RESEARCH ARTICLE
Characteristics of Associated Reactions in People with Hemiplegic Cerebral Palsy Hsiu-Ching Chiu1,2, Louise Ada1*, Jane Butler1 & Susan Coulson1 1
Discipline of Physiotherapy, The University of Sydney, NSW, Australia
2
Department of Physical Therapy, I-Shou University, Kaohsiung, Taiwan
Abstract Purposes. To investigate the relationship between associated reactions and a) spasticity, b) contracture and c) coordination. Methods. Associated reactions were measured as magnitude of muscle activity in the affected limb during a 50% maximum voluntary contraction of muscles in the unaffected limb. Spasticity was measured as hyper-reflexia during passive muscle stretch, coordination as performance during a tracking task, and contracture as loss of range of motion. Chi-square analysis was used to examine the association between associated reactions and spasticity, and linear regression to examine the relationship between associated reactions and spasticity, coordination and contracture. Results. Twenty-three people with hemiplegic cerebral palsy aged from 15 to 47 years (mean [SD]: 29 years [9]) participated. Thirteen participants exhibited spasticity, and six participants exhibited associated reactions. Five of the six participants with associated reactions also had spasticity (χ2 = 2.37, p = 0.12). Associated reactions were highly correlated with spasticity (r = 0.77, p = 0.001), but not with contracture (r = 0.35, p = 0.29) or coordination (r = −0.31, p = 0.30). Conclusions. Although 27% of participants exhibited associated reactions, and these were mostly small, associated reactions appear to be an expression of spasticity in hemiplegic cerebral palsy. Copyright © 2010 John Wiley & Sons, Ltd. Received 2 December 2009; Revised 19 May 2010; Accepted 28 May 2010 Keywords associated reactions; cerebral palsy; hemiplegia *Correspondence Louise Ada, Associate Professor, 75 East Street, Lidcombe, Discipline of Physiotherapy, The University of Sydney, NSW1825, Australia. Email:
[email protected] Published online 22 July 2010 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/pri.487
Introduction In normally developing children, unilateral movement often demonstrates associated movements in the contralateral limb that are considered to be a normal phenomenon (Todor and Lazarus, 1986). However, these movements gradually disappear as myelination occurs in the corpus callosum (Hoy et al., 2004). In particular, the frequency and intensity of associated movements tends to decrease dramatically after nine years of age Physiother. Res. Int. 16 (2011) 125–132 © 2010 John Wiley & Sons, Ltd.
(Cohen et al., 1967). In normal adults, associated movements may still occur, but only in response to performing either unfamiliar or strenuous motor tasks (Hwang et al., 2006a) or using the non-dominant hand (Hwang et al., 2006b). The term ‘associated reactions’ was originally suggested by Walshe (1923) to describe the involuntary associated movements observed in people with neurological conditions. Associated reactions involve involuntary muscle activity occurring in homonymous muscles or in global/mass muscles of the 125
Associated Reactions in Hemiplegic Cerebral Palsy
affected limb when the unaffected limb is active (Walshe, 1923). Although associated reactions have been reported occurring in homologous muscles (Dvir et al., 1996; Ada and O’Dwyer, 2001) and non-homologous muscles (Dickstein et al., 1995; Boissy et al., 1997; Hwang et al., 2005) in stroke, little attention has been paid to people with cerebral palsy. Associated reactions have been seen as a part of the spastic syndrome (Walshe, 1923) and are thought to occur only in the presence of spasticity (Stephenson et al., 1998). That is, associated reactions may be part of the reflex response to the stretch of spastic muscles. For example, when people with spasticity stand up from a chair, the arm straightens, causing stretch of the spastic biceps muscle, which in turn causes a reflex contraction, observed as an associated reaction. The intensity of associated reactions has been seen as a key feature in the measurement of spasticity (Stephenson et al., 1998). Generally, the clinical view has been that associated reactions, like spasticity, interfere with activity, that is, the worse the associated reactions, the worse the performance of daily activities (Dvir et al., 1996). However, Hwang et al. (2005) reported that large associated reactions in people with stroke did not interfere with hand coordination. In addition, associated reactions could theoretically contribute to contracture in people with cerebral palsy. Long-term adaptation may occur if associated reactions cause a muscle to remain in a shortened position (Ada and O’Dwyer, 2001). As there is little research about associated reactions in people with cerebral palsy, the purpose of this study was to investigate 1) the strength of the relationship between associated reactions and spasticity, contracture, and coordination and 2) whether associated reactions are an expression of spasticity. Hemiplegic cerebral palsy was chosen since it is predominately a unilateral neurological condition and the investigation of associated reactions was therefore unlikely to be confounded by significant disability of the unaffected side. Similarly, adults were recruited so that associated reactions were not confounded by associated movements that normally occur in children.
Method Design An observational study of motor impairments was carried out in adolescents and adults with hemiplegic 126
H.-C. Chiu et al.
Figure 1 Experimental set-up for upper limb measurements. This diagram shows the equipment used to measure associated reactions, spasticity, contracture and coordination of the elbow. Three signals were collected: a potentiometer measured joint angle, a load cell measured force and electrodes measured muscle activity (Ada and O’Dwyer, 2001)
cerebral palsy. Four motor impairments were measured using an instrumented arm frame: associated reactions, spasticity, contracture and coordination (Figure 1). The upper limb was investigated, in particular the elbow joint muscles, because they are a common site of associated reactions, spasticity and contracture. All measurements were performed on participants’ affected upper limb, in the same order, and were collected in one session of approximately one hour duration. Participants were blinded to the specific hypotheses of the study and the sequence of measurements. This study was part of a larger project examining the contribution of motor impairments to activity, of which previous results have been published elsewhere (Chiu et al., 2010). It was approved by The Spastic Centre Human Research Ethics Committee and The University of Sydney Human Ethics Committee, and participants gave informed consent before data collection commenced. Participants Volunteers were recruited via advertisements through The Spastic Centre of New South Wales in Australia, through local newspapers, community newsletters and Physiother. Res. Int. 16 (2011) 125–132 © 2010 John Wiley & Sons, Ltd.
H.-C. Chiu et al.
Cerebral Palsy Sport and Recreation Association. Volunteers were scheduled for measurements in order of response. Participants were included in this study if they were diagnosed with hemiplegic cerebral palsy before five years of age, were over 15 years of age now and had sufficient cognition and language to participate. Two functional classifications, the Gross Motor Function Classification System (GMFCS) (Palisano et al., 1997) and the Manual Ability Classification System (MACS) (Eliasson et al., 2006), were employed to describe the characteristics of participants. Associated reactions Associated reactions were measured as the magnitude of abnormal muscle activity in the affected elbow flexors and extensors during a 50% maximum voluntary contraction (MVC) of the unaffected arm. Fifty per cent MVC was chosen in this study because it is reasonably effortful and does not normally produce associated movements (Hwang et al., 2006a), but has been shown to produce associated reactions in people with neurological conditions (Boissy et al., 1997; Ada and O’Dwyer, 2001). Electromyography (EMG) was recorded from bipolar silver-silver surface electrodes (Neotrode) positioned for the affected elbow flexors and extensors according to Basmajian and Blumenstein (1980). A hand-held dynamometer (PowerTrack IITM commander: Australasian Medical & Therapeutic Instruments P/L, Australia; 125 pounds rated capacity, linearity 1%) was used to monitor 50% MVC of the unaffected arm. Participants were seated in a highbacked chair at a height-adjustable table with the affected forearm securely supported by a horizontal arm frame, with the elbow at 60 degrees of flexion and the shoulder at 90 degrees of flexion (Figure 1). The unaffected elbow flexors/extensors were performed against the hand-held dynamometer in order to determine a 50% MVC. In the first condition, participants held a 50% MVC of the unaffected elbow flexors for four seconds using feedback from the dynamometer while the affected limb was instructed to relax in the arm frame. Simultaneously, the EMG activity of both the affected elbow flexors and extensors was recorded. In the second condition, participants held a 50% MVC of the unaffected elbow extensors while EMG activity of both affected elbow flexors and extensors was recorded for four seconds. Since the correlation between the affected elbow flexor and extensor muscle activity Physiother. Res. Int. 16 (2011) 125–132 © 2010 John Wiley & Sons, Ltd.
