Electromyographic analyses of global synkinesis in the paretic upper limb after stroke.

Background and Purpose. Global synkinesis (GS), or motor irradiation, is an involuntary movement associated with the coactivation of numerous muscles in one limb when the opposite limb is active. The electromyographic (EMG) patterns of people with stroke and people who were healthy were analyzed to characterize GS development in relation to joint involvement and to attempt to relate these findings to clinical observations. Subjects and Methods. Twenty patients with stroke, divided into 2 groups with either greater levels of irradiation (SG, n = 10) or lesser levels of irradiation (SL, n = 10), and 20 subjects in a control group were studied. A dynamometer was used to provide resistance for voluntary isometric muscle contractions of the flexor muscle groups of the shoulder, elbow, and wrist. The summated and standardized net EMG amplitudes of 8 principal muscles of the unexercised (paretic) upper extremity were used to characterize intensity and spatial representation of GS. Clinical measurements included the Fugl-Meyer Assessment Scale (FMA), Barthel Index of Activities of Daily Living (BI), and the stage on the Brunnstrom Stages of Motor Recovery Scale (BR). Results. In the SG and control groups, a more substantial GS intensity was associated with muscle contractions of the flexor muscles of the opposite proximal joint than was the case for contractions of the flexor muscles of the distal joint, whereas such a gradient change was absent in the SL group. The corresponding spatial patterns of GS exhibited a predominant cross-excitation over the unexercised pectoralis major and extensor carpi radialis muscles in the control group, contrary to the enhanced activation of the brachioradialis and biceps brachii muscles noted in patients with stroke. The SG group had a better FMA score and a more satisfactory BR stage than did the SL group, and the 2 neurological scores were related to GS intensity for patients with stroke, depending on joint involvement. Discussion and Conclusion. Intensity of GS provided an affiliation with motor deficits and a promising window for poststroke recovery mechanisms. [Hwang IS, Tung LC, Yang JF, et al. Electromyographic analyses of global synkinesis in the Daretic uDDer limb after stroke. Phys Ther. 2005;85:755-765.]

Key Words: Dynamometer, Electromyography, Global synkinesis, Outcome, Stroke.

One of the characteristic outcomes of stroke is the unintended activation of one limb when the homologous part of the opposite limb is active. This phenomenon has long been documented with various associated terms--such as "global synkinesis" (GS), (1,2) "mirror movement," (3-5) "motor overflow," (6) and "contralateral irradiation"7--for the less than-adequate inhibition of the homologous muscle in addition to nonhomologous parts of the unexercised limb. (8-10) Global synkinesis is common to patients with poststroke hemiparesis (11,12) and many other neurological dysfunctions, (13-15) although it traditionally has been considered to represent a trivial phenomenon of relatively little scientific interest or a redundant movement interfering with movement coordination for loss of limb dexterity. (11,16) Global synkinesis also may occur frequently among people without any neurological problems, such as children prior to some maturational changes of the nervous system, (17,18) as well as among adults who are healthy when they are performing unfamiliar or strenuous motor tasks. (19)

Brain images reveal that the presence of GS involves bilateral excitation of the motor cortex (20) so that one hemisphere reduces its inhibitory influence on the opposite hemisphere via transcallosal fibers. (21,22) A well-known example of GS is persistent mirror movements found among patients exhibiting agenesis of the corpus callosum. (23) Following a specific unilateral lesion in one hemisphere, the efficacy of transcallosal connectivity has been reported to have been altered, leading to motor disinhibition of both hemispheres and interhemispheric reorganization (24,25) Recent studies have indicated that modification of cortical excitability subsequent to the emergence of certain brain lesions, especially disinhibition of the unaffected hemisphere, (26) relates to functional restoration following stroke. (25,27,28) Although a great deal of GS research has been conducted to date, (3,4,12,15,26) surprisingly little attention has been paid to the clinical effect of GS on the paretic limb, an activity that changes with cortical reorganization following cerebral lesion. (5)

In a study of patients with stroke, neuronal irradiation was manifested and mutually coupled among the flexor muscles of the affected upper limb, a phenomenon identified as flexion synergy. (11,16) Through observational approaches, the synergistic pattern of GS for patients with stroke appears to be stereotyped and governed by primitive reflexes. (11) Alternatively, GS has been reported to be amendable for people who are healthy, reflecting differences in cortical irradiation patterns with respect to task characteristics. (29) It would appear that such a debatable issue has not been helped by the apparent lack of systematic and quantitative studies investigating whether poststroke GS can be shaped with regard to target movement of different joints.

Using multichannel surface electromyography (EMG), our study will define the characteristics of spatial representation and the intensity of GS in the paretic upper limb during flexion movements of the contralateral arm, and it will attempt to connect these features with functional outcome measures that account for motor impairments across several affected joints. The goals of our study are (1) to uncover whether intensity and spatial representation of GS depend on joint involvement for people who are healthy and patients with poststroke hemiparesis and (2) to reveal the relationship between GS features and the degree of stroke-related motor deficit. Our findings may provide a better insight into GS to help assess functional outcomes after the onset of stroke and to encourage innovative therapeutic interventions for patients with stroke.

