Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2010 Feb 1;588(Pt 6):967–979. doi: 10.1113/jphysiol.2009.185520

Corticospinal contribution to arm muscle activity during human walking

Dorothy Barthelemy 1,2, Jens Bo Nielsen 1
PMCID: PMC2849962  PMID: 20123782

Abstract

When we walk, our arm muscles show rhythmic activity suggesting that the central nervous system contributes to the swing of the arms. The purpose of the present study was to investigate whether corticospinal drive plays a role in the control of arm muscle activity during human walking. Motor evoked potentials (MEPs) elicited in the posterior deltoid muscle (PD) by transcranial magnetic stimulation (TMS) were modulated during the gait cycle in parallel with changes in the background EMG activity. There was no significant difference in the size of the MEPs at a comparable level of background EMG during walking and during static PD contraction. Short latency intracortical inhibition (SICI; 2 ms interval) studied by paired-pulse TMS was diminished during bursts of PD EMG activity. This could not be explained only by changes in background EMG activity and/or control MEP size, since SICI showed no correlation to the level of background EMG activity during static PD contraction. Finally, TMS at intensity below the threshold for activation of corticospinal tract fibres elicited a suppression of the PD EMG activity during walking. Since TMS at this intensity is likely to only activate intracortical inhibitory interneurones, the suppression is in all likelihood caused by removal of a corticospinal contribution to the ongoing EMG activity. The data thus suggest that the motor cortex makes an active contribution, through the corticospinal tract, to the ongoing EMG activity in arm muscles during walking.

Introduction

When we walk, our arms move rhythmically out of phase with the corresponding leg. This is not a simple passive mechanical effect, since several arm muscles, especially at the shoulder joint, show rhythmic alternating activity at specific times during the gait cycle (Elftman, 1939; Ballesteros et al. 1965; see also Dietz, 2002). It has been suggested that this may be reminiscent of quadrupedal locomotion in four-legged animals and that this activity is reflective of spinal rhythm generating networks and spinal reflex mechanisms (Dietz, 2002; Zehr et al. 2004, 2009). The lack of mental effort that is required to elicit rhythmic arm movement and the sense of automaticity that we all experience when walking, are common arguments in favour of this suggestion. However, the sense of voluntary effort and conscious (free) ‘will’ is not equivalent to activity in the motor cortex and the lack of it during walking therefore also cannot be used to infer lack of involvement of cortical control areas (Prochazka et al. 2000; Heisenberg, 2009). Indeed, transcranial magnetic stimulation of the motor cortex is not associated with any conscious perception of movement except that provided by the sensory feedback from the muscle contraction (Ellaway et al. 2004).

The motor cortex and the corticospinal tract appear to be mainly important for visual guidance and motivational adjustments of gait direction and pattern in the cat, but corticospinal neurones are also active during uncomplicated walking, and the threshold to induce muscle activation by intracortical stimulation is in general very low during gait (Armstrong & Drew, 1985; Drew, 1993; Bretzner & Drew, 2005). Furthermore, selective corticospinal lesion leads to drop foot in the cat suggesting that the corticospinal tract makes an active contribution to the muscle activation even during uncomplicated walking (Drew et al. 2002). In humans, lesion of supraspinal motor pathways leads to more severe and lasting functional deficits than in the cat, which probably reflects a more important role of the motor cortex and the corticospinal tract in the control of gait (Nielsen, 2003). It has also been shown that motor evoked potentials (MEPs) induced by transcranial magnetic stimulation (TMS) are greatly modulated during gait and that this modulation is probably partly explained by changes in cortical excitability (Schubert et al. 1997; Capaday et al. 1999). More direct evidence of changes in corticospinal excitability has also been provided by Petersen et al. (1998) who demonstrated that the excitability of monosynaptic corticospinal projections to soleus motoneurones was modulated during the gait cycle. In a subsequent study, it was shown that transient removal of corticospinal activity by activation of intracortical inhibitory neurones resulted in a suppression of EMG activity in ankle muscles during gait (Petersen et al. 2001). This demonstrated that the corticospinal tract contributes directly to the generation of EMG activity in the leg during gait.

The question that we addressed in the present study is whether this is also the case for the muscle activity in the arms. Three different approaches were used: (1) Similar to Schubert et al. (1997) and Capaday et al. (1999), we examined the modulation of MEPs in the arm muscles at different times during the gait cycle; (2) We used paired-pulse TMS to examine changes in intracortical inhibition of arm muscle MEPs during the gait cycle; (3) Similar to Petersen et al. (2001), we investigated whether subthreshold TMS elicits a suppression of the EMG in arm muscles during gait. Part of this work has been published previously in abstract form (Barthélemy et al. 2007a,b; Barthélemy & Nielsen, 2008).

Methods

The experimental protocol was approved by the local ethics committee (Den videnskabsetiske komite for Københavns og Frederiksbergs kommuner; V. 100.1969/91) and was in accordance with the Declaration of Helsinki. Thirty five healthy subjects participated in one or more parts of the study (12 women and 23 men; age 25 to 42). All gave written informed consent to the experimental procedures. None of the subjects had any history of neurological disease.

Electromyographic (EMG) activity was recorded from the right anterior and posterior deltoid (AD and PD), biceps, triceps and – in some experiments – flexor carpi radialis (FCR) and extensor carpi radialis (ECR) muscles. Surface Ag–AgCl disc electrodes were placed over the belly of the muscles. The signals were amplified (×1000–5000), band-pass filtered (25–1000 Hz) then digitized and sampled (2 kHz) to a computer using a micro1401 interface (Cambridge Electronic Design Ltd, Cambridge, UK).

Kinematic analysis

For 10 subjects, arm kinematics were assessed during walking on a treadmill at a speed of 3.4 to 4 km h−1, in order to characterize the arm movements during the gait cycle. This kinematics analysis as well as EMG activity guided us in choosing the time-points used for the TMS measurement. In subjects where kinematics was not recorded, only EMG was used. The time-points chosen were not necessarily at the exact same time in relation to the heel trigger in all subjects, since the EMG activity came at slightly different times in different subjects.

The arm kinematics acquisitions lasted 1 min and were recorded at the beginning of the experimental session, prior to the stimulation. In that way, motor evoked responses did not interfere with the kinematics measurements. In order to record the movements, passive markers were placed on the C7 protuberance, the acromion, the lateral joint line of the elbow and the ulnar head at the wrist. Except for C7, markers were placed bilaterally. The position of the markers were recorded by six infra-red cameras (Pro-reflex; Qualisys, Sweden) connected to a laptop running Qualisys motion capture software. Further analysis (stick figures, markers displacement) were performed in Excel and custom-made softwares (Matlab, Qtools).