Associated Reactions in Hemiplegic Cerebral Palsy
was high within conditions (r > 0.73, p < 0.001) as well as between conditions (r = 0.58, p = 0.003), the average magnitude of muscle activity in the affected arm during 50% MVC across the unaffected elbow flexors and extensors was used to determine the magnitude of associated reactions and reported in uv. Spasticity Spasticity (elbow flexors/extensors) was measured as stretch-induced muscle activity during a fast passive stretch. This is in line with the most widely accepted definition of spasticity that describes it as a motor disorder resulting from hyperexcitability of the stretch reflex that is characterized by exaggerated tendon jerks and velocity-dependent increases in tonic stretch reflexes (Lance, 1980). The load cell (Applied Measurement, Sydney, Australia, 250 N rated capacity, Linearity 0.03%) and a handle were attached to the arm frame. Elbow angle was measured by a potentiometer aligned directly below the elbow joint. Participants relaxed such that there was no muscle activity present and the investigator performed five passive stretches as fast as possible (average peak velocity of 240°/second) across a large amplitude of elbow flexion and extension (average amplitude of 80°). Both the baseline and peak muscle activity during stretch were collected. Normally, no EMG activity is observed when relaxed muscles are stretched, so any stretch-induced muscle activity resulting from hyper-reflexia was taken to represent spasticity. Therefore, the difference between baseline and peak muscle activity during stretch averaged over elbow flexion and extension across the five trials was used to determine the magnitude of spasticity and reported in uv (Chiu et al., 2010). Contracture Contracture was measured as loss of maximum passive elbow joint range with the shoulder in a standardized position. Participants were asked to relax so that there was no muscle activity present. The investigator pushed the elbow joint first to the end of extension and then to the end of flexion. This position was held for four seconds while a moderate torque (2 Nm) was applied (Halar et al., 1978). Meanwhile, the EMG recording was displayed online to check that the participants stayed relaxed. Elbow flexor contracture was calculated as maximum passive extension subtracted from 0 degrees and elbow extensor contracture was calculated as 127
Associated Reactions in Hemiplegic Cerebral Palsy
maximum passive flexion subtracted from 130 degrees. Contracture was determined as the mean elbow flexor and extensor contracture and reported in degrees. Coordination Coordination was measured during a visual tracking task that has been shown to quantify poor muscle coordination (Canning et al., 2000). This task is a credible laboratory model of the requirements of everyday coordination because it incorporates the need to rapidly swap from agonist to antagonist with spatial and temporal accuracy. Participants were required to track a semi-randomly moving target on a computer screen using a 10-degree range of elbow flexion and extension (±5 degrees) around a midpoint of 60 degrees. The tracking task was performed at 0.8 Hz. A ball-bearing joint on the arm frame allowed elbow flexion and extension with minimal strength requirements (less than 1 Nm elbow flexor/extensor torque). The participants were allowed to practice until becoming familiar with the setup, and then performed one minute of the actual task. The overall coherence between the target (semi-random movement) and the response (voluntary elbow flexion and extension between 55–65 degrees) was computed as a quantitative measure of coordination using cross-correlational and spectral analysis (Bendat and Piersol, 1971). Overall coherence was used to determine coordination and was reported as a ratio. Data analysis An amplifier Biopac Systems MP 100A (BIOPAC Systems Inc, Sydney/NSW, Australia) was used for amplifying the EMG signals that were sampled synchronously by a 16-bit A-D converter at 1000 Hz and stored in the Acknowledge Software Package (Version 3.7.3, BIOPAC Systems Inc, Sydney/NSW, Australia). An analysis package MATLAB (Version 4.2c.1, The MathWorks Inc, MA, USA) was used to process these data offline. Integrated EMG data was obtained by high-pass filtering (digital 8th—order Butterworth) at 80 Hz, rectifying and then low-pass filtering the EMG data (digital 8th—order Butterworth) at 5 Hz. This cut-off frequency was chosen because all frequencies of interest less than 5 Hz. Chi-square analysis was used to investigate the association between the presence of associated reactions and the presence of spasticity, contracture and poor coordination. Poor coordination was defined as less 128
H.-C. Chiu et al.
than half of the normal mean (i.e.
