- Email: [email protected]
Kinesthetic motor imagery and spinal excitability: The effect of contraction intensity and spatial localization
Clinical Neurophysiology 119 (2008) 1849–1856 www.elsevier.com/locate/clinph
Kinesthetic motor imagery and spinal excitability: The effect of contraction intensity and spatial localization Patrick M. Cowley*, Brian C. Clark1, Lori L. Ploutz-Snyder Musculoskeletal Research Laboratory, Department of Exercise Science at Syracuse University, 201 Women’s Building, 820 Comstock Ave, Syracuse, NY 13244, USA Accepted 4 April 2008 Available online 16 May 2008
Abstract Objective: Data on whether motor imagery (MI) modulates spinal excitability are equivocal. The purpose of this study was to determine if imagined muscle contractions of the left plantar flexor (PF) alter spinal excitability, and if so, to determine whether this alteration is intensity dependent and/or localized to the target muscles. Our research questions required two experiments. Methods: In experiment 1, 16 healthy volunteers performed imagined muscle contractions using a kinesthetic approach with their left PF at 25% and 100% of imagined effort (IE). The soleus H-reflex was evoked during three conditions, which were separated by about 15 s: rest (preceding MI), during MI, and recovery (following the cessation of MI). In experiment 2, a subset of subjects from experiment 1 performed MI with their left PF at 100% of IE, while either the soleus or flexor carpi radialis (FCR) H-reflex was measured. Results: In experiment 1, we observed a facilitation of soleus H-wave amplitude during MI compared to the rest and recovery conditions (p < 0.05). Furthermore, the soleus H-wave amplitude was greater during 100% than 25% of IE (p < 0.05). In experiment 2, soleus and FCR H-wave amplitude increased during imagined muscle contractions of the left PF (p < 0.05). These changes were independent of voluntary muscle activity. Conclusions: These findings suggest MI can increase spinal excitability by the intensity of imagined effort, but this effect is not fully localized to the task specific muscle. Significance: These data provide evidence that MI can increase spinal excitability in healthy subjects, which suggests future studies are warranted to examine the clinical relevance of this effect. These studies are needed to help establish a therapeutic theory by which to advance motor function rehabilitation using MI. Ó 2008 International Federation of Clinical Neurophysiology All rights reserved. Keywords: Motor imagery; H-Reflex; Spinal excitability
1. Introduction Motor imagery (MI) training has recently intrigued researchers and clinicians for its potential role in mitigating the effects of motor deficits and strength loss following cerebrovascular trauma and periods of prolonged disuse (Butler and Page, 2006; Cicinelli et al., 2006; Clark et al., *
Corresponding author. Tel.: +1 315 443 1411; fax: +1 315 443 9375. E-mail address: [email protected] (P.M. Cowley). 1 Present address: Department of Biomedical Sciences, 211 Irvine Hall, Ohio University, Athens, OH 45701, USA.
2006; Dickstein et al., 2004; Johnson-Frey, 2004; Liu et al., 2004; Nair et al., 2005). It has been proposed that the degree of imagined effort should correspond to the effort during actual performance, but at a lower magnitude (Decety et al., 1993, 1991; Guillot and Collet, 2005). Evidence from functional imaging studies support this viewpoint, as MI and motor task performance have been shown to activate similar areas of the brain (Ehrsson et al., 2003; Gerardin et al., 2000; Hanakawa et al., 2003; Porro et al., 1996; Stephan et al., 1995); the change in functional MRI signal intensity during MI is 30% of that during actual task performance (Porro et al., 1996). MI also
1388-2457/$34.00 Ó 2008 International Federation of Clinical Neurophysiology All rights reserved. doi:10.1016/j.clinph.2008.04.004
1850
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
acutely increases cortical excitability as measured via transcranial magnetic stimulation (Facchini et al., 2002; Hashimoto and Rothwell, 1999; Rossi et al., 1998; Stinear and Byblow, 2003; Stinear et al., 2006; Yahagi and Kasai, 1998; Yahagi et al., 1996). Collectively, these findings suggest MI results in activation of the motor cortex; however, data on whether MI modulates spinal excitability have been equivocal. While some authors have reported an increase in spinal excitability during MI (Bonnet et al., 1997; Gandevia et al., 1997; Hale et al., 2003; Kiers et al., 1997; Rossini et al., 1999), others have not (Facchini et al., 2002; Hashimoto and Rothwell, 1999; Kasai et al., 1997; Stinear and Byblow, 2003; Stinear et al., 2006; Yahagi et al., 1996). For example, using the H-reflex technique Kiers et al. and Hale et al. showed MI increases spinal excitability, but this effect was limited to 50% of subjects and was independent of imagined effort (Hale et al., 2003; Kiers et al., 1997). These findings suggest MI is not only a supraspinal process, but also facilitates spinal excitability. When data from the literature are synthesized it appears that certain methodological differences may account for the lack of consistent findings on the effect of MI on spinal excitability. For example, Hale et al. reported MI increases spinal excitability in the absence of voluntary muscle activity when employing a high-intensity MI task (Hale et al., 2003), whereas many other studies using a low-intensity, fine MI task frequently report no effect on spinal excitability (Bonnet et al., 1997; Hashimoto and Rothwell, 1999; Kasai et al., 1997; Stinear and Byblow, 2003). The primary aim of our study was to determine if MI contraction intensity effects spinal excitability using the H-reflex technique. A secondary aim of this study was to determine if an acute alteration in excitability during MI is localized to the target muscle. Our research questions were addressed through two separate experiments. For our first experiment we examined whether soleus H-wave amplitude was altered during kinesthetic MI at 100% and 25% of imagined effort (IE) using the left plantar flexor (PF) muscle group. We hypothesized that changes in soleus H-wave amplitude would be facilitated during the imagined effort, and that this facilitation would be greatest during the higher imagined contraction intensity. The findings from our first experiment led us to explore if the effect of MI is localized to the target muscle. For this experiment we examined changes in soleus and flexor carpi radialis (FCR) H-wave amplitude during kinesthetically imagined maximal PF muscle contractions. Here, we hypothesized that changes in H-wave amplitude would only be observed in the soleus muscle.
MI trials each subject performed 3 voluntary contractions at 100% and 25% of maximum voluntary contraction (MVC) using the left PF, so they could establish a reference point to relate back to during the MI trials. The time between the voluntary contractions and MI trials was 10 min. A total of 20 trials of MI were performed at the two different intensities (10 per intensity performed in a randomized order), and the soleus H-reflex was elicited at rest, during MI, and at rest following MI (recovery). Fifteen seconds separated H-reflex measurements. For the MI trial each subject was given a verbal command to imagine contracting their left PF at 100% or 25% of IE, which lasted approximately 15 s, and between the 8th and 15th second the soleus H-reflex was elicited. In order to ensure there was no voluntary activation of the left PF during the MI trial the interference electromyogram (EMG) signal was quantified (root mean squared EMG) 50 ms immediately before the H-reflex measurement. Our findings from the first experiment indicated MI can acutely increase soleus H-wave amplitude, and that this was contraction intensity dependent with a slightly higher facilitation during the maximal MI task (see Section 3 for complete details). Therefore, to further investigate these findings we conducted a follow-up experiment to determine if the change in spinal excitability was localized to the soleus. To do this we invited a subset of the subjects from the first experiment back to the laboratory. Subjects in whom soleus H-wave amplitude increased by at least the mean increase (5.1%) were invited to participate in the follow-up experiment (Fig. 1). Of these seven subjects, six volunteered to participate in the second experiment. The decision to use the mean increase of 5.1% in H-wave amplitude as a cut-off point to include subjects in experiment 2 was based on our research question, which was to examine whether the increase in H-wave amplitude was localized to the target muscle. Thus requiring subjects who showed an increase in H-wave in response to MI.
2. Methods 2.1. General overview of experimental design and procedure For the first experiment we examined the effect of imagined PF muscle contractions at 100% and 25% of IE on soleus H-wave amplitude in sixteen subjects. Before the
Fig. 1. Percent change in soleus H20: Mmax from rest to motor imagery for each subject in experiment 1. The line represents the mean increase in soleus H20: Mmax for all subjects (mean increase of 5.1%). Subjects that had greater than a 5.1% increase in soleus H20: Mmax are represented by closed bars, and those that exhibited less than a 5.1% increase are represented by open bars. Subject’s 1, 3, 4, 5, 6, and 7 participated in experiment 2.
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
During this experiment we examined the effect of maximal imagined muscle contractions of the left PF on soleus H-wave amplitude, as well as on FCR (a forearm muscle) H-wave amplitude. The MI trials were preceded by three MVC’s using the left PF to establish a reference point for subjects to relate back to during the MI trials. The MVC’s and MI trials were separated by 10 min. Twelve trials of PF MI were performed with the soleus or FCR H-reflex being randomly assessed (6 per muscle performed in a randomized order) at rest before MI, during MI, and at rest after MI (recovery). Fewer MI trials were used in experiment 2 compared to experiment 1 because the results from experiment 1 indicated H-wave amplitude did not increase with the number of MI trials performed (analysis not shown) as has been previously shown (Hale et al., 2003). Fifteen seconds separated H-reflex measurements. In order to ensure there was no voluntary activation of the left PF or FCR before the H-reflex measurements the interference EMG activity was quantified (RMS EMG) for the 50 ms before the H-reflex measurements. For this experiment, the interference EMG activity of the left soleus is expressed relative to the interference EMG activity recorded during the actual MVC; referred to herein as normalized interference EMG.
1851
lowing the MVC trials in experiment 1 each subject performed three submaximal isometric contractions at 25% MVC. They were asked to match a target force displayed as a horizontal line on the computer monitor and maintain the contraction for 8 s. During the performance of these tasks subjects were asked to remember this feeling and use it during the MI trials. 2.4. Electrical recordings Electrical signals were recorded from both the soleus and FCR muscles using bipolar surface electrodes (Ag– AgCl, Soleus Electrodes: 36 mm diameter, Kendall MediTrace 200; FCR Electrodes: 15 mm diameter, Danlee Medical Products, Inc.). Before electrode placement, the subject’s skin was shaved and then abraded and cleaned with an alcohol pad to minimize skin impedance. Electrodes were placed longitudinally over the muscles with an interelectrode distance of 2 cm. For the soleus recordings a reference electrode was placed on the upper lateral aspect of the calf. For the FCR recordings a reference electrode was placed over the radial styloid. The EMG signals were amplified 500, band-pass filtered between 10 and 500 Hz, and sampled at 2500 Hz (MP150, BioPac Systems Inc., Goleta, CA).
2.2. Subjects 2.5. Electrical stimulation and spinal excitability Sixteen individuals participated in experiment 1 (mean ± standard deviation of 29 yrs ± 8, range of 20–48 yrs; 10 men and 6 women). Six individuals from experiment 1 (mean ± standard deviation of 27.3 yrs ± 4.3, range of 22–33 yrs; 4 men and 2 women) participated in experiment 2. The Institutional Review Board of Syracuse University approved the procedures used in this study, and all subjects gave written informed consent prior to participation. 2.3. Mechanical recordings and voluntary contraction trials Left PF force was assessed while the subject sat in a custom-modified PF dynamometer (Parabody 826, LifeFitness, Schiller Park, IL) equipped with a force transducer (TSD121C, BioPac Systems, Goleta, CA). The left leg of the subject was positioned so that the hip, knee, and ankle joints were at 90°, and a 43-cm computer monitor was positioned 1-m in front of the subject for visual feedback. Prior to the subjects performing the MI tasks, subjects performed voluntary PF contractions. For experiment 1, the subjects performed 3 voluntary contractions with their left PF at 100% and 25% of their MVC with 1–2 min between trials. For experiment 2, the three voluntary contractions were performed using the left PF at 100% MVC. The subjects did not perform a voluntary contraction using their left forearm flexors. For the MVC each subject was asked to gradually increase force and then perform a maximal effort for 2–3 s. The highest of the three was considered the MVC force. During the MVC’s the investigators provided strong verbal encouragement. Fol-
To measure spinal excitability we used the H-reflex technique. The H-reflexes were elicited using procedures previously outlined by our laboratory (Clark et al., 2006, 2007), which we have reported to have high reliability (intraclass correlation coefficient = 0.93) (Clark et al., 2007). In brief, subjects lay prone on an examination table with the ankle joint fixed at 45° of plantarflexion by a cylindrical cushion placed under the talus. The head was down in medial alignment with the torso, and rested in a face rest so that the neck was in a neutral position. The arms and hands were pronated, and placed along the side of the body. To elicit the soleus H-reflex a cathode (Ag–AgCl, 36 mm diameter; Kendal Medi-Trace 200) was placed over the tibial nerve in the popliteal fossa. The cathode site was determined using a hand-held stimulation probe. An anode was placed above the cathode on the posterior thigh (Ag–AgCl, 48 mm diameter; Kendal Medi-Trace 530). The electrical stimulus consisted of a 1 ms square pulse (Grass S88 stimulator coupled with a Grass SIU5 stimulus isolation unit, Astro-Med Inc., West Warwich, RI). The stimulus intensity was continuously increased to obtain a maximal Mwave (Mmax), and then adjusted and continuously monitored to evoke an M-wave equal to 20 ± 2.5% of the Mmax. The corresponding peak-to-peak amplitude of the H-wave was then calculated and expressed relative to the Mmax (H20: Mmax). To elicit the FCR H-reflex a cathode (Ag–AgCl, 15 mm diameter; Kendal Medi-Trace 200) was placed over the median nerve in the bicipital groove after the optimum site
1852
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
contracting, and not a visual approach where one visualizes oneself performing the task. The same investigator gave the instructions to each subject.
for stimulation was located by a hand-held stimulation probe. The anode was placed above the cathode (Ag–AgCl, 36 mm diameter; Kendal Medi-Trace 530). The electrical stimulus consisted of a 1 ms square pulse, and the intensity was continuously increased to obtain a maximal M-wave (Mmax). The stimulus intensity was continuously adjusted and monitored to evoke an M-wave equal to 5 ± 2.5% of the Mmax. The corresponding peak-to-peak amplitude of the H-wave was then calculated and expressed relative to the Mmax(H5: Mmax). To compare and interpret changes in evoked H-reflex responses during the experimental conditions, it was vital that the effective stimulus strength remain similar between the recording sessions, because the size of the H-reflex is heavily modulated by changes in stimulus intensity. As such it has been suggested that when repeated H-reflex measurements are performed, it is advantageous to use a stimulation intensity that produces similar M-wave responses to ensure that the same number of motor axons are recruited in each trial, which indicates that stimulus intensity to the efferent nerve is also kept constant (Aagaard et al., 2002). Based on this logic, we set limits on the evoked M-wave (M-wave equal to 20 ± 2.5% of the Mmax for soleus and 5 ± 2.5% of the Mmax for FCR) to make sure any effect observed was limited to motor imagery and not stimulus intensity. For the H-reflex recordings, if an MI trial was out of range (i.e., greater than or less than the established limits relative to the Mmax) it was not included in the analysis. A complete trial included an H-reflex measurement at rest, during MI, and recovery. Trials that were out of range were not repeated because it has been shown that spinal excitability is modulated by the number of trials performed (i.e., practice) (Hale et al., 2003). Data from one subject was excluded from the analysis for experiment 1 because we were unable to obtain a reasonable number of trials within the given range. For the subjects included in experiment 1, 85% of their trials were complete, and 81% of soleus trials and 92% of FCR trials for experiment 2 were complete.
For experiment 1, to examine the effect of MI on soleus H-wave amplitude a repeated measure analysis of variance was performed with condition (H-reflex at rest, during MI, and recovery) and intensity (100% and 25% of IE) as the within subject factors. To account for the possibility that voluntary muscle activity influenced H-wave amplitude during the MI condition we performed an additional analysis. Separate repeated measure analysis of covariance was performed for the 100% and 25% of IE trials with condition (H-reflex at rest, during MI, and recovery) as the within subject factor and the interference EMG (RMS EMG 50 ms immediately before the MI H-reflex measurement) for the respective intensity as a covariate. For experiment 2, to examine the effect of MI on soleus and FCR H-wave amplitude separate repeated measure analysis of variances was used with condition (H-reflex at rest, during MI, and recovery) as the within subject factors. We also examined the effect of MI on soleus normalized interference EMG and FCR interference EMG (RMS EMG for 50 ms immediately before the H-reflex measurements) by performing separate repeated measure analysis of variance with condition (EMG at rest before MI, during MI, and recovery) as the within subject factor. When a significant main effect or interaction was observed, a Sidak post hoc test was used to determine where the main effect or interaction was significant. The assumption of normality was examined using the Kolmogorov–Smimov Test of Normality and normal probability plots. Data are presented as means ± SEM, unless otherwise stated. The alpha was set at p 6 0.05, and the partial Eta squared effect size statistic (g2) and power of main effects and interactions are reported to further aid in data interpretation. The SPSS (version 13.0, Chicago, IL) statistical package was used for data analysis.
2.6. Kinesthetic motor imagery
3. Results
The verbal directions used during the MI trials are similar to those reported by our laboratory (Clark et al., 2006). Briefly, on a verbal signal to begin, the subject was instructed to ‘‘concentrate on your left calf muscles, and imagine your foot pushing at 100% (or 25%) of your maximum against the footpad. Feel yourself pushing up at 100% (or 25%) of your maximum with your left calf muscles. Keep pushing up at 100% (or 25%) of your maximum with your left calf muscles. Keep imagining you are pushing at 100% (or 25%) of your maximum, keep pushing, keep pushing. . .now stop.” The duration of each imagined contraction was 15 s and the subjects were asked to adopt a kinesthetic imagery approach, in which they were urged to create an internal representation and feeling of their muscle
3.1. Experiment 1
2.7. Statistical analysis
For experiment 1, we observed a significant main effect of condition, and intensity by condition interaction. First, soleus H-wave amplitude increased during MI at 100% and 25% of IE (F(2,28) = 11.1, p 6 0.05, g2 = 0.44, power = 0.93). Further, there was a significant intensity by condition interaction (F(2,28) = 4.1, p 6 0.05, g2 = 0.23, power = 0.56). Subsequent analysis revealed soleus H-wave amplitude was greater during MI at 100% (50.9 ± 3.4 H20: Mmax) than 25% of IE (49.9 ± 3.5 H20: Mmax) compared to the rest (F(1,14) = 5.7, p 6 0.05) and recovery conditions (F(1,14) = 4.8, p 6 0.05) (Fig. 2). An example of the change in soleus H-wave amplitude during MI compared to the
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
rest and recovery conditions is presented in Fig. 3. The interference EMG immediately before the soleus H-reflex measurement during the 100% and 25% of IE trials was 0.002 mV ± 0.00 and 0.002 mV ± 0.00, respectively (means ± standard deviation). The separate repeated measure analysis of covariance performed for the 100% and 25% IE trials using the interference EMG as a covariate showed that voluntary muscle activity did not affect H-wave amplitude (F(2,26) = 4.2, p 6 0.05 and F(2,26) = 3.5, p 6 0.05 for the 100% and 25% IE trials, respectively). Thus, H-wave amplitude increased during MI and was greater at 100% than 25% of IE, and this was not influenced by voluntary muscle activity, as our analysis was significant when we did and did not account for any potential changes in the interference EMG during MI.
1853
Fig. 3. An example of the soleus H-wave amplitude during the rest, motor imagery, and recovery conditions. The stimulus intensity was adjusted to elicit a motor response equal to 20 ± 2.5% of maximum, and the corresponding H-wave amplitude was normalized to the maximum motor response (H20: Mmax).
3.2. Experiment 2 For experiment 2, we observed a significant main effect of condition for the soleus muscle (F(2,10) = 13.0, p 6 0.05, g2 = 0.72, power=0.98) and the FCR muscle (F(2,10) = 8.1, p 6 0.05, g2 = 0.62, power = 0.88). Subsequent analysis revealed soleus H-wave amplitude was higher during MI than during the rest condition (55.3 ± 4.2 vs. 50.3 ± 5.2 H20: Mmax; p 6 0.05), but that there was no difference in soleus H-wave amplitude between the MI and recovery conditions (55.3 ± 4.2 vs. 52.1 ± 4.9 H20: Mmax; p > 0.05). As expected there was no difference in H-wave amplitude between the rest and recovery conditions (p > 0.05). Based on the fact that soleus H-wave amplitude increased during MI compared to rest, with no difference between MI and recovery, and rest and recovery conditions an additional analysis was performed to aid in the interpretation of these data. The values for the rest and recovery conditions were averaged and compared to the value obtained during MI using a
Fig. 2. Soleus H-wave amplitude during the rest, motor imagery (MI), and recovery conditions. Soleus H-wave amplitude increased during MI compared to the rest and recovery conditions, however it increased more during MI at 100% than 25% of imagined effort (p 6 0.05). *Significant increase in soleus H-wave amplitude during MI compared to the rest and recovery conditions. §Significantly greater increase in soleus H-wave amplitude during MI at 100% than 25% of imagined effort.
paired t-test; this analysis was significant (t(1,5) = 3.8, p 6 0.05; 55.3 ± 4.2 vs. 51.2 ± 5.1 H20: Mmax), indicating soleus H-wave amplitude increased during the MI condition compared to the non-MI conditions (Fig. 4). There was no difference in the normalized interference EMG immediately before the soleus H-reflex measurements (F(2,10) = 0.6, p > 0.05; 0.40 ± 0.12% vs. 0.44 ± 0.14% vs. 0.67 ± 0.29% for the rest, MI, and recovery conditions, respectively). Follow-up analysis revealed FCR H-wave amplitude was higher during MI than the recovery condition (22.3 ± 5.0 vs. 20.3 ± 4.7 H5: Mmax; p 6 0.05), but that there was no difference in FCR H-wave amplitude between the MI and rest conditions (22.3 ± 5.0 vs. 20.5 ± 4.7 H5: Mmax; p > 0.05). As expected there was no difference in H-wave amplitude between the rest and recovery conditions (p > 0.05). Based on the fact that FCR H-wave amplitude increased during MI compared to recovery, with no
Fig. 4. Soleus and flexor carpi radialis (FCR) H-wave amplitude during the non-motor imagery conditions (average of the rest and recovery) compared to motor imagery (MI). The amplitude of the soleus and FCR H-waves was expressed relative to Mmax (H20: Mmax for soleus and H5: Mmax for FCR). *Both soleus and FCR H-wave amplitude significantly increased during MI compared to the non-MI conditions.
1854
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
difference between rest and MI, and rest and recovery conditions an additional analysis was performed to aid in the interpretation of these data. The values for the rest and recovery conditions were averaged and compared to the value obtained during MI using a paired t-test; this analysis was significant (t(1,5) = 3.5, p 6 0.05; 22.3 ± 5.0 vs. 20.4 ± 4.8 H20: Mmax), indicating FCR H-wave amplitude increased during the MI condition compared to the nonMI conditions (Fig. 4). There was no difference in the interference EMG immediately before the FCR H-reflex measurements (F(2,10) = 2.1, p > 0.05; 0.0007 ± 0.00 mV vs. 0.0007 ± 0.00 mV vs. 0.0018 ± 0.00 mV for the rest, MI, and recovery conditions, respectively). 4. Discussion Our first experiment examined the effect of imagined PF muscle contractions at 100% and 25% of IE on soleus Hwave amplitude. We hypothesized that changes in soleus H-wave amplitude would occur during MI, and that these changes would be higher during MI at 100% than 25% of IE. Our findings are consistent with this hypothesis, as soleus H-wave amplitude increased during both MI tasks when compared to the rest and recovery conditions; increasing to a greater extent during the maximal MI task. The changes in H-wave amplitude during MI were independent of voluntary muscle activity, as when we covaried for the degree of interference EMG during the MI task no effect was observed. This observation is consistent with other studies (Hale et al., 2003; Kiers et al., 1997). In experiment 2 we examined the effect of imagined maximal PF muscle contractions on soleus and FCR Hwave amplitude. We hypothesized that changes in H-wave amplitude would increase in the soleus, but not in the FCR indicating a localized facilitation and not a systemic response. Soleus and FCR H-wave amplitude increased during MI when compared to the average of the non-MI conditions (rest and recovery). This is in contrast to our hypothesis, as we did not expect there to be any facilitation in the FCR. While it is somewhat difficult to determine the magnitude of each of these responses some insight can be gleaned from the reported effect sizes (partial eta-squared), which represents the proportion of total variation attributable to the factor, partialing out other factors from the total non-error variation. As such, the slightly larger effect size observed for the MI task in the soleus (g2 = 0.72) in comparison to the FCR (g2 = 0.62) suggests that the soleus H-wave amplitude increased slightly more than the FCR’s, and that the increase in spinal excitability during MI is only partially localized to the target muscle. It is important to interpret these data with caution, as the sample size was small (n = 6). However, these findings warrant additional investigations into this effect. In experiment 1 we observed a significant increase in Hwave amplitude during MI, which was primarily due to the large change in H-wave amplitude from 7 of the 15 subjects (illustrated in Fig. 1). A potential reason why some sub-
jects’ spinal excitability did not increase to as great an extent of those in Fig. 1 may be because PF force and the H-reflex measurements were obtained in different positions (sitting for PF force vs. lying prone for the H-reflex measurement). However, this observation is consistent with that of Kiers et al. (Kiers et al. (1997)), as only 4 of the 8 subjects in their study increased spinal excitability during MI. For the 6 subjects that completed both experimental paradigms, the change in soleus H20: Mmax during MI compared to the rest condition was 11% during experiment 1 and 10% during experiment 2. Collectively, these findings suggest MI is mainly a supraspinal process; however, it can also affect excitability at the level of the spinal cord. Modulation of H-wave amplitude can result from a number of factors. These factors can include: (1) presynaptic inhibition of Ia fibers, (2) variation in the amount of Ia neurotransmitter release, and (3) changes in excitability of the motorneuron due to changes in excitatory or inhibitory inputs (Clark et al., 2006). We cannot explain which of these factors effects H-wave amplitude during MI; thus, we will refer to these changes herein as changes in spinal excitability. 4.1. MI can increase soleus spinal excitability by the intensity of imagined effort It has previously been shown that MI shares many of the same neural substrates as actual performance (Ehrsson et al., 2003; Gerardin et al., 2000; Hanakawa et al., 2003; Porro et al., 1996; Stephan et al., 1995); therefore, it has been proposed that the degree of imagined effort should correspond to the effort during actual performance, but at a lower magnitude (Decety et al., 1991, 1993; Guillot and Collet, 2005). These data are consistent with this suggestion, as MI can acutely increase spinal excitability in an intensity dependent fashion, with a maximal imagined contraction being greater than an imagined contraction at an intensity equal to 25% of maximum. We are aware of only two studies that have examined the effect of imagined effort on soleus spinal excitability. Bonnet et al. assessed spinal excitability using the H- and tendon-reflex (T-reflex) techniques by having subjects imagine either ‘‘weak” (2% of IE) or ‘‘strong” (10% of IE) PF isometric contractions (Bonnet et al., 1997). They found the change in T-reflex amplitude during MI, but not H-reflex amplitude intensity dependent. We cannot compare our results to those of Bonnet et al. as interference EMG activity increased during their MI condition (which we did not observe). In addition, the contraction intensities used by Bonnet et al. during the MI condition were extremely low (2% vs. 10%). Thus, it is possible that subtle differences in spinal excitability during MI at very low contraction intensities cannot be discerned using the H-reflex technique. More recently, Hale et al. assessed soleus H-reflex excitability during MI by having subjects perform imagined PF contractions at 40%, 60%, 80%, and 100% of IE (Hale et al., 2003). They showed spinal excitability increased with
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
1855
the number of MI trials performed, but not by the intensity of imagined effort. Differences in the intensity of the imagined contraction may explain the differing results. We obtained our measurements at 25% and 100% of IE, while the lowest measured intensity by Hale and colleagues was 40%. Therefore, because we used a lower contraction intensity this may have allowed us to better delineate the effect of MI intensity on spinal excitability (25% vs. 100% compared to 40% vs. 100%). Our data suggest MI can modulate spinal excitability in an intensity dependent fashion, independent of voluntary muscle activity.
(measured using the H-reflex technique) in select subjects performing imagined MVC’s of their left PF. The clinical significance of MI training enhancing spinal excitability is uncertain at this time. However, these data provide a necessary first step in examining this effect in healthy subjects. Future studies examining this effect in clinical populations to determine any effect on clinical outcomes are warranted. These studies are needed to help establish a therapeutic theory by which to approach motor function rehabilitation following cerebrovascular trauma or during periods of prolonged disuse (e.g., bed rest, casting, spaceflight).
4.2. Changes in spinal excitability are not localized to the target muscle
References
Our data indicate MI increases soleus spinal excitability during imagined PF contractions, but based on our observation of a concurrent facilitation in the FCR spinal excitability it appears that the effect is not fully localized to the target (soleus) muscle. Based on the reported effect sizes (g2 = 0.72 for soleus, and g2 = 0.62 for FCR), however, it appears the larger facilitation occurred at the soleus. This is in contrast to our hypothesis, as we did not expect there to be any facilitation in the FCR. These findings are consistent with that of Marconi et al., who found imagined PF muscle contractions increased cortical excitability of the resting FCR in lower limb amputees and control subjects without amputation (Marconi et al., 2007). We extend the findings of Marconi et al. by showing that spinal excitability increases in the FCR during imagined PF contractions as well. This effect was first observed during voluntary muscle contractions of the foot, which caused reciprocal changes in the excitability of the resting wrist muscles (Baldissera et al., 2002; Borroni et al., 2004). Recent evidence indicates something of central origin modulates this effect (Borroni et al., 2004; Byblow et al., 2007; Cerri et al., 2003). Byblow et al. recently showed neural networks involving secondary motor areas, specifically the dorsal premotor cortex, and the primary motor cortex facilitate these reciprocal changes in excitability of the resting wrist muscles during voluntary contraction of the foot (Byblow et al., 2007). The clinical relevance of this finding remains to be determined. However, Byblow et al. suggested that voluntary movement of the leg may induce effects on upper limb primary motor cortex excitability, and this may be of considerable value given the importance of the dorsal premotor cortex during recovery from stroke (Byblow et al., 2007). 5. Conclusion We examined the effect of MI on spinal excitability. Our data indicate that MI can increase soleus spinal excitability in an intensity dependent fashion when performing a highintensity MI plantarflexion task in select subjects, but the entire effect is not fully localized to the soleus muscle. Collectively, these findings suggest kinesthetic MI is not solely a supraspinal process, but modulates spinal excitability
Aagaard P, Simonsen EB, Andersen JL, Magnusson P, Dyhre-Poulsen P. Neural adaptation to resistance training: changes in evoked V-wave and H-reflex responses. J Appl Physiol 2002;92:2309–18. Baldissera F, Borroni P, Cavallari P, Cerri G. Excitability changes in human corticospinal projections to forearm muscles during voluntary movement of ipsilateral foot. J Physiol 2002;539:903–11. Bonnet M, Decety J, Jeannerod M, Requin J. Mental simulation of an action modulates the excitability of spinal reflex pathways in man. Brain Res Cogn Brain Res 1997;5:221–8. Borroni P, Cerri G, Baldissera F. Excitability changes in resting forearm muscles during voluntary foot movements depend on hand position: a neural substrate for hand–foot isodirectional coupling. Brain Res 2004;1022:117–25. Butler AJ, Page SJ. Mental practice with motor imagery: evidence for motor recovery and cortical reorganization after stroke. Arch Phys Med Rehabil 2006;87:S2–S11. Byblow WD, Coxon JP, Stinear CM, Fleming MK, Williams G, Muller JF, et al. Functional connectivity between secondary and primary motor areas underlying hand–foot coordination. J Neurophysiol 2007;98:414–22. Cerri G, Borroni P, Baldissera F. Cyclic H-reflex modulation in resting forearm related to contractions of foot movers, not to foot movement. J Neurophysiol 2003;90:81–8. Cicinelli P, Marconi B, Zaccagnini M, Pasqualetti P, Filippi MM, Rossini PM. Imagery-induced cortical excitability changes in stroke: a transcranial magnetic stimulation study. Cereb Cortex 2006;16:247–53. Clark BC, Cook SB, Ploutz-Snyder LL. Reliability of techniques to assess human neuromuscular function in vivo. J Electromyogr Kinesiol 2007;17:90–101. Clark BC, Manini TM, Bolanowski SJ, Ploutz-Snyder LL. Adaptations in human neuromuscular function following prolonged unweighting: II. Neurological properties and motor imagery efficacy. J Appl Physiol 2006;101:264–72. Decety J, Jeannerod M, Durozard D, Baverel G. Central activation of autonomic effectors during mental simulation of motor actions in man. J Physiol 1993;461:549–63. Decety J, Jeannerod M, Germain M, Pastene J. Vegetative response during imagined movement is proportional to mental effort. Behav Brain Res 1991;42:1–5. Dickstein R, Dunsky A, Marcovitz E. Motor imagery for gait rehabilitation in post-stroke hemiparesis. Phys Ther 2004;84:1167–77. Ehrsson HH, Geyer S, Naito E. Imagery of voluntary movement of fingers, toes, and tongue activates corresponding body-part-specific motor representations. J Neurophysiol 2003;90:3304–16. Facchini S, Muellbacher W, Battaglia F, Boroojerdi B, Hallett M. Focal enhancement of motor cortex excitability during motor imagery: a transcranial magnetic stimulation study. Acta Neurol Scand 2002;105:146–51. Gandevia SC, Wilson LR, Inglis JT, Burke D. Mental rehearsal of motor tasks recruits alpha-motoneurones but fails to recruit human fusimotor neurones selectively. J Physiol 1997;505(Pt 1):259–66.
1856
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
Gerardin E, Sirigu A, Lehericy S, Poline JB, Gaymard B, Marsault C, et al. Partially overlapping neural networks for real and imagined hand movements. Cereb Cortex 2000;10:1093–104. Guillot A, Collet C. Contribution from neurophysiological and psychological methods to the study of motor imagery. Brain Res Brain Res Rev 2005;50:387–97. Hale BS, Raglin JS, Koceja DM. Effect of mental imagery of a motor task on the Hoffmann reflex. Behav Brain Res 2003;142:81–7. Hanakawa T, Immisch I, Toma K, Dimyan MA, Van Gelderen P, Hallett M. Functional properties of brain areas associated with motor execution and imagery. J Neurophysiol 2003;89:989–1002. Hashimoto R, Rothwell JC. Dynamic changes in corticospinal excitability during motor imagery. Exp Brain Res 1999;125:75–81. