| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
MS 11267 Received 19 June 2000; accepted after revision 31 October 2000.
 |
ABSTRACT |
|---|
 TopAbstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sensory feedback from the contracting muscles has been demonstrated to contribute to the muscle activation during walking both in the cat (Andersson et al. 1981; Grillner & Zangger, 1984; Hiebert & Pearson, 1999) and in man (Sinkjær et al. 2000). In addition, sudden perturbations of the ankle joint may lead to reflex activation of the muscles, which is greatly modulated during the walking cycle (Yang et al. 1991; Sinkjær et al. 1996). It has recently been suggested that the main role of the stretch reflex activation is to ensure the stability of the supporting leg, since stretch of the ankle plantar flexors evokes large responses in the stance phase of walking, but has no or only a minor effect in the swing phase (Zehr & Stein, 1999). However, this idea is based solely on observations regarding soleus stretch reflexes - and the modulation of stretch reflexes in this muscle could easily be explained by the fact that the muscle is active in stance and silent in swing. The hypothesis would therefore be strengthened if it were observed that stretch reflex activity is also largest in the ankle dorsiflexors in the stance phase, although these muscles are silent at that time. The first purpose of the present study was to investigate this by applying stretches to the ankle dorsiflexors at different times in the walking cycle by a portable stretching device (Andersen & Sinkjær, 1995) and to record the evoked stretch reflex activity in the tibialis anterior (TA) muscle.
Since muscle stretch generally evokes several different reflex bursts, which are believed to be mediated by different central mechanisms (Petersen et al. 1998), the second purpose of the study was to provide evidence about the nature of the responses observed during walking.
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
General experimental set-up
The experiments were performed on 17 healthy subjects (aged 23-40 years), some of whom participated for two or more sessions. According to the Helsinki declaration, all subjects gave informed written consent to the experimental procedure, which was approved by the local ethics committee.
The main part of the experiment was performed during walking on a treadmill at a speed of 3·5-4·0 km h-1 according to the subjects' preferred speed. For all subjects, a stretching device consisting of a mechanical joint mounted on a level with the axis of rotation in the ankle joint (for details, see Andersen & Sinkjær, 1995), was placed on the left leg in order to stretch the ankle dorsiflexors at different times throughout the walking cycle. The muscle activity was measured with bipolar Ag-AgCl electrodes (1 cm2, 1 cm inter-electrode distance) placed on the tibialis anterior (TA) and soleus muscle (SOL). The electromyograms (EMGs) were amplified, filtered from 20 to 1000 Hz (DISA, model 15C01) and sampled together with the force and positional records (for details, see Andersen & Sinkjær, 1995). A heel contact placed in the shoe of the left leg was used to identify the onset of the stance phase.
Stretch reflexes
Stretching the TA was done by imposing a quick plantarflexion with an amplitude of 8 deg, a hold phase of 100-200 ms, and velocities ranging from 100 to 500 deg s-1. The stretch velocity was measured as the velocity over the first 20 ms after stretch onset. The walking cycle was divided into four to six different time points. The points were defined in relation to the heel contact (HC) and occurred in early stance (
HC + 100 ms), mid stance (HC + 400 ms), early swing (HC + 800 ms) and mid swing (HC + 1000 ms). In a few subjects, points later in both stance and swing were examined, but in general this was not possible because of a built-in security stop preventing the stretch device from delivering stretches when too close to the limits of the ankle position. Stretches with identical stretch amplitude and velocity were applied throughout the walking-cycle in 11 subjects plus in early stance and early swing only for six additional subjects. In eight of the subjects, similar stretches were applied during tonic dorsiflexion with the same set-up as during walking. The subjects were sitting and performed the contraction against a lever.
Transcranial stimulation
Transcranial magnetic stimulation (TMS; Magstim 200, Magstim Company Ltd, UK) was combined with stretch of the TA in early stance and early swing. First, the optimal position of the coil (a figure-of-eight prototype coil; diameter of the wing 9 cm) for eliciting a response in TA was found with the subject standing relaxed. In general, the optimal position was 1-2 cm lateral to the vertex contralateral to the stimulated leg (right hemisphere). The coil was then kept in this position throughout the rest of the experiment by way of a specially built harness (see Schubert et al. 1997). The threshold for a motor-evoked potential (MEP) in the muscle was found and a stimulation of 1·2-1·3 × MEP threshold was then used. The size of the resulting MEP was adjusted to the same size in both stance and swing phase, which occasionally required a slightly increased stimulation strength in the stance phase, probably because of low, or absent, muscle activity in this phase. Finally, TMS was conditioned by a stretch at conditioning-test intervals from 20 ms to 120 ms. The following alternatives were used: six different conditioning-test intervals with (1) combined stretch and TMS, (2) TMS alone, (3) stretch alone and (4) background EMG alone. These were randomized and in general 10-20 trials of each alternative were recorded.
In some experiments transcranial magnetic and electrical stimulation (TES) were compared. For the electrical stimulation, the cathode was placed 5 cm in front of, and the anode 2 cm lateral to, the vertex (Nielsen et al. 1995). A digitimer D180A stimulator (Digitimer Co., Ltd, Welwyn Garden City, UK) was used. The intensity was adjusted in the same way as for the magnetic stimulation, and to a size comparable with the size of the MEP from the magnetic stimulation. An increase of the stimulation intensity in the stance phase was especially necessary for TES in order to obtain this.
Ischaemia
Ischaemia of the lower leg was produced in five subjects by placing a cuff around the thigh just above the knee, inflating it and waiting in general 20 min for the effect. The effect of the ischaemia was checked by electrical stimulations to the posterior tibial nerve (PTN) in fossa poplitea eliciting H-reflexes and maximal M-responses in soleus.