Method

Subjects

Twenty people (13 men, 7 women; mean age=51.3, SD=7.5) with a cerebrovascular accident (CVA) were recruited from a pool of appropriate patients at the Rehabilitation Hospital of Chung Shan Medical University. Seven of the 20 people with a CVA demonstrated left-sided weakness. Specific stroke subtypes for these patients were identified using a computed tomographic or magnetic resonance image. For people with a CVA, the time subsequent to the lesion's detection exceeded 1 month (range=15-24 months). None of the subjects with a CVA were taking any antispastic medication or had any limitations in passive range of motion. The motor recovery status of those patients with hemiparesis was between stages II and VI in the Brunnstrom Stages of Motor Recovery Scale (BR). (11) The inclusion criterion for the subjects with a CVA was the presence of a unilateral motor deficit in the upper and lower extremities. Those subjects who demonstrated deficits in language, attention, or cognition were excluded from the study because of their possible inability to follow experimental instructions. Characteristics of the subjects with a CVA are listed in Table 1. The 20 subjects (12 men, 8 women; mean age=49.7 years, SD=9.5) of the control group were volunteers from a local community. All study participants were right-handed by self-report and signed written informed consent forms in order to participate in the study.

Instrumentation and Data Acquisition

A Biodex dynamometer (Multi-Joint System 3 Pro)* was used to provide appropriate resistance against voluntary isometric muscle contractions of muscle groups of the shoulder, elbow, and wrist, and the electrical signal derived from the torque output of the dynamometer was connected to an oscilloscope (Instek GOS-620), ([dagger]) which served as an immediate visual feedback to the participants. Eight bipolar surface electrodes ([double dagger]) (electrode spacing=2.5 cm; diameter-l.1 cm, with a gain of 365 and a common mode rejection ratio of 102 dB) were used to record GS activity of the biceps brachii, brachioradialis, flexor carpi radialis, middle deltoid, triceps brachii, pronator teres, extensor carpi radialis, and pectoralis major muscles contralateral to the exercised limb. The electrodes were applied in parallel with the respective muscle bellies according to the technique suggested by Cram et al. (30) The recorded myosignals were conditioned using analog low-pass filters (with cutoff frequencies set at 400 Hz) and were then digitized at 1 kHz. Three separate clinical evaluations were obtained for each subject with stroke, including an arm motor score on the Fugl-Meyer Assessment Scale (FMA) (31) (range=0-66), a score on the Barthel Index of Activities of Daily Living (BI) for independence in mobility and personal care (32) (range=0-20), and a BR stage for motor recovery (11) (range=I-VI). Before dynamometer measurements were taken, a licensed and appropriately trained physical therapist administered the scales.

The subjects, who were not aware of the focus of our research interests, performed a set of voluntary isometric muscle contractions with test trials randomly ordered across the group of subjects. A total of 3 maximal voluntary isometric muscle contractions (MVIC) were used in the study, including shoulder, elbow, and wrist flexor muscle groups. The subjects with a CVA completed the target tasks using the joints of the unaffected extremity. Similarly, to prevent GS laterality between experimental and control groups, (13,17) 7 subjects in the control group performed identical tasks using the right upper extremity, and the other 13 subjects performed the same tasks with the left upper extremity. The positions in which the exercised upper extremity were held for the purposes of performing the isometric muscle contractions were standard positions documented in the user manual of the Biodex system (33) (Appendix), and the relaxed (affected) upper extremity hung without support parallel to the trunk. For these positions, subjects remained seated in the dynamometer chair with their hips, knees, and ankles stabilized at approximately 90 degrees to the direction of isometric flexor muscle contractions. Each subject's target joints were aligned with the axis of the dynamometer, with the trunk and lower extremities secured to the testing chair. Before the experiment, the subject was instructed to relax completely. Background activity of all muscles then was meticulously controlled by keeping the value of its root mean square (RMS) below 4.8 [micro]V before activity was recorded 3 times for a period of 3 seconds. We verbally encouraged the subjects to perform voluntary isometric muscle contractions and to relax the unexercised upper extremity. When steady torque output at the desired exertion level was achieved, as determined by monitoring the computer display, irradiated muscle activity was recorded for 3 seconds. The RMS was determined based on the EMG signal during that period. Each subject performed an isometric muscle contraction 3 times consecutively with a 2-minute rest between trials. The recorded myosignals were conditioned using analog low-pass filters (cutoff frequencies were set at 400 Hz) within a distribution box and then digitized at 1 kHz using a computer program constructed on a LabVIEW platform ([section]) (version 6.1).

Data Analysis

Off-line analyses included removal of the linear trend from the raw EMG results and also removal of possible artifacts using a 10th-order digital Butterworth band-pass filter (cutoff frequency=40-400 Hz) and the calculation of the RMS value from the conditioned EMG data. The mean RMS value was determined by averaging the RMS values of the 3 trials for each task. In terms of RMS, we defined the standardized net excitation level of an irradiated muscle (SNEi, i=1-8) by first subtracting the value of the RMS for irradiated EMG from its background activity and then normalizing the value with the RMS of the background activity (Fig. 1). The summation of the SNEi from all recorded muscles resulted in standardized net excitation (SNE), the GS intensity. The spatial representation of the GS pattern was characterized with respect to the relative excitation (RE) level by dividing the SNEi of an individual muscle by the corresponding SNE value. Namely, RE is the proportion of irradiation of a muscle that is attributed to GS intensity.

