Stimulation-induced changes in lower limb corticomotor excitability during treadmill walking in humans
1 Sensory Motor Performance Program, Rehabilitation Institute of Chicago, Chicago, IL, USA
2 Department of Physical Therapy, University of Illinois at Chicago, Department of Physical Medicine and Rehabilitation, Northwestern University, Chicago, IL, USA
Abstract
Magnetic stimulation of human primary motor cortex (M1) paired with electrical stimulation of a peripheral motor nerve has been used to produce a lasting modulation of corticomotor (CM) excitability. This ‘paired associative stimulation’ (PAS) protocol has been used to induce bidirectional changes in excitability in upper limb CM pathways. The present study tested the hypothesis that temporally dependent PAS applied to the common peroneal nerve during the swing phase of walking would induce bidirectional changes in CM excitability consistent with the Hebbian principle of activity-dependent plasticity. Fourteen subjects with no known neurological disorder participated in two data collection sessions each. PAS was delivered as a single block of 120 pairs of stimuli delivered in a 10 min period during treadmill walking at 4.0 km h–1. Changes in CM excitability were assessed by examining the size of motor potentials evoked by transcranial magnetic stimulation prior to and following PAS. Tibialis anterior motor-evoked potentials amplitude increased to 121% over baseline when the magnetic stimulus was delivered over M1 after the estimated arrival time of the afferent volley in sensorimotor cortex and decreased to 83% of baseline when the magnetic stimulus was delivered prior to the estimated afferent volley arrival. This extent of modulation was undiminished following a further 10 min period of walking without stimulation. The temporal nature of the bidirectional effects following PAS, their rapid evolution and subsequent persistence supported the study's hypothesis and were similar to the effects described by others in quiescent muscles of the upper limb.
, http://www.100md.com
* Corresponding author J. W. Stinear: Sensory Motor Performance Program, 345 E. Superior St, Chicago, IL 60611, USA.
Introduction
Modulation of corticomotor (CM) excitability can be induced in humans using a variety of neural stimulation techniques which invoke principles of activity- (or spike-timing-) dependent neural plasticity. For example, repetitive transcranial magnetic stimulation (rTMS) applied over human primary motor cortex (M1) has been used in a number of protocols to reveal a decrease in CM excitability when the frequency of stimulation was 1 Hz (Muellbacher et al. 2000), and an increase in CM excitability when the frequency of stimulation was 5 Hz or more (Peinemann et al. 2004). Pharmacological agents have been shown to increase (e.g. D-amphetamine, a noradrenaline (norepinephrine) agonist) or decrease CM excitability (e.g. lorazepam, an allosteric modulator of the GABAA receptor) (see Ziemann, 2004 for a review). Ischaemic nerve block of the upper limb in combination with rTMS has also been used to demonstrate stimulation-induced alterations in the size of motor-evoked potentials (MEPs) (Ziemann et al. 2002). More recently, a technique based on Hebbian principles of synaptic plasticity has been used to increase (Stefan et al. 2002; Wolters et al. 2003), and decrease the excitability of CM pathways to the upper limb (Wolters et al. 2003). This technique, called ‘paired associative stimulation’ (PAS), follows the association principle first proposed by D. O. Hebb, whereby changes in synaptic weighting are induced by a presynaptic neurone taking part in the repetitive firing of a postsynaptic cell (Hebb, 1949).
, 百拇医药
Specifically, Wolters et al. (2003) observed increases in the size of motor-evoked potentials (MEPs) elicited in the abductor pollicis brevis when suprathreshold TMS was delivered to M1 within 6 ms of sensory-evoked potentials arriving in sensory cortex. In contrast, decreases in MEP size were detected when TMS was delivered within 10 ms prior to the afferent volley arriving at the somatosensory cortex. The effects evolved rapidly, were long lasting, yet reversible.
, 百拇医药
Studies of PAS involving CM pathways to lower limb motoneurones are limited (Uy et al. 2003), and no studies have been reported where changes in CM excitability have been induced during muscle activation. The hypothesis adopted for the present experiment predicted that temporally dependent PAS applied to tibialis anterior (TA) pathways during the swing phase of walking would induce bidirectional changes in CM excitability consistent with the Hebbian principle of activity-dependent plasticity. We chose TA as the target muscle because the effects of PAS are likely to be more readily revealed in pathways that are known to have strong CM connections, and the functional importance of the TA motor pool is particularly evident in individuals following neurological injury (Said et al. 2001). Studies have revealed the relative strength of CM pathways to TA and soleus (Sol); for example, in response to TMS (Capaday et al. 1999) constructed input–output curves of CM pathways to TA and to Sol muscles. Pathways to Sol were less excitable during the stance phase of walking than during a tonic voluntary contraction. However, they found no difference between the excitability of projections to TA during the swing phase of walking and during a voluntary tonic contraction of TA. In a similar study Bawa et al. (2002) used surface electromyography (EMG) and single motor unit responses to provide further evidence that CM projections to TA are relatively stronger compared with projections to Sol.
, 百拇医药
During motor re-training, the inclusion of PAS to manipulate the excitability of injured CM pathways may be of value in the application of walking re-training regimens. Because PAS has been reported previously only in pathways to muscles at rest, and because of our interest in developing adjuncts to walking training protocols, the aim of the present experiment was to determine if bidirectional changes in healthy individuals' TA CM excitability could be induced with PAS applied during walking. If this aim is fulfilled, we intend to investigate the mechanisms of PAS further and assess its value as an adjunct to walking training following neural injury.
, 百拇医药
Methods
Subjects
Healthy subjects with no known neurological disorder were recruited from the staff of the Rehabilitation Institute of Chicago. The sample size required to reject the null hypothesis was calculated using preliminary data and PS v. 2.1.31 software (Dupont & Plummer, 1990), where the difference in means was 0.25 and the standard deviation was 0.30. The significance level was set at = 0.05 and the power at 1 –= 0.80, yielding a sample size of 13. Fourteen (6 female, 8 male) subjects volunteered and were recruited for the study. Their ages ranged from 24 to 58 years (mean 31). The study was approved by the Northwestern University Institutional Review Board. All methods conformed to the Declaration of Helsinki and all participants provided written informed consent.
, http://www.100md.com
Stimulation and recording
M1 was stimulated by a Magstim 2002 unit (Magstim, Dyfed, Wales, UK) via a Magstim double cone coil (see below for details regarding intensity setting). A tight fitting linen cap was tied on the subject's head with the vertex marked on the cap at the intersection of the inion–nasion and interaural lines. The coil was suspended over the subject's head from an overhead gantry and attached to the cap using Velcro tapes so that the junction of the coils was located in the mid-sagittal plane 1 cm anterior to the vertex. The large magnetic field generated by the double cone coil and the close proximity of homologous lower limb cortical representations allowed the coil to be fixed in the mid-sagittal plane thus eliminating the need to shift the coil to an optimum site for each TA. Minor adjustments were made to the coil position until optimal-sized MEPs of similar size could be elicited from both TA muscles during the respective limb's late swing phase of treadmill walking. A chin strap attached to the coil and small foam pads were used to increase coil stability. Subjects found this to be a comfortable arrangement, enabling them to adopt their preferred gait pattern, and move their head with ease between trials without causing a shift in the location of the coil. The coil and cap positions were checked repeatedly throughout data collection and coil position did not change in any of the sessions.
, 百拇医药
Peripheral nerve stimulation was delivered from a Model DS7A constant current stimulator (Digitimer Ltd, Hertfordshire, UK) via a Meditrace 133 surface Ag–AgCl electrode (Tyco Healthcare, Ludlow, Chicopee, MD, USA) used as a cathode placed over the deep peroneal nerve (PN) just distal and anterior to the fibula head at a position where maximum TA muscle contraction and minimum peroneal muscle contraction was elicited. The anode was a 25 cm2 Dura-Soft (Pain Management Technologies, Inc., Akron, OH, USA) electrode placed over the ipsilateral patella.
, 百拇医药
The timing of stimulator trigger pulses was achieved using custom-built heel switches and a two channel electronic delay circuit. The cycle phase was established using trigger pulses and EMG bursts displayed on an oscilloscope. The target cycle phase was the middle of the large TA burst that occurs in late swing as the ankle is dorsi-flexed prior to heel strike. The interstimulus interval (ISI) between Digitimer and Magstim units was set using a second component of the delay circuit.
, 百拇医药
Surface EMG was recorded from TA and Sol via Ag–AgCl EMG electrodes (ConMed SureTrace, Utica, NY, USA) following standard skin preparation. Signals were amplified (x 1000), sampled at 2000 Hz and band-pass filtered (10–500 Hz) using a MyoSystem 1400 (Noraxon, USA, Inc., Scottsdale, AZ). A Tektronix TDS 2014 storage oscilloscope (Tektronix Inc., Beaverton, OR, USA) was used to monitor EMG in real time. Data were stored to computer disk for later analysis.
Experimental protocol
, 百拇医药
An approximate motor threshold for PN stimulation was determined by finding the lowest current level at which a twitch in TA could be detected by inspection and palpation. The current level was then set at 120% of this threshold to ensure a modest but distinct twitch was elicited. PN stimulation was always delivered to the self-reported ‘dominant’ lower limb, i.e. the limb each subject said they preferred to use to kick a ball. This group of subjects all reported their right lower limb as being dominant. The TMS intensity that was used for the PAS intervention and for measures of CM excitability was set to give MEP amplitudes of 1.0–1.5 mV peak to peak in the TA during late swing. The ISI for the pairs of electrical and magnetic stimuli was determined by first estimating the latency of the MEPs recorded from the stimulated limb's TA. The ISI was then set as follows. To achieve PAS-induced facilitation (PASfac) of CM pathways, the ISI was set so that the TMS was delivered at an ISI equivalent to the estimated MEP latency plus 5 ms. This interval was consistent with the intervals selected by others for achieving facilitation using PAS applied to upper limb pathways (Stefan et al. 2002; Wolters et al. 2003). To achieve PAS-induced inhibition (PASinhib) of CM pathways, the ISI was set so that the TMS was delivered at an ISI that was 10 ms shorter than the estimated MEP latency. Previous work in the upper extremity (Stefan et al. 2002; Wolters et al. 2003) and preliminary experiments in this laboratory confirmed that this process reliably achieved facilitation and inhibition, respectively.
, 百拇医药
Upon recruitment, subjects were assigned consecutive Roman numerals (to XIV) for identification. To control for order effects, even-numbered subjects received PASfac in the first of two sessions and odd-numbered subjects received PASinhib. The two sessions were scheduled at least 1 week apart. In the second session the subject-stimulation order was reversed. In each session subjects walked on the treadmill at 4.0 km h–1 for a period of 20 min during which 120 pairs of stimuli were applied during the first 10 min. The schematics in Fig. 1 illustrate the protocols used in the main experiments and the control experiments.
, 百拇医药
The main experimental sessions are illustrated in A, where in the first session, odd-numbered subjects received PASinhib and even-numbered subjects received PASfac. In the second session odd-numbered subjects received PASfac and even-numbered subjects received PASinhib. Time points when Pre, Post0 and Post10 measures were made during the 20 min period of treadmill walking are illustrated in B. Similarly, the time points when measures were made during the 50 min period of walking for the control experiments are illustrated in C. After Pre measures were made, control sessions always began with a 10 min period of walking without stimulation, followed immediately by a 10 min period of PN stimulation only for even-numbered subjects and TMS only for odd-numbered subjects. The type of stimulation was reversed for the third 10 min period. During the fourth and fifth 10 min periods, PAS was applied in the same manner as the main experiment. In the first control session even-numbered subjects received PASfac and odd-numbered subjects received PASinhib. In the second control session even-numbered subjects received PASinhib and odd-numbered subjects received PASfac.
, http://www.100md.com
Measures of CM excitability were taken during walking prior to the intervention, at the end of the 10 min period of intervention and at the end of the 20 min period of walking from EMG data collected from the TA and Sol of both limbs. A set of 12 TMS-evoked responses were collected (1 every 5 s) during late swing of each limb. The mean MEP size was taken as a pre-intervention (Pre) measure of CM excitability. After the initial 10 min period of PAS, the mean MEP size of a further set of 12 TMS-evoked responses were taken as a measure of post-intervention CM excitability (Post0). This process was repeated after a further 10 min period of walking during which no stimulation was delivered. The mean MEP size of this set of responses was taken as a 10 min post-intervention (Post10) measure of CM excitability.
, http://www.100md.com
Control experiments
Four subjects from the main experiment (VI, VII, IX and XIII) were subsequently recruited to participate in three control experiments. Two of these experiments assessed the effects of walking alone, electrical stimulation of the peroneal nerve during walking alone, or TMS during walking alone, on MEP amplitude recorded from TA only (see Fig. 1). In one experiment, PASfac was applied immediately following the three control conditions. In another experiment, PASinhib was applied immediately following the three control conditions. All data were collected during one continuous period of walking. The order of PN stimulation alone and TMS alone was alternated to control for order effects. In a third experiment, PASfac was applied as subjects voluntarily held a tonic contraction of TA. Subjects were sitting on a chair and weights were added to the dorsum of the right foot so that dorsiflexion produced a level of EMG similar to that generated by the individual during the late swing phase of walking. The PASfac protocol for this control experiment was the same as in the main experiment.
, 百拇医药
Analysis
Peak-to-peak MEP amplitude, rectified MEP area, and root mean square (r.m.s.) pre-MEP EMG amplitudes were calculated. Up to 20% of responses were discarded to bring the background EMG r.m.s. amplitude means for the three conditions for each subject to within ±2%. The remaining MEP values were retained for analysis. Subject means of Pre, Post0 and Post10 MEP amplitude and area were then calculated from individual response values. The Post0 means and the Post10 means were expressed separately as a percentage of Pre means. Group means were tested for differences from 100% (unity) using one-tailed t tests, and also tested using separate two-way repeated measures ANOVAs. The factors were ‘pathway’, stimulated and non-stimulated (Stim, Non-Stim), and ‘time’ (Post0, Post10) as factors. A significance level of 0.05 was adopted for all statistical analyses.
, 百拇医药
Data from the first two control sessions were treated in a similar manner. Mean MEP size following walking alone (Postwalk), following PN stimulation during walking (PostPN), and following TMS alone (PostTMS) were expressed as a percentage of pre-intervention mean MEP size (CPre). For statistical purposes post-PAS unsigned changes in MEP amplitude were pooled from the four PASfac and four PASinhib trials. Means of MEP amplitude at Post0 and Post10 from the third control experiment (tonic contraction) were pooled and the differences between means derived during walking and tonic contraction were tested with a paired t test. For all control experiments differences between means and unity were tested for significance as conducted in the main experiment.
, 百拇医药
Results
An example of a 4 s simultaneous recording of EMG from right TA and left Sol is shown in Fig. 2 during late swing and late stance, respectively, from Subject V during treadmill walking. The MEPs have been extracted from the EMG to illustrate the typically triphasic TA MEP, the typically polyphasic Sol MEP, and post-MEP EMG silent periods.
Four seconds of EMG from Subject V showing bursts in right TA (top trace) and left Sol (middle trace) during treadmill walking at 4.0 km h–1. The bottom trace indicates when the TMS trigger occurred during late swing and late stance, respectively. Simultaneous MEPs (including the post-MEP silent period) can be distinguished in each trace at the time of the trigger. These MEPs are illustrated on a larger scale in the boxes below. Left box, TA MEP; right box, Sol MEP. The TA MEP is typically triphasic and the Sol MEP is typically polyphasic.
, 百拇医药
Temporally dependent and persistent modulations of MEP size were evident
Figure 3 illustrates individual subject's mean MEP amplitudes in stimulated TA CM pathways following PASfac and PASinhib. MEP amplitude at Post0 and Post10 is expressed as a percentage of Pre. The main result was that group mean facilitation and inhibition was 121% and 83%, respectively, and that no reduction in these means was evident between Post0 and Post10. The figure illustrates the general effects and variability of PAS on stimulated pathways. Twelve subjects demonstrated facilitation and 12 subjects demonstrated inhibition. Two subjects (II, XII) failed to demonstrate facilitation, and two failed to demonstrate inhibition (XI, XIV).
, 百拇医药
Individual responses for fourteen subjects in stimulated TA CM pathways following PASfac (A) and PASinhib (B) showing persistent modulation in group means illustrated by the thick continuous line. All values are expressed as a percentage of Pre.
Opposite sign modulation was evident in non-stimulated pathways
Of particular interest was that inhibition was evident in non-stimulated (opposite limb) TA pathways when facilitation was induced in stimulated limb TA pathways. Figure 4 illustrates subjects' mean MEP amplitudes in stimulated and non-stimulated TA CM pathways following PASfac. Note the four subjects highlighted with heavy dashed lines (I, III, IV, VIII) where inhibition was evident in their contralateral non-stimulated TA pathways. Eight subjects revealed an increase in CM excitability in stimulated and non-stimulated pathways, with greater facilitation on the stimulated side except for subject VII who demonstrated greater facilitation in non-stimulated pathways.
, http://www.100md.com
Means for individual subjects are illustrated in stimulated limb and non-stimulated limb TA CM pathways. Stimulated (Stim) and non-stimulated (Non-Stim) values are subject's means of responses recorded immediately after a 10 min period of stimulation (Post0) and at the end of a further 10 min of walking without stimulation (Post10). The group means are illustrated by the thick continuous line. The horizontal thin dashed line represents Pre values. Values for four subjects (I, III, IV, VIII) that are highlighted with thick dashed lines and large symbols revealed inhibition in non-stimulated TA CM pathways.
, http://www.100md.com
Table 1 sets out the group means derived from Post0 and Post10 TA MEP amplitude and MEP area data following PASfac and PASinhib expressed as a percentage of Pre. The P values indicate whether the means differed from unity (100%).
Persistent modulation was evident in stimulated pathways
Persistent stimulated pathway modulation was revealed in the analysis of TA MEP amplitude following PASfac using a 2-way repeated measures ANOVA. This analysis revealed an effect of ‘pathway’ (Stim versus Non-Stim, F1,13= 20.83, P < 0.001), but no effect of ‘time’ (P > 0.3) (see Fig. 5A). Similarly, the analysis of TA MEP area following PASfac revealed an effect of ‘pathway’ (Stim versus Non-Stim, F1,13= 14.54, P= 0.002), but no effect of ‘time’ (P > 0.5) (see Fig. 5B). The analysis of TA MEP amplitude and area following PASinhib using a 2-way, repeated measures ANOVA did not reveal an effect of ‘pathway’ or of ‘time’ (see Fig. 6)
, 百拇医药
The comparison of mean MEP size (A, amplitude; B, rectified area) during late swing recorded from stimulated TA pathways and during late swing recorded from non-stimulated TA pathways (contralateral). MEP size is expressed as a percentage of pre-intervention (Pre). Open bars represent the means of Post0, and filled bars represent the means of Post10 responses. The horizontal dashed line represents the Pre mean. Non-bracketed stars indicate the significance of the difference of the means from 100% (one-tailed t tests). Bracketed stars indicate the significance of the difference between Stim and Non-Stim means (2-way repeated measures ANOVA). *P < 0.05; **P < 0.01; ***P < 0.001. Error bars represent 1 S.E.M.
, http://www.100md.com
The comparison of mean MEP size (A, amplitude; B, rectified area) during late swing recorded from stimulated (Stim) TA pathways and during late swing recorded from non-stimulated (Non-Stim) TA pathways (contralateral). MEP size is expressed as a percentage of pre-intervention (Pre). Open bars represent the means of Post0, and filled bars represent the means of Post10 responses. The horizontal dashed line represents the Pre mean. Stars indicate the significance of the difference of the means from 100% (one-tailed t tests). *P < 0.05; **P < 0.01; ***P < 0.001, d.f. 13. Error bars represent 1 S.E.M.
, 百拇医药
Some subjects revealed robust bidirectional modulation
An inspection of the extent of facilitation and inhibition in each subject's stimulated TA pathways (see Fig. 7) suggested there was a correlation between the extent of facilitation and inhibition. While a regression analysis did not reveal a significant correlation, when the analysis was restricted to the seven ‘best modulators’ (I, V, VI, VIII, IX, XI, XIII) who demonstrated the largest percentage differences between facilitation and inhibition (> 40), a significant positive correlation emerged (r2= 0.88, P= 0.005). This analysis indicated that although these subjects demonstrated robust modulation in both directions, greater facilitation was associated with less inhibition.
, 百拇医药
A comparison of the effects of PASfac and PASinhib in each subject. A, averaged Post0 and Post10 values of individual subject's means. The thick line represents the group mean. B, a scatter plot of means from subjects I, V, VI, VIII, IX, XI and XIII with fac-inhib differences of > 40 revealing that although greater facilitation was associated with less inhibition, subjects who demonstrated robust facilitation in one session also demonstrated robust inhibition in the other session (P= 0.005).
, 百拇医药
Modulation was also revealed in soleus pathways
Following PASfac applied to TA pathways, no effects were revealed by the ANOVA of Sol responses, although group mean Sol MEP area differed from 100% in the stimulated limb at Post10 (Fig. 8). Sol responses were collected during late stance. Data from four subjects were omitted from the following analyses because either motor threshold for Sol MEPs was not reached (recall that TA was the target muscle and parameters were established to optimize TA responses), or background Sol EMG could not be matched within an individual across the three trials. ‘Stim’ refers to means of Sol responses recorded from the muscle ipsilateral to stimulated TA pathways, ‘Non-Stim’ refers to means of Sol responses recorded from the muscle contralateral to stimulated TA pathways. Of the eight PASfac Sol means (amplitude, area; ‘Stim’, ‘Non-Stim’; Post0, Post10), only ‘Stim’ Post10 area differed from 100% (mean, 112; P= 0.044) (see Fig. 8B). The remaining means ranged from 94 to 107. The ANOVA for amplitude and area did not reveal effects of ‘pathway’ or ‘time’ (all P > 0.1). Following PASinhib (not illustrated), two of the eight Sol means differed from 100%: the ‘Stim’ amplitude mean (122%) at Post10 (P= 0.029) and the ‘Stim’ area mean (116%) at Post0 (P= 0.041). The ANOVA for amplitude revealed an effect of ‘time’ where ‘Stim’ Post10 (mean, 121%) was greater than Post0 (mean, 109%), F1,10= 5.7, P= 0.038).
, http://www.100md.com
The comparison of mean MEP size (A, amplitude; B, rectified area) during late stance recorded from the Sol ipsilateral to the stimulated TA pathways (‘Stim’) and during late stance recorded from the Sol ipsilateral to the non-stimulated TA pathways (‘Non-Stim’). MEP size expressed as a percentage of pre-intervention (Pre). Open bars represent the means of Post0, and filled bars represent the means of Post10 responses. The horizontal dashed line represents the Pre mean. The star indicates that the mean differed from 100% (one-tailed t test), P < 0.05. Error bars represent 1 S.E.M.
, 百拇医药
Robust modulation was not evident with peripheral and cortical stimulation alone
Only PAS induced a significant modulation in MEP amplitude. Analysis of control data recorded from TA during the eight sessions conducted on four subjects did not reveal an effect of walking alone, PN stimulation alone, or TMS alone, but did reveal an effect of PAS. The PostPN and PostTMS means (101 and 105%, respectively) did not differ from 100% using one-tailed t tests (all P > 0.3). The CPost0, and CPost10 conditions (equivalent to the intervention in the main experiment) differed from 100%, where the CPost0 mean was 114% (P= 0.034), and the CPost10 mean was 110% (P= 0.011). A one-way ANOVA revealed an effect of condition (F4= 3.0, P= 0.033), and post hoc Tukey HSD tests (to detect ‘honestly significant differences’) revealed an effect of PAS where the CPost0 mean (114%) was greater than the Postwalk mean (96%) (P < 0.05). For illustrative purposes Fig. 9 separates the control condition means of the four control sessions that were concluded with PASfac from the four that were concluded with PASinhib.
, http://www.100md.com
, the means of the four control subjects (VI, VII, IX, XIII) during the session that culminated in PASfac. , the means of the four control subjects during the session that culminated in PASinhib. Filled symbols represent means expressed as a percentage of Prewalk. Open symbols represent means expressed as a percentage of either PostTMS or PostPN. Stars indicate that means of unsigned changes in MEP amplitude following PASfac and PASinhib that were grouped for statistical analysis differed from 100% (see text for details). Error bars are 1 S.E.M.
, http://www.100md.com
PAS applied during tonic contraction also induced facilitation
In the last control experiment, the extent of facilitation of TA pathways was similar to that derived from the main experiment (123%). The Post0 and Post10 MEP amplitudes for the four control subjects (% Pre) were pooled (n= 8) separately for data collected during walking and during tonic contraction. The tonic contraction mean (118%) differed from unity (P= 0.002). The tonic contraction mean did not differ from the walking mean (120%) using a paired t test (P= 0.79).
, http://www.100md.com
Discussion
The main finding was that PAS induced changes in dorsiflexor CM excitability of healthy subjects during walking. Bidirectional changes in lower limb CM excitability induced by PAS during walking have not previously been reported. Four criteria characterizing the physiological profile of PAS-induced phenomena in the human motor system have previously been described by Stefan et al. (2000, 2002); and Wolters et al. (2003). Three of these were evident in the present data. The first is the temporal rule, in which the same group of subjects demonstrated an increase in CM excitability when the paired ISI was 5 ms longer than the estimated MEP latency, and also demonstrated a decrease in CM excitability when the paired ISI was 10 ms shorter than the estimated MEP latency. The second criterion is rapid evolution, also demonstrated in the present experiment, where 120 pairs of stimuli were delivered over a 10 min period (cf. 90 pairs delivered over a 30 min period: Stefan et al. 2000, 2002; Wolters et al. 2003). While the relationship between stimulation period and the number of pairs of stimuli has yet to be examined, the present data clearly illustrate a rapid evolution of bidirectional changes in CM excitability. The third criterion is persistence of effects following paired stimulation. The present group data revealed no reduction in facilitation or inhibition at Post10 compared with Post0 (see Figs 5 and 6). The fourth criterion is reversibility which was not assessed in the present experiment.
, 百拇医药
Stefan et al. (2000, 2002) and Wolters et al. (2003) referred to PAS-induced phenomena as ‘long-term potentiation-like’ (LTP-like) and ‘long-term depression-like’ (LTD-like). Their choice of terminology was based on the characteristics of the temporal nature of the phenomena, its rapid evolution, persistence, and reversibility. Further, these authors found that PAS-induced facilitation was blocked by dextromethorphan, an N-methyl-D-aspartate receptor antagonist known to block long-term potentiation in animal tissue slice preparations. Despite these similarities, a clear relationship between the effects of PAS and the in vitro induction of LTP has not been established.
, 百拇医药
Inspection of Fig. 3 reveals that during PASfac four subjects continued to increase MEP amplitude (Fig. 3A), and during PASinhib five subjects continued to decrease MEP amplitude (Fig. 3B) in the 10 min period of walking immediately following PAS termination. This finding and the between-subject variability is consistent with the previous results from Stefan et al. (2000). Further, it was also noted that PAS failed to elicit MEP changes in a few subjects during either conditioning protocol (Fig. 3), indicating that some subjects were amenable to both PASfac and PASinhib and others were not amenable to either. This suggestion is also supported by the regression analysis of the seven ‘best modulators’ (Fig. 7B). If the difference in the extent of modulation between subjects had been merely the result of inappropriate setting of the interstimulus interval between stimulus-pairs no correlation between robust facilitation and robust inhibition would have been evident. However, inappropriate setting of parameters as the cause of the between-subject variability in the extent of modulation across the two sessions cannot be entirely ruled out. More robust effects of PAS may be evident if afferent conduction time was estimated from sensory-evoked potentials (SEP) recorded from the scalp in response to electrical stimulation of the peroneal nerve. There are no reports of PAS where specific interstimulus intervals have been based on SEP latency of individual subjects.
