Increased H_max:M_max ratio in community walkers poststroke without increase in ankle plantarflexion during walking☆☆☆★★★♢

Article Outline

• Abstract
• Method
  • Subjects
  • H-reflex evoking techniques
  • Knee kinematics
  • Functional walking ability at home and in the community
  • Statistical analysis
• Results
• Discussion
  • H-reflex and ankle control during walking
  • Reliability of Mmax responses
  • Peripheral versus central control
  • Implications for retraining of functional walking ability
• Conclusions
• Acknowledgements
• References
• Copyright

Abstract 

Garrett M, Caulfield B. Increased H_max:M_max ratio in community walkers poststroke without increase in ankle plantarflexion during walking. Arch Phys Med Rehabil 2001;82:1066-72. Objective: To investigate whether changes in H-reflex response at midswing and midstance are related to excessive plantarflexion during walking in community walkers poststroke compared with control subjects without stroke. Design: Survey of functional walking handicap in a random sample of an annual stroke cohort followed by H-reflex and M_max testing of a smaller sample. Setting: Community and laboratory testing. Participants: Forty individuals with stroke (IWS group) completed the functional walking handicap survey, 10 of whom agreed (with 10 age-matched controls) to enroll in a study of of the H_max:M_max ratio in soleus during walking. Intervention: Electromyography during treadmill walking. Main Outcome Measures: Functional Walking Handicap Scale, soleus H_max:M_max ratio, and the ankle joint's angle of displacement. Results: Nine of the 10 stroke patients were community walkers. All had significantly (p < .05) more variable ankle movement during walking than the controls. The H_max:M_max ratio was significantly (p < .01) increased in the IWS group because of a decrease in M_max response without significant (p > .05) increase in H_max response. Conclusions: Individuals with community-level walking ability after stroke have significantly (p < .05) less repeatability of ankle joint movement than controls at both midswing and midstance. Simultaneous soleus H_max and M_max testing showed a significant (p < .01) reduction in the H_max and H_max:M_max ratio at midswing in controls only. This inhibition at midswing was lost by the IWS group without significant increase in H_max, suggesting that central synaptic excitability was within the normal range, and possibly accounting for the absence of excessive ankle plantarflexion during walking in the IWS group with community level walking ability. © 2001 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation

Keywords:  Gait disorders, neurologic, Hemiplegia, H-reflex, Rehabilitation

ANKLE JOINT MOVEMENT disorder arising from hyperactive muscle response is generally considered a major problem in restoring functional walking ability after stroke. One neurophysiologic mechanism¹ postulated to explain hyperactive muscle responses to stretch is increased central synaptic excitability (CSE) from primary (IA) spindle afferents to motoneurons; another is muscle spindle sensitivity.¹ The soleus H-reflex has previously been used to investigate the CSE of spinal motoneurons in normal subjects² and in patients with diseases such as stroke³, ⁴, ⁵ and spinal cord injury⁶, ⁷ (SCI) under static (lying, sitting) conditions. However, a preliminary investigation⁸ of stroke patients during walking showed that an increase in H_max:M_max ratio was not associated with increased plantarflexion. Keenan et al⁹ suggested that poor motor control rather than spasticity was the major factor affecting ambulation and balance of persons with stroke. In the present investigation, we used the H-reflex to describe changes in excitability at the central synapse between the IA spindle afferents and motoneurons of persons with stroke during walking, comparing the results with those of age-matched controls.

The H-reflex is monosynaptic and is generally considered to be the electric analog of the stretch reflex.¹ It is evoked by electric stimulation of lowest threshold muscle spindle afferents in the tibial nerve at the knee at a latency of about 30ms. At higher intensity stimulation, an M response is recorded at a shorter latency of about 5ms. The ratio of maximal H-reflex amplitude to maximal M-response amplitude is thought to represent the number of motoneurons recruited through the monosynaptic reflex as a proportion of the motoneuron pool.¹⁰ Angel and Hofmann³ found an increase in the H_max:M_max ratio in stroke patients in static conditions. Increased H_max:M_max ratio was also found in static spinal cord-injured patients.⁷ Little and Halar⁷ recorded how increased H-response amplitude and H_max:M_max ratio developed along with spasticity in 6 SCI patients during months 1 to 3 postinjury. In recent years, the H-reflex has been examined during walking in healthy subjects¹⁰, ¹¹, ¹² and in patients with spinal cord lesions.¹³ The increase in H-reflex amplitude in the ankle plantarflexors of SCI patients during walking (found by Yang et al¹³) is in agreement with earlier studies³, ⁴, ⁵ of stroke patients who were in static (lying) condition. Yang¹³ concluded that an association may exist between increased H-response amplitude and hyperactive electromyographic responses to stretch of the soleus muscle during walking in spinal cord-injured patients. However, the precise effect of such a change in the H-reflex on ankle joint movement has not been objectively described. Because ambulatory ability is likely to vary with degree of impairment of ankle joint movement in stroke, determining the character and extent of impairment and categorizing patients accordingly is fundamental to understanding the functional consequences of H-reflex change. In the present study, we categorized functional walking ability of the patient cohort by using the method described by Perry et al.¹⁴

