measure the movement trajectory of the hemiparetic arm while performing three tasks: affected limb reaching forward; two-limb reaching forward; and two-limb reaching forward with inertia loading of 25% upper limb weight on the unaffected limb, respectively. Kinematical parameters were utilized to quantify the reaching performance of the affected arm. Results. No matter whether loading was applied on the unaffected arm or not, the bilateral reaching task did not significantly facilitate smoother and faster movement. Furthermore, during bilateral reaching task with/without loading on the unaffected arm, stroke patients showed slower movement, lower maximal movement velocity, feedback control dominant and discontinuous movements in the affected arm than the same task with unilateral reaching. Subjects showed the greatest active upper extremity range of motion in proximal joints during the bilateral reaching task without unaffected arm loading. The amount of trunk movement also increased during bilateral reaching either with or without loading on the unaffected arm. Patients with moderate upper extremity motor impairment performed more discontinuous movements and less active elbow range of motion during bilateral reaching tasks; however, those with mild upper extremity motor impairment performed smoother movements and demonstrated greater active elbow range of motion during bilateral reaching tasks. Conclusions. Bilateral reaching tasks with/without loading on the unaffected arm could be considered as adding challenges during motor control training. Training with bilateral arm movements may be considered as a treatment strategy, and can be incorporated in stroke rehabilitation to facilitate greater arm active movement and improve motor control performance in the affected arm. Keywords: Bilateral movement, motor control, stroke Introduction Stroke describes a variety of disorders characterized by the sudden onset of neurological deficits caused by vascular injury to the brain, resulting in many neurological deficits. One of the major problems of stroke patients is motor deficit, especially in upper extremity movements. However, many patients still do not regain full movement of the upper extremities. One research study found that 69% of patients admitted to a rehabilitation unit following stroke had mild to severe upper extremity dysfunction, but only 14 16% of stroke survivors with initial upper extremity hemiparesis regained complete or near-complete motor function [1]. Hence, recovery of upper extremity function is an important aspect of retraining the patient with impaired motor control, and falls within the areas of rehabilitation. The model of brain plasticity or reorganization is useful to develop a conceptual approach to understand motor recovery after stroke. Previous studies, in the neurobiological theory of stroke recovery, have Correspondence: F. C. Su, PhD, Institute of Biomedical Engineering, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan. Tel: þ886 6 276 0665. Fax: þ886 6 234 3270. E-mail: fcsu@mail.ncku.edu.tw ISSN 0963-8288 print/issn 1464-5165 online ª 2006 Informa UK Ltd. DOI: 10.1080/09638280600646060

1508 J. J. Chang et al. shown that neural reorganization might have a longlasting effect on recovery [2]. It is believed that rehabilitation therapy might affect the brain reorganization, such as sprouting of new synapses [3,4], enhancing cortical activity [5,6], and inducing brain plasticity [7], but these changes may depend on the additional activity of the unimpaired limbs [5,8]. The rationale behind bilateral arm movement training is that concurrent movement of both arms may further facilitate activation of dormant control centers [9,10]. The intact corpus callosum plays a role in interlimb spatial coupling, providing a channel for the contralateral influences. Furthermore, not all the projections from a hemisphere terminate contralaterally. Approximately 5 7% of the supplemental motor area project ipsilaterally bypassing the primary motor cortex. Approximately 25% of the primary motor cortex projections do not decussate but merge with the ipsilateral corticopinal trace [11]. It is believed that these contralateral and ipsilateral projections from each respective hemisphere may contribute to interlimb coupling. The use of bilateral movement training is one of the recent approaches for enhancing upper arm movement recovery in stroke rehabilitation. Researches found that bilateral movement and stimulation produced higher EMG activation in the wrist and finger extensors than unilateral movement and stimulation for stroke subjects [12,13], and bilateral arm movement training could perhaps assist stroke patients to develop a strategy via ipsilateral input from the undamaged hemisphere, to improve control of muscle activation on the weaker, affected side [9]. Studies of Walter and Swinnen [14,15] demonstrated torque contributes to the degree of interlimb attraction emerging during bimanual actions in healthy subjects. Hatzitaki and Mckinley found that addition of an inertial load, for normal subjects, resulted in an increased movement time and concomitant decrease in maximal velocity of both the upper arm and forearm of only the loaded limb during bilateral arm reaching [16]. Additionally, Cunningham et al. suggested that inertial loading on the uninvolved arm in the bilateral movements, with upper arm supported, probably could benefit the hemiplegic arm in some stroke patients [17]. However, in a realistic environment, most functional reaching and aiming movements are performed without arm support and the effect of interlimb coupling with loading to the unaffected upper limb following stroke is still not clear enough. In this study, we hypothesized that unsupported bilateral arm