Although assessment of rhythmic arm movement during locomotion was the aim of this study, subjects were not encouraged to use their arms during locomotion. Electrodes were put on their arms, so they knew that arm activity was recorded, but we never referred to the movement of their arms or asked them to move their arm a certain way during the experiment. They were just asked to walk at a comfortable pace, as they would do overground.

Cortical stimulation

In all experiments we used a Magstim Rapid Rate stimulator (Magstim Company Ltd, Dyfed, UK) for magnetic stimulation of the motor cortex and a figure-of-eight coil (loop diameter 9 cm). At the beginning of each experiment, we determined the optimal location for evoking a motor response (MEP) in the shoulder muscles (Fig. 1A). Overall this was 4 to 5 cm lateral to the vertex (contralateral to the side under investigation). The coil was held in a fixed position in relation to the head by a specially designed harness (Balgrist Tec, Zurich, Switzerland; see Schubert et al. 1997; Petersen et al. 1998), or a custom-made helmet. Additionally, the position of the coil was regularly inspected throughout the experiments in relation to markers drawn on a head-fitted hat with a grid, and the position of the coil for each stimulus was monitored and recorded by a frameless stereotaxic device (Brainsight, Magstim Co., UK).

Figure 1. Methods.

Figure 1

Open in a new tab

A, transcranial magnetic stimulation (TMS) was applied over the arm area of the motor cortex and motor evoked responses (MEP) were recorded in the EMG signal of the arm muscles. B, modulation of the amplitude of the MEP (marked by arrows) in posterior deltoid was assessed by applying TMS at 1.2 × MEP threshold (T) at different times during gait cycle. Each trace starts at the ipsilateral heel contact, and the delay is calculated from that heel contact. C, short interval intracortical inhibition (SICI) was tested by using a conditioning-test design, where the test stimulus (1.2 T) was given alone (control; 0 ms) or with a conditioning stimuli below threshold (0.7 T) through the same coil at a short interstimulus interval (ISI) of 2, 3, 4, 5 or 6 ms (see lower panel). The largest inhibition was obtained with ISI of 2 ms. Value at each data point is presented as mean ±s.e.m. D, stimulation of the motor cortex by subthreshold TMS. At this intensity which is below threshold to induce an MEP, an inhibition is demonstrated (here in triceps) when comparing the stimulated trace (dark) with the EMG activity without stimulation (control, light grey). The start and end of the inhibition is marked by vertical lines. The time of stimulation is represented by the arrow.

Similar to the majority of studies making use of magnetic and/or electrical cortical stimulation (e.g. Liepert et al. 1998), the MEP threshold was defined as the intensity at which an MEP was clearly distinguishable from the background activity in 3 out of 5 trials, during an isometric contraction equivalent to 10% of the maximal voluntary contraction (MVC). The latter was measured while the subjects were seated with the arm in line with the trunk and the elbow bent at 90 deg. They were asked to make a maximal and isometric contraction of the PD muscle against a solid resistance (wooden board) and the level of EMG activity of the posterior deltoid was recorded. They made three distinct maximal contractions and the amplitude of the highest contraction was kept as the MVC.

A pressure-sensitive resistor was placed under the right heel of the subject's shoe, ipsilateral to the assessed arm. The signal elicited triggered recording of EMG activity. Each acquisition sweep lasted from 1.5 to 5 s, depending on the protocol used (modulation of MEP, SICI or subthreshold TMS; cf. following 4 sections). This signal was also used to activate the magnetic stimulator at the appropriate delay. Control steps, without any TMS, were randomly alternated with steps in which TMS was given by means of CED software (Signal).

Modulation of the MEPs

This protocol was assessed in seven subjects. Stimuli (1.2 T) were given at each 0.1 s from 0.1 to 1 s after heel strike and at 1.2 s after heel strike (Fig. 1B). In one condition, no stimulation was given at all (control trial). For each of the 11 stimulation times, 10 stimuli were applied in a random manner. Each recorded sweep lasted 2 s. Peak-to-peak amplitude of the MEP was measured for a window of 50 ms after the stimulus for each trial, and then averaged for each delay. The corresponding background EMG was measured in the same time-window in the control trials.

All measurements were normalized to MVC rather than maximal M wave (Mmax). Evoking a maximal response in PD muscle by stimulating Herb's point requires a large intensity and such stimulation will activate other muscles of the arm and forearm, making the stimulation very uncomfortable for most subjects. Normalising to MVC will most probably not change the correlations and differences seen in the studies, but it might make group averages more variable (higher standard error of the mean).

Short-latency intracortical inhibition

The modulation of short-latency intracortical inhibition (SICI) during gait was tested in 16 subjects, but since no inhibition was observed in 2 of them, we present results from only 14 subjects. In 13 of these subjects the amount of SICI was compared between walking, while the arm is near its maximal forward excursion (or maximal forward movement) and matched EMG activity during static contraction of the posterior deltoid muscle. In 4 additional subjects, measurements were only done at maximal forward movement during walking and were compared to static contraction at matched EMG activity (for a total of 17 subjects for the comparison static vs. walking). EMG levels were considered to be similar if the difference in their normalized amplitude was less than 5% MVC. Pairs of conditioning (0.7 T) and test (1.2 T) stimuli were applied through the same figure-of-eight coil at an interval of 5 s. A timecourse of the effect of the conditioning (subthreshold) stimulation on the test stimulus (suprathreshold) was obtained in the majority of subjects and the interval at which the largest inhibition was observed was chosen (Fig. 1C). This interval was usually 2 ms and it was used for all experiments, except for one subject, in whom 3 ms was used. Ten test responses without and 10 responses with the preceding conditioning stimulus were recorded. The peak-to-peak amplitude of the motor evoked potential was measured and averaged. The size of SICI is the size of the conditioned MEP as a percentage of the control MEP. During locomotion the test stimuli were given at the same time-points as what was described for modulation of MEP in the previous section.

SICI was further tested during tonic contraction at different EMG levels in five subjects to determine the correlation between background EMG and percentage of inhibition.