The present study was undertaken to determine the relation between change in H-reflex response and ankle movement disorder at 2 points of the walking cycle in stroke patients categorized by this walking ability. We evaluated patients at midswing, when foot clearance is essential to achieve a new forward base, and at midstance, when ankle dorsiflexion is essential for the body to progress forward over the supporting foot. The null hypothesis was that changes in ankle plantarflexors' H-reflex would not be directly associated with changes in angular displacement at the ankle joint at either midswing or midstance.

Back to Article Outline

Method 

Subjects 

By using a restricted, randomized procedure, we selected 77 patients from an annual cohort of 168 stroke patients discharged from the Mater Hospital, Dublin. The inclusion criteria were: patient must be on the hospital medical records computer database with an International Classification of Diseases (ICD)¹⁵ code of 436 (which includes cerebrovascular accident [CVA]), and at least 3 months must have elapsed since the CVA. All patients were contacted 1-year postdischarge. Thirty patients had died since discharge (1-yr mortality rate, 43%) and 7 could not be traced. The remaining 40 patients (IWS group) were contacted by letter and by telephone. All consented to participate in a preliminary assessment of functional walking ability¹⁴ performed during a home visit. All were asked to participate in H-reflex testing in the clinical movement analysis laboratory, 10 of whom (5 men, 5 women; age range, 55-80yr) agreed to participate in the study. Ten healthy elderly persons (8 men, 2 women; age range, 55-80yr), served as controls. Ethics committee approval and informed consent were obtained in all cases.

H-reflex evoking techniques 

Soleus muscle response to electric stimulus during walking in stroke patients compared with healthy elderly subjects was the basis of the present investigation, by using the method previously described.¹⁶ Subjects were placed in the prone position. The popliteal crease was marked and the nerve to the soleus muscle was identified by applying a 1-cm cathodal disk electrode at 1-cm intervals from the medial border of the popliteal fossa. Elicitation of an H response, at a latency of about 30ms, without an M response identified the optimal stimulating position.² In preparation for the walking trial, the hand-held stimulating electrode was replaced by a spring-loaded saline pad electrode, supported in a plastic housing, which was secured by thick elastic bands to the posterior aspect of the subject's knee. The track of the bands on the skin was protected with padded adhesive gauze. A 50-cm² anodal disk electrode was placed on the posterior aspect of the thigh, 100mm proximal to the cathode. A grounding electrode (100cm²) was positioned on the contralateral thigh. Prior electromyography testing¹⁶ showed that this placement reduced the stimulus artifact comparably to that obtained by placing the ground between the stimulating and recording electrodes, and it reduced the instruments required on the test leg. In the stroke group, we tested the soleus muscle of the affected leg and, in the control group, we tested the right soleus muscle.

The recording surface Ag/AgCl electrodes, 1-cm diameter, were placed 0.8cm apart, 3cm above the musculotendinous junction of the soleus muscle in the midline of the calf.¹⁷ Stimulation duration was 0.5ms with a rate greater than 0.2Hz. A custom device¹⁸ was used to process input from footswitches on subjects' heels and toes to time delivery of the stimulus to soleus at midstance and midswing. A full H-reflex recruitment curve in standing, and at midstance and midswing, was obtained by using the method first described by Garrett et al¹⁰ based on that described by Hugon² in the seated subject. Stimuli were applied at ≥ 5-ms intervals during a period of treadmill walking until a full H-reflex recruitment curve was obtained. The maximal M response was determined by increasing stimulus intensity to the point at which no further increase in direct motor response was obtained and the H response was abolished. Five H-reflexes were elicited and recorded at this stimulus intensity. The stimulus intensity was then decreased by 10% and 5 more H-reflexes were elicited and recorded. This procedure was repeated until no H and no M responses were present. The stimulus site chosen provided repeatable waveforms at midswing and midstance, indicating that responses were obtained from the same motoneurons throughout the test. Maximal M and H responses obtained during the different experimental conditions were monitored and checked for similarity in configuration. Similarity of responses throughout each testing session indicated that, in each experimental condition, the electromyographic responses recorded were from the same motor units. Patients and controls were given ample rests between tests to minimize the effects of fatigue.