reaching with/without loading manipulation on the unaffected arm may facilitate movement control on the affected upper limb in strokes with different levels of motor impairment. Methods Subjects We chose a convenient sample in this study and 20 stroke patients were enrolled. Subjects were all from the Department of Rehabilitation Medicine in Kaohsiung Medical Hospital, Taiwan. There were 17 males and three females aged from 37 75 years and the mean (+ Standard deviation) of age, body weight and height were 56 + 10.54 years, 70.00 +10.68 kg and 168.95 + 6.48 cm respectively. Three of the 20 patients were left-hand dominant. The duration of post incidence of stroke was from 12 days to 6 years. Ratings of functional ability of the affected upper limb of each subject were assessed prior to the study using the upper extremity subtest of the Fugl-Meyer Motor Function Assessment (FMA) [18] to examine for the presence of synergistic and isolated movement pattern and grasp. The scoring in upper extremity subtest of FMA is from 0 66, with 66 indicating nearly normal function. Muscle tone at the elbow joint was evaluated using a six-point scale (0 ¼ normal tone, 5 ¼ severe spasticity) based on the Modified Ashworth Scale [19]. Criteria for inclusion in this study included: (i) computed tomography or magnetic resonance imaging evidence of singlehemisphere involvement; (ii) demonstrating arm reaching ability with FMA score was greater than 30; (iii) no perceptual-cognitive dysfunction, such as loss of arm proprioception, apraxia, hemispatial neglect, which limits comprehension of the experimental task; and (iv) no severe concurrent medical problems, such as shoulder pain, or other neurological and orthopedic conditions affecting the arm or trunk movements. Demographic and clinical data of the individual subjects are summarized in Table I. All subjects gave informed consent and the institutional review board of Kaohsiung Medical University approved all procedures. Experimental protocol The subject was seated in front of a rectangular experimental table with a seat-belt on his or her waist to protect their sitting safety and the seat height was adjustable so that the subject s feet were flat on the floor and the knees and hips were flexed at 908. At the beginning of the experimental task, the subject had to put his or her upper limbs on standard initial positions on the table, flex both elbows at 908, place wrists in a neutral position, and then the tasks required the rapid forward projection of unilateral or bilateral arms so as to reach the cups at the shoulder height level in front of the subject. The position of the cups was adjustable to accommodate the arm length of each subject.

Table I. Demographic and clinical data of the individual subject. Subject id Sex Age (years) Dominant hand Days after onset (days) Side of stroke Type of stroke FMA MAS 1 F 37 R 2281 R hemorrhagic 32 2 2 M 57 L 425 L infarct 35 1 þ 3 M 49 R 327 R infarct 38 2 4 M 41 R 29 L infarct 40 1 þ 5 M 66 R 111 L infarct 42 0 6 M 66 R 765 R hemorrhagic 47 1 þ 7 M 51 R 250 L infarct 48 1 þ 8 F 56 L 1341 R infarct 51 1 þ 9 M 63 R 421 R infarct 52 1 10 M 59 R 964 L infarct 54 1 11 F 52 R 377 R infarct 54 1 12 M 75 R 36 L infarct 59 1 13 M 70 R 98 L infarct 60 1 14 M 65 R 12 L infarct 61 0 15 M 48 R 122 R infarct 61 0 16 M 46 R 12 R hemorrhagic 62 0 17 M 43 R 283 L infarct 62 1 18 M 57 R 25 L infarct 62 0 19 M 50 L 192 R infarct 63 0 20 M 69 R 23 L infarct 63 0 M ¼ male, F ¼ female; R ¼ right, L ¼ left. Bilateral arm reaching and motor control 1509 A three-dimensional optical motion capture system (Visualeyez TM Hardware, PhoeniX Technologies Inc., Canada) was used to collect the movement trajectories of the arm and trunk in this study. Infrared light-emitting diodes were positioned on the anatomic landmarks of affected arm and trunk. The selected 14 anatomic landmarks were as follows: The metacarpophalangeal joint of the index finger, the metacarpophalangeal joint of the fifth finger, the middle of the 3rd metacarpal, radial and ulnar styloid processes, the middle portion of forearm, lateral epicondyle of the elbow, a triangular frame with three-markers on upper arm, ipsilateral and contralateral acromion processes, top of sternum and xyphoid process. The positions of markers on the arm and trunk were recorded at a sampling rate of 70 Hz and digitally filtered by using a low pass 2nd order forward and backward Butterworth filter with cut-off frequency at 5 Hz. Each subject was requested to perform three movement tasks, at a random sequence (i) reaching forward with the affected limb only (Uni); (ii) reaching forward with both limbs simultaneously (Bil); and (iii) reaching forward with both limbs simultaneously while adding a load of 25% upper limb inertia to the unaffected limb (Bil þ 25%). We assigned random testing order to the subjects. This was used to control for potential practice effects by counterbalancing order. Each experimental condition had to be performed for five trials and a 5-min rest period was administered between each experimental movement task. Reaching movement was chosen in these tasks because it involved the coordination of multiple joints and represented a basic and functional movement in daily activities. In addition, it represented modeling a synergistic arm movement pattern that would be expected during recovery after stroke. We used the weightadjustable wrist weights added to the unaffected wrist joint of subjects. The moment of inertia of the arm was referenced from the database of anthropological measurement of the Institute of Occupational Safety and Health (IOSH) in Taiwan and estimated from the bodyweight and unaffected arm length (shoulder acromion process to the tip of middle finger) of each subject. It was suggested