[FIGURE 1 OMITTED]

Because of the substantial heterogeneity of the variance in SNE values for the subjects with stroke and subjects who were healthy, we adopted nonparametric analyses in this study and dichotomized the 20 subjects with stroke by ordering their level of muscle irradiation into high irradiation (SG, n=10) and low irradiation (SL, n=10) groups according to the overall SNE level for the 3 target movements. The clinical assessments (FMA, BI, and BR) of the 2 groups of subjects with stroke were compared with the Mann-Whitney U test, and the relationship between the clinical assessment and GS intensity was examined using the Spearman rank correlation coefficient ([rho]). In order to compare the effect of joint involvement on the relative intensity of the GS pattern, the specific SNE values for the flexor muscle groups of the shoulder, elbow, and wrist at MVIC were examined using the Friedman 2-way analysis of variance by rank test for all 3 groups (control, SG, and SL). Another feature of the resulting GS response, the spatial pattern, was examined by comparing the respective RE for each recorded muscle among the control, SG, and SL groups using Kruskal-Wallis 1-way analysis of variance. The level of significance for the determination of difference for the nonparametric analyses was .05. Signal processing and statistical analyses were completed using Matlab version 6.0??? and the Statistical Package for Social Sciences (SSPS for Windows, version 10.0). (#)

Results

In general, GS activity was observed to spread through the muscles of the contralateral upper extremity in association with target movements (Fig. 2). Figure 3 displays the means and standard deviations of the 3 clinical assessments for the 2 stroke groups included in our study. Application of the Mann-Whitney U test indicated that there were significant differences in FMA score (H=77.5, P=-.035) and BR stage (H=77.5, P=.035) between the 2 stroke groups, whereas the subjects in the SG group had better neurological scores than did those in the SL group. The BI scores for both stroke groups were not different (H=101.5, P=.789). Table 2 displays the correlations between SNE level and FMA score, BI score, and BR stages for 3 different modes of maximal isometric flexion for subjects with a CVA. The Spearman correlation indicated that the SNE value for contralateral elbow flexion correlated significantly (P<.05) with all clinical assessments, whereas the SNE value for the contralateral wrist flexion was not related to neurological scores (FMA and BR) or activities of daily living (BI) (P>.05). For the contralateral shoulder flexion, a significant relationship existed only between SNE level and the corresponding neurological scores, both FMA and BR. More specifically, the SNE level for elbow and shoulder flexion movements correlated only with FMA subscale scores for motor function of the shoulder/ elbow/forearm (P<.005), but was poorly correlated with the FMA subscale scores for wrist and hand, respectively (P>.05).

[FIGURES 2-3 OMITTED]

During contralateral voluntary isometric muscle contractions of the flexors at MVIC, the Friedman test indicated that the effect of joint involvement on GS intensity, in terms of SNE value, was evident in the control group ([[chi].sub.r.sup.2](2) =12.4, P=-.002) and the SG group ([[chi.sub.r.sup.2](2)=11.4, P=.003), but the effect of joint involvement on GS intensity was not evident in subjects in the SL group ([[chi].sub.r.sup.2](2) =2.6, P=.273). Figure 4 shows that the SNE level for shoulder flexor muscle contraction was greater than the corresponding values for elbow and wrist flexor muscle contractions in the control and SG groups; however, the change in SNE level that resulted from an effect of joint involvement on GS intensity was not present in subjects in the SL group. Besides, each subject's SNE level was dependent on the group, with a greater SNE level for those subjects in the control and SG groups than for those in the SL group during MVIC of the shoulder ([[chi].sup.2](2)=8.16, P=017) and elbow ([chi square](2)=16.83, P<.001) flexors. No statistical difference in SNE level, however, was found during MVICs of the contralateral wrist flexors ([chi square](2)=0.572, P=.751) when the 3 groups were compared.

[FIGURE 4 OMITTED]

The spatial representation of the irradiated muscle activity was represented as the relative contribution to the overall GS intensity from each recorded muscle. Based on the corresponding RE level, Figure 5 depicts the spatial representation of GS associated with isometric shoulder, elbow, and wrist flexor movements. Significant differences in RE level for a specific muscle between control and stroke (SG and SL) groups are labeled with asterisks (P<.05). Despite considerable interindividual variability in SNE level, the Kruskal-Wallis test suggested that the RE level for the brachioradialis and the biceps brachii muscles for the SG group was greater than it was for the control group, regardless of joint involvement (Fig. 5). For the SL group, we noted that only the RE values for the biceps brachii muscle during elbow flexor muscle contractions and the flexor carpi radialis muscle during wrist flexor muscle contractions were relatively enhanced; this contrasts with the corresponding values for the control group (Fig. 5B and C). The RE value for the pectoralis major muscle for both stroke groups, conversely, was smaller than that of the control group for most of the flexor muscle contractions. Similarly, a pronounced RE value for the pronator teres and extensoring elbow and wrist flexor movements (Fig. 5B and C). This was not the case for people with a CVA. In brief, members of both stroke groups unanimously demonstrated uncharacteristic spatial alterations, with remarkable cross-excitation for the brachioradialis and biceps brachii muscles (flexors) in the SG group, but consistent decline in cross-excitation for the pectoralis major and extensor carpi radialis muscles (adductor and extensor) for both SG and SL groups.

