1972a, b; Neilson and McCaughey 1981,. 1982) have quantified abnormalities in the amplitude and timing of tonic stretch reflexes in spastic cerebral palsy.
REDUCING SPASTICITY TO CONTROL MUSCLE CONTRACTURE OF CHILDREN WITH CEREBRAL PALSY Janet Nash Peter D. Neilson Nicholas J. O’Dwyer
Disorders of muscle tone in cerebral palsy arise from abnormalities of stretch reflexes and involuntary muscle contractions or spasms. Previous investigations in this laboratory (e.g. Neilson 1972a, b; Neilson and McCaughey 1981, 1982) have quantified abnormalities in the amplitude and timing of tonic stretch reflexes in spastic cerebral palsy. Stretch applied to a spastic muscle is vigorously opposed by a hypersensitive tonic stretch reflex response, causing the muscle to feel tight and stiff. Spasticity and spasm both resist stretch and tend to maintain the muscle in a shortened position. Stretch is an important stimulus to muscle growth (Holly et al. 1980) and its absence or diminution appears to account for diminished growth and consequent contracture in spastic muscle (Ziv et al. 1984). Progressive plaster-casting for prevention or correction of muscle contracture in cerebral palsy attempts to promote muscle growth by lengthening the muscle, but an increased range of joint movement following casting may be the result of lengthening of either muscle or tendon, or both (Tardieu et al. 1977, 1982b), whereas only muscle lengthening increases the range over which the muscle can actively generate tension. This point is relevant to surgical interventions for muscle contracture, as well as to casting.
Moreover, recurrence of symptoms may be a problem following either treatment, since the involuntary contractions due to spasticity and spasm usually persist after treatment. The best treatment, as suggested by Tabary et al. (1981), would be suppression of the abnormal muscle contractions responsible for muscle shortening. This would prevent not only the development of contracture but also the disruption of limb function caused by both the contracture and the abnormal muscle contractions. In previous experimental investigations (Neilson and Lance 1978, Neilson and McCaughey 1982) we have shown that adults with cerebral palsy can be trained to reduce spasticity in elbow muscles. Moreover, reduced spasticity was accompanied by a reduction in involuntary muscle contractions (spasms). This training used biofeedback of the gain of the tonic stretch reflex in elbow flexor muscles to facilitate control of reflex sensitivity. The aim of the present study was to determine whether these techniques could be adapted for cerebral-palsied children at risk for contracture of the calf muscles. If some control of spasticity and spasm could be achieved, it might prevent the development of muscle contracture and achieve some more normal muscle activity.
weeks measurement days were alternated with training days to minimise the effect of withdrawal of feedback. The final two weeks were dedicated solely to measurement. Training was provided for one leg only (whichever was more impaired), but two sessions were allotted to measurements from the other leg during both the pre- and post-training assessment.
Fig. 1. Sketch of therapist, child and equipment. Microcomputer and amplifier for potentiometer output are on lower table; EMG amplifier with meter display and cassette player are on higher table, with video cassette recorder on top of television set. Potentiometer and surface electrode signals are led first to their respective amplifiers, then to joystick inputs on rear of microcomputer.
Method
4 72
Subjects The subjects were three spastic diplegic children with normal intelligence. They were two boys aged 4 years 1 1 months at the beginning of the study and one girl aged 8 years 3 months. One boy was ambulatory and the other was confined to a wheelchair: both had difficulty in lowering their heels to the ground when standing. The girl was ambulatory only with a walking frame and tended to drag her feet. Informed consent was obtained from the parents. The co-operation of the children was elicited without difficulty. The children attended measurement/ training sessions for four days each week over 18 weeks, each session lasting approximately one hour. In order to obtain a reliable and stable pre-training assessment, baseline measurements were carried out for the first four weeks, when no feedback was provided. Training was then conducted for the next 10 weeks. Each child's family also was given instructions and provided with the necessary equipment for training at home, the frequency being left up to the individual families. Extensive training was provided by the family of one of the boys, but was more intermittent for the other two children. Post-training assessmerit was carried Out Over the last four weeks, but during the first two of these
Equipment The major features of the measurement/ training situation are illustrated in Figure 1. The children sat in a three-cornered posture chair with hips at 90" and knees extended. A two-armed angle-measuring device, instrumented with a potentiometer, was strapped around the foot and lower leg to record the ankle angle. Silver-silver-chloride surface electrodes were placed over the lateral head of the gastrocnemius muscle. To ensure that the electrodes were placed in the same position each day, the position for each electrode was marked on the skin with a coloured marking pen or stamp. This ensured reproducibility of the electromyogram (EMG) signal from day to day in each child. The EMG was amplified (bandwidth ~OO-~OOHZ), full-wave rectified and low-pass filtered (time constant 0.25 sec) to obtain a DC voltage (IEMG) proportional to the contraction level of the muscle*. This IEMG signal is functionally related to muscle tension, the relation being linear under isometric conditions or during contractions of constant velocity (e.g. Inman el al. 1952, Lippold 1952, Bigland and Lippold 1954, Gottlieb and Agarwal 197 1, Lindstrom et al. 1974, Milner-Brown and Stein 1975, Hof and van den Berg 1977). The IEMG and ankle-angle signals were connected via the joystick socket to a widely available home microcomputer (Tandy TRS-80) and sampled by a 6-bit A-D converter at 20 samples/sec. Measurement During the pre-
and
post-training
*This signal processing was performed with a lowcost EMG amplifier (CI Myograph) manufactured to our design specifically for this purpose by Centre Industries, which is the industrial adult-training division of the Spastic Centre of New South Wales.
assessment periods, two sets of measurements were made of the IEMG and ankle-angle signals. First the range of voluntary joint rotation was measured. After placement of the electrodes the children were required to move the foot through its maximum range, first by plantarflexion and then by dorsiflexion. On reaching the end of the range in each direction they were encouraged to try with maximum effort to reach even further. The highest level of IEMG activity recorded during this manoeuvre was designated the maximum level of voluntary contraction. These maximum contractions usually were reproducible from day to day, with a small range of variability. In the illustrations in the Results section the IEMG signals are calibrated in terms of average maximum voluntary contraction, so that some appreciation may be gained of the over-all level of muscle activity involved. Following measurement of the maximum range of joint rotation, the degree of spasticity was assessed. The children were asked to relax and t o keep their muscles as 'soft' and loose as possible while sitting in the posture chair and watching cartoons on television. The experimenter applied a quasi-sinusoidal stretch to the muscle by passively moving the foot back and forth. The ankle angle was displayed on a monitor so that ankle position and the amplitude and frequency of displacement could be controlled. The knees tended to flex, especially at extreme muscle lengths, so they were held down with small sandbags to ensure the legs remained straight. The size of IEMG activity elicited by each stretch provided a measure of muscle spasticity. The gain of the EMG amplifier was adjusted to provide a wide range of deflection on the meter display during stretching. This display was monitored to ensure that the IEMG activity remained time-locked to muscle stretch and varied about a stable background level. If shifts in the average level of activity were detected because of muscle spasms or other voluntary or involuntary background activity, stretching was halted until the child had relaxed the muscle. Ankle angle and IEMG signals obtained during sampling periods of 22 seconds
were stored in computer files for later analysis. The amplitude of IEMG activity was normalised with respect to a constant gain setting of the EMG amplifier. Reflex sensitivity scores were computed for each 22-second sampling period by dividing the standard deviation of the IEMG activity by the standard deviation of the ankle angle. This provided a measure of the amount of reflex EMG activity relative to the size of muscle stretch. The amplitude of displacement of the ankle was always 25". Mean ankle positions of 122,112 and 98" and stretching frequencies of 20, 30 and 60/min were used. In addition to the above measurement during continuous joint rotation, the response of the muscle t o sustained stretch was examined on several occasions during the pre- and post-training assessment periods. Ankle angle and IEMG signals were recorded while the muscle was stretched slowly from rest through to 85" at the ankle, then held in that position.
%
4
-
G m
& z
s
,o
j 9
B -3
P Q .s $j
P
c
2 ti
% ,o
&
Training No measurements were taken during the training period of the study, the sessions being devoted entirely to reducing spasticity. The microcomputer was programmed so that simple graphic displays on the computer screen reflected the level of input from the EMG amplifier while the child's ankle was rotated by the experimenter. For example the size of the smile on a face was programmed to be inversely proportional to the average level of EMG-a low EMG input produced a big smile and higher EMG levels produced progressively smaller smiles. A threshold level could be set so that when the smile reached a particular size the remote control switch of a cassette player was activated to play music or childrens' stories. The cassette player would stop if the smile was reduced below the threshold level. The aim was to maintain a big smile, with the cassette playing continuously, so success depended on a continuous reduction in the level of stretch-evoked EMG activity. For variety, many other games were programmed. For example, in a game of 'squash', the racquet accurately tracked the ball and hit it against a wall if the
473
was low enough. As the EMG increased the racquet missed the ball by a proportionately greater distance. A score was displayed for the number of consecutive hits before the ball was missed, thus providing an incentive to try for a higher score next time. Success in the games could be influenced by the experimenter by changing either the threshold level for success or the gain of the EMG amplifier. Threshold and gain were adjusted so that success was achieved about 75 per cent of the time; without this high rate of success the children rapidly became disheartened and lost their motivation to continue. Once improvement had occurred, either the threshold was lowered or the EMG amplifier gain was increased so that a greater reduction in EMG level was needed to achieve success. The children were aware that these changes were being made and their aim became not only to achieve a high score in a game, but also to work with a high gain on the EMG amplifier. Greater difficulty also was introduced by stretching the muscle further and faster. Thus at any stage when reduction in EMG was achieved, the difficulty in controlling the display was increased to promote further EMG reduction. After some weeks of training it became more difficult to maintain the interest of two of the children in the games, so a video cassette recorder was introduced. The microcomputer was connected to the pause switch on the remote control of the video recorder and this switch was activated if the EMG rose above the threshold level. By reducing EMG levels below the required threshold, the children could watch recordings of their favourite cartoons. Motivation, particularly in one child, was much improved. EMG
50%-
IEMG
-
-=