that moving to real objects might produce better performance in stroke patients than rote and meaningless tasks [20], therefore we used cups to mark the end positions, where subjects had to reach. In order to control the learning effect, the counterbalancing order was also administered in our experimental procedures. At the beginning of the experiment, subjects received a brief description of the study. During the tasks, the examiner provided positive and supportive verbal feedback (e.g., speed up a bit, you re really doing a good job...) to each subject and encouraged the subject to perform as quickly as possible. The typical length of an experimental session was approximately 40 min, and no evidence of fatigue was observed or reported from any subject after finishing the three experimental movement tasks. Data analysis Kinematic data from the arm and trunk movements were analysed by the VZAnalyzer software,

1510 J. J. Chang et al. V3.0 (PhoeniX Technologies Inc., Canada). VzAnalyzer software gave a three-dimensional reconstruction of the marker positions. The upper limb was modeled as three rigid-bodies at the upper arm, forearm and hand. A relative velocity above or below 3% of the maximum movement velocity on the sagittal plane (Z-axis), which was parallel to the reach movement direction, was used to detect the start and end of each reaching movement. The following kinematic-dependent variables were derived from the marker position to examine and quantify the affected arm movement: maximal velocity (cm/s) (MV), percentage of reach where maximal velocity occurs (%) (PRMVO), movement time (s) (MT), number of movement units (NMU), and normalized jerk score of movement (NJSM). Maximal velocity, the highest instantaneous velocity during the reaching movement, is regarded as being correlated with the force of a movement (Nelson 1983). Movement time, the duration of execution of a movement, reflects the overall speed of a movement, as a faster movement would result in shorter movement time. Both NMU and NJSM were used to quantify the movement smoothness [17,21,22]. NMU was determined by the number of peaks presenting in the velocity profile of the paretic arm. This provides information about the smoothness and efficiency of a movement [21,23]. Fewer movement units indicate a smoother and more efficient reaching movement. To obtain the NJSM, a mathematical formula was used compute the integrated squared jerk with the unit of distance/time 3 [23]. Since integrated squared jerk increases dramatically with movement duration and the distance traveled during the movement, it was useful to normalize this quantity in time and distance [24,25]. This was done by introducing the term t 5 /s 2 into the formula for normalized jerk score. The formula was taken: vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Z 1 d NJSM ¼ 3 x 2 2 dt 3 þ d3 y 2! u dt 3 þ d3 z 2 t t dt 3 dt 5 s 2 Where x: the position of the hand rigid body on X-axis; y: the position of the hand rigid body on Y-axis; z: the position of the hand rigid body on Z-axis; t: movement time; s: movement distance of hand. Other parameters regarding range of motion in elbow and shoulder joints were computed by dot product of vectors defined from markers on anatomical body landmarks, which are referenced to measure joint range of motion. These variables were elbow flexion-extension range (deg) (EFER), and shoulder flexion-extension range (deg) (SFER). These two variables were used to present the main active range of motion of upper extremity joints during reaching. Trunk involvement was measured as the linear line value of the marker located on the top of the sternum in the sagittal plane. Trunk linear line value (cm) (TLLV) represented the straight-line distance between the initial point of the movement and its end point. Statistical analysis Statistical analyses were performed with SPSS software (V11.0). Taking the difference of upper extremity function level into consideration, we classified the subjects into two groups according to their scores of FMA. Thirteen subjects in group 1 scored between 50 and 66, indicating a mild motor deficit, and seven subjects in group 2 scored between 30 and 49, indicating a moderate motor deficit. Dependent variables were tested by using 3(tasks) 6 2(groups) analysis of variance (ANOVA) with repeated measures on both factors. The first factor included the three experimental tasks (unilateral, bilateral, bilateral with 25% inertia loading), and the second factor referred to the mild group and moderate group. Mauchly s test was conducted to test the equality of variances of the differences between levels of the repeated measures factor, and the Greenhouse-Geisser correction was applied for violations of sphericity. If significance existed, post hoc comparison was conducted on the means by using least significant difference pairwise multiple comparison tests. The a-level was set at 0.05 for all tests. Results Statistical results of all dependent variables are presented in Table II. The impact of different task conditions on the changes of paretic arm motor control and trunk movement are shown by the following results. Movement time and velocity profiles Significantly shorter movement time was found in unilateral movement tasks [F (2, 36) ¼ 4.54, p 5 0.05]. Post hoc testing revealed that the mean MT of the unilateral task in mild and moderate groups was the smallest among the three tasks (p 5 0.05). Subjects in both groups took a longer time to perform bilateral reaching movement tasks. A significant within-subjects effect was shown in MV means [F (2, 36) ¼ 8.92, p 5 0.001]. Post hoc analysis indicated that the mean MV of both groups in the unilateral movement task was the greatest among the three movement tasks ( p 5 0.05).