Discussion

Global synkinesis was present in approximately 40% to 90% of the people who had sustained a CVA, depending on the choice of motor task, (5) the sensitivity of the GS detection, (28) and the severity of stroke-elicited motor deficits. (34) Although several studies have described GS quantitatively using surface EMG, force (or torque) measurements, and modern brain imaging techniques, to the best of our knowledge, the relationship between specific GS characteristics and the extent of motor function impairment for patients with stroke has seldom been addressed. (5) In our study, people with a CVA who were in the SG group had more satisfactory FMA and BR scores than did those in the SL group. Those people with a CVA and had less disability generally had a greater level of cross-excitation of the paretic upper limb during contralateral movements. Because the BI measures a patient's general functional ability, including the functional synergy of the lower extremity with sphincter control, the limited relationship between GS intensity in the upper extremity and BI scores did not come as a surprise.

Although GS intensity symbolized the neurological status of the affected upper extremity, the exact pathophysiological mechanisms for the development of GS over the paretic upper extremity remain current subjects of debate. The most plausible hypothesis for the pathophysiological mechanisms underlying the development of GS involves the ipsilateral corticospinal tract of the irradiated limb and the relative effectiveness of the transcallosal connectivity from the opposite (affected) hemisphere that inhibits the ipsilateral fast-conducting corticospinal tract of the working (intact) hemisphere. (21,22) Using transcranial magnetic stimulation, several researchers (24-26) have demonstrated a reduced level of intracortical inhibition in the intact hemisphere and facilitation of the ipsilateral descending pathway following stroke, arising as a consequence of the release of transcallosal inhibition from the affected hemi sphere. This disinhibition in the ipsilateral motor cortex relates strongly to the persistence of poststroke GS.

An ever-increasing wealth of evidence favors GS as an extensive compensatory mechanism for better restoration of motor function. (5,28,34) For instance, clinical assessment of trunk function and lingual movement has been correlated with ipsilateral motor-evoked potentials, as determined by stimulation of the unaffected hemisphere. (10,27) Furthermore, Nelles et al (5) related mirror movements to hand performance and reported that people with a CVA who exhibited mirror movements in the paretic hand demonstrated better motor function than did people with a CVA who did not exhibit these mirror movements. Moreover, our study indicated that global patterns of mirror movement were specifically pertinent to motor functions of the proximal arm following stroke, because we found that the SNE level related to the FMA subscales for motor function of shoulder and elbow muscle groups. The motor function of the wrist and hand graded with the FMA subscales for the wrist and hand, however, was not dependent on SNE level to any significant level (Tab. 2). Selective linkage of GS in the paretic limb with shoulder and elbow functions coincided with known anatomical findings that axial and proximal muscles tend to have a richer supply of the ipsilateral corticospinal fibers and brain-stem descending systems than do distal muscles. (35-37) The neural mechanism underlying functional recovery of the distal arm may differ from that of the proximal arm (38) and may rely primarily on the successful reorganization of the contralateral corticospinal pathway. (37,39,40)

It was also very interesting to observe the lack of a gradient change in SNE level with respect to joint involvement for people with a CVA and poor motor recovery (Fig. 4). Contralateral flexor muscle contractions about proximal joints were characteristically associated with the development of a greater SNE level than was the case for the contralateral muscle contractions about distal joints. In this case, such a patterned change in GS intensity could be explained by the rule of "perceived effort" (19) because exercising the proximal muscles generally demands greater muscle torque production than does exercising the muscles about the distal joint. In addition, acknowledging a comparatively sparser transcallosal projection to the region of the distal limb relative to the proximal limb tends to support our present finding from an anatomical standpoint, (41,42) the transcallosal connectivity presumably having been involved in the development of GS. As observed in our study, the lack in gradient change in GS intensity for people with a CVA and a pronounced disability is due primarily to the absence of significant cross-excitation with shoulder flexion of the uninvolved limb (Fig. 4). We argue that the absence of joint-dependent GS intensity for patients from the SL group was evidence of maladaptive suppression in the ipsilateral corticospinal tract of the intact hemisphere, which is attributable to topological impairment of transcallosal connectivity. Therefore, for patients with a poor motor recovery after stroke, severe damage in the anatomical realm could lead to an atypically low level of cortical excitation of the intact hemisphere even when intensively facilitated by strenuous contraction of the unaffected proximal arm.