For each subject we calculated an amplitude ratio (H_max:M_max ratio) defined as the amplitude of the maximal H response as a ratio of the maximal M response. This ratio abolished differences arising from intrasubject variation in electric resistance. The maximal M response represents the total motor response of the soleus muscle.² Constancy of the maximal M response showed that the stimulating electrode retained a constant relationship with the nerve to soleus during walking. All values in the present article are mean ± standard deviation (SD).

The investigation started with a period of habituation to treadmill walking as previously described.¹⁹ During the walking trial period of typically 1.5 hours, the H-reflex at midstance of treadmill walking was elicited first. The foot-switch signal was used to time delivery of a stimulus at midstance or midswing. We obtained a full H-reflex recruitment curve while simultaneously recording ankle joint movements.

Knee kinematics 

Before testing, all subjects were habituated to walking on the treadmill, wearing their customary walking shoes.¹⁹ A comfortable walking speed was identified for each patient. The controls walked within the range identified by Perry²⁰ as normal (men's norm, 1.3-1.5m/s; women's, 1.2-1.4m/s). The stroke patients' mean walking velocity was .75 ± .17m/s. Ankle joint angular displacement was measured (resolution ± .22°) by electrogoniometer^a and was relayed to the personal computer through the third multiplexor channel. Zero position of the test ankle was set for each subject as that of normal bipedal standing, generally accepted as 3° to 5° of dorsiflexion.²⁰

Functional walking ability at home and in the community 

The level of functional walking ability of each person in the stroke group (n = 40) was evaluated by means of the walking handicap scale developed by Perry et al.¹⁴ Patients were assigned to 1 of 6 grades based on their level of functional walking ability at home and in the community. These comprised 3 categories of community-level walking ability (unlimited, least limited, most limited), and 3 categories of household walking ability (limited, unlimited, physiologic). Patients were evaluated on a 0 to 4 scale based on their ability to ambulate in the following areas: bedroom, bathroom, entering and exiting home, curbs, local grocery store, uncrowded shopping center, and crowded shopping center (scale: 0 = unable, 1 = wheelchair, 2 = assisted, 3 = supervised, 4 = independent). Assignment to a functional walking ability classification was achieved by evaluation of these scores by a computer-based matrix calculation.¹⁴

Statistical analysis 

The analyses were performed separately on the stroke patients and on the controls. Descriptive analyses were performed to quantify the differences between the groups. Assumptions of normality and homogeneity of variance were met and allowed the use of parametric tests. Two-way analysis of variance (ANOVA) was used to identify the effect of stroke on H_max:M_max ratio and on joint angular displacement at midswing and midstance. The significance level was set at .05.

Back to Article Outline

Results 

Functional walking ability of the IWS group is outlined in table 1.

Table 1: Functional Walking Ability of the IWS Group (n = 40)

Twenty-seven patients (67%) had attained some degree of functional walking ability in the community. Thirteen patients (33%) were household walkers. Four of these were physiologic walkers without functional walking ability who walked for therapeutic exercise only.

H-reflexes were measured during walking in the 10 stroke patients and 10 age-matched controls, and all data were used in the analysis. Nine of the stroke patients were community walkers with functional walking ability adequate for independent unsupervised mobility in high-challenge community activities such as visiting a crowded shopping center (table 2).

Table 2: Functional Walking Handicap Scale Scores for the IWS Group (n = 10)

Abbreviations: I, independent; S, supervised; A, assisted.

Their mean walking velocity was .77 ± .15m/s. One stroke patient was a least-limited household walker with a walking velocity of 0.5m/s. He required supervision while walking in some household activities and when entering and exiting the home and mounting and dismounting a sidewalk. He required the assistance of 1 person when walking in community locations.

We found no significant between-group difference between the mean angular position at the ankle joint at midstance or midswing (IWS group, 2.9° ± 4.2° and 3.5° ± 5°, respectively; control group, 0.1° ± 6.9° and 1.3° ± 3.4° respectively). Substantial intersubject differences (table 3) existed between the mean angular displacement at the ankle joint in the controls.