Bilateral arm reaching and motor control 1511 Table II. Outcome measure of the dependent variables (mean + SD) calculated from the three movement tasks. Mild group (n ¼ 13) Moderate group (n ¼ 7) Variables Task Uni Bil Bil þ 25% Uni Bil Bil þ 25% MT* (s) 0.66 + 0.33 0.77 + 0.43 0.74 + 0.39 0.924 + 0.27 0.97 + 0.26 1.02 + 0.27 MV* (cm/s) 127.95 + 48.45 113.27 + 48.10 114.40 + 48.26 88.78 + 25.78 79.51 + 18.70 80.22 + 21.60 PRMVO* (%) 41.20 + 5.81 38.40 + 7.87 39.30 + 7.60 45.84 + 15.41 39.14 + 8.81 34.59 + 8.91 NMU? 1.05 + 0.18 1.05 + 0.18 1.03 + 0.09 1.62 + 1.11 2.10 + 0.90 2.00 + 0.94 NJSM* 29.43 + 34.51 43.50 + 69.86 39.97 + 67.85 73.81 + 43.85 97.64 + 63.26 95.94 + 55.67 EFER*,? (deg) 59.67 + 16.25 65.97 + 16.32 62.29 + 13.04 44.81 + 10.21 47.54 + 14.76 45.13 + 12.10 SFER* (deg) 52.09 + 14.83 58.81 + 16.70 55.16 + 15.15 46.08 + 7.82 52.89 + 10.42 49.98 + 14.47 TLLV* (cm) 2.33 + 1.55 3.77 + 2.64 3.58 + 2.45 4.20 + 1.70 4.16 + 1.53 5.29 + 3.14 *Significant with-subjects effect existed across reaching tasks;? Significant between-subjects effect existed between mild and moderate groups; MT, movement time; PRMVO, percentage of reach where maximal velocity occurs; MV, maximal velocity; NMU, number of movement units; NJSM, normalized jerk score of movement; EFER, elbow flexion-extension range; SFER, shoulder flexion-extension range; TLLV, Trunk linear line value. The mean MV significantly decreased in both types of bilateral movement tasks. The PRMVO means showed a significant within-subjects effect among three tasks [F (2, 38] ¼ 4.18, p 5 0.05). The mean PRMVO of both groups in the unilateral movement task was the greatest among the three movement tasks (p 5 0.05). Smoothness of the movement As for the movement smoothness, Mauchly s Test of Sphericity for the NMU means reached a significant level (X 2 ¼ 16.57, p 5 0.05) and Greenhouse- Geisser correction was applied. NMU for the hemiplegic hand showed no considerable differences among the three movement tasks [F (1.233, 22.185) ¼ 0.847, p 4 0.05]. Different experiment conditions did not markedly influence the number of movement units. However, there was a significant between-subjects effect in the NMU means [F (1, 18) ¼ 14.85, p 5 0.001]. The NMU mean in moderate group was larger than that in mild group. NJSM data for the involved limb revealed a significant within-subjects effect [F (2, 36) ¼ 3.36, p 5 0.05] among the three movement tasks. Follow-up post hoc testing revealed the NJSM of bilateral tasks performed without loading was significantly greater than that of unilateral tasks ( p 5 0.05). The consequence implied that stroke patients moved less smoothly when reaching targets with both arms. As to the between-subjects effect in the NJSM mean, the significance reached a borderline level [F (1, 18) ¼ 3.98, p ¼ 0.068]. Trunk and upper arm active movement during reaching With respect to the upper extremity active range of motion and trunk movement, a borderline significant within-subjects effect was shown in EFER means [F (2, 36) ¼ 3.20, p ¼ 0.05]. Subjects performed the greatest elbow extension range of motion in the bilateral movement task without loading. However, they decreased the elbow ROM during bilateral movement tasks with 25% inertia loads on the uninvolved hand. Furthermore, there was a between-subjects effect in the EFER means [F (1, 18) ¼ 7.12, p 5 0.01]. The results implied that stroke patients with mild upper extremity impairment had greater active motion range of the elbow than those with moderate impairment level. Within-subjects effect in the SFER means appeared to be significant [F (2, 36) ¼ 5.78, p 5 0.01]. In post hoc comparison tests, subjects in the bilateral movement task without loading performed greater SFER than in the unilateral movement task ( p 5 0.05), which meant that stroke patients performed greater active shoulder flexion-extension range of motion in bilateral movements without inertia loading than in that of unilateral ones. Test of with-subjects effects in the TLLV means was significant [F (2, 36) ¼ 3.53, p 5 0.05]. Post hoc comparison tests indicated that the mean of TLLV in the unilateral movement task was less than that in the bilateral movement task with loading (p 5 0.05). As results showed, subjects recruited more trunk movement during bilateral reaching tasks. Reaching performances between different levels of arm motor impairment Velocity trajectory profiles of the affected hand movement were further analysed to study the effect of bilateral reaching between different upper extremity impairment levels on the quality of paretic arm motor control. The typical velocity trajectory of one subject (ID ¼ 2) with moderate upper extremity impairments (FMA score ¼ 35) showed marked skewed velocity profiles (Figure 1). The skewed