The spatial patterns of GS for people with a CVA (both SG and SL groups) differed from the corresponding patterns of the control group as characterized by the RE level (Fig. 5). Based on the quantitative relations among EMG activity for different limb muscles, a common feature for people with a CVA was the relative waning in RE level for the pectoralis major muscle, but with rather exaggerated relative activation of the brachioradialis and biceps brachii muscles, regardless of the joint involvement. To our knowledge, no previous study has focused on the atypical irradiation of the pectoralis major muscle in people with a CVA, even though irradiation of the pectoralis major muscle typically represents marked cross-excitation scenario for people who are healthy. It may be that pectoralis major muscle activation is a purposely associated reaction that ensures shoulder and trunk stabilization against inertia driven by contralateral movements. (2,8) Acknowledging the relative absence of poststroke spatial patterning from supraspinal control systems for affected subjects, people with a CVA instead tend to develop an enhanced EMG activation of the paretic elbow flexors during GS, similar to the demonstration of flexor synergy during voluntary movement. (11)

Compared with the RE level for members of the control group, the deviation from normality of spatial patterns of GS for people with a CVA was rather manifest in association with the additional stroke-related reduction in RE activity of the extensor carpi radialis muscle during contralateral movements of the elbow and wrist joints (Fig. 5B and C). In contrast to the GS intensity, however, GS spatial patterns appeared to reveal little in the way of a discernible relationship with the degree of poststroke motor impairment because of the apparent absence of any significant difference in RE representation between the SG and SL groups. Furthermore, we also observed a sizeable variation in RE value in people with a CVA, symbolizing an apparent consequence of individualized cortical reorganization following a brain lesion. Recent studies (43,44) have pointed to the fact that brain activation patterns during paretic hand movement for people with a CVA could involve either the recruitment of ipsilateral activity or a focus on contralateral primary motor cortex activity. Perhaps individualized cortical reorganization also accounts for the observation of a rather good poststroke motor recovery associated with the demonstration of a rather slight GS intensity among a small number of patients with stroke and vice versa.

In the patients with poststroke hemiparesis, contralateral muscle contractions produced a widespread GS throughout the affected upper limb. Because GS intensity was especially remarkable for the patients with better motor recovery, our finding appears to encourage the use of GS as a means of facilitating recovery of unresponsive muscles in the paretic limb. (11) Our study, however, found little difference in GS spatial representation between patients in the SG and SL groups. Because GS spatial representation was not evidently related to motor recovery, it would be too early to claim that the patients could benefit from GS-based treatments. In addition, the muscles responding to contralateral isometric flexor muscle contractions in both patient groups deviated from the typical activation found in subjects who are healthy, regardless of joint involvement (Fig. 5). Another critical issue of supporting GS-based treatments is whether GS spatial representation can be shaped by movement patterns of the contralateral limb. Otherwise, it remains uncertain whether patterns of facilitation by means of GS selectively strengthen paretic muscles. Therapists who favor Bobaths' concepts argued that stereotyped association movements following stroke hamper functional use of the affected limb. (45)

An important methodological concern of our study was Normalization of EMG activity. Under the condition of consistent background activity across muscles, we preferred to express irradiated activity with the SNE value rather than as a percentage of maximal voluntary EMG activity. The reason for this choice is that the maximal torque development of a target movement, such as an elbow flexor movement can hardly be achieved by a single muscle, requiring instead coactivation of several functional synergists. Moreover, for people with a CVA, the strength (ie, force-generating capacity) of both clinically involved and uninvolved limbs is affected. (46) Normalization with maximal voluntary EMG activity of the involved or uninvolved limb, therefore, would be functionally questionable. In contrast, SNE allows us to realize the multiples of net irradiation level relative to background activity with ease and accounts for the contribution of undersized irradiation in some muscles, if background activity level is strictly controlled. The interpretation of GS characteristics based on SNE and RE, however, should be performed cautiously. Standardized net excitation and RE cannot provide a sufficient basis to contrast the excitation level among different irradiated muscles caused by contralateral movements; therefore, we focused the population effect on SNE and RE among the SG, SL, and control groups.

Conclusion

Global synkinesis depended geometrically on contralateral muscle contraction, with greater intensity being associated with contraction of the muscles about the proximal joint in subjects who were healthy and those people with a CVA who had a good recovery than was the case for the other study participants. The level of GS intensity in the paretic arm related to functional outcomes of patients with poststroke hemiparesis, especially when GS was triggered by contraction of the contralateral elbow flexors. For muscle contractions of the flexor muscles of the contralateral joints in the upper extremity, the resulting GS spatial pattern for people with a CVA appeared to be atypical, but seemingly unrelated to the degree of poststroke recovery. For subjects who were healthy, the EMG activity of the elbow flexors, rather than the shoulder adductors, became the predominant manifestation of the demonstrated GS spatial pattern. Because GS patterns may reflect cortical reorganization following stroke, it might be valuable to attempt to determine the temporal evolution of GS features after stroke and to attempt to relate these features to clinical assessments and brain imagery in order to attempt to enhance the course of poststroke recovery.

This study was approved by the institutional review board of Chung Shan Medical University Rehabilitation Hospital.

This research was in part supported by a grant from the National Science Council, Taiwan, under grant no. NSC 91-2314-B006-141 and NSC 92-2314-B040-005.

This article was submitted March 3, 2004, and was accepted January 7, 2005.

* Biodex Medical Systems Inc, 20 Ramsay Rd, PO Box 702, Shirley, NY 11967.

([dagger]) Good Will Instrument Co Ltd, No. 95-11, Pao-Chung Rd, Hsin-Tien City, Taipei Hsien, Taiwan.