Analysis
The stretch responses were identified by subtracting the background EMG (in the control step) from the evoked responses. The latencies for the responses to stretch were measured from the onset of stretch, which normally started 10-20 ms after the trigger because of inertia in the system. In all figures (except Fig. 1) zero time corresponds to stretch onset. The amplitudes of the signals were measured from the baseline to the peak of the rectified signal. The MEPs were measured as the amplitude in a 30 ms time window starting from the onset of the response. In some experiments area measurements were used in order to check that this did not result in any qualitative differences. All values are expressed as means ± 1 S.D. A one-way ANOVA test was used to determine any statistical significant difference between responses. In a few cases (when noted in the text) Student's t test was used.
 |
View larger version [in this window] [in a new window] |
|
|
Stretches (amplitude 8 deg, velocity 250 deg s-1 and a hold phase of ~120 ms) were applied to the ankle dorsiflexors by a portable stretching device at different times during the walking cycle. In A the EMG pattern in TA (upper traces) and soleus (middle traces) together with the changes in ankle joint position (lower traces) during the full walking cycle are shown. The arrows mark the time of the stretches described in B. The EMG responses in TA to the stretches are shown in the top traces in B (thick lines). The onset of the responses are marked by arrows together with the latency in milliseconds. The changes in the position of the ankle joint are shown below. The thin lines show the EMG activity and the ankle joint position in control steps without stretch. The stretches were applied at different delays in relation to the time of heel contact. In the graphs to the far left the stretch was applied 100 ms after heel contact (early stance); in the following graphs it was applied 300 ms after heel contact (mid stance), then 800 ms after heel contact (early swing) and finally 1000 ms after heel contact in the graphs to the far right (mid swing). All the traces consist of an average of 10 sweeps, while time zero in B corresponds to the triggering of the stretch device. PF and DF (plantar and dorsiflexion, respectively) signify the movement direction. | ||
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Modulation of TA stretch reflexes during walking
The responses observed in the TA EMG following stretch of the ankle dorsiflexors varied considerably during the walking cycle. In general, a single large response at a relatively long latency was observed in the stance phase, whereas several small variable responses at different latencies were observed in the swing phase.
This is illustrated for a single subject in Fig. 1. Figure 1A shows the EMG pattern in TA and soleus, together with the position record (bottom part) during the walking cycle, while Fig. 1B shows the responses evoked by ankle dorsiflexor stretch (amplitude 8 deg, velocity 250 deg s-1 and a hold phase of 120 ms) at four different times in the walking cycle (from left to right: early and mid stance and early and mid swing). The responses in the stance phase, where the TA was inactive, were largest and had the shortest onset latency in early stance (76 ms as compared with 90 ms in mid stance), but the latencies to the peak of the responses were almost the same (104 ms as compared with 107 ms in mid stance). In the soleus muscle (not shown) a depression of the ongoing activity in the stance phase was seen, as reported in Sinkjær et al. (2000). The large responses in TA observed in stance were in contrast to the small responses in the swing phase, although the TA muscle was active in this phase of walking and the length of the muscle was comparable (see positional records in Fig. 1A). In the subject used for the illustration only two responses were observed in early swing (onset latencies of 44 and 80 ms, respectively) and in mid swing only the early response followed by a depression was observed.
For the population of subjects (n = 17) the most consistent response in the (early and mid) swing phase was the early response, which was seen in all but one subject. The mean onset latency of this response was 41 ± 5 ms. In four of the subjects two later reflex components could be observed at onset latencies of 76 ± 4 and 90 ± 4 ms (latency of peak 103 ± 6 ms), respectively. In ten other subjects the only identifiable component had an onset latency which corresponded to the first of these components (75 ± 3 ms), but the peak had a latency (104 ± 4 ms) which corresponded to the latency of the last reflex component. In the remaining three subjects only the short latency response was observed. The short latency response was equally common in all parts of the swing phase and without any clear difference in amplitude, whereas the long latency responses tended to be more common in early swing (observed in 14 out of 17 as compared with only 4 out of 11 in mid swing). A depression after the short latency response as in Fig. 1 was seen in mid and late swing in four out of five subjects.
The large late response in stance was seen in all subjects both in the early and the late part of the stance phase. In 11 subjects the response had a relatively short latency to onset (76 ± 5 ms), whereas it had a longer latency in the remaining six subjects (92 ± 4 ms). However, the latency to the peak of the response was the same for the two groups (110 ± 10 ms).
In experiments on six subjects changes in the size of stretch responses observed as a function of the stretch velocity were investigated. The responses generally had a threshold at a stretch velocity around 100 deg s-1 and then increased rapidly with increased stretch velocity. The largest responses were seen with stretch velocities >250 deg s-1 without any clear increase in the size of the responses with faster stretches. With fast stretches oscillations in the position occurred (especially in the swing phase) and this may influence the responses because of subsequent unloading and re-stretching. The stretch velocity was therefore in general reduced in order to avoid the oscillations and the effect of stretch velocities ranging from the velocity resulting in a maximum response (>250 deg s-1) to the velocity where the response disappeared (< 100 deg s-1 ) was systematically examined. The same responses that were observed with fast stretches were also observed with slow stretches (although the responses generally had a lower amplitude). The oscillations which were only present with the higher velocities thus could not have been responsible for the responses of longer latency that followed. This also applies to the experiment in Fig. 1, where small oscillations are clearly present (see position record in Fig. 1B, early and mid swing). The illustrated responses were still observed when slower stretches were used.
Comparison of TA stretch responses during walking and tonic dorsiflexion
When the ankle dorsiflexors are stretched during tonic dorsiflexion in a sitting subject three distinct reflex bursts, which have been denoted M1, M2 and M3, are observed (Toft et al. 1989). Do the responses observed during walking correspond in latency to any of these responses?