Table 3: Angular Displacement (deg) at the Ankle Joint Over 5 Random Steps During the Walking Cycle

NOTE. Range is the difference between the maximal and minimal angular positions at the ankle joint over 5 random steps at midstance and at midswing for both Control and IWS Group subjects.

However, the angular displacement of each successive step in any control appeared highly reproducible. This was not the case in the IWS group. The mean range of angular displacement (table 4), obtained by calculating the average difference between the maximal and minimal angular position at the ankle joint over 5 random steps at midstance and at midswing for each individual in both groups, showed that the ankle joint position was significantly (p < .05) more variable in the IWS group. In the control group, an individual's ankle joint position at either midswing or midstance was repeated with an accuracy of 3° in succeeding walking cycles (table 4). In the IWS group, the ankle position varied within a range of 8° at midswing and 9° at midstance.

Table 4: Reproducibility of Angular Displacement (deg) at the Ankle Joint During the Walking Cycle in the IWS Group (n = 10) and Age-Matched Controls (n = 10)

NOTE. Values are group means ± SDs of the range of angular position at the ankle joint.

Abbreviation: NS, not significant.

The mean ± SD for absolute amplitude of maximal H and maximal M responses and H_max:M_max ratios are in table 5. Examples of typical M_max and H_max responses at midswing and midstance are shown in figures 1A (IWS group) and 1B (control group).

• 
  • View Large Image
  • Download to PowerPoint

• Fig. 1. 

  Maximal M and H responses obtained at (A) midstance and (B) midswing by an IWS group stroke patient and an age-matched control subject.

Two-way ANOVA showed that the largest difference in the H_max:M_max ratio between the stroke patients and their age-matched controls occurred at midswing. The highly significant (p < .0001) reduction to .12 in the controls, compared with .48 in the IWS group, was caused by a marked (p < .001) reduction in the peak-to-peak amplitude of the maximal H-reflex at midswing in the age-matched controls (0.7 ± 0.4). It was not because of increased H_max amplitude in the IWS group (see table 5). The H_max amplitudes (table 5) in the IWS group at midstance, midswing, and standing (2.8 ± 0.9, 2.1 ± 1.0, 2.2 ± 1.0, respectively) were not significantly different from those observed at midstance and standing in the control group (2.2 ± 0.8 and 1.5 ± 0.8). Despite this, the H_max:M_max ratio at midstance in soleus motoneurons in the IWS group was .67 ± .17 (p < .01) greater than that observed in the control group. However, the M_max amplitude at midstance was significantly lower in the IWS group compared with the controls. At midstance, this barely significant decrease in M response appeared to be the major factor leading to the significant increase in the H_max:M_max ratio.

Table 5: Changes in Maximal M Wave, Maximal H Wave, and H_max:M_max Ratio at Midstance, Midswing, and Standing in the IWS Group and Age-Matched Controls

NOTE. Values are the group means ± SDs.

We found no association between joint angular position and H_max:M_max ratios at midswing in the control group (r = .04) or in the IWS group (r = .02). A slight association (r = .32) was observed between joint repeatability and H_max:M_max ratio at midstance in the control subjects. This association did not occur in the IWS group. The coefficient of determination (r²) suggests that approximately 10% of the repeatability observed at midstance in control subjects without stroke could be attributed to the ratio between H_max and M_max. A correlation coefficient of r equal to 0.4 indicated a modest association between the maximal M response and variability of ankle joint displacement at midstance in the IWS group that was not apparent in the controls. This finding suggests that between 12% to 15% of the loss of joint angular repeatability at midstance in the IWS group could be attributed to change in the maximal M response. A similar relationship was seen between maximal H response and joint angular variability in the IWS group.

