The authors have declared that no competing interests exist.
Manuscript revision and criticism: GA. Conceived and designed the experiments: L. Marinelli. Performed the experiments: L. Marinelli L. Mori GV. Analyzed the data: L. Marinelli FC ET. Contributed reagents/materials/analysis tools: GV. Wrote the paper: L. Marinelli CT.
In a clinical setting, where motor-driven systems are not readily available, the major difficulty in the assessment of the stretch reflex lies in the control of passive limb displacement velocity. A potential approach to this problem arises from the use of manual sinusoidal movements (made by continuous alternating flexions and extensions) paced by an external stimulus. Unfortunately, there are conditions in which sinusoidal movements induce interfering phenomena such as the shortening reaction or postactivation depression. In the present paper, a novel manual method to control the velocity of passive linear movements is described and the results obtained from both healthy subjects and spastic patients are reported. This method is based on the synchronisation of movements with tones played by a metronome at different speeds. In a first set of experiments performed in healthy subjects, we demonstrated consistent control of velocity during passive limb movements using this method. Four joints usually examined during muscle tone assessment were tested: wrist, elbow, knee and ankle joints. Following this, we conducted a longitudinal assessment of the stretch reflex amplitude in wrist flexor muscles in patients with spasticity treated with botulinum toxin type A. The evaluators were not only able to vary the movement velocity based on the metronome speed, but also could reproduce the respective speeds two weeks later, despite the changing degree of hypertonia. This method is easy to perform in a clinical setting and hardware requirements are minimal, making it an attractive and robust procedure for the widespread clinical assessment of reflex hypertonia.
Studies utilizing motor-driven systems that are capable of producing passive limb displacements at predetermined velocities have clearly shown that increased excitability of the stretch reflex (SR) is a primary effector of spasticity
Movements executed by motor-driven devices to elicit the SR can be either sinusoidal or “ramp and hold” displacements (herein referred as “linear movements”). Sinusoidal displacements consist of rhythmic and consecutive flexion and extension movements, while linear movements involve displacement of a limb from one position to another and stopping without a consequent movement.
In a clinical setting where motor-driven systems are not readily available, hypertonia is commonly evaluated by means of subjective clinical scales incapable of differentiating between the reflex and non-reflex components
Without the aid of motor-driven systems, the major difficulty in SR assessment lies in the control of passive limb displacement velocity. A potential approach to this problem arises from the use of manual sinusoidal movements paced by an external stimulus. The evaluator can take advantage of these sinusoidal movements to displace the joint at consistent and accurate velocities, indicating a possible clinical strategy for SR evaluation
In the present paper, a novel manual method to control the velocity of passive linear movements is described and the results obtained from both healthy subjects and spastic patients are reported. This method is based on the synchronisation of the movements with tones played by a metronome at different speeds. In a first set of experiments performed in healthy subjects, we demonstrated consistent control of velocity during passive limb movements using this method. Four joints usually examined during muscle tone assessment were tested: wrist, elbow, knee and ankle joints. In a second set of experiments performed in patients with spasticity, the validity of this method was demonstrated in a longitudinal SR assessment performed in wrist flexor muscles before and after the injection of botulinum toxin type A (BoNT-A). Part of these findings have been presented at the 6th World Congress of Neurorehabilitation
The first set of experiments was performed on 6 right-handed healthy subjects [two women, age: 30±3 years (mean±SD), range: 26–34 years].
The second set of experiments was performed on 12 patients [two women, age 66±9 (mean±SD), range 48–78 years] with post-stroke hemiparesis who were selected according to the following criteria: (1) clinical presentation of a hemispheric stroke leading to unilateral motor deficit at least 9 months prior to participation; (2) CT or MRI documenting a single vascular lesion in the middle cerebral artery territory; (3) a stable clinical picture in the last three months; (4) presence of spasticity in the wrist flexor muscles with a Modified Ashworth Scale (MAS) score ranging from 1 to 3 (1, 1+, 2, 3); (5) no previous treatment with BoNT-A. All subjects gave informed consent according to the Declaration of Helsinki. The participants provided their written informed consent to participate in this study, which was notified to the local ethical committee “Comitato Etico IRCCS Azienda Ospedaliera Universitaria San Martino – IST”, according to the “Determinazione AIFA 20 Marzo 2008” (
Subject | Age/sex | Lesioned hemisphere | BoNT-A dosage | Baseline MAS | Test MAS |
1 | 75 M | L | 30U | 1+ | 0 |
2 | 64 M | R | 70U | 2 | 2 |
3 | 68 M | R | 50U | 2 | 1 |
4 | 48 M | R | 70U | 3 | 2 |
5 | 63 F | R | 50U | 2 | 1+ |
6 | 65 M | L | 50U | 2 | 2 |
7 | 67 M | L | 40U | 2 | 0 |
8 | 78 M | L | 75U | 3 | 1+ |
9 | 71 M | R | 30U | 1 | 1 |
10 | 56 M | R | 40U | 1+ | 1 |
11 | 61 F | L | 50U | 2 | 1 |
12 | 75 M | R | 75U | 3 | 2 |
Passive movements on the six healthy subjects were performed by six evaluators [two male medical doctors, two female physiotherapists and two medical students (one woman); age: 33±10 years (mean±SD), range: 23–50 years], one for each tested subject.