Subthreshold TMS

Subthreshold TMS was tested for the PD muscle in 25 subjects during walking and assessment of subthreshold inhibition was made during both walking and tonic contraction in 16 of them. We first determined the period of activity of the muscle by having subjects walk for 1 min on the treadmill. An appropriate delay from the heel trigger was then decided to ensure that the stimulation was applied during the EMG activity. The rectified EMG signal was averaged and monitored on a computer for on-line inspection of the effect of the stimulation. With this technique, we used a large number of stimulated and non-stimulated sweeps (50 to 75 sweeps for the stimulated condition and the equivalent number of sweeps for the control condition) to investigate the effect of TMS. The average of such a large number of trials revealed effects of TMS that are not obvious in single trials (see also Petersen et al. 2001). To facilitate comparisons with other studies we used the term MEP threshold in the usual way, as described previously (cf. Cortical stimulation section), whereas we used the term facilitation threshold to denote the intensity at which a clear facilitation (this could be a small MEP, but mostly it is below the traditional MEP threshold) was observed in the average of all trials.

To reveal the inhibition we first stimulated at around MEP threshold during the EMG burst of the PD and then decreased the intensity. At first, both MEP facilitation and the subsequent inhibition can be observed. When the intensity was below the facilitation threshold, only the inhibition was observed (Fig. 1D). Different delays from the time of heel contact were examined. The exact onset of the stimulation depended on the stride length (speed) chosen by the subject and corresponded to the time of activity of the muscle. Sweeps of 1.5 to 2 s in length were recorded from the time of heel contact. Multiple delays could be evaluated simultaneously, and since each delay requires the averaging of 50–75 sweeps, a total of 100 to 250 sweeps could be collected during one acquisition sequence.

Analysis of subthreshold TMS

The average of sweeps recorded for trials with stimulation was superimposed on the average of sweeps recorded for trials without stimulation. The onset and end of the facilitation and suppression were estimated by visual inspection of the recordings and more precisely when the stimulated trace was below or above a difference of one standard deviation (s.d.) from the control trace. The presence of facilitation was determined if the mean amplitude of the stimulated trace was larger than one s.d. from the mean amplitude of the control EMG activity at the expected latency (around the latency of the MEP). For quantification of the suppression, the mean area of EMG activity was measured between two cursors placed at the onset and end of the suppression. The size of the suppression was expressed as the size of this area as a percentage of control EMG area. The control EMG area was measured in the same time-window (between the cursors) during unstimulated trials. The measurements were made only for averages where facilitation was not observed.

Isometric contraction

In order to compare MEP, SICI and subthreshold inhibition obtained during locomotion to the same measurements recorded during a non-locomotor task, all subjects were seated and asked to isometrically contract the PD muscle against a solid resistance (wooden board; see section Cortical stimulation) to match the level of EMG activity recorded during gait (difference < 5% MVC). The contraction was made when the arm was in neutral position. This corresponds to the shoulder angle at the time of the first burst of PD EMG activity during gait. Subjects were given feedback of their performance using an oscilloscope displaying the rectified and integrated (time constant: 200 ms) EMG activity of the PD. Stimuli (whether for MEP, SICI or subthreshold) were applied in a randomized manner (minimal interstimulus interval of 1.5 s) and a time-window of 50 ms pre-stimulus was chosen to measure the background EMG level in the isometric control experiments.

Statistics

Differences in EMG area between stimulated and non-stimulated trials were compared statistically for subthreshold TMS. For TMS measures of MEP and SICI, most of the recorded data had a normal distribution and Student's paired t test was performed for comparison of those parameters between the two tasks (walking and tonic contraction). For data that were not normally distributed, a non-parametric test, the Rank sum test, was applied. Repeated measures ANOVA (RM ANOVA) was used to determine significance of comparison of the MEP and SICI modulation during locomotion. All RM ANOVA tests used in this study were followed by pairwise multiple comparison procedures (Student–Newman–Keuls method). For all tests, a difference was significant if P < 0.05. The Holm–Sidak method was used in order to correct the P value for multiple comparisons. Averages are presented as mean ±s.e.m. (standard error of the mean).

Results

Arm muscle EMG activity during gait

All 35 investigated subjects showed EMG activity in one or more of the proximal arm muscles when walking on the treadmill (3.4–4 km h−1). Note that no encouragements for using the arms were given during the experiments and the subjects were only instructed to ‘walk at a comfortable pace’. Consequently, the EMG pattern observed in the arm muscles during walking varied from subject to subject, even between subjects walking at the same preferred speed. EMG activity was seen most consistently in the posterior deltoid (n= 35) and the triceps brachii muscles (n= 19/31). The anterior deltoid and the biceps muscles were only clearly active in 5 and 10 subjects, respectively out of the 31 subjects in which it was tested. The extensor and flexor carpi radialis muscles were active in 5 and 8 subjects out of the 13 subjects assessed.

In the majority of subjects (n= 25), an EMG activity pattern similar to that of the subject illustrated in Fig. 2A–C was observed. In this subject, the elbow joint marker remained behind the shoulder joint marker, along the antero-posterior axis (Fig. 2A and B), although the wrist marker moved ahead of the shoulder marker during the forward movement (Fig. 2B). The posterior deltoid (PD) muscle was active both at the transition from forward to backward swing of the arm and, to a lesser degree, the transition from backward to forward swing. In the majority of these subjects the triceps was also active but the anterior deltoid (AD) and biceps showed weaker EMG activity (Fig. 2C). In a smaller group of subjects (n= 5), the excursion of movement at the shoulder was significantly larger and a flexion of the shoulder joint occurred beyond the neutral position (arm along the trunk). The AD showed significant activity, having two bursts of EMG: at the transition from backward to forward swing and, to a lesser degree, at the transition between forward and backward swing. Although there was EMG activity in posterior deltoid in the remaining subjects, clear bursts were difficult to distinguish. The activity pattern of the triceps muscle was similar to that of the PD and alternated with EMG of the biceps. Since the PD was the only muscle consistently active in all participants, the subsequent analysis will mainly concentrate on findings for that muscle.

Figure 2. EMG and kinematics.

Figure 2

Open in a new tab

A, stick figures representing the right arm and forearm of a subject during locomotion at 3.4 km h−1. B, shoulder (black), elbow (dark grey) and wrist (light grey) marker excursion of the same subject walking on the treadmill in the transversal plane, movement along the anteroposterior axis. The origin of the y-axis (0) was set at the front of the treadmill belt. The x-axis represents the time of the gait cycle and corresponds to the EMG averaging of the same subject walking on the treadmill, shown in C. Heel trigger corresponds to the time of the heel contact of the ipsilateral leg, which represents the beginning of the gait cycle. The stance phase (pale grey) and swing phase (dark grey) are also displayed.