1512 J. J. Chang et al. velocity trajectory of the affected hand showed discontinuous movement with more feedback control during unilateral and bilateral reaching tasks. Typical velocity trajectories of one subject (ID ¼ 16) with mild upper extremity impairment level (FMA score ¼ 62) showed nearly bell-shaped velocity profiles (Figure 2). The trajectories of the velocity profiles showed that for only one movement unit Figure 1. Typical velocity trajectory of a representative subject who was moderate motor deficit among three movement tasks. The arrows show the MV and PRMVO in each task. The profile suggests that the subject performed less smooth reaching with more visual guidance in bilateral tasks than in unilateral task. Figure 2. Typical velocity trajectory of a representative subject who was mild motor deficit among three movement tasks. The arrows show the MV and PRMVO in each task. The subject performed nearly bell-shaped velocity profiles in all three tasks.

Bilateral arm reaching and motor control 1513 used across the three reaching tasks, feed forward control and smooth movement were noted. In addition, both of the two subjects showed a decrease in MV during bilateral reaching tasks, with/without loading on affected arm. Discussion The objectives of this study were to examine the motor control of paretic arm movements following a unilateral and bilateral unsupported reaching movement, to investigate whether an inertial load applied to the uninvolved arm facilitated bilateral reaching performance, and to explore the reaching performance, under different reaching tasks, between strokes with different levels of motor impairment. As expected for the effect of the unilateral reaching compared to bilateral reaching, our results suggested that bilateral movement tasks with/without loading did not benefit, immediately, stroke subjects in quantitative reaching performance. Does bilateral movement facilitate affected arm movement control immediately? Previous findings suggested that bimanual reaching in the healthy subjects had longer movement duration, lower maximal velocity, and longer deceleration phases than unimanual reaching [26,27]. They described these differences as an effect of two target cost. This effect has generally been attributed to a limitation in attentional resources. We also found the same results in stroke patients: that longer movement time and lower maximal velocity occurred during bilateral reaching directed towards two separate objects. Taking a closer look at the compared results of bilateral movement task without loading and bilateral movement task with loading, this study did not support the hypothesis that a loading manipulation with bilateral reaching benefits better movement in stroke victims. Findings in this study showed bilateral movement with loading on the unaffected arm might not seemingly facilitate smoother and faster affected arm movements concurrently, and feedback control was more prominent during bilateral reaching. We propose that bilateral reaching, with/without loading on the uninvolved arm, presents an additional challenge in attention and executing movement. When increasing the difficulty applied to the task, subjects will change in movement kinematics and strategy to reduce the error of movement performance. However, our results did not support previous findings in the effects of bilateral reaching on motor control. The reasons may be as follows. Firstly, motor system deficits might deeply reduce arm motor behaviours. Thus, a functional coupling of motor control across limbs during bilateral movements still might not overcome the residual deficits of the affected pyramidal motor system [28]. Instead, the resultant effect of motor extinction would deteriorate the motor control of the paretic limb during bilateral reaching with/without loading on unaffected limb [29,30]. Secondly, the experimental task condition was different from other related studies. In the present study, subjects were asked to reach to targets without the assistance of arm support. However, participants performed bilateral movements on arm supports in other studies [14,15,17]. Thirdly, heterogeneity in stroke subjects and sample size was different than that of other studies. Cunningham et al. [17] reported that bilateral movement with inertial loading on the uninvolved arm may improve arm motor control, but there were only six subjects recruited; in addition, not all of the subjects had significantly impacted performance in the experimental task conditions. Therefore, the notion that loading the uninvolved limb serves as a useful adjunctive method to bilateral training in stroke still needs further experimental evidence. Our findings showed that bilateral movement tasks with and without loading did not immediately facilitate more smooth reaching movements nor improve better movement quality in kinematics variables. How is it therefore, that other studies still suggest the use of bilateral activation therapies in stroke patients and find it effective? Whitall et al. [10] reported significant increases in strength, range of motion and standardized motor assessment scales following six weeks of a simultaneous bilateral arm training protocol. Mudie and Matyas [31] also demonstrated an enhancement in task performance after a three-week period of bilateral