Figure 2 illustrates data from one of eight subjects in whom the stretch responses during walking and tonic dorsiflexion were compared. The purpose of the comparison was to look for peaks with latencies similar to the peaks seen in the responses from tonic contraction (M1, M2 and M3), while peaks occurring at longer latencies were not included. It is seen that the latencies of the short and long reflex responses observed in the swing phase (Fig. 2B) correspond well to the latencies of the M1 and M3 reflex bursts observed during tonic dorsiflexion (Fig. 2C). The large response in the stance phase had an onset which was slightly shorter than the onset of M3 during tonic dorsiflexion. The small increase in EMG activity over the background level at latencies corresponding to M1 and M2 is probably explained simply by variability since it was not a consistent observation.
 |
View larger version [in this window] [in a new window] |
|
|
In A-D data from a single subject are shown. In A the response in the TA EMG evoked by stretch in the early stance phase (100 ms after heel contact; stretch velocity 109 deg s-1) is shown, whereas B and C show the responses to stretch in the early swing phase (700 ms after heel contact; stretch velocity 161 deg s-1) and during tonic dorsiflexion (stretch velocity 114 deg s-1). The positional records are shown in D. The tonic dorsiflexion was recorded with the subject sitting down and performing a dorsiflexion against a resistance. The subject was asked to produce a similar amount of TA EMG activity to that in early swing. The stretches applied all had an amplitude of ~8 deg and a hold phase ~200 ms. All traces are the average of 10 sweeps. Time zero corresponds to stretch onset. | ||
In all the subjects the short and long latency responses observed in the swing phase corresponded almost exactly to the latencies of M1 and M3 observed during tonic dorsiflexion (42 ± 5 vs. 43 ± 3 ms in swing and during tonic contraction, respectively). It would be reasonable to assume that the response observed in the swing phase at an intermediate latency would correspond to M2, but the latencies did not fit all that well (75 ± 6 ms for the response during walking as compared with 67 ± 4 ms for M2).
The onset latency of the response observed in the stance phase corresponded to the onset latency of M3 in two of the eight subjects compared, but in five of the remaining six subjects the latency was shorter than the latency of M3 and longer than the latency of M2 (75 ± 4 ms for stance vs. 65 ± 3 ms for M2 and 91 ± 3 ms for M3). For the last subject the latency was identical to M2. Furthermore, in the five subjects a second deflection in the stretch response was seen with a peak at a latency of 94 ± 4 ms. Figure 3A shows an example of this, whereas Fig. 3B shows data from a subject in whom the response in the stance phase had a latency in the range of M3. In all the subjects, the latency of the peak of the response in the stance phase (105 ± 8 ms) corresponded to the latency of the peak of M3 during tonic dorsiflexion (105 ± 9 ms). This close correlation is illustrated in Fig. 3C. It thus seems reasonable to suggest that a main part of the response observed in the stance phase may correspond to M3.
 |
View larger version [in this window] [in a new window] |
|
|
In A and B the TA EMG responses evoked by stretch of the ankle dorsiflexors from two different subjects in early stance are shown together with the positional records (100 ms after heel contact; all traces are the average of 10 sweeps and time zero corresponds to stretch onset). The subject in A had a latency to onset of the response of 73 ms (stretch velocity 119 deg s-1) and peaks at both 95 and 107 ms (in this case a broad peak extending to 117 ms), while the subject in B had a latency to onset of 89 ms (stretch velocity 156 deg s-1) and a peak at 108 ms. In C, data from the eight subjects in whom a comparison between walking and tonic dorsiflexion were made, are shown. The graph shows the latency of the peak of the response recorded in the early stance phase as compared with the latency of the peak of M3 in each of the subjects. The line is a regression line with a slope of 1 and origin in (90,90), showing only minor deviations from identical latencies. | ||
Effect of ischaemia on the TA stretch response in stance
To investigate which sensory afferents are responsible for the response observed in the stance phase, we induced ischaemia in the lower leg by inflating a cuff placed above the patella (5 subjects). In the subject used for the illustration in Fig. 4, the soleus H-reflex evoked by stimulation of the tibial nerve distal to the cuff disappeared 20 min after inflation of the cuff, suggesting that transmission in Ia afferents was effectively blocked at this time. As can be seen from Fig. 4 the TA response evoked by stretch of the ankle dorsiflexors in the early stance phase was also completely abolished at the same time (compare responses before and during ischaemia in Fig. 4A and B, respectively). The stretch response was similarly abolished or strongly diminished at the same time as the H-reflex was abolished in the other four subjects (average size of stretch response before and during ischaemia: 100 ± 46 µV as compared with 29 ± 26 µV (n = 5); P < 0·02). All the subjects were able to walk at the same speed and with the same stride length with and without ischaemia. The pattern and amplitude of the background EMG activity in the TA and soleus muscles were also unchanged by ischaemia.
 |
View larger version [in this window] [in a new window] |
|
|
The data are from a single subject. Ischaemia was induced by inflating a cuff, placed above the patella, to ~240 mmHg. A comparison of the response in the TA EMG to stretch of the ankle dorsiflexors before (A; pre-ischaemia) and 20 min after inflation of the cuff (B; ischaemia) is shown. The thick lines are the average of 10 sweeps with a stretch of 8 deg, a hold phase of ~120 ms and stretch velocities of 419 and 415 deg s-1, respectively. The thin lines are the background EMG. Time zero corresponds to stretch onset. | ||
Stretch-induced facilitation of TA motor-evoked potentials (MEPs) elicited by TMS
In the study by Petersen et al. (1998) it was found that TA MEPs evoked by TMS were facilitated at a latency corresponding to M3 during tonic dorsiflexion in sitting subjects. If the major part of the TA stretch response observed in the stance phase as well as the long latency response observed in the swing phase correspond to M3, a similar facilitation of TA MEPs should be seen corresponding to these responses. As shown in Fig. 5 this was the case. Figure 5A and C demonstrate data from the early stance phase, whereas Fig. 5B and D demonstrate data from the early swing phase. Figure 5A and B demonstrate data from a single subject, whereas Fig. 5C and D demonstrate pooled data from all nine investigated subjects.