Back to Article Outline

Discussion 

The patients included in the present investigation were drawn from a sample of an annual cohort of stroke patients whose level of functional walking ability was known. All were at least 1-year poststroke, by which time their medical condition was considered to have stabilized. All but one had full community-walking ability. Although it might be suggested that the number of subjects is low, the number tested represents a response rate of 25% of the sample group of the annual cohort of stroke patients at our institution. Furthermore, this is the largest group of stroke patients of whom we are aware whose members have undergone H_max and M_max testing during walking. Because all but 1 participant were community-level walkers, the results are not readily generalizable to all persons with chronic stroke. Although our subjects had a high level of walking function for stroke survivors, their achievement did not represent normal walking. The group mean velocity of .77 ± .15m/s was approximately half normal walking velocity. Our findings suggest that—to the extent that the behavior of the H-reflex represents that of the monosynaptic reflex during walking—reflex modulation plays a subordinate role in control of joint angular position during normal walking. Its contribution appeared to be function-related control of the proportion of motoneurons reflexively recruited. During the stance phase a greater proportion of motoneurons (approximately 50%) were available. During swing, this proportion dropped to approximately 12%. Our stroke subjects with community-level walking ability lost the reduction in CSE seen at midswing in normals. They did not, however, appear to sustain an absolute increase in CSE at midswing or midstance sufficient to cause hyperactive muscle responses with excessive ankle plantarflexion.

H-reflex and ankle control during walking 

We quantified the function-related changes in the input-output properties of the central synapse. The increase in H_max:M_max ratio to approximately 40% of the motoneuron pool at midstance (see table 5) indicated an increase in the input-output property of the soleus motoneuron pool at midstance appropriate to the increased loading on the ankle at midstance. The reduction in H_max:M_max ratio, observed in age-matched controls at midswing and previously documented in young adults,¹⁰, ¹⁶ indicates that the number of motoneurons available for recruitment through the monosynaptic reflex was reduced. The reduction in potential force of the reflex contraction to approximately 10% of the motoneuron pool reflects the unloading of the ankle and minimizes the possibility of reflexive ankle plantarflexion at midswing. This finding, which suggests a control mechanism that offers availability of the spinal reflex in leg extensors at midstance but seriously curtails its availability at midswing, is supported by earlier results²¹, ²², ²³ showing a polysynaptic rather than a spinal reflex response to momentary resistance during swing phase of walking. In control subjects without stroke, we found a very modest correlation (r = .32) between the H_max:M_max ratio and the repeatability of joint-angle position at midstance but not at midswing.

In the present study, we found that in control subjects the reflex modulation during walking was lost and no association existed between the H_max:M_max ratios and joint-angle repeatability at either midstance or midswing. The components of the H-reflex appeared to be released from central control with associated adverse effects on angle repeatability at midstance. The increase in the H_max response at midstance to 2.8 ± 0.9mV did not achieve significance (p < .07) compared with control subjects without stroke; it was, nevertheless, found to have a modest correlation (r = 0.3) with loss of joint-angle repeatability. The associated increase in CSE may account for 12% of the variability in ankle joint position. The barely significant decrement in M_max response at midstance had a modest correlation (r = 0.4) with loss of repeatability of ankle angle at midstance in the IWS group. This association indicates that decrement in the M_max response at midstance may contribute 15% to the loss of repeatability. Further studies with more subjects are required to support these correlations.

At midswing, the H_max response in the IWS group lost the reduction observed in control subjects and was not significantly different from values observed in standing (see table 5). Although this loss of inhibition was not related to joint changes, it does provide CSE-related explanation for the finding of Faist et al²⁴ that the magnitude of the knee's extensor reflex response to tendon tap during swing was approximately the same as that found during standing. The increase in CSE seen in spinal cord-injured patients¹³ did not occur in the present study's IWS group. Increased CSE cannot, therefore, account for soleus hypertonicity with an associated tendency for ankle plantarflexion during walking in the stroke population we tested. Increased CSE in the ankle dorsiflexors leading to cocontraction at the ankle is unlikely, considering that (1) CSE is not raised in the soleus reflex loop and (2) neither clinical nor electromyographic studies of stroke patients have shown ankle dorsiflexor spasticity as a characteristic impairment in walking disablement.²⁰, ²⁵ Furthermore, the prior population sampling suggests that the functional walking abilities of the 10 stroke patients in the IWS group are reasonably representative of those experienced by 1-year survivors of stroke. However, the stroke subjects we tested were those who had the highest present daily mobility. It is likely that values of H_max and H_max:M_max ratios may be critical in IWS group patients who are household ambulators.

Reliability of M_max responses 

In the present study, the decrease in M_max responses in the IWS group was not from electrode shift during walking, because the values for M_max response amplitudes are virtually identical (midstance, 4.2 ± 1.0mV; midswing, 4.4 ± 1.1mV; standing 4.3 ± 1.1mV). Similar consistency of M_max response was seen in the control subjects. Also, the control group's H_max:M_max ratios (midstance, 41 ± .14; midswing, .12 ± .06) were close to those found in a previous investigation¹⁶ in healthy young adults (midstance, .47 ± .11; midswing, .13 ± .11). The use of voltage-controlled stimuli in the interests of comfort in the elderly frail subject evidently does not affect the H_max:M_max ratios obtained. These observations may be useful in future investigations of the input-output properties of the motoneuron pools during walking in persons with stroke and their age-matched controls.

Yang et al²⁶ showed that changes in H-response amplitude in control subjects without stroke during walking are not governed by background electromyography. The findings of the present investigation show that this is also the case in persons with stroke. Electric silence of the soleus muscle was confirmed by oscilloscope monitoring during the acquisition of standing H_max responses. The H_max responses at midstance and midswing in the IWS group, and at midstance in age-matched controls, showed no significant difference compared with standing responses (see table 5).

Peripheral versus central control 

Mechanisms postulated to explain changes in CSE include decreased pre- or postsynaptic inhibition, collateral sprouting, and synaptogenesis by spared afferents. Reduction of presynaptic inhibition, which was previously shown in the stationary stroke patient,²⁷, ²⁸, ²⁹ most likely contributes to the loss of modulation seen in persons with stroke. Recent work¹⁵ has shown that the significant reduction in both the H_max amplitude at midswing and the H_max:M_max ratio in control subjects without stroke cannot be accounted for by lost peripheral afferent input from the mechanoreceptors of the leg extensor after reduction in knee-angle velocity.³⁰, ³¹

Little and Halar⁷ attributed the increase in H_max response in spinal cord-injured patients to collateral sprouting and synaptogenesis by spared afferents. Although branching of corticospinal axons occurs after damage to the central nervous system (CNS), this phenomenon declines with age. This has been shown by Carr et al³² who found evidence of corticospinal reorganization in persons with cerebral palsy hemiplegia when their CNS damage was prenatal but not when it had occurred postnatally. We therefore believe that this mechanism is unlikely to account for the changes observed in the IWS group patients in the present study.

The reduction in the M response may be from preferential disuse atrophy of muscle fibers belonging to the largest motor units (type II atrophy). A previous investigation¹⁴ showed that reduced power of the leg muscles is a critical factor in determining level of functional walking ability in persons with stroke. Failure of these motor units to participate in the H-reflex because of their high threshold would result in a lower M-response amplitude but a higher H_max:M_max ratio. However, no differences were seen in the number or proportion of soleus muscle fiber types in rats after surgical occlusion of the middle cerebral artery.³³ Somewhat surprisingly, human studies of the enzyme-histochemical and morphologic characteristics of the soleus after stroke have not been performed.

It may be that descending tracts exert a differential effect on low- and high-threshold motoneurons, thereby increasing the functional thresholds of some high-threshold motoneurons. This occurrence would effectively reduce the motoneuron pool and the M_max response. Afferent input in stroke patients would then excite the same number of motoneurons as in healthy subjects, making their maximal H response the same as healthy patients' and would result in an increased H_max:M_max ratio. This explanation fits the concept of a raised motoneuronal threshold consequent to damage to corticospinal tracts as a result of pathologic changes associated with stroke.

Implications for retraining of functional walking ability 

The findings of the present study suggest that impaired muscle function in 1-year survivors of stroke cannot be accounted for simply in terms of spasticity. Disorders in repeatability of joint movement and in control of motor force may affect the level of functional walking ability a person achieves after stroke. Rosenfalck and Andreassen³⁴ showed impaired regulation of force and motor unit firing patterns in patients with pathology of the CNS, including stroke. For clinicians to be cautious about the effects of strength training for spastic stroke patients is justified.³⁵ In the present study, even a modest increase in the maximal H response appears to influence the regulation of joint-angle position. However, the findings also suggest that increased CSE may not be a characteristic of stroke patients who have community-walking ability. The success of therapeutic exercise programs in improving muscle strength without increasing spasticity suggests that, for some stroke patients, this approach is effective and safe. Further studies are necessary to understand whether H-reflex testing during walking could identify patients with increased CSE and to determine if these patients are at risk for increasing their spasticity as a result of strength training programs.