Passive movements on the 12 stroke patients were performed by two evaluators [two medical doctors (one woman); age: 30–35 years], herein referred to as evaluator A and evaluator B. Both evaluators tested all 12 patients. Therefore, each patient was tested twice.
To ensure a standardized level of training, all evaluators were properly instructed in a 15–20 minute training session.
For the kinematic recording of the passive movements and the assessment of the passive range of movement (ROM), we used two independent methods: the optoelectronic system “ELITE” (BTS S.p.a., Milan, Italy) and the Biopac MP100 data acquisition system connected to a TSD130B twin-axis electronic goniometer (Biopac Systems Inc, USA).
The ELITE system consists of six infrared cameras (100 Hz sampling rate) that track the motion of passive reflective markers in 3D coordinates. Synchronised acquisition and data processing was performed using Analyser software (BTS S.p.a.). Three reflective markers (15 mm in diameter) were positioned for each investigated joint: the first on the proximal segment, the second near the joint pivot and the third on the distal segment.
The goniometer was placed across the investigated joint in order to optimally record the angle during the joint displacements; a sampling rate of 2 KHz was used.
EMG activity was recorded by surface electrodes (TSD150B, Biopac Systems Inc, USA) placed over the muscle belly. The signal was acquired by a MP100 unit (Biopac Systems Inc, USA) with a 2 KHz sampling rate and underwent a Blackman −61dB 80–300 Hz band-pass filter for offline processing (AcqKnowledge 3.8.1 software by Biopac Systems Inc, USA). A constant-current stimulator (model DS7A, Digitimer, UK) was used to stimulate the median nerve.
Movement timing was paced with a software emulated metronome.
The method consisted of 4 phases. During all of the phases, the subject was instructed to stay relaxed and to avoid resisting or facilitating the joint movements applied by the evaluator. Upper limb joints were assessed with the subject in a supine position while lower limb joints were evaluated with the subject in a prone position, with the feet off the surface of the examination table.
An audio tone frequency was set on a metronome. For instance, the frequency of 60 beats per minute (BPM) was set. This means that the interval between two consecutive tones (tone interval, TI) corresponded to 1000 ms. Only the evaluator could perceive the tones through earphones.
While trying to cover the entire ROM, the evaluator applied continuous manual sinusoidal joint displacements, moving the distal segment of the joint smoothly. These movements were done at a constant pace so that the distal segment of joint would arrive at the extreme flexed and extended positions in synchrony with consecutive tones. In this way, the evaluator related the motion with the rhythm of the metronome.
When the evaluator had comfortably synchronized the movement with the tones, the sinusoidal movement was stopped at one extreme position (flexed or extended). At this point, the evaluator continued to conceptualize the previously executed movement using first perspective motor imagery
After at least 15 seconds of motor imagery, when the evaluator felt prepared to follow the pace of the metronome accurately, he moved the distal segment of the joint so that it arrived at the opposite position in synchrony with the following tone; then he stopped again (
During phase 4 the evaluator performs a smooth extension movement which starts and ends in synchrony with the metronome tones. The mean velocity is derived from the resulting velocity profile.
Through this procedure, linear movement was derived from sinusoidal movement while retaining the feature of velocity control. This was made possible by the utilization of both an external audio cue as well as first person motor imagery. This procedure was repeated several times, in order to collect several flexion and extension linear movements.
The method was applied to wrist, elbow, knee and ankle joints all examined on the right side. Before the application of the method, the passive ROM of the tested joint was evaluated. In the assessment of wrist, elbow and knee joints, the metronome was set at the following BPM values: 40BPM (TI = 1500 ms), 60BPM (TI = 1000 ms), 120BPM (TI = 500 ms) and 180BPM (TI = 330 ms). Since the ankle is the joint with the smallest ROM, we used the BPM values of 60BPM, 120BPM, 180BPM and 240BPM (TI = 250 ms) in order to achieve joint velocities comparable with the other three joints. At each BPM value, 15 flexion and 15 extension linear movements were collected. Flexion was considered the movement reducing the angle between skeletal members.
The method was applied to the wrist joint of the affected side before (baseline condition) and 15 days after (test condition) the injection of BoNT-A in the
The M-wave was recorded in the FCR through supramaximal electrical stimulation of the median nerve in the cubital fossa.
Wrist flexor muscle hypertonus was clinically rated at baseline condition and test condition according to the MAS. BoNT-A was dosed from 30 to 75 Xeomin® Units, according to the clinically appropriate dose (
Passive ROM evaluated before the application of the method was assessed as the difference between joint angles at the maximal flexed and extended positions.