isokinematic therapy. In these previous studies, bilateral movement task was provided to stroke patients as a training program for a long period of training duration (6*8 weeks). These authors investigated the improvement of subjects affected arm function by analysing the results of pretest and post-test. Lewis and Byblow [32] reported that short-term bilateral training following unilateral training may have limited effectiveness in enhancing upper limb motor performance in acute and chronic individuals post stroke. Therefore, the effect of bilateral movement training to arm recovery may need a long period of practice. Besides, the repetitions of training movement and intensity of therapy were also important to the motor rehabilitation of the paretic hand [33,34]. Different to previous studies, we investigated the immediate effect of bilateral reaching movement tasks and found subjects seemingly had discontinuous movement and movement quality that became worse in reaching kinematics during performing bilateral movement tasks with/without loading to

1514 J. J. Chang et al. the unaffected upper limb. However, this does not mean that bilateral reaching task has no benefits for stroke patients. Staines et al. reported that bilateral movement may enhance activation in the primary motor cortex of the affected hemisphere compared with unilateral paretic hand movement [5]. Luft et al. reported that changes in activation were observed in the contralesional cerebrum and ipsilesional cerebellum after repetitive bilateral training in chronic stroke [35]. If training provides enough intensive and repetitive practice of bilateral movement, stroke patients will cope with the attention-demand and learn motor control under appropriate challenges, and progressively enhance reorganization in contralesional cerebrum and ipsilesional cerebellum motor networks. Thus, stroke patients receiving bilateral movement training, with enough intensity and period, would potentially demonstrate effective improvement in motor control [31]. Bilateral arm reaching may enhance increasing affected arm active range of motion In this study, we also found that subjects performed the greatest upper extremities active ROM in bilateral reaching task without loading. This result can be indirectly supported by some previous observations. Di Stefano et al. [36] have reported that motor systems subserving control of the trunk and of the proximal limb musculature are organized on a bilateral basis. Steenbergen et al. [37] suggested that the activation of the proximal musculature is closely synchronized. In this study, reaching movements were major by proximal limbs. As a result, interlimb coupling in proximal limbs may be a reason why subjects have the greatest active ROM in bilateral movement task. On the other hand, visual guiding in bilateral movement tasks may be the other reason for this result. Franz and Packman [38] have examined the effects of a mirror reflection of hand circle-drawing movements on the motor output of a bimanual task. They suggested that the feedback of movement of one hand was used to actively guide movement of the other hand and a more rapid and direct form of coupling occurred between the visual information and the processes governing motor output of the hidden hand. However, bilateral reaching movements are just like the effect of mirror reflection. The affected arm may perform greater active ROM by visual guiding of the simultaneous unaffected arm movement. Bilateral arm reaching may induce more compensatory trunk movement Although subjects performed more active ROM in bilateral movement tasks, they also used more trunk movements to adapt to different task demands. Van Roon et al. [39] have reported that increasing the accuracy constraint of the task caused an increased trunk displacement for both healthy participants and tetraparetic participants. This larger trunk involvement may reflect a flexible adaptation to the disorder and the task constraints. Bilateral movement tasks without arm supports are more difficult than unilateral movement tasks for stroke patients. Besides, bilateral movement tasks with loading may increase the task demand for subjects. Therefore, subjects may perform more trunk movement to adapt to the more difficult task demands present in bilateral movement tasks. Michaelsen and Levin found that practices of unilateral reaching with trunk restraint would contribute the recovery of reaching movements in stroke patients [40,41]. Thus, we suggest further studies are needed to testify whether practices of bilateral reaching with trunk restraint would have additional benefits for the recovery of reaching movements in stroke patients. Effects of bilateral arm reaching may be different by levels of motor impairment Different levels of upper extremity motor impairment demonstrated specific differences in reaching kinematics across different reaching tasks. Subjects who have mild upper arm motor deficits had smoother movement trajectories, more normal velocity profile, and greater elbow active