Back to Article Outline

Conclusions 

Changes in the control group's H_max:M_max ratio suggest that the reflex modulation occurs in the soleus muscle at midswing, rather than at midstance. The proportion of the motoneuron pool available for reflex contraction is decreased during midswing and is increased at midstance while maintaining a reserve of 60% of the motoneuronal pool throughout the cycle. The IWS group patients, who were competent community walkers and at least 1 year poststroke, had lost this modulation and showed an increase in H_max:M_max ratio at midswing and midstance. However, because they did not have the significant CSE-related increase in H_max response found in SCI patients, it is logical that they did not exhibit the ankle plantarflexion generally considered characteristic of stroke.¹² Hypertonicity arising from hyperactive stretch reflexes cannot account for impaired muscle function during walking at all levels of functional walking ability poststroke. The joint disturbance observed in patients with a good level of functional return was loss of repeatability of ankle flexion and extension at midswing and midstance. The H-reflex changes do not contribute dominantly to the changes observed in joint-angle displacement poststroke in patients who have good return of walking ability. However, the association between loss of repeatability of joint movement with the very modest increase in H_max and decrease in M_max suggests that changes in the H-reflex components may contribute powerfully to disturbed joint movement during walking in the less-competent walker. Further studies into the magnitude of the change in the H-reflex in poststroke in persons with a lower level of functional walking ability are likely to reveal a greater disturbance of reflex function.

Back to Article Outline

Acknowledgements 

The authors thank the physicians of the Mater Misericordiae Hospital, Dublin, for allowing access to their patients.