([double dagger]) IOMED Inc, 2441 S 3850 W, Ste A, Salt Lake City, UT 84120.

([section]) National Instruments Inc, 11500 N Mopac Expressway, Austin, TX 78759-3504.

([parallel]) The MathWorks Inc, 3 Apple Hill Dr, Natick, MA 01760-2098.

(#) SPSS Inc, 233 S Wacker Dr, Chicago, IL 60606.

References

(1) Boissy P, Bourbonnais D, Gravel D, et al. Effects of upper and lower limb static exertions on global synkineses in hemiparetic subjects. Clin Rehabil. 2000;14:393-401.

(2) Boissy P, Bourbonnais D, Kaegi C, et al. Characterization of global synkineses during hand grip in hemiparetic patients. Arch Phys Med Rehabil. 1997;78:1117-1124.

(3) Green JB. An electromyographic study of mirror movements. Neurology. 1967;17:91-94.

(4) Radhakrishnan K, Koshy E, Prakash C. Ataxic hemiparesis and mirror movements. J Neurol Neurosurg Psychiatry. 1981;44:190.

(5) Nelles G, Cramer SC, SchaechterJD, et al. Quantitative assessment of mirror movements after stroke. Stroke. 1998;29:1182-1187.

(6) Lazarus JA, Whitall J. Motor overflow and children's tracking performance: is there a link? Dev Psychobiol. 1999;35:178-187.

(7) Hopf HC, Schlegel HJ, Lowitzsch K. Irradiation of voluntary activity to the contralateral side in movements of normal subjects and patients with central motor disturbances. Eur Neurol. 1974;12:142-147.

(8) Hwang IS, Abraham LD. Quantitative EMG analysis to investigate synergistic coactivation of ankle and knee muscles during isokinefic ankle movement, part 2: time frequency analysis. J Electromyogr Kinesiol. 2001;11:327-335.

(9) Dimitrijevic MR, McKay WB, Sarjanovic I, et al. Co-activation of ipsiand contralateral muscle groups during contraction of ankle dorsiflexors. J Neurol Sci. 1992;109:49-55.

(10) Fujiwara T, Sonoda S, Okajima Y, Chino N. The relationships between trunk function and the findings of transcranial magnetic stimulation among patients with stroke. J Rehabil Med. 2001;33: 249 -255.

(11) Sawner KA, LaVigne JM. Brunnstrom's Movement Therapy in Hemiplegia: A Neurophysiological Approach. 2nd ed. Philadelphia, Pa: Lippincott; 1992.

(12) Lazarus JC. Associated movement in hemiplegia: the effects of force exerted, limb usage and inhibitory training. Arch Phys Med Rehabil. 1992;73:1044-1049.

(13) Edwards JM, Elliott D. Asymmetries in intermanual transfer of training and motor overflow in adults with Down's syndrome and nonhandicapped children. J Clin Exp Neuropsychol. 1989;11:959-966.

(14) Kuhtz-Buschbeck JP, Sundhohn LK, Eliasson AC, Forssberg H. Quantitative assessment of mirror movements in children and adolescents with hemiplegic cerebral palsy. Dev Med Child Neurol. 2000;42: 728 -736.

(15) Vidal JS, Derkinderen P, Vidailhet M, et al. Mirror movements of the non-affected hand in hemiparkinsonian patients: a reflection of ipsilateral motor overactivity? J Neurol Neurosurg Psychiatry. 2003;74: 1352-1353.

(16) Zulch KJ, Muller N. Associated movements in man. In: Vinken PJ, Bruyn GW, eds. Handbook of Clinical Neurology. Amsterdam, the Netherlands: North Holland Publishing Co; 1968:404-426.

(17) Parlow SE. Asymmetrical movement overflow in children depends on handedness and task characteristics. J Clin Exp Neuropsychol. 1990; 12:270-280.

(18) Mayston MJ, Harrison LM, Stephens JA. A neurophysiological study of mirror movements in adults and children. Ann Neurol 1999;45:583-594.

(19) Armatas CA, Summers JJ, Bradshaw JL. Mirror movements in normal adult subjects. J Clin Exp Neuropsychol. 1994;16:405-413.

(20) Kristeva R, Cheyne D, Deecke L. Neuromagnetic fields accompanying unilateral and bilateral voluntary movements: topography and analysis of cortical sources. Electroencephalogr Clin Neurophysiol. 1991;81:284-298.

(21) Meyer BU, Roricht S, Grafin von Einsiedel H, et al. Inhibitory and excitatory interhemispheric transfers between motor cortical areas in normal humans and patients with abnormalities of the corpus callosum. Brain. 1995;118(pt 2):429-440.

(22) Schnitzler A, Kessler KR, Benecke R. Transcallosally mediated inhibition of interneurons within human primary motor cortex. Exp Brain Res. 1996;112:381-391.

(23) Dennis M. Impaired sensory and motor differentiation with corpus callosum agenesis: a lack of callosal inhibition during ontogeny? Neuropsychologia. 1976;14:455-469.