ROM than those who have moderate upper arm motor deficits in bilateral reaching. In addition, we found the significant within-subjects effect of PRMVO was mainly from the treatment of moderate impaired group. Subjects, with moderate levels of impairment, may use more visual feedback to correct movement errors in the deceleration phase with resulting less smooth velocity trajectories as task demand increases from unilateral to bilateral with/without loading on the un-affected arm. Movement smoothness is a result of learned interjoint coordination. Previous studies report that movement smoothness will increase with motor recovery for subjects with movement disorders, and movement smoothness is related to the levels of motor impairment [22,42]. A higher level of brain damage might demonstrate more imbalanced hemisphere activation and show motor extinction, which would cause moderately impaired subjects to perform more discontinuous movements and exhibit a smaller active elbow range of motion during performing reaching tasks [29,43]. These above might lead the subject to recruit more trunk movement in order to adapt to the demands of the tasks [39]. As the task demand increases for bilateral reaching movement without arm supports and loading to

Bilateral arm reaching and motor control 1515 the unaffected arm, the bell-shaped velocity profile became skewed and maximal velocity occurred earlier in movement. The present study supported the hypothesis that the use of movement strategies would be different between different levels of upper limb motor impairment during performing reaching tasks with increasing demands [44,45]. Limitation of this study Small sample size and heterogeneity of our subjects with factors such as age, chronicity, lesion type and extend of lesion, will limit the generalization of our findings. However, the use of quantitative technique in motion analysis could play a complementary role to clinical assessments, provide kinematical descriptions of reaching in the paretic arm, and assist in understanding the mechanisms underlying altered reaching ability following experimental interventions. In future studies, use of electromyography combined with kinematical analysis to investigate the upper extremity muscle activation and movement kinematics during bilateral movement tasks and following bilateral movement training should be carried out, thereby contributing to the establishment of evidence-based practise of the bilateral movement therapies. Acknowledgements This study was supported by the National Science Council, Taiwan, R.O.C., under grants NSC92-2320-B-037-033 and NSC 92-2851-C-037-004-B. References 1. Gowland C, debruin H, Basmajian JV, Plews N, Burcea I. Agonist and antagonist activity during voluntary upper-limb movement in patients with stroke. Phys Ther 1992;72: 624 633. 2. Small SL, Hlustik P, Noll DC, Genovese C, Solodkin A. Cerebellar hemispheric activation ipsilateral to the paretic hand correlates with functional recovery after stroke. Brain 2002;125:1544 1557. 3. Jones TA, Schallert T. Overgrowth and pruning of dendrites in adult rats recovering from neocortical damage. Brain Res 1992;581:156 160. 4. Trombly CA. Deficits of reaching in subjects with left hemiparesis: a pilot study. Am J Occup Ther 1992;46:887 897. 5. Staines WR, McIlroy WE, Graham SJ, Black SE. Bilateral movement enhances ipsilesional cortical activity in acute stroke: a pilot functional MRI study. Neurology 2001;56: 401 404. 6. Johansen-Berg H, Dawes H, Guy C, Smith SM, Wade DT, Matthews PM. Correlation between motor improvements and altered fmri activity after rehabilitative therapy. Brain 2002;125:2731 2742. 7. Nelles G, Jentzen W, Jueptner M, Muller S, Diener HC. Arm training induced brain plasticity in stroke studied with serial positron emission tomography. Neuroimage 2001;13: 1146 1154. 8. Jones TA, Schallert T. Use-dependent growth of pyramidal neurons after neocortical damage. J Neurosci 1994;14: 2140 2152. 9. Mudie MH, Matyas TA. Can simultaneous bilateral movement involve the undamaged hemisphere in reconstruction of neural networks damaged by stroke? Disab Rehab 2000;22: 23 37. 10. Whitall J, McCombe Waller S, Silver KH, Macko RF. Repetitive bilateral arm training with rhythmic auditory cueing improves motor function in chronic hemiparetic stroke. Stroke 2000;31:2390 2395. 11. Brinkman C, Porter R. Supplementary motor area in the monkey: Activity of neurons during performance of a learned motor task. J Neurophysiol 1979;42:681 709. 12. Cauraugh JH, Kim S. Two coupled motor recovery protocols are better than one: Electromyogram-triggered neuromuscular stimulation and bilateral movements. Stroke 2002;33:1589 1594. 13. Cauraugh JH, Kim SB. Chronic stroke motor recovery: duration of active neuromuscular stimulation. J Neurol Sci 2003;215:13 19. 14. Walter CB, Swinnen SP. Kinetic attraction during bimanual coordination. J Mot Behav 1990;22:451 473. 15. Walter CB, Swinnen SP. Asymmetric interlimb interference during the performance of a dynamic bimanual task. Brain Cogn 1990;14:185 200. 