, 百拇医药
An intriguing observation made from comparing the extent of facilitation during PASfac in stimulated pathways compared with non-stimulated pathways (Fig. 4) was that four subjects who demonstrated robust facilitation also demonstrated robust inhibition in non-stimulated pathways. Although data from these four subjects in the present experiment cannot provide an explanation for this observation, if group data in a future experiment were to reveal a similar opposite-side-effect of PAS, the observation would warrant further investigation.
, http://www.100md.com
In the present experiment CM pathways to TA were chosen as the target. All stimulation parameters for probing CM excitability and for PAS were set for the target pathways. Therefore Sol data may be disregarded or interpreted with great caution. Notwithstanding this caveat, it is interesting to note that Sol MEP amplitude increased to 122% in the Post10 condition of the stimulated limb when PASinhib was being applied to the target TA pathway. The finding may indicate a reciprocal inhibitory mechanism for this agonist–antagonist pair of ankle effectors. However, during PASfac, a reciprocal effect was not evident and a modest 112% facilitation (not inhibition) was evident at Post10 in the stimulated limb Sol (Fig. 8B). At best, these data might indicate a non-specific facilitation of Sol pathways. The data warrant the design of an experiment that would examine the specificity of PAS on CM pathways to TA and to Sol separately. Changes in CM excitability in pathways to muscles of the leg as well as muscles of the shank could be included and optimal stimulation parameters could be established for each muscle pathway to enhance the validity of observed changes in non-target CM excitability.
, 百拇医药
An important issue addressed in the present experiment is whether any of the components of the intervention applied during walking modulate CM excitability on their own. The eight sessions conducted with four subjects revealed no statistically significant effects on MEP amplitude of walking alone, electrical stimulation of the peroneal nerve during walking alone, or TMS during walking alone. Significant effects were only revealed following PAS applied at the end of the control data collection sessions that involved walking. An inspection of Fig. 9 reveals that some modulation of MEP amplitude was taking place over the course of the control sessions; however, the extent of the modulation was small and failed to reach statistical significance compared with the modulation that followed PAS. These findings suggest that components of the PAS protocol as used in the present experiment do not significantly modulate CM excitability on their own. Finally, the results of the third control condition where subjects held a tonic TA contraction during the application of PASfac were consistent with the findings of Capaday et al. (1999) and Bawa et al. (2002) who showed that CM projections to TA are relatively stronger compared with projections to Sol. Capaday et al. (1999) found no difference between the extent of TA pathway excitability during walking and during a voluntary tonic contraction, and we found no difference between these two tasks in the extent of modulation induced by PAS.
, 百拇医药
The level of the neuroaxis at which the effects of PAS are occurring is assumed to be the M1 region because the temporal relationship between the paired stimuli is set to occur when the induced afferent volleys are estimated to have arrived in the cortex. This hypothesis has previously been supported by Stefan et al. (2000) who provided evidence of motoneurone stability despite significant increases in MEP amplitude following PAS. They reported no significant change in F-wave or M-wave amplitude or in the amplitude of potentials evoked by electrical stimulation at the level of the brainstem as a result of PAS. They also found that when they applied PAS to increase MEP amplitude there was an increase in the length of the post-MEP EMG silent period, suggesting the cortical component of the silent period had increased in length due to an increase in cortical input to motoneurone pools. Furthermore, Wolters et al. (2003) demonstrated that significant decreases in MEP amplitude induced by PAS were not evident in F-wave amplitude or the amplitude of potentials evoked by electrical stimulation at the level of the brainstem. In the present experiment silent periods were not examined because silent periods are difficult to assess in an EMG burst during walking where EMG amplitude is changing as a function of time. Since the stimuli were delivered in the present experiment toward the end of the swing phase, the return of voluntary EMG occurred at the offset of the EMG burst. A demonstration of where on the neuroaxis the effects of PAS were taking place was not attempted because the present experiment was primarily designed to assess the ability of PAS to modulate the excitability of the CM pathway as a whole during walking. However, the time window in which paired stimuli were delivered (–10 ms to 5 ms), and the lack of an effect of peripheral nerve stimulation alone suggests a supraspinal site. Future experiments to assess intracortical excitability and the excitability of spinal circuitry may provide further evidence of the cortex as the site where effects of PAS occur during walking.
, 百拇医药
Although PAS used to increase CM excitability during walking is unique to the present study, other authors have reported the use of functional electrical stimulation to increase CM excitability. For example, Knash et al. (2003) demonstrated increased MEP size in the TA with repetitive electrical stimulation of the common peroneal nerve. Significant facilitation was observed within the first 10 min of stimulation, and lasted up to 30 min following cessation of stimulation. Further, Khaslavskaia et al. (2002) delivered repetitive electrical stimulation to the peroneal nerve of healthy subjects and found that CM excitability was increased to a greater extent when stimulation was combined with voluntary dorsiflexor activation than when the muscle was at rest. Such augmented effects of repetitive electrical stimulation on CM excitability with combined voluntary effort may play a role in the restoration of motor behaviours following neurological injury (Field-Fote & Tepavac, 2002; Yan et al. 2005).
, http://www.100md.com
In conclusion, the hypothesis adopted for the present experiment was supported by evidence that PAS applied to ankle dorsiflexor CM pathways during the swing phase of walking induced temporally dependent bidirectional changes in CM excitability consistent with the Hebbian principle of activity-dependent plasticity. Similarities between these data derived from the lower limb, and data derived from CM pathways to hand muscles (Stefan et al. 2000, 2002; Wolters et al. 2003) suggest that the same neural mechanisms were involved. Of particular importance was the demonstration of these phenomena during walking which has not previously been reported. Walking training protocols applied to neurologically impaired individuals where CM excitability is suppressed, may benefit from the addition of a training adjunct that increases CM excitability.
, 百拇医药
References
Bawa P, Chalmers GR, Stewart H & Eisen AA (2002). Responses of ankle extensor and flexor motoneurons to transcranial magnetic stimulation. J Neurophysiol 88, 124–132.
Capaday C, Lavoie BA, 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. J Neurophysiol 81, 129–139.
, 百拇医药
Dupont WD & Plummer WD (1990). Power and sample size calculations: a review and computer program. Controlled Clin Trials 11, 116–128.
Field-Fote EC & Tepavac D (2002). Improved intralimb coordination in people with incomplete spinal cord injury following training with body weight support and electrical stimulation. Phys Ther 82, 707–715.
Hebb DO (1949). The Organization of Behavior. Wiley, New York.
, 百拇医药
Khaslavskaia S, Ladouceur M & Sinkjaer T (2002). Increase in tibialis anterior motor cortex excitability following repetitive electrical stimulation of the common peroneal nerve. Exp Brain Res 145, 309–315.
Knash ME, Kido A, Gorassini M, Chan KM & Stein RB (2003). Electrical stimulation of the human common peroneal nerve elicits lasting facilitation of cortical motor-evoked potentials. Exp Brain Res 153, 366–377.
Muellbacher W, Ziemann U, Boroojerdi B & Hallett M (2000). Effects of low-frequency transcranial magnetic stimulation on motor excitability and basic motor behavior. Clin Neurophysiol 111, 1002–1007.
, 百拇医药
Peinemann A, Reimer B, Loer C, Quartarone A, Munchau A, Conrad B & Siebner HR (2004). Long-lasting increase in corticospinal excitability after 1800 pulses of subthreshold 5 Hz repetitive TMS to the primary motor cortex. Clin Neurophysiol 115, 1519–1526.
Said CM, Goldie PA, Patla AE & Sparrow WA (2001). Effect of stroke on step characteristics of obstacle crossing. Arch Phys Mederehabil 82, 1712–1719.
Stefan K, Kunesch E, Benecke R, Cohen LG & Classen J (2002). Mechanisms of enhancement of human motor cortex excitability induced by interventional paired associative stimulation. J Physiol 543, 699–708.
, 百拇医药
Stefan K, Kunesch E, Cohen LG, Benecke R & Classen J (2000). Induction of plasticity in the human motor cortex by paired associative stimuli. Brain 123, 572–584.
Uy J, Ridding MC, Hillier S, Thompson PD & Miles TS (2003). Does induction of plastic change in motor cortex improve leg function after strokeNeurology 61, 982–984.
Wolters A, Sandbrink F, Schlottmann A, Kunesch E, Stefan K, Cohen LG, Benecke R & Classen J (2003). A temporally asymmetric Hebbian rule governing plasticity in the human motor cortex. J Neurophysiol 89, 2339–2345.
, http://www.100md.com
Yan T, Hui-Chan CWY & Li LSW (2005). Functional electrical stimulation improves motor recovery of the lower extremity and walking ability of subjects with first acute stroke: a randomized placebo-controlled trial. Stroke 36, 80–85.
Ziemann U (2004). TMS and drugs. Clin Neurophysiol 115, 1717–1729.