 |
View larger version [in this window] [in a new window] |
|
|
A and B show data from a single subject, whereas C and D show mean data from all nine investigated subjects. A and C show early stance phase (100-200 ms after heel contact), whereas B and D show the early swing phase (700-800 ms after heel contact). In all cases TMS was adjusted to evoke an MEP in the TA EMG, which was just above threshold. The stretches used had an amplitude of 8 deg, a hold phase of ~200 ms and velocities of ~300 deg s-1 in both the stance and swing phase. The upper traces in A and B show the response in the TA EMG to combined stretch and TMS (thick line) and the response to TMS alone (thin line) when the stretch preceded TMS by 30 ms (left) and 76 ms (right). The arrows indicate the time of stretch onset (filled arrow) and TMS (open arrow). The lower traces show the response to stretch when applied alone (thick traces) as compared with the background EMG activity (thin traces). In all nine investigated subjects the size of the response to combined stretch and TMS was expressed as a percentage of the algebraic sum of the responses to separate stretch and TMS for each investigated interval between the stretch and TMS. The mean data from all subjects were calculated. The graphs in C and D show the result of this for each conditioning-test interval in the stance (C) and swing phase (D), respectively. The vertical bars are S.D. | ||
TMS was applied at different intervals in relation to the onset of stretch in the two phases of walking. When TMS was applied at a short interval after the stretch (conditioning-test interval: 30 ms) the combined response to stretch and TMS was only slightly larger than when TMS was applied alone (compare thick and thin lines in the upper traces to the left in Fig. 5A and B for the stance and swing phases, respectively). However, when TMS was applied at a longer interval (conditioning-test interval: 76 ms), so that the MEP occurred in the TA EMG around the onset of the late responses in the stance and swing phases (upper right traces in Fig. 5A and B, respectively), the combined response to stretch and TMS was much larger than the response to TMS alone. The lower traces in both Fig. 5A and B show the response to stretch when applied alone (thick lines) as compared with the background EMG activity (thin lines).
In the nine subjects, six intervals between the conditioning stretch and the test MEP were investigated. For each subject and each conditioning-test interval the size of the response to combined stretch and TMS was expressed as a percentage of the algebraic sum of the responses to separate stretch and TMS. As can be seen from the pooled data in Fig. 5C and D, there was no facilitation of the MEP until a conditioning- test interval of 60 ms in either of the two phases of walking. However, when the MEP was evoked at a latency corresponding to the late stretch responses, a significant extra-facilitation of the MEP was observed. The peak of the facilitation corresponded to the peak of the late responses (conditioning-test interval 80 ms; latency of MEP 30 ms; latency of peak-of-late response 110 ms). In subjects who had a relatively short latency to onset of the late response (around 75 ms), no facilitation was seen until conditioning- test intervals corresponded to the late part of the response.
Comparison between transcranial magnetic and electrical stimulation
Petersen et al. (1998) provided evidence that M3 observed during tonic dorsiflexion in sitting subjects may be at least partly mediated by a transcortical reflex, since only muscular responses evoked by TMS were facilitated by prior stretch, whereas responses evoked by TES were not. At intensities just above the threshold for the MEP, TES seems to penetrate deep into the brain and activate the axons of the corticospinal cells at some distance from the cell body, whereas TMS activates the corticospinal cells either indirectly (trans-synaptically) or directly at a site close to the cell body (see Rothwell, 1997 for review). Responses evoked by TMS are therefore sensitive to cortical excitability changes, whereas responses evoked by TES at just-above-threshold intensities are generally not. To obtain evidence of whether a transcortical reflex pathway contributes to the late responses to stretch in the early stance and swing phases of walking, we consequently compared the effects of prior stretch on MEPs evoked by TMS and TES in these two phases of the movement.
Figure 6 shows data from a single subject. In early swing the MEP evoked by TMS was facilitated by prior stretch corresponding to the late stretch response (Fig. 6B; compare thick and thin lines), whereas MEPs evoked by TES were not (Fig. 6D). As in the study by Petersen et al. (1998) this thus provides evidence that the late response observed in swing may be at least partly mediated by a transcortical reflex pathway.
 |
View larger version [in this window] [in a new window] |
|
|
TA MEPs were evoked by transcranial magnetic (A and B) and electrical (C and D) stimulation in the early stance (A and C) and early swing (B and D) phases of walking. The intensity of the stimuli were adjusted to evoke MEPs with almost equal amplitudes in both stance and swing. The intensity of the electrical stimulus was 22 % of the maximal stimulator output in swing and 31 % in stance, whereas the intensity of the magnetic stimulus was 50 % in both phases. The traces show the average of 10 sweeps following either combined stretch and transcranial stimulation (thick lines) or transcranial stimulation alone (thin lines). Conditioning-test intervals of 70 and 72 ms were used for TMS and TES, respectively, as marked by the open arrows. Time zero corresponds to stretch onset (filled arrows). The stretch had an amplitude of 8 deg, a hold phase of ~200 ms and velocities of ~280 deg s-1 in both the stance and swing phase. | ||
In the early stance phase it was necessary to increase the intensity of TES significantly in order to obtain an MEP of the same size as in the swing phase (from 22 to 31 % of maximal stimulator output in this subject). An increase of the intensity of TMS was not necessary. In contrast to what was observed in the swing phase, the MEPs evoked by both TMS and TES were strongly facilitated by prior stretch in the stance phase (Fig. 6A and C, respectively; compare thick and thin lines). It is, thus, not possible unequivocally to argue for a transcortical contribution to the responses observed in the stance phase of walking. However, as will be pointed out in Discussion this possibility also cannot be disregarded, since relatively strong electrical stimuli had to be used in the stance phase in order to evoke clear MEPs, which may consequently have been sensitive to cortical excitability changes.