16. Hatzitaki V, McKinley P. Effect of single-limb inertial loading on bilateral reaching: interlimb interactions. Exp Brain Res 2001;140:34 45. 17. Cunningham CL, Stoykov ME, Walter CB. Bilateral facilitation of motor control in chronic hemiplegia. Acta Psychol (Amst) 2002;110:321 337. 18. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. A method for evaluation of physical performance. Scand J Rehab Med 1975;7:13 31. 19. Bohannon RW, Smith MB. Interrater reliability of a modified Ashworth scale of muscle spasticity. Phys Ther 1987;67: 206 207. 20. Wu C, Trombly CA, Lin K, Tickle-Degnen L. Effects of object affordances on reaching performance in persons with and without cerebrovascular accident. Am J Occup Ther 1998;52:447 456. 21. Richardson MJ, Flash T. Comparing smooth arm movements with the two-thirds power law and the related segmentedcontrol hypothesis. J Neurosci 2002;22:8201 8211. 22. Chang JJ, Wu TI, Wu WL, Su FC. Kinematical measure for spastic reaching in children with cerebral palsy. Clin Biomech (Bristol, Avon) 2005;20:381 388. 23. Flash T, Inzelberg R, Schechtman E, Korczyn AD. Kinematic analysis of upper limb trajectories in Parkinson s disease. Exp Neurol 1992;118:215 226. 24. Schneider K, Zernicke RF. Jerk-cost modulations during the practice of rapid arm movements. Biol Cybern 1989;60: 221 230. 25. Teulings HL, Contreras-Vidal JL, Stelmach GE, Adler CH. Parkinsonism reduces coordination of fingers, wrist, and arm in fine motor control. Exp Neurol 1997;146:159 170. 26. Jackson GM, Jackson SR, Kritikos A. Attention for action: Coordinating bimanual reach-to-grasp movements. Br J Psychol 1999;90(Pt 2):247 270. 27. Jackson GM, German K, Peacock K. Functional coupling between the limbs during bimanual reach-to-grasp movements. Hum Mov Sci 2002;21:317 333. 28. Platz T, Bock S, Prass K. Reduced skilfulness of arm motor behaviour among motor stroke patients with good clinical recovery: Does it indicate reduced automaticity? Can it be improved by unilateral or bilateral training? A kinematic motion analysis study. Neuropsychologia 2001;39:687 698.

1516 J. J. Chang et al. 29. Mattingley JB, Phillips JG, Bradshaw JL. Impairments of movement execution in unilateral neglect: A kinematic analysis of directional bradykinesia. Neuropsychologia 1994; 32:1111 1134. 30. Robertson IH, North NT. One hand is better than two: Motor extinction of left hand advantage in unilateral neglect. Neuropsychologia 1994;32:1 11. 31. Mudie MH, Matyas TA. Responses of the densely hemiplegic upper extremity to bilateral training. Neurorehabil Neural Repair 2001;15:129 140. 32. Lewis GN, Byblow WD. Neurophysiological and behavioural adaptations to a bilateral training intervention in individuals following stroke. Clin Rehab 2004;18:48 59. 33. Butefisch C, Hummelsheim H, Denzler P, Mauritz KH. Repetitive training of isolated movements improves the outcome of motor rehabilitation of the centrally paretic hand. J Neurol Sci 1995;130:59 68. 34. Parry RH, Lincoln NB, Vass CD. Effect of severity of arm impairment on response to additional physiotherapy early after stroke. Clin Rehab 1999;13:187 198. 35. Luft AR, McCombe-Waller S, Whitall J, Forrester LW, Macko R, Sorkin JD, Schulz JB, Goldberg AP, Hanley DF. Repetitive bilateral arm training and motor cortex activation in chronic stroke: A randomized controlled trial. Jama 2004;292: 1853 1861. 36. Di Stefano M, Morelli M, Marzi CA, Berlucchi G. Hemispheric control of unilateral and bilateral movements of proximal and distal parts of the arm as inferred from simple reaction time to lateralized light stimuli in man. Exp Brain Res 1980;38:197 204. 37. Steenbergen B, Hulstijn W, de Vries A, Berger M. Bimanual movement coordination in spastic hemiparesis. Exp Brain Res 1996;110:91 98. 38. Franz EA, Packman T. Fooling the brain into thinking it sees both hands moving enhances bimanual spatial coupling. Exp Brain Res 2004;157:174 180. 39. van Roon D, Steenbergen B, Meulenbroek RG. Trunk recruitment during spoon use in tetraparetic cerebral palsy. Exp Brain Res 2004;155:186 195. 40. Michaelsen SM, Luta A, Roby-Brami A, Levin MF. Effect of trunk restraint on the recovery of reaching movements in hemiparetic patients. Stroke 2001;32:1875 1883. 41. Michaelsen SM, Levin MF. Short-term effects of practice with trunk restraint on reaching movements in patients with chronic stroke: a controlled trial. Stroke 2004;35: 1914 1919. 42. Rohrer B, Fasoli S, Krebs HI, Hughes R, Volpe B, Frontera WR, Stein J, Hogan N. Movement smoothness changes during stroke recovery. J Neurosci 2002;22: 8297 8304. 43. Stinear JW, Byblow WD. Rhythmic bilateral movement training modulates corticomotor excitability and enhances upper limb motricity poststroke: a pilot study. J Clin Neurophysiol 2004;21:124 131. 44. Levin MF. Interjoint coordination during pointing movements is disrupted in spastic hemiparesis. Brain 1996;119: 281 293. 45. Cirstea MC, Ptito A, Levin MF. Arm reaching improvements with short-term practice depend on the severity of the motor deficit in stroke. Exp Brain Res 2003;152:476 488.