Ziemann U, Wittenberg GF & Cohen LG (2002). Stimulation-induced within-representation and across-representation plasticity in human motor cortex. J Neurosci 22, 5563–5571., 百拇医药(James W. Stinear and T. G)
2 Department of Physical Therapy, University of Illinois at Chicago, Department of Physical Medicine and Rehabilitation, Northwestern University, Chicago, IL, USA
Abstract
Magnetic stimulation of human primary motor cortex (M1) paired with electrical stimulation of a peripheral motor nerve has been used to produce a lasting modulation of corticomotor (CM) excitability. This ‘paired associative stimulation’ (PAS) protocol has been used to induce bidirectional changes in excitability in upper limb CM pathways. The present study tested the hypothesis that temporally dependent PAS applied to the common peroneal nerve during the swing phase of walking would induce bidirectional changes in CM excitability consistent with the Hebbian principle of activity-dependent plasticity. Fourteen subjects with no known neurological disorder participated in two data collection sessions each. PAS was delivered as a single block of 120 pairs of stimuli delivered in a 10 min period during treadmill walking at 4.0 km h–1. Changes in CM excitability were assessed by examining the size of motor potentials evoked by transcranial magnetic stimulation prior to and following PAS. Tibialis anterior motor-evoked potentials amplitude increased to 121% over baseline when the magnetic stimulus was delivered over M1 after the estimated arrival time of the afferent volley in sensorimotor cortex and decreased to 83% of baseline when the magnetic stimulus was delivered prior to the estimated afferent volley arrival. This extent of modulation was undiminished following a further 10 min period of walking without stimulation. The temporal nature of the bidirectional effects following PAS, their rapid evolution and subsequent persistence supported the study's hypothesis and were similar to the effects described by others in quiescent muscles of the upper limb.
, http://www.100md.com
* Corresponding author J. W. Stinear: Sensory Motor Performance Program, 345 E. Superior St, Chicago, IL 60611, USA.
Introduction
Modulation of corticomotor (CM) excitability can be induced in humans using a variety of neural stimulation techniques which invoke principles of activity- (or spike-timing-) dependent neural plasticity. For example, repetitive transcranial magnetic stimulation (rTMS) applied over human primary motor cortex (M1) has been used in a number of protocols to reveal a decrease in CM excitability when the frequency of stimulation was 1 Hz (Muellbacher et al. 2000), and an increase in CM excitability when the frequency of stimulation was 5 Hz or more (Peinemann et al. 2004). Pharmacological agents have been shown to increase (e.g. D-amphetamine, a noradrenaline (norepinephrine) agonist) or decrease CM excitability (e.g. lorazepam, an allosteric modulator of the GABAA receptor) (see Ziemann, 2004 for a review). Ischaemic nerve block of the upper limb in combination with rTMS has also been used to demonstrate stimulation-induced alterations in the size of motor-evoked potentials (MEPs) (Ziemann et al. 2002). More recently, a technique based on Hebbian principles of synaptic plasticity has been used to increase (Stefan et al. 2002; Wolters et al. 2003), and decrease the excitability of CM pathways to the upper limb (Wolters et al. 2003). This technique, called ‘paired associative stimulation’ (PAS), follows the association principle first proposed by D. O. Hebb, whereby changes in synaptic weighting are induced by a presynaptic neurone taking part in the repetitive firing of a postsynaptic cell (Hebb, 1949).
, 百拇医药
Specifically, Wolters et al. (2003) observed increases in the size of motor-evoked potentials (MEPs) elicited in the abductor pollicis brevis when suprathreshold TMS was delivered to M1 within 6 ms of sensory-evoked potentials arriving in sensory cortex. In contrast, decreases in MEP size were detected when TMS was delivered within 10 ms prior to the afferent volley arriving at the somatosensory cortex. The effects evolved rapidly, were long lasting, yet reversible.
, 百拇医药
Studies of PAS involving CM pathways to lower limb motoneurones are limited (Uy et al. 2003), and no studies have been reported where changes in CM excitability have been induced during muscle activation. The hypothesis adopted for the present experiment predicted that temporally dependent PAS applied to tibialis anterior (TA) pathways during the swing phase of walking would induce bidirectional changes in CM excitability consistent with the Hebbian principle of activity-dependent plasticity. We chose TA as the target muscle because the effects of PAS are likely to be more readily revealed in pathways that are known to have strong CM connections, and the functional importance of the TA motor pool is particularly evident in individuals following neurological injury (Said et al. 2001). Studies have revealed the relative strength of CM pathways to TA and soleus (Sol); for example, in response to TMS (Capaday et al. 1999) constructed input–output curves of CM pathways to TA and to Sol muscles. Pathways to Sol were less excitable during the stance phase of walking than during a tonic voluntary contraction. However, they found no difference between the excitability of projections to TA during the swing phase of walking and during a voluntary tonic contraction of TA. In a similar study Bawa et al. (2002) used surface electromyography (EMG) and single motor unit responses to provide further evidence that CM projections to TA are relatively stronger compared with projections to Sol.
, 百拇医药
During motor re-training, the inclusion of PAS to manipulate the excitability of injured CM pathways may be of value in the application of walking re-training regimens. Because PAS has been reported previously only in pathways to muscles at rest, and because of our interest in developing adjuncts to walking training protocols, the aim of the present experiment was to determine if bidirectional changes in healthy individuals' TA CM excitability could be induced with PAS applied during walking. If this aim is fulfilled, we intend to investigate the mechanisms of PAS further and assess its value as an adjunct to walking training following neural injury.
, 百拇医药
Methods
Subjects
Healthy subjects with no known neurological disorder were recruited from the staff of the Rehabilitation Institute of Chicago. The sample size required to reject the null hypothesis was calculated using preliminary data and PS v. 2.1.31 software (Dupont & Plummer, 1990), where the difference in means was 0.25 and the standard deviation was 0.30. The significance level was set at = 0.05 and the power at 1 –= 0.80, yielding a sample size of 13. Fourteen (6 female, 8 male) subjects volunteered and were recruited for the study. Their ages ranged from 24 to 58 years (mean 31). The study was approved by the Northwestern University Institutional Review Board. All methods conformed to the Declaration of Helsinki and all participants provided written informed consent.
, http://www.100md.com
Stimulation and recording
M1 was stimulated by a Magstim 2002 unit (Magstim, Dyfed, Wales, UK) via a Magstim double cone coil (see below for details regarding intensity setting). A tight fitting linen cap was tied on the subject's head with the vertex marked on the cap at the intersection of the inion–nasion and interaural lines. The coil was suspended over the subject's head from an overhead gantry and attached to the cap using Velcro tapes so that the junction of the coils was located in the mid-sagittal plane 1 cm anterior to the vertex. The large magnetic field generated by the double cone coil and the close proximity of homologous lower limb cortical representations allowed the coil to be fixed in the mid-sagittal plane thus eliminating the need to shift the coil to an optimum site for each TA. Minor adjustments were made to the coil position until optimal-sized MEPs of similar size could be elicited from both TA muscles during the respective limb's late swing phase of treadmill walking. A chin strap attached to the coil and small foam pads were used to increase coil stability. Subjects found this to be a comfortable arrangement, enabling them to adopt their preferred gait pattern, and move their head with ease between trials without causing a shift in the location of the coil. The coil and cap positions were checked repeatedly throughout data collection and coil position did not change in any of the sessions.
, 百拇医药
Peripheral nerve stimulation was delivered from a Model DS7A constant current stimulator (Digitimer Ltd, Hertfordshire, UK) via a Meditrace 133 surface Ag–AgCl electrode (Tyco Healthcare, Ludlow, Chicopee, MD, USA) used as a cathode placed over the deep peroneal nerve (PN) just distal and anterior to the fibula head at a position where maximum TA muscle contraction and minimum peroneal muscle contraction was elicited. The anode was a 25 cm2 Dura-Soft (Pain Management Technologies, Inc., Akron, OH, USA) electrode placed over the ipsilateral patella.
, 百拇医药
The timing of stimulator trigger pulses was achieved using custom-built heel switches and a two channel electronic delay circuit. The cycle phase was established using trigger pulses and EMG bursts displayed on an oscilloscope. The target cycle phase was the middle of the large TA burst that occurs in late swing as the ankle is dorsi-flexed prior to heel strike. The interstimulus interval (ISI) between Digitimer and Magstim units was set using a second component of the delay circuit.
, 百拇医药
Surface EMG was recorded from TA and Sol via Ag–AgCl EMG electrodes (ConMed SureTrace, Utica, NY, USA) following standard skin preparation. Signals were amplified (x 1000), sampled at 2000 Hz and band-pass filtered (10–500 Hz) using a MyoSystem 1400 (Noraxon, USA, Inc., Scottsdale, AZ). A Tektronix TDS 2014 storage oscilloscope (Tektronix Inc., Beaverton, OR, USA) was used to monitor EMG in real time. Data were stored to computer disk for later analysis.