Similar findings were obtained in all five investigated subjects. The amount of extra-facilitation in the stance phase (calculated in the same way as in Fig. 5) of the MEP evoked by TMS was 192 ± 47 % (P < 0·00003), whereas the MEP evoked by TES was facilitated by 206 ± 88 % (P < 0·0003). In the swing phase the MEP evoked by TMS was facilitated by 189 ± 64 % (P < 0·008), whereas there was a non-significant decrease of the MEP evoked by TES (group mean 64 ± 34 %). On average the intensity of TES was increased from 24 ± 3 % in the swing phase to 33 ± 4 % of the maximal stimulator output in the stance phase. It was not necessary to change the intensity for TMS. The MEPs evoked by TES in the stance phase lasted significantly longer than the MEPs evoked in the swing phase (29 ± 6 vs. 19 ± 3 ms; P < 0·02, Student's t test). The MEPs evoked by TMS had the same duration in stance as in swing. The TMS was adjusted to be just above threshold and the MEPs to have amplitudes of similar size when comparing TMS with TES (66 ± 24 vs. 77 ± 29 µV for stance and 98 ± 48 vs. 102 ± 47 µV for swing, respectively).
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
It has been demonstrated in the present study that stretch responses in the TA muscle are strongly modulated during the walking cycle. In the swing phase when the muscle is active, two to three small and variable responses are seen, whereas a large, but late response is consistently seen in the stance phase when the muscle is silent.
Which mechanisms are responsible for the TA reflex responses in swing?
It seems reasonable to assume that the first response observed following stretch in the swing phase is equivalent to M1 and like M1 it is mediated mainly by the monosynaptic Ia pathway to the spinal motoneurones. It would be difficult to explain the very short latency of the response by any other mechanism. It also seems reasonable to assume that the latest of the responses, which had an onset latency similar to that of M3, is mediated, at least in part, by a transcortical reflex. As observed previously for M3 (Petersen et al. 1998), MEPs evoked by TMS were facilitated at latencies corresponding to this late response, but not corresponding to the earlier stretch responses. The onset of this facilitation in relation to the stretch (60 ms) corresponded exactly to the onset of the facilitation observed in sitting subjects by Petersen et al. (1998). They showed that an afferent volley evoked by the stretch would need at least 47 ms to reach the cortex and if a central delay for processing the signal in the cortex of around 10 ms is added, a value corresponding quite well to the onset of the facilitation is reached. Furthermore, a similar facilitation was not seen for MEPs evoked by transcranial electrical stimulation. As argued also by Petersen et al. (1998) this difference in the behaviour of MEPs evoked by TMS and TES is most easily explained by an increased cortical excitability evoked by the stretch. MEPs evoked by TES at just above threshold intensities are less sensitive to cortical excitability changes because the electrical stimulus at such intensities primarily activates the axons of the corticospinal cells at some distance from the cell body (Edgley et al. 1990; Burke et al. 1993; Nielsen et al. 1995). TMS on the other hand activates the corticospinal cells either directly close to the cell body or indirectly via projections onto the corticospinal cells and MEPs evoked by TMS are therefore generally sensitive to cortical excitability changes.
The mechanism responsible for the middle latency response in the swing phase is an enigma. When present it had a latency which was almost 10 ms longer than the M2 response and almost 20 ms shorter than the M3 response observed during tonic dorsiflexion. This may suggest that the response is mediated by a pathway, which is open during walking, but not during tonic contraction. However, we find it more likely that the response corresponds to M2, and that the longer latency is explained by the small and variable size of the response, which makes an exact measurement difficult. It may also be that longer processing of the transmission in the pathway is necessary during walking. If the response does correspond to M2, the smaller size of the response during walking would indeed suggest that transmission in the pathway is depressed, which might lead to some prolongation of the onset of the response. Corna et al. (1995) have suggested that the TA medium latency response observed in standing subjects perturbed by rotation of a platform is mediated by group II afferents. This medium latency response is possibly equivalent to M2 and like Schieppati & Nardone (1997) we have found in preliminary experiments that M2 during tonic dorsiflexion as well as the middle latency response observed during walking are delayed more than M1 during cooling of the leg (M. Grey, J. Andersen & T. Sinkjær, unpublished observations). This is consistent with the idea that both responses are mediated by group II afferents.
The mechanism for the depression of the EMG activity following the initial stretch responses, as seen in late swing, is unclear. One possibility is that the small unloading of the dorsiflexors, which was seen immediately after the stretch in the swing phase, could result in decreased muscle afferent activity (cf. position records in Fig. 1). In this case the depression might reflect removal of a contribution of muscle afferents to the TA EMG activity. However, the depression of the EMG activity is also observed after a stretch input where no unloading takes place (Sinkjær et al. 1988; Toft et al. 1991). It may be an effect of the synchronized afferent input to the motoneurones caused by the fast stretch, but this does not explain why no depression was seen in early swing. The occurrence of several peaks of, in turn, short and long latency was probably not due to the oscillations in the position induced by the stretches, since the characteristics of the responses remained the same when the stretch velocity was reduced so that the oscillations disappeared.
Which mechanisms are responsible for the TA stretch response in stance?
The lack of short latency responses to stretch in the TA during the stance phase is not surprising, since the muscle is silent at this time and the motoneurones are therefore not very excitable. It is likely that inhibitory mechanisms such as disynaptic reciprocal Ia inhibition are responsible for depressing the excitability of the motoneurones (Petersen et al. 1999). The necessity for increasing the intensity of TES in the stance phase in order to obtain an MEP is consistent with a low motoneuronal excitability. It is unclear whether presynaptic inhibition of TA Ia afferents is increased in the stance phase as is the case for presynaptic inhibition of soleus Ia afferents (Capaday & Stein, 1986), but this would seem possible, since presynaptic inhibition of antagonistic Ia afferents has been shown to be increased in relation to other types of movement (Crone & Nielsen, 1989). If so, this would add further to the depression of the short latency responses.