Back to Article Outline

References 

1. Britton TC. Abnormalities of muscle tone and movement. In:  Stoke M editors. Neurological Physiotherapy. London (UK): : Mosby; 1998;p. 57–65
   • View In Article
2. Hugon M. Methodology of the Hoffmann Reflex in man. In:  Desmedt JE editors. New developments in electromyography and clinical neurophysiology. Vol 3:Basel: : Karger; 1973;p. 277–293
   • View In Article
3. Angel RW, Hofmann WW. The H-reflex in normal, spastic, and rigid subjects. Arch Neurol. 1963;8:591–596
   • View In Article
4. Garcia-Mullin R, Mayer RF. H reflexes in acute and chronic hemiplegia. Brain. 1972;95:559–572
   • View In Article
   • MEDLINE
5. Somerville J, Ashby P. Stroke spasticity: neurophysiologic studies. Arch Phys Med Rehabil. 1978;59:592–596
   • View In Article
   • MEDLINE
6. Ashby P, Verrier M, Lightfoot E. Segmental reflex pathways in spinal shock and spinal spasticity in man. J Neurol Neurosurg Psychiatry. 1974;37:1352–1360
   • View In Article
   • MEDLINE
   • CrossRef
7. Little JW, Halar EM. H-reflex changes following spinal cord injury. Arch Phys Med Rehabil. 1985;66:19–22
   • View In Article
   • MEDLINE
8. Garrett M, Shakoor SB. Changes in the excitability of the Hoffman reflex during walking in the hemiplegic man. [abstract] J Physiol. 1991;432:29P
   • View In Article
9. Keenan MA, Perry J, Jordan C. Factors affecting balance and ambulation after stroke. Clin Orthop. 1984;182:165–171
   • View In Article
10. Garrett M, Luckwill RG, McAleer JJ. The sensitivity of the monosynaptic reflex arc in the leg extensor muscles of the walking man. In:  Morecki A,  Fidelus K,  Krzysztof K,  Wit A editor. International series on biomechanics. Biomechanics VIIa. Vol 3a:Baltimore (MD): : University Park Pr; 1981;p. 95–99
    • View In Article
11. Crenna P, Frigo C. Excitability of the soleus H-reflex arc during walking and stepping in man. Exp Brain Res. 1987;66:49–60
    • View In Article
    • MEDLINE
12. Capaday C, Stein RB. Amplitude modulation of the soleus H-reflex in the human during walking and standing. J Neurosci. 1986;6:1308–1313
    • View In Article
    • MEDLINE
13. Yang JF, Fung J, Edamura M, Blunt M, Stein RB, Barbeau H. H-reflex modulation during walking in spastic paretic subjects. Can J Neurol Sci. 1991;18:443–452
    • View In Article
    • MEDLINE
14. Perry J, Garrett M, Gronley JK, Mulroy SJ. Classification of walking handicap in the stroke population. Stroke. 1995;26:982–989
    • View In Article
    • MEDLINE
    • CrossRef
15. World Health Organization (WHO) . International classification of disease. Geneva: : WHO; 1975;
    • View In Article
16. Garrett M, Kerr T, Caulfield B. Phase-dependent inhibition of H-reflexes during walking is independent of reduction in knee angular velocity. J Neurophysiol. 1999;82:747–753
    • View In Article
    • MEDLINE
17. Maryniak O, Yarowski R. H-reflex: optimum location of recording electrodes. Arch Phys Med Rehabil. 1987;68:798–802
    • View In Article
    • MEDLINE
18. Murray CD, Delaney JD, de Paor AM, Reilly RB. Stimulation, capture and display for the Hoffmann reflex. Innov Technol Biol Med. 1992;13:641–652
    • View In Article
19. Garrett M, van Hemelrijk B, Caulfield B. Establishment of steady-state treadmill walking in hemiplegics. [abstract] J Physiol (Lond). 1997;501:43P
    • View In Article
20. Perry J. Gait analysis: normal and pathological function. Thorofare (NJ): : Slack; 1992;
    • View In Article
21. Eng JJ, Winter DA, Patla AE. Strategies for recovery from a trip in early and late swing during human walking. Exp Brain Res. 1994;90:404–414
    • View In Article
    • MEDLINE
22. Dietz V, Quintern J, Berger W. Obstruction of the swing phase during gait: phase-dependent bilateral muscle co-ordination. Brain Res. 1986;384:166–169
    • View In Article
    • MEDLINE
    • CrossRef
23. Garrett M, Luckwill RG. Role of reflex responses of knee musculature during the swing phase of walking in man. Eur J Appl Physiol. 1983;52:36–41
    • View In Article
24. Faist M, Ertel M, Berger W, Dietz V. Impaired modulation of quadriceps tendon jerk reflex during spastic gait: differences between spinal and cerebral lesions. Brain. 1999;122:567–579
    • View In Article
    • CrossRef
25. Knutsson E, Richards C. Different types of disturbed motor control of gait in hemiparetic patients. Brain. 1979;102:405–430
    • View In Article
    • MEDLINE
26. Yang JF, Stein RB, James KB. Contribution of peripheral afferents to the activation of the soleus muscle during walking in humans. Exp Brain Res. 1991;87:679–687
    • View In Article
    • MEDLINE
27. Delwaide PJ. Human monosynaptic reflexes and presynaptic inhibition. An interpretation of spastic hyperreflexia. In:  Desmedt JE editors. New developments in electromyography and clinical neurophysiology. Vol 3:Basel: : Karger; 1973;p. 508–522
    • View In Article
28. Milanov I. A comparative study of methods for estimation of presynaptic inhibition. J Neurol. 1992;239:287–292
    • View In Article
    • MEDLINE
29. Harburn KL, Vandervoort AA, Helewa A, Goldsmith CH, Kertesz A, Teasell RW, et al.  A reflex technique to measure presynaptic inhibition in cerebral stroke. Electromyogr Clin Neurophysiol. 1995;35:149–163
    • View In Article
    • MEDLINE
30. Brooke JD, Cheng J, Misiaszek JE, Lafferty K. Amplitude modulation of the soleus H-reflex in the human during active and passive stepping movements. J Neurophysiol. 1995;73:102–111
    • View In Article
    • MEDLINE
31. Brooke JD, Cheng J, Collins DF, McIlroy WE, Misiaszek JE, Staines WR. Sensori-afferent conditioning with leg movement: gain control in spinal reflex and ascending paths. Prog Neurobiol. 1997;51:393–421
    • View In Article
    • MEDLINE
    • CrossRef
32. Carr LJ, Harrison LM, Evans AL, Stephens JA. Patterns of central motor reorganization in hemiplegic cerebral palsy. Brain. 1993;116:1223–1247
    • View In Article
33. Ansved T, Ohlsson AL, Jakobsson F, Johansson BB. Enzyme-histochemical and morphological characteristics of fast-twitch and slow-twitch skeletal muscle after brain infarction in the rat. J Neurol Sci. 1996;144:14–20
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (1309 KB)
    • CrossRef
34. Rosenfalck A, Andreassen S. Impaired regulation of force and firing pattern of single motor units in patients with spasticity. J Neurol Neurosurg Psychiatry. 1980;43:907–916
    • View In Article
    • MEDLINE
    • CrossRef
35. Teixeira-Salmela LF, Olney SJ, Nadeau S, Brouwer B. Muscle strengthening and physical conditioning to reduce impairment and disability in chronic stroke survivors. Arch Phys Med Rehabil. 1999;80:1211–1218
    • View In Article
    • Abstract
    • Full-Text PDF (1203 KB)
    • CrossRef