Modulation of MEP

In order to investigate the excitability of the corticospinal projections to arm muscles at different times during the gait cycle, TMS was applied over the cortical arm area at an intensity of 1.2 T in seven subjects. With this stimulus intensity, MEPs were observed in all recorded arm muscles. The modulation of the PD MEP is shown for a single subject in Fig. 3A and for the population of subjects in Fig. 3B. To determine modulation, the assessment of MEP was made at three different time-points during the gait cycle (in Fig. 3A, arrows specify the 3 time-points assessed in 1 subject), based on the kinematics of the arm and the EMG activity of PD (right side of Fig. 3B). The MEPs were generally very small during the middle of the forward movement, when the muscle was less active (MEP = 3.9 ± 1.4% MVC) and increased considerably in size during the two bursts of EMG activity at the time of maximal forward (MEP = 10.7 ± 1.6% MVC) and backward excursion (MEP = 8.4 ± 2% MVC) of the arm. The difference in amplitude of MEP between the middle point of the forward movement and the maximal backward and forward movement (1st and 2nd EMG burst) was statistically significant (RM ANOVA, P < 0.001 and P= 0.02, respectively). Figure 3C shows the amplitude of the MEPs during walking (grey columns) compared to the amplitude of the MEPs during tonic contraction of the PD (black columns) at the same background EMG level (right side of the graph) and at similar shoulder angle (at the neutral position, when the arm was along the trunk). No statistically significant difference was found between the MEPs during walking and static contraction (Student's paired t test, P= 0.2).

Figure 3. Modulation of MEP in PD.

Figure 3

Open in a new tab

A, timecourse of MEP amplitude at different times during gait cycle of a single subject. Lower panel is the rectified EMG averaging of the PD muscle. X-axis represents the time between heel strike (0) and the application of the stimulation. Upper panel displays the amplitude of MEP as % of maximal voluntary contraction (MVC) at each stimulation point. The dashed lines mark the onset of EMG bursts. Arrows point to the three time-points assessed during gait to analyse the modulation of PD MEP amplitude. B, averaged PD MEP amplitude and background EMG during walking for all subjects at 3 different time-points during gait: when the PD EMG activity was low, i.e. while the arm was mid-way during the forward movement (200–300 ms after heel strike, varying between subjects; black column, middle forward movement). The grey column represents the amount of MEP during the bursts of EMG activity at the maximal forward movement (400–700 ms after heel strike; maximal forward movement). The white column is at the time of the maximal backward excursion of the arm (900 to 1200 ms after heel strike; maximal backward movement. One subject did not exhibit activity in the PD during that period). C, comparison of MEP amplitude during locomotion (taken at maximal forward movement; grey) and tonic contraction (black) for all subjects. The comparison was made at similar shoulder angle and similar background EMG (right axis). MEP_N: normalized MEP; B-EMG_N: normalized background EMG. Amplitude is presented as mean ±s.e.m.

Short-latency intracortical inhibition

SICI was evoked during locomotion by conditioning the PD MEP by subthreshold TMS (0.7 T) at an interval of 2 ms (Fig. 1C) in 14 subjects. In the subject displayed in Fig. 4A, the MEP was depressed to less than 30% of its control size by the conditioning stimulation in the phase of the gait cycle when the PD was mainly silent. During the two bursts of activity at the maximal excursions of the arm, the amount of inhibition decreased considerably so that the MEP was depressed to only 99% of its control size for the first burst and 60% for the second burst. A similarly decreased inhibition was observed during the EMG bursts in all 14 subjects. Figure 4B shows the average amount of SICI for all subjects, along with the size of MEP and background EMG. The inhibition was higher during the middle of the forward movement (MEP was reduced to 69 ± 5% of its original size) than during the maximal backward and forward movement (MEP was 94 ± 7% and 86 ± 9% of their respective original size). The difference between the inhibition obtained during the middle of the forward movement, and the maximal forward movement (first EMG burst) was statistically significant (RM ANOVA P= 0.007), but less so between the middle of the forward movement and the maximal backward movement (RM ANOVA P= 0.04, not significant when correction for multiple comparison is applied). The inhibition obtained during maximal backward and maximal forward movement was not statistically different from each other (RM ANOVA, P= 0.3).

Figure 4. Modulation of SICI in PD.

Figure 4

Open in a new tab

A, timecourse of SICI at different time-points during the gait cycle of a single subject. Amplitude of SICI activity is plotted as a percentage of control MEP. X-axis represents the time between heel strike (0) and the application of the stimulation. B, group comparison of SICI, test MEP and background EMG at 3 different times in gait for all subjects: middle point of the forward movement (black column), maximum forward excursion of the arm (grey column), and maximal backward excursion of the arm (white) On the left axis is the amplitude of SICI and on the right axis, the amplitude of test MEP (MEP normalized; MEP_N) and background EMG (normalized; B-EMG_N). C, comparison of SICI during locomotion (grey) and tonic contraction (black) was made at the same amplitude of background EMG, MEP and shoulder angle. D, amplitude of SICI during tonic contraction was plotted against different levels of background EMG. In B and C, amplitude is presented as mean ±s.e.m.

In Fig. 4C, the level of SICI at the maximal forward movement was compared to the level of SICI obtained during tonic contraction, while the arm was in a similar position, in 17 subjects. SICI was not statistically different between tonic contraction and locomotion (n= 17, Student's paired t test, P= 0.2) at a comparable level of background EMG (4.9 ± 1% MVC for locomotion and 5.7% MVC for tonic contraction; Rank sum test, P= 0.2), test MEP (15.6 ± 2% MVC for locomotion and 15 ± 2% for tonic contraction; Rank sum test, P= 0.97) and arm position (in line with the trunk).

In five subjects, different levels of background EMG from 0 to levels above that observed during locomotion were tested, and SICI was evaluated at each of those levels. Figure 4D shows that no direct correlation could be drawn between SICI and the level of EMG activity obtained during static contraction. Furthermore, no direct correlation between SICI and MEP size was observed. Thus, the modulation of SICI observed during gait was not solely due to changes in the background EMG.