It would seem as if at least two different mechanisms were involved in the generation of the long latency stretch response that was observed in the stance phase. We may with certainty conclude that at least the initial part of the response in the 11 subjects, in whom the response had a relatively short latency (i.e. around 75 ms), is not mediated by a transcortical mechanism, since such a pathway would require a transmission time of at least 79 ms according to the calculations by Petersen et al. (1998). However, in the remaining subjects the response had a similar latency to onset as M3 during tonic dorsiflexion and as argued in the case of M3 (Petersen et al. 1998), this latency corresponds well to that expected for a transcortical reflex pathway. Given that a second deflection was seen at the latency of M3 in several of the subjects in whom the response had a relatively short latency and that the peak of the response corresponded to that of M3 in all investigated subjects, this suggests that a similar mechanism may contribute to the later part of the response in all subjects. One possibility would be that a mechanism similar to that responsible for M2 contributed to the initial part of the response in some subjects. It might be that a rather long processing time is necessary before the motoneurones are discharged in the stance phase, because of their low excitability in this phase of walking as compared with tonic dorsiflexion when the muscle is active. M1 similarly has a longer latency when it is evoked at rest than when it is evoked during voluntary contraction (Toft et al. 1989).
The observation that all the response was most often completely abolished by ischaemia along with the soleus H-reflex, however, makes it rather unlikely that group II afferents, which are assumed to be responsible for M2 (see earlier and Schieppati & Nardone, 1997), should contribute to the response. Based on this observation we find it likely that group I afferents are responsible for all the response. We cannot exclude the probability that skin afferents were involved in the generation of the response, since the thickest skin afferents are as thick as group I afferents and therefore also equally sensitive to ischaemia. All subjects also reported anaesthesia of the foot when ischaemia was at its maximum.
The time course of the facilitation of the TMS-induced MEP is consistent with the idea that the late part of the response is mediated by a transcortical reflex pathway, whereas such a mechanism cannot contribute to the earliest part of the response. There was thus no facilitation of the MEPs at latencies corresponding to the initial part of the response in any of the subjects, in whom it had a (relatively) short latency. The facilitation only occurred when the MEP was evoked at latencies corresponding to the later part of the response - that is at latencies comparable with M3 (i.e. a conditioning-test interval of 60 ms, corresponding to a latency of around 90 ms for the response, since the MEP latency of 30 ms has to be added). If nothing else, this facilitation of the MEP signifies that activity in the corticospinal tract was of significant importance for the response. However, since a facilitation of MEPs evoked by both TMS and TES was observed in stance, we cannot conclude that the MEP facilitation reflects an increased cortical excitability as is the case for the MEP facilitation during tonic dorsiflexion (Petersen et al. 1998) and in the swing phase. As already pointed out in Results and in the previous section a different behaviour of MEPs evoked by TMS and TES may be used to provide evidence of the transcortical nature of the stretch responses. Without this different behaviour of the MEPs following stretch in the stance phase, the case for a transcortical contribution to the TA stretch responses in this phase of walking is weaker. Despite this, we would like to argue in favour of such a contribution. Firstly, the time course of the facilitation of the MEP was identical in stance, swing and during tonic dorsiflexion, which makes it likely that the same mechanism is at play in all three cases (i.e. a transcortical reflex pathway). Secondly, it was necessary to increase the intensity of TES quite substantially in order to evoke an MEP in the stance phase, possibly due to low excitability of the spinal TA motoneurones in this phase of walking (cf. above). It has been demonstrated that TES at strong intensities also activates the corticospinal cells indirectly (Rothwell, 1997). It is thus only when weak electrical stimuli are applied that the MEPs may be evoked exclusively by activation of the corticospinal axons. With stronger stimuli indirect volleys (I-waves) contribute to the activation of the motoneurones and the MEPs evoked by such stimuli may therefore be sensitive to cortical excitability changes. Although the evidence is only indirect, we believe that it adequately explains our observations and that the MEP facilitation observed in stance also reflects increased cortical excitability following the stretch and thus makes it likely that a transcortical reflex pathway contributes to the later part of the stretch response.
Two further observations lend some support to this. Firstly, Capaday et al. (1999) demonstrated that TA MEPs evoked by TMS are hardly depressed in the stance phase of walking. This was confirmed in the present study by the observation that we did not have to increase the intensity of TMS in order to evoke an MEP of equal size in stance to that in swing. Since we did have to increase the intensity of TES, however, we may suggest in accordance with Capaday et al. (1999) that this lack of depression of TA MEPs evoked by TMS despite decreased spinal motoneuronal excitability is explained by a high excitability of corticospinal cells projecting onto TA motoneurones in the stance phase. Given this high cortical excitability it is not surprising to see large transcortical reflexes being elicited by stretch of the TA. Secondly, we (T. Sinkjær, J. B. Andersen, J. Nielsen & M. Ladouceur, unpublished observation) have recently investigated TA stretch reflexes in the stance phase of walking in patients with multiple sclerosis (MS). In only one out of eight patients was a response similar to that described here in healthy subjects observed. Since these MS patients all had evidence of lesion of the pyramidal tract, we think that this observation supports the fact that the corticospinal tract is engaged in the generation of the response.
Responses evoked by TA stretch and responses to backwards displacement of the body are not similar
The responses that we have described here differ in many ways from the responses to horizontal perturbations of the supporting leg that have been described previously by Dietz et al. (1984, 1987). They studied responses in the medial gastrocnemius and TA muscles following sudden accelerations and decelerations of the treadmill on which the subjects were walking. The resulting stretch of the ankle dorsiflexors when the mass of the body was displaced backwards relative to the supporting foot was seen to evoke responses in both muscles which counteracted the perturbation. In contrast to the responses that we have described in the present paper, the responses described by Dietz et al. (1984, 1987) had a shorter latency, changed latency as a function of displacement velocity and were resistant to ischaemia. As suggested already by Dietz et al. (1987) the latter finding makes it likely that their responses were mediated by group II afferents and are equivalent to M2. Why backwards translation of the body, which results in stretch of the ankle dorsiflexors, should lead to preferential activation of group II afferents and mainly spinal reflex activity, whereas primary stretch of the muscle with a resulting vertical translation of the body should result in activation of group I afferents and mainly long latency stretch reflex activity is unclear. In order to clarify this, comparisons including the same subjects in both setups should be made.
Functional significance
As pointed out in Introduction the observation that there are large stretch responses in the TA muscle in the stance phase when the muscle is silent and small responses in the swing phase when the muscle is active, supports the idea that stretch reflexes, at least in the ankle joint muscles, are mainly of importance in the stance phase (Zehr & Stein, 1999). It does indeed seem likely that the high reflex activity in both the antagonistic muscles around the ankle (in this case expressed by soleus and TA) in the stance phase is involved in securing the stability around the ankle in case of external perturbations. We thus believe that the high excitability of TA corticospinal neurones (Capaday et al. 1999) and the large TA stretch responses during the stance phase, which are probably partly mediated by a transcortical pathway, have to be seen in relation to the need to be ready to respond to any instability of the supporting ground. The reaction in other muscles around the ankle (some of which may have a significant importance for the stabilization), was not examined and we therefore cannot say anything about the role or origin of any possible responses in these.
By integrating the response to the somatosensory input at a cortical level, it may be possible to adjust the reaction according to visual and motivational influences. This solution may be unique for bipedal man given the increased demand on the stability of the supporting leg in bipedal walking as compared with quadrupedal walking, but there is in fact evidence that responses to stretch are also mediated at least partly through transcortical reflex mechanisms in the cat. Corticospinal cells thus respond to peripheral nerve stimulation during precision walking (Marple-Horvat & Armstrong, 1999) and following perturbations in early stance (Marple-Horvat et al. 1993). In the experiments by Marple-Horvat et al. (1993) the cats were walking on a horizontal ladder where one of the steps suddenly gave way under the weight of the cat. This resulted in discharge of corticospinal cells, probably in order to lift the paw up and secure the continued walking movements.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| Andersen, J. B. & Sinkjaer, T. (1995). An actuator system for investigating electrophysiological and biomechanical features around the human ankle joint during gait. Transaction on Rehabilitation Engineering 3, 299-306<ÏTD> | |
| Andersson, O., Forssberg, H., Grillner, S. & Wallen, P. (1981). Peripheral feedback mechanisms acting on the central pattern generators for locomotion in fish and cat. Canadian The Journal of Physiology and Pharmacology, 59, 713-26 | |
| Burke, D., Hicks, R., Gandevia, S. C., Stephen, J., Woodforth, I. & Crawford, M. (1993). Direct comparison of corticospinal volleys in human subjects to transcranial magnetic and electrical stimulation. The Journal of Physiology 470, 383-393. | [Abstract] |
| Capaday, C., Lavoie, B. A., Barbeau, H., Schneider, C. & Bonnard, M. (1999). Studies on the corticospinal control of human walking. I. Responses to focal transcranial magnetic stimulation of the motor cortex. Journal of Neurophysiology 81, 129-139 | [Abstract/Full Text] |
| Capaday, C. & Stein, R. B. (1986). Amplitude modulation of the soleus H-reflex in the human during walking and standing. Journal of Neuroscience 6, 1308-1313. | [Abstract] |
Corna, S., Grasso, M., Nardone, A. & Schieppati, M. (1995). Selective depression of medium-latency leg and foot muscle responses to stretch by an  2-agonist in humans. The Journal of Physiology 484, 803-809. |
[Abstract] |
| Crone, C. & Nielsen, J. (1989). Spinal mechanisms in man contributing to reciprocal inhibition during voluntary dorsiflexion of the foot. The Journal of Physiology 416, 255-272. | [Abstract] |
| Dietz, V., Quintern, J. & Berger, W. (1984). Corrective reactions to stumbling in man: functional significance of spinal and transcortical reflexes. Neuroscience Letters 10, 131-135. | |
| Dietz, V., Quintern, J. & Sillem, M. (1987). Stumbling reactions in man: significance of proprioceptive and pre-programmed mechanisms. The Journal of Physiology 386, 149-163. | [Abstract] |
| Edgley, S. A., Eyre, J. A., Lemon, R. N. & Miller, S. (1990). Excitation of the corticospinal tract by electromagnetic and electrical stimulation of the scalp in the macaque monkey. The Journal of Physiology 425, 301-320. | [Abstract] |
| Grillner, S. & Zangger, P. (1984). The effect of dorsal root transection on the efferent motor pattern in the cat's hindlimb during locomotion. Acta Physiological Scandinavica 120, 393-405. | |