Experimental protocol
, 百拇医药
An approximate motor threshold for PN stimulation was determined by finding the lowest current level at which a twitch in TA could be detected by inspection and palpation. The current level was then set at 120% of this threshold to ensure a modest but distinct twitch was elicited. PN stimulation was always delivered to the self-reported ‘dominant’ lower limb, i.e. the limb each subject said they preferred to use to kick a ball. This group of subjects all reported their right lower limb as being dominant. The TMS intensity that was used for the PAS intervention and for measures of CM excitability was set to give MEP amplitudes of 1.0–1.5 mV peak to peak in the TA during late swing. The ISI for the pairs of electrical and magnetic stimuli was determined by first estimating the latency of the MEPs recorded from the stimulated limb's TA. The ISI was then set as follows. To achieve PAS-induced facilitation (PASfac) of CM pathways, the ISI was set so that the TMS was delivered at an ISI equivalent to the estimated MEP latency plus 5 ms. This interval was consistent with the intervals selected by others for achieving facilitation using PAS applied to upper limb pathways (Stefan et al. 2002; Wolters et al. 2003). To achieve PAS-induced inhibition (PASinhib) of CM pathways, the ISI was set so that the TMS was delivered at an ISI that was 10 ms shorter than the estimated MEP latency. Previous work in the upper extremity (Stefan et al. 2002; Wolters et al. 2003) and preliminary experiments in this laboratory confirmed that this process reliably achieved facilitation and inhibition, respectively.
, 百拇医药
Upon recruitment, subjects were assigned consecutive Roman numerals (to XIV) for identification. To control for order effects, even-numbered subjects received PASfac in the first of two sessions and odd-numbered subjects received PASinhib. The two sessions were scheduled at least 1 week apart. In the second session the subject-stimulation order was reversed. In each session subjects walked on the treadmill at 4.0 km h–1 for a period of 20 min during which 120 pairs of stimuli were applied during the first 10 min. The schematics in Fig. 1 illustrate the protocols used in the main experiments and the control experiments.
, 百拇医药
The main experimental sessions are illustrated in A, where in the first session, odd-numbered subjects received PASinhib and even-numbered subjects received PASfac. In the second session odd-numbered subjects received PASfac and even-numbered subjects received PASinhib. Time points when Pre, Post0 and Post10 measures were made during the 20 min period of treadmill walking are illustrated in B. Similarly, the time points when measures were made during the 50 min period of walking for the control experiments are illustrated in C. After Pre measures were made, control sessions always began with a 10 min period of walking without stimulation, followed immediately by a 10 min period of PN stimulation only for even-numbered subjects and TMS only for odd-numbered subjects. The type of stimulation was reversed for the third 10 min period. During the fourth and fifth 10 min periods, PAS was applied in the same manner as the main experiment. In the first control session even-numbered subjects received PASfac and odd-numbered subjects received PASinhib. In the second control session even-numbered subjects received PASinhib and odd-numbered subjects received PASfac.
, http://www.100md.com
Measures of CM excitability were taken during walking prior to the intervention, at the end of the 10 min period of intervention and at the end of the 20 min period of walking from EMG data collected from the TA and Sol of both limbs. A set of 12 TMS-evoked responses were collected (1 every 5 s) during late swing of each limb. The mean MEP size was taken as a pre-intervention (Pre) measure of CM excitability. After the initial 10 min period of PAS, the mean MEP size of a further set of 12 TMS-evoked responses were taken as a measure of post-intervention CM excitability (Post0). This process was repeated after a further 10 min period of walking during which no stimulation was delivered. The mean MEP size of this set of responses was taken as a 10 min post-intervention (Post10) measure of CM excitability.
, http://www.100md.com
Control experiments
Four subjects from the main experiment (VI, VII, IX and XIII) were subsequently recruited to participate in three control experiments. Two of these experiments assessed the effects of walking alone, electrical stimulation of the peroneal nerve during walking alone, or TMS during walking alone, on MEP amplitude recorded from TA only (see Fig. 1). In one experiment, PASfac was applied immediately following the three control conditions. In another experiment, PASinhib was applied immediately following the three control conditions. All data were collected during one continuous period of walking. The order of PN stimulation alone and TMS alone was alternated to control for order effects. In a third experiment, PASfac was applied as subjects voluntarily held a tonic contraction of TA. Subjects were sitting on a chair and weights were added to the dorsum of the right foot so that dorsiflexion produced a level of EMG similar to that generated by the individual during the late swing phase of walking. The PASfac protocol for this control experiment was the same as in the main experiment.
, 百拇医药
Analysis
Peak-to-peak MEP amplitude, rectified MEP area, and root mean square (r.m.s.) pre-MEP EMG amplitudes were calculated. Up to 20% of responses were discarded to bring the background EMG r.m.s. amplitude means for the three conditions for each subject to within ±2%. The remaining MEP values were retained for analysis. Subject means of Pre, Post0 and Post10 MEP amplitude and area were then calculated from individual response values. The Post0 means and the Post10 means were expressed separately as a percentage of Pre means. Group means were tested for differences from 100% (unity) using one-tailed t tests, and also tested using separate two-way repeated measures ANOVAs. The factors were ‘pathway’, stimulated and non-stimulated (Stim, Non-Stim), and ‘time’ (Post0, Post10) as factors. A significance level of 0.05 was adopted for all statistical analyses.
, 百拇医药
Data from the first two control sessions were treated in a similar manner. Mean MEP size following walking alone (Postwalk), following PN stimulation during walking (PostPN), and following TMS alone (PostTMS) were expressed as a percentage of pre-intervention mean MEP size (CPre). For statistical purposes post-PAS unsigned changes in MEP amplitude were pooled from the four PASfac and four PASinhib trials. Means of MEP amplitude at Post0 and Post10 from the third control experiment (tonic contraction) were pooled and the differences between means derived during walking and tonic contraction were tested with a paired t test. For all control experiments differences between means and unity were tested for significance as conducted in the main experiment.
, 百拇医药
Results
An example of a 4 s simultaneous recording of EMG from right TA and left Sol is shown in Fig. 2 during late swing and late stance, respectively, from Subject V during treadmill walking. The MEPs have been extracted from the EMG to illustrate the typically triphasic TA MEP, the typically polyphasic Sol MEP, and post-MEP EMG silent periods.
Four seconds of EMG from Subject V showing bursts in right TA (top trace) and left Sol (middle trace) during treadmill walking at 4.0 km h–1. The bottom trace indicates when the TMS trigger occurred during late swing and late stance, respectively. Simultaneous MEPs (including the post-MEP silent period) can be distinguished in each trace at the time of the trigger. These MEPs are illustrated on a larger scale in the boxes below. Left box, TA MEP; right box, Sol MEP. The TA MEP is typically triphasic and the Sol MEP is typically polyphasic.
, 百拇医药
Temporally dependent and persistent modulations of MEP size were evident
Figure 3 illustrates individual subject's mean MEP amplitudes in stimulated TA CM pathways following PASfac and PASinhib. MEP amplitude at Post0 and Post10 is expressed as a percentage of Pre. The main result was that group mean facilitation and inhibition was 121% and 83%, respectively, and that no reduction in these means was evident between Post0 and Post10. The figure illustrates the general effects and variability of PAS on stimulated pathways. Twelve subjects demonstrated facilitation and 12 subjects demonstrated inhibition. Two subjects (II, XII) failed to demonstrate facilitation, and two failed to demonstrate inhibition (XI, XIV).
, 百拇医药
Individual responses for fourteen subjects in stimulated TA CM pathways following PASfac (A) and PASinhib (B) showing persistent modulation in group means illustrated by the thick continuous line. All values are expressed as a percentage of Pre.
Opposite sign modulation was evident in non-stimulated pathways
Of particular interest was that inhibition was evident in non-stimulated (opposite limb) TA pathways when facilitation was induced in stimulated limb TA pathways. Figure 4 illustrates subjects' mean MEP amplitudes in stimulated and non-stimulated TA CM pathways following PASfac. Note the four subjects highlighted with heavy dashed lines (I, III, IV, VIII) where inhibition was evident in their contralateral non-stimulated TA pathways. Eight subjects revealed an increase in CM excitability in stimulated and non-stimulated pathways, with greater facilitation on the stimulated side except for subject VII who demonstrated greater facilitation in non-stimulated pathways.
, http://www.100md.com
Means for individual subjects are illustrated in stimulated limb and non-stimulated limb TA CM pathways. Stimulated (Stim) and non-stimulated (Non-Stim) values are subject's means of responses recorded immediately after a 10 min period of stimulation (Post0) and at the end of a further 10 min of walking without stimulation (Post10). The group means are illustrated by the thick continuous line. The horizontal thin dashed line represents Pre values. Values for four subjects (I, III, IV, VIII) that are highlighted with thick dashed lines and large symbols revealed inhibition in non-stimulated TA CM pathways.
, http://www.100md.com
Table 1 sets out the group means derived from Post0 and Post10 TA MEP amplitude and MEP area data following PASfac and PASinhib expressed as a percentage of Pre. The P values indicate whether the means differed from unity (100%).
Persistent modulation was evident in stimulated pathways
Persistent stimulated pathway modulation was revealed in the analysis of TA MEP amplitude following PASfac using a 2-way repeated measures ANOVA. This analysis revealed an effect of ‘pathway’ (Stim versus Non-Stim, F1,13= 20.83, P < 0.001), but no effect of ‘time’ (P > 0.3) (see Fig. 5A). Similarly, the analysis of TA MEP area following PASfac revealed an effect of ‘pathway’ (Stim versus Non-Stim, F1,13= 14.54, P= 0.002), but no effect of ‘time’ (P > 0.5) (see Fig. 5B). The analysis of TA MEP amplitude and area following PASinhib using a 2-way, repeated measures ANOVA did not reveal an effect of ‘pathway’ or of ‘time’ (see Fig. 6)
, 百拇医药
The comparison of mean MEP size (A, amplitude; B, rectified area) during late swing recorded from stimulated TA pathways and during late swing recorded from non-stimulated TA pathways (contralateral). MEP size is expressed as a percentage of pre-intervention (Pre). Open bars represent the means of Post0, and filled bars represent the means of Post10 responses. The horizontal dashed line represents the Pre mean. Non-bracketed stars indicate the significance of the difference of the means from 100% (one-tailed t tests). Bracketed stars indicate the significance of the difference between Stim and Non-Stim means (2-way repeated measures ANOVA). *P < 0.05; **P < 0.01; ***P < 0.001. Error bars represent 1 S.E.M.
, http://www.100md.com
The comparison of mean MEP size (A, amplitude; B, rectified area) during late swing recorded from stimulated (Stim) TA pathways and during late swing recorded from non-stimulated (Non-Stim) TA pathways (contralateral). MEP size is expressed as a percentage of pre-intervention (Pre). Open bars represent the means of Post0, and filled bars represent the means of Post10 responses. The horizontal dashed line represents the Pre mean. Stars indicate the significance of the difference of the means from 100% (one-tailed t tests). *P < 0.05; **P < 0.01; ***P < 0.001, d.f. 13. Error bars represent 1 S.E.M.
, 百拇医药
Some subjects revealed robust bidirectional modulation
An inspection of the extent of facilitation and inhibition in each subject's stimulated TA pathways (see Fig. 7) suggested there was a correlation between the extent of facilitation and inhibition. While a regression analysis did not reveal a significant correlation, when the analysis was restricted to the seven ‘best modulators’ (I, V, VI, VIII, IX, XI, XIII) who demonstrated the largest percentage differences between facilitation and inhibition (> 40), a significant positive correlation emerged (r2= 0.88, P= 0.005). This analysis indicated that although these subjects demonstrated robust modulation in both directions, greater facilitation was associated with less inhibition.
, 百拇医药
A comparison of the effects of PASfac and PASinhib in each subject. A, averaged Post0 and Post10 values of individual subject's means. The thick line represents the group mean. B, a scatter plot of means from subjects I, V, VI, VIII, IX, XI and XIII with fac-inhib differences of > 40 revealing that although greater facilitation was associated with less inhibition, subjects who demonstrated robust facilitation in one session also demonstrated robust inhibition in the other session (P= 0.005).
, 百拇医药
Modulation was also revealed in soleus pathways
Following PASfac applied to TA pathways, no effects were revealed by the ANOVA of Sol responses, although group mean Sol MEP area differed from 100% in the stimulated limb at Post10 (Fig. 8). Sol responses were collected during late stance. Data from four subjects were omitted from the following analyses because either motor threshold for Sol MEPs was not reached (recall that TA was the target muscle and parameters were established to optimize TA responses), or background Sol EMG could not be matched within an individual across the three trials. ‘Stim’ refers to means of Sol responses recorded from the muscle ipsilateral to stimulated TA pathways, ‘Non-Stim’ refers to means of Sol responses recorded from the muscle contralateral to stimulated TA pathways. Of the eight PASfac Sol means (amplitude, area; ‘Stim’, ‘Non-Stim’; Post0, Post10), only ‘Stim’ Post10 area differed from 100% (mean, 112; P= 0.044) (see Fig. 8B). The remaining means ranged from 94 to 107. The ANOVA for amplitude and area did not reveal effects of ‘pathway’ or ‘time’ (all P > 0.1). Following PASinhib (not illustrated), two of the eight Sol means differed from 100%: the ‘Stim’ amplitude mean (122%) at Post10 (P= 0.029) and the ‘Stim’ area mean (116%) at Post0 (P= 0.041). The ANOVA for amplitude revealed an effect of ‘time’ where ‘Stim’ Post10 (mean, 121%) was greater than Post0 (mean, 109%), F1,10= 5.7, P= 0.038).
, http://www.100md.com
The comparison of mean MEP size (A, amplitude; B, rectified area) during late stance recorded from the Sol ipsilateral to the stimulated TA pathways (‘Stim’) and during late stance recorded from the Sol ipsilateral to the non-stimulated TA pathways (‘Non-Stim’). MEP size expressed as a percentage of pre-intervention (Pre). Open bars represent the means of Post0, and filled bars represent the means of Post10 responses. The horizontal dashed line represents the Pre mean. The star indicates that the mean differed from 100% (one-tailed t test), P < 0.05. Error bars represent 1 S.E.M.
, 百拇医药
Robust modulation was not evident with peripheral and cortical stimulation alone
Only PAS induced a significant modulation in MEP amplitude. Analysis of control data recorded from TA during the eight sessions conducted on four subjects did not reveal an effect of walking alone, PN stimulation alone, or TMS alone, but did reveal an effect of PAS. The PostPN and PostTMS means (101 and 105%, respectively) did not differ from 100% using one-tailed t tests (all P > 0.3). The CPost0, and CPost10 conditions (equivalent to the intervention in the main experiment) differed from 100%, where the CPost0 mean was 114% (P= 0.034), and the CPost10 mean was 110% (P= 0.011). A one-way ANOVA revealed an effect of condition (F4= 3.0, P= 0.033), and post hoc Tukey HSD tests (to detect ‘honestly significant differences’) revealed an effect of PAS where the CPost0 mean (114%) was greater than the Postwalk mean (96%) (P < 0.05). For illustrative purposes Fig. 9 separates the control condition means of the four control sessions that were concluded with PASfac from the four that were concluded with PASinhib.
, http://www.100md.com
, the means of the four control subjects (VI, VII, IX, XIII) during the session that culminated in PASfac. , the means of the four control subjects during the session that culminated in PASinhib. Filled symbols represent means expressed as a percentage of Prewalk. Open symbols represent means expressed as a percentage of either PostTMS or PostPN. Stars indicate that means of unsigned changes in MEP amplitude following PASfac and PASinhib that were grouped for statistical analysis differed from 100% (see text for details). Error bars are 1 S.E.M.
, http://www.100md.com
PAS applied during tonic contraction also induced facilitation
In the last control experiment, the extent of facilitation of TA pathways was similar to that derived from the main experiment (123%). The Post0 and Post10 MEP amplitudes for the four control subjects (% Pre) were pooled (n= 8) separately for data collected during walking and during tonic contraction. The tonic contraction mean (118%) differed from unity (P= 0.002). The tonic contraction mean did not differ from the walking mean (120%) using a paired t test (P= 0.79).
, http://www.100md.com
Discussion
The main finding was that PAS induced changes in dorsiflexor CM excitability of healthy subjects during walking. Bidirectional changes in lower limb CM excitability induced by PAS during walking have not previously been reported. Four criteria characterizing the physiological profile of PAS-induced phenomena in the human motor system have previously been described by Stefan et al. (2000, 2002); and Wolters et al. (2003). Three of these were evident in the present data. The first is the temporal rule, in which the same group of subjects demonstrated an increase in CM excitability when the paired ISI was 5 ms longer than the estimated MEP latency, and also demonstrated a decrease in CM excitability when the paired ISI was 10 ms shorter than the estimated MEP latency. The second criterion is rapid evolution, also demonstrated in the present experiment, where 120 pairs of stimuli were delivered over a 10 min period (cf. 90 pairs delivered over a 30 min period: Stefan et al. 2000, 2002; Wolters et al. 2003). While the relationship between stimulation period and the number of pairs of stimuli has yet to be examined, the present data clearly illustrate a rapid evolution of bidirectional changes in CM excitability. The third criterion is persistence of effects following paired stimulation. The present group data revealed no reduction in facilitation or inhibition at Post10 compared with Post0 (see Figs 5 and 6). The fourth criterion is reversibility which was not assessed in the present experiment.
, 百拇医药
Stefan et al. (2000, 2002) and Wolters et al. (2003) referred to PAS-induced phenomena as ‘long-term potentiation-like’ (LTP-like) and ‘long-term depression-like’ (LTD-like). Their choice of terminology was based on the characteristics of the temporal nature of the phenomena, its rapid evolution, persistence, and reversibility. Further, these authors found that PAS-induced facilitation was blocked by dextromethorphan, an N-methyl-D-aspartate receptor antagonist known to block long-term potentiation in animal tissue slice preparations. Despite these similarities, a clear relationship between the effects of PAS and the in vitro induction of LTP has not been established.
, 百拇医药
Inspection of Fig. 3 reveals that during PASfac four subjects continued to increase MEP amplitude (Fig. 3A), and during PASinhib five subjects continued to decrease MEP amplitude (Fig. 3B) in the 10 min period of walking immediately following PAS termination. This finding and the between-subject variability is consistent with the previous results from Stefan et al. (2000). Further, it was also noted that PAS failed to elicit MEP changes in a few subjects during either conditioning protocol (Fig. 3), indicating that some subjects were amenable to both PASfac and PASinhib and others were not amenable to either. This suggestion is also supported by the regression analysis of the seven ‘best modulators’ (Fig. 7B). If the difference in the extent of modulation between subjects had been merely the result of inappropriate setting of the interstimulus interval between stimulus-pairs no correlation between robust facilitation and robust inhibition would have been evident. However, inappropriate setting of parameters as the cause of the between-subject variability in the extent of modulation across the two sessions cannot be entirely ruled out. More robust effects of PAS may be evident if afferent conduction time was estimated from sensory-evoked potentials (SEP) recorded from the scalp in response to electrical stimulation of the peroneal nerve. There are no reports of PAS where specific interstimulus intervals have been based on SEP latency of individual subjects.
, 百拇医药
An intriguing observation made from comparing the extent of facilitation during PASfac in stimulated pathways compared with non-stimulated pathways (Fig. 4) was that four subjects who demonstrated robust facilitation also demonstrated robust inhibition in non-stimulated pathways. Although data from these four subjects in the present experiment cannot provide an explanation for this observation, if group data in a future experiment were to reveal a similar opposite-side-effect of PAS, the observation would warrant further investigation.
, http://www.100md.com
In the present experiment CM pathways to TA were chosen as the target. All stimulation parameters for probing CM excitability and for PAS were set for the target pathways. Therefore Sol data may be disregarded or interpreted with great caution. Notwithstanding this caveat, it is interesting to note that Sol MEP amplitude increased to 122% in the Post10 condition of the stimulated limb when PASinhib was being applied to the target TA pathway. The finding may indicate a reciprocal inhibitory mechanism for this agonist–antagonist pair of ankle effectors. However, during PASfac, a reciprocal effect was not evident and a modest 112% facilitation (not inhibition) was evident at Post10 in the stimulated limb Sol (Fig. 8B). At best, these data might indicate a non-specific facilitation of Sol pathways. The data warrant the design of an experiment that would examine the specificity of PAS on CM pathways to TA and to Sol separately. Changes in CM excitability in pathways to muscles of the leg as well as muscles of the shank could be included and optimal stimulation parameters could be established for each muscle pathway to enhance the validity of observed changes in non-target CM excitability.
, 百拇医药
An important issue addressed in the present experiment is whether any of the components of the intervention applied during walking modulate CM excitability on their own. The eight sessions conducted with four subjects revealed no statistically significant effects on MEP amplitude of walking alone, electrical stimulation of the peroneal nerve during walking alone, or TMS during walking alone. Significant effects were only revealed following PAS applied at the end of the control data collection sessions that involved walking. An inspection of Fig. 9 reveals that some modulation of MEP amplitude was taking place over the course of the control sessions; however, the extent of the modulation was small and failed to reach statistical significance compared with the modulation that followed PAS. These findings suggest that components of the PAS protocol as used in the present experiment do not significantly modulate CM excitability on their own. Finally, the results of the third control condition where subjects held a tonic TA contraction during the application of PASfac were consistent with the findings of Capaday et al. (1999) and Bawa et al. (2002) who showed that CM projections to TA are relatively stronger compared with projections to Sol. Capaday et al. (1999) found no difference between the extent of TA pathway excitability during walking and during a voluntary tonic contraction, and we found no difference between these two tasks in the extent of modulation induced by PAS.
, 百拇医药
The level of the neuroaxis at which the effects of PAS are occurring is assumed to be the M1 region because the temporal relationship between the paired stimuli is set to occur when the induced afferent volleys are estimated to have arrived in the cortex. This hypothesis has previously been supported by Stefan et al. (2000) who provided evidence of motoneurone stability despite significant increases in MEP amplitude following PAS. They reported no significant change in F-wave or M-wave amplitude or in the amplitude of potentials evoked by electrical stimulation at the level of the brainstem as a result of PAS. They also found that when they applied PAS to increase MEP amplitude there was an increase in the length of the post-MEP EMG silent period, suggesting the cortical component of the silent period had increased in length due to an increase in cortical input to motoneurone pools. Furthermore, Wolters et al. (2003) demonstrated that significant decreases in MEP amplitude induced by PAS were not evident in F-wave amplitude or the amplitude of potentials evoked by electrical stimulation at the level of the brainstem. In the present experiment silent periods were not examined because silent periods are difficult to assess in an EMG burst during walking where EMG amplitude is changing as a function of time. Since the stimuli were delivered in the present experiment toward the end of the swing phase, the return of voluntary EMG occurred at the offset of the EMG burst. A demonstration of where on the neuroaxis the effects of PAS were taking place was not attempted because the present experiment was primarily designed to assess the ability of PAS to modulate the excitability of the CM pathway as a whole during walking. However, the time window in which paired stimuli were delivered (–10 ms to 5 ms), and the lack of an effect of peripheral nerve stimulation alone suggests a supraspinal site. Future experiments to assess intracortical excitability and the excitability of spinal circuitry may provide further evidence of the cortex as the site where effects of PAS occur during walking.
, 百拇医药
Although PAS used to increase CM excitability during walking is unique to the present study, other authors have reported the use of functional electrical stimulation to increase CM excitability. For example, Knash et al. (2003) demonstrated increased MEP size in the TA with repetitive electrical stimulation of the common peroneal nerve. Significant facilitation was observed within the first 10 min of stimulation, and lasted up to 30 min following cessation of stimulation. Further, Khaslavskaia et al. (2002) delivered repetitive electrical stimulation to the peroneal nerve of healthy subjects and found that CM excitability was increased to a greater extent when stimulation was combined with voluntary dorsiflexor activation than when the muscle was at rest. Such augmented effects of repetitive electrical stimulation on CM excitability with combined voluntary effort may play a role in the restoration of motor behaviours following neurological injury (Field-Fote & Tepavac, 2002; Yan et al. 2005).
, http://www.100md.com
In conclusion, the hypothesis adopted for the present experiment was supported by evidence that PAS applied to ankle dorsiflexor CM pathways during the swing phase of walking induced temporally dependent bidirectional changes in CM excitability consistent with the Hebbian principle of activity-dependent plasticity. Similarities between these data derived from the lower limb, and data derived from CM pathways to hand muscles (Stefan et al. 2000, 2002; Wolters et al. 2003) suggest that the same neural mechanisms were involved. Of particular importance was the demonstration of these phenomena during walking which has not previously been reported. Walking training protocols applied to neurologically impaired individuals where CM excitability is suppressed, may benefit from the addition of a training adjunct that increases CM excitability.
, 百拇医药
References
Bawa P, Chalmers GR, Stewart H & Eisen AA (2002). Responses of ankle extensor and flexor motoneurons to transcranial magnetic stimulation. J Neurophysiol 88, 124–132.
Capaday C, Lavoie BA, 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. J Neurophysiol 81, 129–139.
, 百拇医药
Dupont WD & Plummer WD (1990). Power and sample size calculations: a review and computer program. Controlled Clin Trials 11, 116–128.
Field-Fote EC & Tepavac D (2002). Improved intralimb coordination in people with incomplete spinal cord injury following training with body weight support and electrical stimulation. Phys Ther 82, 707–715.
Hebb DO (1949). The Organization of Behavior. Wiley, New York.
, 百拇医药
Khaslavskaia S, Ladouceur M & Sinkjaer T (2002). Increase in tibialis anterior motor cortex excitability following repetitive electrical stimulation of the common peroneal nerve. Exp Brain Res 145, 309–315.
Knash ME, Kido A, Gorassini M, Chan KM & Stein RB (2003). Electrical stimulation of the human common peroneal nerve elicits lasting facilitation of cortical motor-evoked potentials. Exp Brain Res 153, 366–377.
Muellbacher W, Ziemann U, Boroojerdi B & Hallett M (2000). Effects of low-frequency transcranial magnetic stimulation on motor excitability and basic motor behavior. Clin Neurophysiol 111, 1002–1007.
, 百拇医药
Peinemann A, Reimer B, Loer C, Quartarone A, Munchau A, Conrad B & Siebner HR (2004). Long-lasting increase in corticospinal excitability after 1800 pulses of subthreshold 5 Hz repetitive TMS to the primary motor cortex. Clin Neurophysiol 115, 1519–1526.
Said CM, Goldie PA, Patla AE & Sparrow WA (2001). Effect of stroke on step characteristics of obstacle crossing. Arch Phys Mederehabil 82, 1712–1719.
Stefan K, Kunesch E, Benecke R, Cohen LG & Classen J (2002). Mechanisms of enhancement of human motor cortex excitability induced by interventional paired associative stimulation. J Physiol 543, 699–708.
, 百拇医药
Stefan K, Kunesch E, Cohen LG, Benecke R & Classen J (2000). Induction of plasticity in the human motor cortex by paired associative stimuli. Brain 123, 572–584.
Uy J, Ridding MC, Hillier S, Thompson PD & Miles TS (2003). Does induction of plastic change in motor cortex improve leg function after strokeNeurology 61, 982–984.
Wolters A, Sandbrink F, Schlottmann A, Kunesch E, Stefan K, Cohen LG, Benecke R & Classen J (2003). A temporally asymmetric Hebbian rule governing plasticity in the human motor cortex. J Neurophysiol 89, 2339–2345.
, http://www.100md.com
Yan T, Hui-Chan CWY & Li LSW (2005). Functional electrical stimulation improves motor recovery of the lower extremity and walking ability of subjects with first acute stroke: a randomized placebo-controlled trial. Stroke 36, 80–85.
Ziemann U (2004). TMS and drugs. Clin Neurophysiol 115, 1717–1729.
Ziemann U, Wittenberg GF & Cohen LG (2002). Stimulation-induced within-representation and across-representation plasticity in human motor cortex. J Neurosci 22, 5563–5571., 百拇医药(James W. Stinear and T. G)