Suppression of the arm EMG activity by subthreshold TMS

Stimulation of the motor cortex by subthreshold TMS evoked a suppression of the rectified averaged PD EMG activity in 20 out of 25 subjects in whom it was tested.

Data obtained from a single subject are shown in Fig. 5. TMS was applied 600 ms after ipsilateral heel contact, which corresponds to the burst of PD EMG activity during maximal forward excursion of the arm. At an intensity of 72% of maximal stimulator output (0.95 T; Fig. 5A), a short-latency facilitation was observed at a latency of 15 ms. This facilitation in all likelihood reflects activation of descending corticospinal tract fibres to the active PD motoneurones (Davey et al. 1994; Petersen et al. 2001). The facilitation was followed by suppression at a latency of 29 ms. When the stimulus intensity was decreased to 55% of maximum stimulator output (0.72 T; Fig. 5B), the facilitation was decreased but the suppression could still be seen. At 52% the facilitation was no longer observed, suggesting that the stimulus intensity was now below threshold for activation of corticospinal tract fibres (0.68 T; Fig. 5C). However, the later suppression could still be observed (marked by black arrow in Fig. 5C). Inhibition of the PD EMG without any prior facilitation was observed in 11 of the 20 subjects in whom inhibition was observed. In these 11 subjects, the average latency of the suppression was 25 ± 2 ms. The latency of the facilitation in the same subjects was on average 14 ± 1 ms and the size of the suppression was 22 ± 2%, i.e. the EMG area during the inhibition was 78% of the control EMG area.

Figure 5. Subthreshold inhibition in PD.

Figure 5

Open in a new tab

A, B and C, traces of PD EMG during locomotion when subthreshold TMS was applied 600 ms after heel strike (dark trace) or in the control state (light grey) at different intensities. Vertical lines show the onset and offset of inhibition. The white arrows point to the facilitation and the black arrows point to the inhibited portion of the EMG.

Inhibition was observed in other muscles such as triceps (n= 7), biceps (n= 5), AD (n= 2) and FCR and/or ECR (n= 5 for both) when those muscles were active. Examples of the EMG suppression in these muscles are shown for the triceps (Fig. 6A), the biceps (Fig. 6B) and the ECR (Fig. 6C).

Figure 6. Subthreshold inhibition in other arm muscles.

Figure 6

Open in a new tab

Inhibition of ongoing EMG evoked by subthreshold TMS in triceps (A), biceps (B) and ECR (C).

Subthreshold TMS was also applied during static voluntary contraction of the PD in 16 of the subjects. A suppression of EMG activity was observed in 15 subjects, and suppression without prior facilitation could be observed in 11 of them. The latency of the suppression was 25 ± 2 ms and the latency of the facilitation obtained at higher intensity was 16 ± 1 ms. In seven subjects, an inhibition without prior facilitation could be shown both during locomotion and tonic contraction at similar EMG levels. Overall, in these subjects, a larger suppression during voluntary static contraction as compared with walking was observed at similar background EMG. Figure 7 shows that area corresponded to 85 ± 1% of the control EMG during walking (decrease of 15%) and 74 ± 1% during static contraction (decrease of 26%; Student's paired t test P= 0.03). The TMS intensity (mean of 50 ± 1% for sitting vs. 48 ± 1% for walking, Rank sum test, P= 0.5) was similar in the two situations.

Figure 7. Subthreshold inhibition – group results.

Figure 7

Open in a new tab

The inhibited area of EMG for individuals (grey) as well as the group (mean ±s.e.m.; black) is plotted as a percentage of background EMG during control trials, in both locomotion and tonic contraction. The y-axis depicts the area of the EMG during the inhibitory period as a percentage of the control EMG in steps without stimulation. n= 7.

Discussion

We have shown in this study that PD motor evoked potentials are modulated with the background EMG activity during walking, that short-latency intracortical inhibition (SICI) is diminished during bursts of PD EMG activity and finally that subthreshold TMS evokes a suppression of the PD EMG activity during gait. These observations support the idea that the motor cortex and the corticospinal tract are an integrated part of the central network involved in the generation of EMG activity in arm muscles during human walking, as previously demonstrated for the leg.

Modulation of motor evoked potentials during walking

Modulation of MEPs during walking has been reported previously for leg muscles by Schubert et al. (1997), Capaday et al. (1999) and Bonnard et al. (2002). Although this modulation may partly be explained by changes in the excitability of corticospinal neurones (for review and discussion: Rothwell, 1997; Petersen et al. 2003), it is also clear that changes in the excitability of spinal interneurones and motoneurones contribute importantly (Nielsen et al. 1999; Schneider et al. 2004). Most of the modulation of the PD MEPs in the present study and of the leg muscle MEPs in the studies by Schubert et al. (1997), Capaday et al. (1999) and Bonnard et al. (2002) is thus most probably mainly caused by the rhythmic changes in spinal motoneuronal excitability during the gait cycle. The lack of difference in MEPs recorded during gait and during static contraction at similar background EMG levels, and thus presumably similar spinal motoneuronal excitability, could be taken to indicate that there were no major differences in the excitability of corticospinal neurones during the two tasks. However, the MEPs are also influenced by transmission in indirect pathways to the spinal motoneurones and changes in the excitability of, for example, spinal interneurones, may also contribute to differences in the MEPs in different tasks (Nielsen et al. 1999). Furthermore, a comparable background EMG activity during two different tasks does not rule out the possibility of significant changes in the recruitment gain within the motoneuronal pool, which may dramatically change the responsiveness of the pool to inputs (Kernell & Hultborn, 1990; Nielsen & Kagamihara, 1993). Finally, it is unclear whether the descending volley elicited by TMS will always increase with the level of cortical activity, which is the prerequisite for using MEPs as a measure of ‘cortical excitability’ (Matthews, 1999; see, however, Di Lazzaro et al. 1998a).