(24) Shimizu T, Hosaki A, Hino T, et al. Motor cortical disinhibition in the unaffected hemisphere after unilateral cortical stroke. Brain. 2002;125(pt 8):1896-1907.

(25) Manganotti P, Patuzzo S, Cortese F, et al. Motor disinhibition in affected and unaffected hemisphere in the early period of recovery after stroke. Clin Neurophysiol. 2002;113:936-943.

(26) Liepert J, Hamzei F, Weiller C. Motor cortex disinhibition of the unaffected hemisphere after acute stroke. Muscle Nerve. 2000;23: 1761-1763.

(27) Muellbacher W, Artner C, Mamoli B. The role of the intact hemisphere in recovery of midline muscles after recent monohemispheric stroke. J Neurol. 1999;246:250-256.

(28) Kim YH, Jang SH, Chang Y, et al. Bilateral primary sensori-motor cortex activation of post-stroke mirror movements: an fMRI study. Neuroreport. 2003;14:1329-1332.

(29) Armatas CA, Summers JJ. The influence of task characteristics on the intermanual asymmetry of motor overflow. J Clin Exp Neuropsyehol. 2001;23:557-567.

(30) Cram JR, Kasman GS, Holtz J. Introduction to Surface Electromyography. Gaithersburg, Md: Aspen Publishers; 1998.

(31) Fugl-Meyer AR, Jaasko L, Leyman I, et al. The post-stroke hemiplegic patient, 1: a method for evaluation of physical performance. Scand J Rehabil Med. 1975;7:13-31.

(32) Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J. 1965;14:61-65.

(33) Biodex Multi-Joint System 3 Pro Manual: Applications/Operations. Shirley, NY: Biodex Medical Systems Inc.

(34) Mulley G. Associated reactions in the hemiplegic arm. Scand J Rehabil Med. 1982;14:117-120.

(35) Matsunami K, Hamada I. Characteristics of the ipsilateral movement-related neuron in the motor cortex of the monkey. Brain Res. 1981;204:29-42.

(36) Chen R, Gerloff C, Hallett M, Cohen LG. Involvement of the ipsilateral motor cortex in finger movements of different complexities. Ann Neurol. 1997;41:247-254.

(37) Turton A, Wroe S, Trepte N, et al. Contralateral and ipsilateral EMG responses to transcranial magnetic stimulation during recovery of arm and hand function after stroke. Electroencephalogr Clin NeurophysioL 1996;101:316-328.

(38) Jung HY, Yoon JS, Park BS. Recovery of proximal and distal arm weakness in the ipsilateral upper limb after stroke. NeuroRehabilitation. 2002;17:153-159.

(39) Nass R. Mirror movement asymmetries in congenital hemiparesis: the inhibition hypothesis revisited. Neurology. 1985;35:1059-1062.

(40) Cramer SC, Nelles G, Benson RR, et al. A functional MRI study of subjects recovered from hemiparetic stroke. Stroke. 1997;28:2518-2527.

(41) Killackey HP, Gould HJ III, Cusick CG, et al. The relation of corpus callosum connections to architectonic fields and body surface maps in sensorimotor cortex of new and old world monkeys. J Comp Neurol. 1983;219:384-419.

(42) Rouiller EM, Babalian A, Kazennikov O, et al. Transcallosal connections of the distal forelimb representations of the primary and supplementary motor cortical areas in macaque monkeys. Exp Brain Res. 1994;102:227-243.

(43) Feydy A, Carlier R, Roby-Brami A, et al. Longitudinal study of motor recovery after stroke: recruitment and focusing of brain activation. Stroke. 2002;33:1610-1617.

(44) Ward NS, Brown MM, Thompson AJ, Frackowiak RS. Neural correlates of outcome after stroke: a cross-sectional fMRI study. Brain. 2003;126:1430-1448.

(45) Bobath B. Adult Hemiplegia: Evaluation and Treatment. 3rd ed. Oxford, United Kingdom: Heinemann Medical Books; 1990.

(46) McCrea PH, Eng JJ, Hodgson AJ. Time and magnitude of torque generation is impaired in both arms following stroke. Muscle Nerve. 2003;28:46-53.

IS Hwang, PT, PhD, is Associate Professor, School of Physical Therapy and Institute of Allied Health Sciences, National Cheng Kung University, Tainan, Taiwan.

LC Tung, MD, is Attending Physician, Department of Physical Medicine and Rehabilitation, Chi Mei Hospital, Liouying Township, Tainan County, Taiwan.

JF Yang, PT, MS, is Lecturer, School of Physical Therapy, National Cheng Kung University.

YC Chen, PT, MS, is Assistant Professor, School of Physical Therapy, Chung Shan Medical University, Taichung, Taiwan.

CY Yeh, PT, PhD, is Lecturer, School of Physical Therapy, Chung Shan Medical University.

CH Wang, PT, BS, is Associate Professor, School of Physical Therapy, Chung Shan Medical University, and Department of Physical Therapy, Chung Shan Medical University Rehabilitation Hospital, No. 110, 1 Sec., Chieh-Kuo N. Rd, Taichung 402, Taiwan. Address all correspondence to Mr Wang ([e-mail]).