For each linear movement obtained in phase 4, onset and termination times were calculated. With the ELITE system, these times were identified on the velocity profile, considering the 1% of the peak velocity amplitude as threshold. With the electronic goniometer, they were visually detected, using a display gain of 20°/1cm and a temporal window of 340 ms/1cm.
Onset and termination times were used to calculate the following two parameters: 1)
Since a preliminary analysis showed no difference between the kinematic parameters acquired using the optoelectronic system and the electronic goniometer, only the latter are shown in the results.
The SR amplitude was measured as the mean amplitude of the rectified EMG during the passive displacements. The SR amplitude was divided by the amplitude of the M-wave elicited at baseline condition to obtain SR/M ratio.
In the normal subject group, the comparison between ROM and
In the patients groups
All analyses were considered significant for p<0.05. All the measures of variability are expressed as standard deviation.
ROM | ||||||
|
|
|
|
|
||
|
141±4 | 119±10 | 113±12 | 112±6 | 119±7 | |
|
135±8 | 127±7 | 127±7 | 133±9 | 136±7 | |
|
127±9 | 107±17 | 105±17 | 107±19 | 118±15 | |
|
47±3 | 35±4 | 35±4 | 37±5 | 41±3 |
Mean-V has a direct linear relationship with BPM for all the tested joints (
Wrist: y = 1.8*x+3.8, elbow: y = 1.8*x+1.3, knee: y = 1.3*x+29.5, ankle: y = 7.9*x–15.1.
ROM | |||||
|
|
||||
A | B | A | B | ||
|
125±17 | 110±25 | 113±24 | 113±23 | 111±33 |
|
133±16 | 115±17 | 117±20 | 116±19 | 114±21 |
The filled circles connected with dashed line represent the evaluator A, while the filled squares connected with a solid line represent the evaluator B. Mean-V is higher at 60 BPM without any difference between baseline and test conditions or between evaluator A and B.
In all the patients, passive wrist extension movements evoked a SR on the FCR. No EMG activity was recorded while the muscle was in the shortened position before the passive extension. SR/M ratios were higher during baseline condition compared to the test condition (F [1,44] = 93.1, p<0.0001). No difference was found between evaluators (F [1,44] = 0.03, p = 0.9) and between BPM (F [1,44] = 1.2, p = 0.3) (
The trend toward an increased SR/M at 60 BPM does not reach significance. No difference can be found between the two evaluators.
For all the joints, the
The most important finding was that the mean-V increased linearly with progressive BPM values for all the tested joints. Furthermore, the gap between the
The range of BPM was large enough to produce mean-V values from about 30 to 400°/s, covering all the angular velocities used in a clinical setting
These results prompted us to investigate this method’s efficacy in assessing the SR in patients with spasticity.
In both the baseline and test conditions, the mean-V obtained at 60 BPM was higher than that obtained at 40 BPM. Furthermore, there was no difference in mean-V values between the two evaluators as well as between the two conditions. Therefore, the passive movements obtained with our method demonstrated all the features necessary for a longitudinal SR assessment.
In the case that the treatment with BoNT-A had induced an increase of the
Even if the treatment had resulted in an increased ROM, two strategies would have been used to achieve comparable velocities between baseline and test conditions. The first strategy would have been to predetermine and limit the joint displacement instead of moving throughout the entire ROM. The second strategy would have consisted of decreasing the BPM value during the test condition. Further studies performed in patients with severe spasticity and ROM limitations are needed to test these procedures.
A SR was evoked in each patient during the baseline condition, reflecting the presence of reflex hypertonia in all 12 tested subjects. The SR amplitude was highly variable among the 12 patients as has been reported in the literature
To reduce the inter-subject variability, we divided the SR amplitude by the M-wave amplitude obtaining the SR/M ratio
SR/M ratios did not differ between the two evaluators. Furthermore, SR/M ratios did not differ between the two BPM values. Based on the findings in healthy subjects, we decided to use 40 and 60 BPM values to test the accuracy of our method in producing passive displacements at different velocities. We can now say that a gap of 500 ms between the two tested TIs (40BPM = TI 1500; 60BPM = TI 1000) yielded two appreciably distinct movement velocities. As the difference between the two velocities was small (31°/s), we did not find any difference between the SR/M ratios obtained at the two BPM values. This is consistent with previous results obtained from motor-driven systems
For both the evaluators, the SR/M ratios decreased after BoNT-A treatment. In each patient, the SR/M ratio at test condition was less the 50% of the ratio at baseline condition. Again, these results are corroborated by previous findings obtained with motor-driven systems
The method was consistent at achieving discrete velocities during linear passive movements in both healthy subjects and stroke patients. In the latter, the method was useful in detecting the reflex component of hypertonia as well as the effects induced by BoNT-A treatment.
The method is easy to perform in a clinical setting and the hardware requirements are minimal, making it an attractive and robust procedure for widespread clinical assessment of reflex hypertonia.
We would like to thank William Chiang for the language revision and for the valuable criticism.