Modulation of intracortical inhibition during walking

SICI was reduced during the two bursts of EMG activity in the PD muscle towards the maximal forward and backward excursion of the arm during gait (Fig. 4). It seems likely that this reflects a reduced excitability of the involved inhibitory cortical interneurones, possibly in order to ensure an unhindered activity of corticospinal neurones, similar to what has been argued for the reduction of SICI during voluntary muscle activation as compared with rest (Kujirai et al. 1993; Ziemann et al. 1996). If so, this would again imply that the corticospinal neurones are active and contribute to the EMG activity in the PD muscle during gait. The conditioning stimulation in SICI is thought to activate intracortical inhibitory circuits, which will then suppress the subsequent test stimulus when given at a short interval (2–3 ms; Kujirai et al. 1993). It might be argued that the reduction of SICI could be explained by the larger control MEPs during the bursts of EMG activity, but this is unlikely since the size of the control MEP was found to have little influence on SICI during tonic contraction. It seems more likely that the phasically reduced SICI during the bursts of EMG activity during gait reflects the need of reducing intracortical inhibition of the corticospinal drive to the spinal motoneurones.

Subthreshold TMS and modulation of CST excitability

That corticospinal drive indeed does seem to contribute directly to the bursts of PD EMG activity during gait was shown by the observation that subthreshold TMS produced a suppression of the ongoing EMG activity. It has been demonstrated previously that TMS at these intensities does not produce any descending activity, but only activates cortical inhibitory interneurones (Di Lazzaro et al. 1998b). Stimulation of the corticospinal tract below the cortex also fails to produce a similar inhibition (Davey et al. 1994; Petersen et al. 2001). The suppression must therefore reflect the inhibitory effect of the cortical inhibitory interneurones on the activity of the corticospinal neurones. The observation that a suppression could be produced during walking consequently demonstrates – by inference – that corticospinal neurones contribute to the EMG (see also Davey et al. 1994; Petersen et al. 2001).

The suppression of the PD EMG activity was observed without prior facilitation in 11 of the 20 subjects in whom it was investigated during walking. This does not imply that corticospinal neurones do not contribute to the EMG activity in the remaining subjects. Demonstration of the suppression rests on the ability to activate the inhibitory cortical interneurones at a lower intensity than that required to activate the corticospinal neurones by TMS (see Davey et al. 1994; Petersen et al. 2001). The threshold of these two populations of neurones is likely to be very close to each other especially during the PD EMG bursts during walking. The large size of the MEPs (Fig. 3) suggests that the excitability of the corticospinal neurones is quite high and the small size of SICI suggests that the excitability of the inhibitory interneurones is quite low (Fig. 4). In view of this, it is remarkable that we were nevertheless able to demonstrate the suppression in so many subjects.

The suppression was also observed in several other arm muscles (AD, biceps, triceps, ECR) which were active at the same or different time as PD, although we did not specifically search for it in these muscles. There was a small tendency for the suppression to be larger in the more distal muscles and if so this might be taken as evidence for a more significant corticospinal drive to these muscles. However, it might also simply reflect a difference in the excitability of the cortical inhibitory neurones and/or the different location of the respective motor representations with respect to the coil.

The size of the suppression was larger during tonic voluntary contraction than during walking at matched background EMG activity. This does not necessarily indicate a difference in corticospinal drive, but could simply reflect a difference in the excitability of the cortical inhibitory interneurones (if the excitability is low, fewer interneurones would be activated and they would produce less inhibition of the corticospinal neurones). As already discussed, SICI was found to be quite small in the present study during walking and we may therefore assume that the interneurones were not as efficient in producing inhibition of the corticospinal neurones in this situation as they were during tonic contraction. The observation of a sizable suppression of the EMG despite low excitability of the inhibitory interneurones may thus conceal a powerful corticospinal drive during walking, but we have no way of judging this. The suppression of the EMG can thus only be used to provide positive confirmation of the contribution of corticospinal drive to the ongoing EMG activity, but it cannot be used to draw conclusions regarding the size of that drive, how significant it is for the EMG activity or indeed for differences in the drive during two different tasks.

Central control of arm muscle activity during walking

It has been demonstrated in the cat and other vertebrates that the isolated spinal cord has the capacity to generate the basic locomotor activity in the absence of any descending motor control (Grillner, 1981). Indeed, the degree of walking ability that may be obtained in animals with lesions of the motor cortex and/or corticospinal tract or other supraspinal pathways is remarkable and underscores the capacity of these subcortical circuitries in controlling complex multi-limb motor behaviours (Grillner, 1981; Rossignol et al. 1996). However, such observations do not rule out that the motor cortex and corticospinal tract under normal circumstances in the intact animal contributes in an important way to the control of walking. It seems likely that this is even more so in human subjects where the corticospinal tract appears to be of generally larger importance than in the cat (Lemon & Griffiths, 2005). In our opinion, the motor cortex, the corticospinal tract and the spinal cord circuitries, including a possible central pattern generator in the spinal cord, should not be seen as separate entities in the intact human subjects, but rather as integrated circuitries that interact heavily to ensure an optimal control under all circumstances. As already mentioned, the present study does not determine how important the corticospinal drive to the arm muscles is during gait. It is probably only one of several inputs which all contribute to some extent to raise the excitability of the neurones above their threshold and control their firing frequency in the appropriate phases of the gait cycle.

We have no way of telling from these experiments how important this drive is or whether the corticospinal drive is caused by input to the motor cortex from premotor areas in the cortex, from the basal ganglia, from the cerebellum or possibly from sensory afferent inputs as part of transcortical reflex loop activity (Christensen et al. 1999, 2001). The EMG activity in the PD muscle was observed towards the maximal excursion of the forward and backward arm movement during gait (Fig. 2) and it is therefore possible that some of it was created by stretch reflex activity. However, since a significant part of the EMG activity was seen prior to the maximal backwards arm excursion when the PD muscle was shortening, this is unlikely to explain all of the EMG activity.

It was not the aim of the present study to investigate the functional significance of the EMG activity in the arm muscles during walking. However, the characteristic activity pattern in the deltoid muscle suggests that the swing of the arms may not be primarily generated by the EMG activity, but rather may be similar to a mechanical pendulum counteracting the movement of the torso and lower limb (Lieberman et al. 2007, 2008; Pontzer et al. 2009). Indeed, the EMG activity might be more involved in curtailing the forward and backward swing of the arm and initiating the phase shift from flexion into extension and vice versa. In this way it may act to ensure that the swing of the arms contributes to maintaining a low metabolic cost of walking (Collins et al. 2009). The observations in the present study extend previous findings showing coupling and parallel control of arm and leg movements during walking (Dietz et al. 2001; Dietz, 2002; Zehr et al. 2004, 2009) to the level of the motor cortex and corticospinal tract. This emphasizes the need of considering the corticospinal input to the cervical and lumbar spinal locomotor circuitries and the interaction between arm and leg movements when interpreting the recovery of walking ability following lesions of descending motor tracts. It also opens up the possibility of taking advantage of the corticospinal drive to the cervical spinal cord and the activation of the arm muscles in facilitating recovery of gait ability following lesions.