Dr Hwang, Dr Tung, and Mr Wang provided concept/idea/research design, writing, and institutional liaisons. Ms Chen and Dr Yeh provided data collection, and Mr Yang and Dr Hwang provided data analysis. Mr Wang, Ms Chen, and DrYeh provided subjects and clerical support. Dr Hwang,

Dr Tung, and Mr Wang provided project management, facilities/equipment, and fund procurement. Mr Yang, Ms Chen, and Dr Yeh provided consultation (including review of manuscript before submission).

Appendix.

Standard Testing Positions Recommended in the Biodex System for
Voluntary Isometric Muscle Contractions of the Flexors of Different
Target Joints (33)

           Voluntary Isometric
           Muscle Contraction    Homologous   Effort
           (MVIC) (Seated)       Muscle (a)   Level

Shoulder   Flexors               BI           MVIC

Elbow      Flexors               BI and BRA   MVIC

Wrist      Flexors               FCR          MVIC

Shoulder   Recommended Testing
           Position in the Upper
           Extremity (Biodex)

Elbow      Shoulder flexion: 90 [degrees]
           Elbox flexion: 0 [degrees]
           Wrist: neutral position
           Seatback tilt: 70 [degrees]-85 degrees]

Wrist      Shoulder flexion: about 20 [degrees]-30 [degrees]
           Elbow flexion: about 60 [degrees]
           Wrist: neutral position
           Seatback tilt: 85 [degrees]
           Shoulder flexion: 0 [degrees]
           Elbow flexion: 90 [degrees]
           Wrist pronation: about 90 [degrees]
           Seatback tilt: 85 [degrees]

(a) Bt=biceps brachii muscle, BRA=brachuiradials muscle, FCR=flexor
carpi radialis muscle.

Table 1.

Characteristics of Subjects With a Cerebrovascular Accident (a)

        Age         Affected   Onset
Group   (y)   Sex   Side       Time         Type

SG      27    F     Right      3 mo         I
SG      54    M     Right      4 mo         I
SG      41    M     Right      6 mo         I
SG      65    F     Right      1 y, 5 mo    I
SG      47    M     Right      1 y, 1 mo    I
SG      51    M     Left       1 y, 2 mo    I
SG      52    F     Right      2 y          H
SG      55    M     Right      1 y          H
SG      45    M     Right      1 y, 11 mo   H
SG      53    M     Left       10 mo        H
SL      56    F     Right      2 y          I
SL      49    M     Right      1 y, 9 mo    I
SL      43    F     Left       1 y, 4 mo    I
SL      62    M     Left       1.5 mo       I
SL      58    M     Right      8 mo         H
SL      72    M     Right      1 y, 4 mo    H
SL      33    F     Right      10 mo        H
SL      52    M     Left       11 mo        H
SL      60    M     Left       1 y, 10 mo   H
SL      51    F     Left       3 mo         H

        Main
Group   Location         FMA   BR    BI

SG      MCA              25    IV    17
SG      Thalamus         52    VI    11
SG      MCA              59    VI    19
SG      MCA              53    V     19
SG      Brain stem       52    V     10
SG      MCA              23    III   19
SG      Basal ganglion   18    III   12
SG      Thalamus         59    VI    20
SG      Basal ganglion   10    IV    20
SG      Basal ganglion   23    IV    13
SL      MCA               9    III   11
SL      Basal ganglion    9    III   19
SL      MCA              24    III   19
SL      MCA              21    IV    16
SL      MCA              51    V     20
SL      Basal ganglion   26    IV    13
SL      MCA              32    IV    20
SL      Basal ganglion   10    III   12
SL      Putamen           7    III   11
SL      MCA               8    III   13

(a) FMA=Fugl-Meyer Assessment Scale arm motor score, BR=Brunnstrom
Stages of Motor Recovery Scale, BI= Barthel Index of Activities of
Daily Living, SG=subjects with stroke and a greater level of
irradiation, SL-subjects with stroke and a lesser level of
irradiation, F=female, M= in male, I=infraction, H=hemorrhage,
MCA=middle cerebral artery.

Table 2.

Spearman Correlation Coefficients (p) for Standardized Net Excitation
(SNE) Level and Clinical Assessments (a)

            Shoulder Flexion  Elbow Flexion      Wrist Flexion

SNE(N=20)   [rho]   P         [rho]   P          [rho]   P

FMA         .530    .016 *   .640     .002**    .409    .073
BTI         .422    .064     .492     .027*    -.050    .883
BR          .539    .014 *   .575     .008*     .434    .056
FMA_SE      .604    .005 **  .682    <.001 **   .421    .064
FMA_W       .374    .105     .320     0.169     .340    .143
FMA_H       .238    .312     .393     0.087     .365    .114

(a) Probability scores are evaluated as being not significant or
significant (* : P<.05) (** : P=.005). FMA=Fugl-Meyer Assessment Scale
at in motor score, BI: Barthel Index of Activities of Daily Living,
BR=Brunnstrom Stages of Motor Recovery Scale stage, FMA_SL=Fugl-Meyer
Assessment arm motor score for shoulder/elbow/forearm, FMA W=Fugl-Meyer
Assessment Scale arm motor score for wrist, FMA H=Fugl-Meyer arm motor
score for hand.