| Hiebert, G. W. & Pearson, K. G. (1999). Contribution of sensory feedback to the generation of extensor activity during walking in the decerebrate cat. Journal of Neurophysiology 81, 758-770 | [Abstract/Full Text] |
| Marple-Horvat, D. E., Amos, A. J., Armstrong, D. M. & Criado, J. M. (1993). Changes in the discharge patterns of cat motor cortex neurones during unexpected perturbations of on-going locomotion. The Journal of Physiology 462, 87-113. | [Abstract] |
| Marple-Horvat, D. E. & Armstrong, D. M. (1999). Central regulation of motor cortex neuronal responses to forelimb nerve inputs during precision walking in the cat. The Journal of Physiology 519, 279-299. | [Abstract/Full Text] |
| Nielsen, J., Petersen, N. & Ballegaard, M. (1995). Latency of effects evoked by electrical and magnetic brain stimulation in lower limb motoneurones in man. The Journal of Physiology 484, 791-802. | [Abstract] |
| Petersen, N., Christensen, L. O., Morita, H., Sinkjaer, T. & Nielsen, J. (1998). Evidence that a transcortical pathway contributes to stretch reflexes in the tibialis anterior muscle in man. The Journal of Physiology 512, 267-276. | [Abstract/Full Text] |
| Petersen, N., Morita, H. & Nielsen, J. (1999). Modulation of reciprocal inhibition between ankle extensors and flexors during walking in man. The Journal of Physiology 520, 605-619. | [Abstract/Full Text] |
| Rothwell, J. C. (1997). Techniques and mechanisms of action of transcranial stimulation of the human motor cortex. Journal of Neuroscience Methods 74, 113-122 | [Medline] |
| Schieppati, M. & Nardone, A. (1997). Medium-latency stretch reflexes of foot and leg muscles analysed by cooling the lower limb in standing humans. The Journal of Physiology 503, 691-698. | [Abstract] |
| Schubert, M., Curt, A., Jensen, L. & Dietz, V. (1997). Corticospinal input in human gait: modulation of magnetically evoked motor responses. Experimental Brain Research 115, 234-246. | [Medline] |
| Sinkjaer, T., Andersen, J. B., Ladouceur, M., Christensen, L. O. D. & Nielsen, J. (2000). Sensory feedback contributes importantly to soleus EMG activity in the stance phase of walking in man. The Journal of Physiology 523, 817-827. | [Abstract/Full Text] |
| Sinkjaer, T., Andersen, J. B. & Larsen, B. (1996). Soleus stretch reflex modulation during gait in humans. Journal of Neurophysiology 76, 1112-1120 | [Medline] |
| Sinkjaer, T., Toft, E., Andreassen, S. & Hornemann, B. C. (1988). Muscle stiffness in human ankle dorsiflexors: intrinsic and reflex components. Journal of Neurophysiology 60, 1110-1121 | [Medline] |
| Toft, E., Sinkjaer, T. & Andreassen, S. (1989). Mechanical and electromyographic responses to stretch of the human anterior tibial muscle at different levels of contraction. Experimental Brain Research 74, 213-219. | [Medline] |
| Toft, E., Sinkjaer, T., Andreassen, S. & Larsen, K. (1991). Mechanical and electromyographic responses to stretch of the human ankle extensors. Journal of Neurophysiology 65, 1402-1410 | [Medline] |
| Yang, J. F., Stein, R. B. & James, K. B. (1991). Contribution of peripheral afferents to the activation of the soleus muscle during walking in humans. Experimental Brain Research 87, 679-687. | [Medline] |
| Zehr, E. P. & Stein, R. B. (1999). What functions do reflexes serve during human locomotion? Progress in Neurobiology 58, 185-205 | [Medline] |
This work received support from the Danish Health Research Council, the Danish Sports Research Council, The Novo Nordisk Foundation and the Danish Society of Multiple Sclerosis. Lars O. D. Christensen received a stipend funded under a joint research programme from the Danish Health and Technical Research Councils.
Corresponding author
L. O. D. Christensen: The Panum Institute, Blegdamsvej 3, 2200 Copenhagen N, Denmark.
Email: lodc{at}mfi.ku.dk
This article has been cited by other articles:
|
|
 |
|
  W. Taube, C. Leukel, M. Schubert, M. Gruber, T. Rantalainen, and A. Gollhofer Differential Modulation of Spinal and Corticospinal Excitability During Drop Jumps J Neurophysiol, March 1, 2008; 99(3): 1243 - 1252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P.H.J.A. Nieuwenhuijzen and J. Duysens Proactive and Reactive Mechanisms Play a Role in Stepping on Inverting Surfaces During Gait J Neurophysiol, October 1, 2007; 98(4): 2266 - 2273. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Courtine, S. J. Harkema, C. J. Dy, Y. P. Gerasimenko, and P. Dyhre-Poulsen Modulation of multisegmental monosynaptic responses in a variety of leg muscles during walking and running in humans J. Physiol., August 1, 2007; 582(3): 1125 - 1139. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Melnyk, M. Faist, M. Gothner, L. Claes, and B. Friemert Changes in Stretch Reflex Excitability Are Related to "Giving Way" Symptoms in Patients With Anterior Cruciate Ligament Rupture J Neurophysiol, January 1, 2007; 97(1): 474 - 480. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Taube, M. Schubert, M. Gruber, S. Beck, M. Faist, and A. Gollhofer Direct corticospinal pathways contribute to neuromuscular control of perturbed stance J Appl Physiol, August 1, 2006; 101(2): 420 - 429. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Schneider, B. A. Lavoie, H. Barbeau, and C. Capaday Timing of cortical excitability changes during the reaction time of movements superimposed on tonic motor activity J Appl Physiol, December 1, 2004; 97(6): 2220 - 2227. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Courtine and M. Schieppati Tuning of a Basic Coordination Pattern Constructs Straight-Ahead and Curved Walking in Humans J Neurophysiol, April 1, 2004; 91(4): 1524 - 1535. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Nakazawa, N. Kawashima, M. Akai, and H. Yano On the reflex coactivation of ankle flexor and extensor muscles induced by a sudden drop of support surface during walking in humans J Appl Physiol, February 1, 2004; 96(2): 604 - 611. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
,a statistically significant difference (P < 0·05) between the response to combined stretch and TMS as compared with the algebraic sum of the responses to separate stretch and TMS.