Acknowledgments

This work was supported by Canadian Institutes of Health Research (CIHR), the Danish Medical Research council and The Elsass foundation.

Glossary

Abbreviations

AD

anterior deltoid

ECR

extensor carpi radialis

FCR

flexor carpi radialis

MEP

motor evoked potentials

Mmax

maximal M wave

MVC

maximal voluntary contraction

PD

posterior deltoid

RM ANOVA

repeated measures ANOVA

SICI

short latency intracortical inhibition

T

threshold

TMS

transcranial magnetic stimulation

Author contributions

Conception, design, analysis and interpretation of data were done by both D.B. and J.B.N. Drafting the article was done by D.B. and revising was done by both authors. Final approval of the version to be published was done by both authors. All experiments were done in Copenhagen, Denmark.

References

  1. Armstrong DM, Drew T. Forelimb electromyographic responses to motor cortex stimulation during locomotion in the cat. J Physiol. 1985;367:327–351. doi: 10.1113/jphysiol.1985.sp015827. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ballesteros ML, Buchthal F, Rosenfalck P. The pattern of muscular activity during the arm swing of natural walking. Acta Physiol Scand. 1965;63:296–310. doi: 10.1111/j.1748-1716.1965.tb04069.x. [DOI] [PubMed] [Google Scholar]
  3. Barthélemy D, Korchi M, Nielsen JB. Society for the Neural Control of Movement. Spain: 2007a. Corticospinal drive contributes to arm muscle activity during human walking (abstract) [Google Scholar]
  4. Barthélemy D, Korchi M, Nielsen JB. International Society for Posture and Gait Research (ISPGR) USA: 2007b. Corticospinal drive contributes to arm muscle activity during human walking (abstract) [Google Scholar]
  5. Barthélemy D, Nielsen JB. Third International Conference on Transcranial Magnetic and Direct Current Stimulation. Germany: 2008. Contribution of the corticospinal tract to arm muscle activity during human walking (abstract) [Google Scholar]
  6. Bonnard M, Camus M, Coyle T, Pailhous J. Task-induced modulation of motor evoked potentials in upper-leg muscles during human gait: a TMS study. Eur J Neurosci. 2002;11:2225–2230. doi: 10.1046/j.1460-9568.2002.02295.x. [DOI] [PubMed] [Google Scholar]
  7. Bretzner F, Drew T. Contribution of the motor cortex to the structure and the timing of hindlimb locomotion in the cat: a microstimulation study. J Neurophysiol. 2005;94:657–672. doi: 10.1152/jn.01245.2004. [DOI] [PubMed] [Google Scholar]
  8. Capaday C, Lavoie BA, Barbeau H, Schneider C, Bonnard M. Studies on the corticospinal control of human walking. I. Responses to focal transcranial magnetic stimulation of the motor cortex. J Neurophysiol. 1999;81:129–139. doi: 10.1152/jn.1999.81.1.129. [DOI] [PubMed] [Google Scholar]
  9. Christensen LO, Andersen JB, Sinkjaer T, Nielsen J. Transcranial magnetic stimulation and stretch reflexes in the tibialis anterior muscle during human walking. J Physiol. 2001;531:545–557. doi: 10.1111/j.1469-7793.2001.0545i.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Christensen LO, Morita H, Petersen N, Nielsen J. Evidence suggesting that a transcortical reflex pathway contributes to cutaneous reflexes in the tibialis anterior muscle during walking in man. Exp Brain Res. 1999;124:59–68. doi: 10.1007/s002210050600. [DOI] [PubMed] [Google Scholar]
  11. Collins SH, Adamczyk PG, Kuo AD. Dynamic arm swinging in human walking. Proc Biol Sci. 2009;276:3679–3688. doi: 10.1098/rspb.2009.0664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Davey NJ, Romaiguere P, Maskill DW, Ellaway PH. Suppression of voluntary motor activity revealed using transcranial magnetic stimulation of the motor cortex in man. J Physiol. 1994;477:223–235. doi: 10.1113/jphysiol.1994.sp020186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Di Lazzaro V, Restuccia D, Oliviero A, Profice P, Ferrara L, Insola A, Mazzone P, Tonali P, Rothwell JC. Effects of voluntary contraction on descending volleys evoked by transcranial stimulation in conscious humans. J Physiol. 1998a;508:625–633. doi: 10.1111/j.1469-7793.1998.625bq.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Di Lazzaro V, Restuccia D, Oliviero A, Profice P, Ferrara L, Insola A, Mazzone P, Tonali P, Rothwell JC. Magnetic transcranial stimulation at intensities below active motor threshold activates intracortical inhibitory circuits. Exp Brain Res. 1998b;119:265–268. doi: 10.1007/s002210050341. [DOI] [PubMed] [Google Scholar]
  15. Dietz V. Do human bipeds use quadrupedal coordination? Trends Neurosci. 2002;25:462–467. doi: 10.1016/s0166-2236(02)02229-4. [DOI] [PubMed] [Google Scholar]
  16. Dietz V, Fouad K, Bastiaanse CM. Neuronal coordination of arm and leg movements during human locomotion. Eur J Neurosci. 2001;14:1906–1914. doi: 10.1046/j.0953-816x.2001.01813.x. [DOI] [PubMed] [Google Scholar]
  17. Drew T. Motor cortical activity during voluntary gait modifications in the cat. I. Cells related to the forelimbs. J Neurophysiol. 1993;70:179–199. doi: 10.1152/jn.1993.70.1.179. [DOI] [PubMed] [Google Scholar]
  18. Drew T, Jiang W, Widajewicz W. Contributions of the motor cortex to the control of the hindlimbs during locomotion in the cat. Brain Res Brain Res Rev. 2002;40:178–191. doi: 10.1016/s0165-0173(02)00200-x. [DOI] [PubMed] [Google Scholar]
  19. Elftman H. The function of the arms in walking. Hum Biol. 1939;11:529–535. [Google Scholar]
  20. Ellaway PH, Prochazka A, Chan M, Gauthier MJ. The sense of movement elicited by transcranial magnetic stimulation in humans is due to sensory feedback. J Physiol. 2004;556:651–660. doi: 10.1113/jphysiol.2003.060483. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Grillner S. Control of locomotion in bipeds, tetrapods, and fish. In: Brookhart JM, Mountcastle VB, editors. Handbook of Physiology. The Nervous System II. Bethesda, MD: Am. Physiol Soc; 1981. pp. 1179–1236. [Google Scholar]
  22. Heisenberg M. Is free will an illusion? Nature. 2009;459:164–165. doi: 10.1038/459164a. [DOI] [PubMed] [Google Scholar]
  23. Kernell D, Hultborn H. Synaptic effects on recruitment gain: a mechanism of importance for the input-output relations of motoneurone pools? Brain Res. 1990;507:176–179. doi: 10.1016/0006-8993(90)90542-j. [DOI] [PubMed] [Google Scholar]
  24. Kujirai T, Caramia MD, Rothwell JC, Day BL, Thompson PD, Ferbert A, Wroe S, Asselman P, Marsden CD. Corticocortical inhibition in human motor cortex. J Physiol. 1993;471:501–519. doi: 10.1113/jphysiol.1993.sp019912. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lemon RN, Griffiths J. Comparing the function of the corticospinal system in different species: organizational differences for motor specialization? Muscle Nerve. 2005;32:261–279. doi: 10.1002/mus.20333. [DOI] [PubMed] [Google Scholar]
  26. Lieberman DE, Bramble DM, Raichlen DA. Integration of the head and forelimb in bipedal hominins. J Morphol. 2007;268:1099. [Google Scholar]
  27. Lieberman DE, Bramble DM, Raichlen DA, Whitcome KW. Functional, developmental and morphological integration: the case of the head and forelimb in bipedal hominins. Am J Phys Anthropol. 2008;135(S46):140–141. [Google Scholar]
  28. Liepert J, Classen J, Cohen LG, Hallett M. Task-dependent changes of intracortical inhibition. Exp Brain Res. 1998;118:421–426. doi: 10.1007/s002210050296. [DOI] [PubMed] [Google Scholar]
  29. Matthews PB. The effect of firing on the excitability of a model motoneurone and its implications for cortical stimulation. J Physiol. 1999;518:867–882. doi: 10.1111/j.1469-7793.1999.0867p.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Nielsen J, Kagamihara Y. Differential projection of the sural nerve to early and late recruited human tibialis anterior motor units: change of recruitment gain. Acta Physiol Scand. 1993;147:385–401. doi: 10.1111/j.1748-1716.1993.tb09515.x. [DOI] [PubMed] [Google Scholar]
  31. Nielsen J, Morita H, Baumgarten J, Petersen N, Christensen LO. On the comparability of H-reflexes and MEPs. Electroencephalogr Clin Neurophysiol Suppl. 1999;51:93–101. [PubMed] [Google Scholar]
  32. Nielsen JB. How we walk: central control of muscle activity during human walking. Neuroscientist. 2003;9:195–204. doi: 10.1177/1073858403009003012. [DOI] [PubMed] [Google Scholar]
  33. Petersen N, Christensen LO, Nielsen J. The effect of transcranial magnetic stimulation on the soleus H reflex during human walking. J Physiol. 1998;513:599–610. doi: 10.1111/j.1469-7793.1998.599bb.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Petersen NT, Butler JE, Marchand-Pauvert V, Fisher R, Ledebt A, Pyndt HS, Hansen NL, Nielsen JB. Suppression of EMG activity by transcranial magnetic stimulation in human subjects during walking. J Physiol. 2001;537:651–656. doi: 10.1111/j.1469-7793.2001.00651.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Petersen NT, Pyndt HS, Nielsen JB. Investigating human motor control by transcranial magnetic stimulation. Exp Brain Res. 2003;152:1–16. doi: 10.1007/s00221-003-1537-y. [DOI] [PubMed] [Google Scholar]
  36. Pontzer H, Holloway JH, Raichlen DA, Lieberman DE. Control and function of arm swing in human walking and running. J Exp Biol. 2009;212:523–534. doi: 10.1242/jeb.024927. [DOI] [PubMed] [Google Scholar]
  37. Prochazka A, Clarac F, Loeb GE, Rothwell JC, Wolpaw JR. What do reflex and voluntary mean? Modern views on an ancient debate. Exp Brain Res. 2000;130:417–432. doi: 10.1007/s002219900250. [DOI] [PubMed] [Google Scholar]
  38. Rossignol S, Chau C, Brustein E, Belanger M, Barbeau H, Drew T. Locomotor capacities after complete and partial lesions of the spinal cord. Acta Neurobiol Exp (Wars) 1996;56:449–463. doi: 10.55782/ane-1996-1148. [DOI] [PubMed] [Google Scholar]
  39. Rothwell JC. Techniques and mechanisms of action of transcranial stimulation of the human motor cortex. J Neurosci Methods. 1997;74:113–122. doi: 10.1016/s0165-0270(97)02242-5. [DOI] [PubMed] [Google Scholar]
  40. Schneider C, Lavoie BA, Barbeau H, Capaday C. Timing of cortical excitability changes during the reaction time of movements superimposed on tonic motor activity. J Appl Physiol. 2004;97:2220–2227. doi: 10.1152/japplphysiol.00542.2004. [DOI] [PubMed] [Google Scholar]
  41. Schubert M, Curt A, Jensen L, Dietz V. Corticospinal input in human gait: modulation of magnetically evoked motor responses. Exp Brain Res. 1997;115:234–246. doi: 10.1007/pl00005693. [DOI] [PubMed] [Google Scholar]
  42. Zehr EP, Carroll TJ, Chua R, Collins DF, Frigon A, Haridas C, Hundza SR, Thompson AK. Possible contributions of CPG activity to the control of rhythmic human arm movement. Can J Physiol Pharmacol. 2004;82:556–568. doi: 10.1139/y04-056. [DOI] [PubMed] [Google Scholar]
  43. Zehr EP, Hundza SR, Vasudevan EV. The quadrupedal nature of human bipedal locomotion. Exerc Sport Sci Rev. 2009;37:102–108. doi: 10.1097/JES.0b013e31819c2ed6. [DOI] [PubMed] [Google Scholar]
  44. Ziemann U, Rothwell JC, Ridding MC. Interaction between intracortical inhibition and facilitation in human motor cortex. J Physiol. 1996;496:873–881. doi: 10.1113/jphysiol.1996.sp021734. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES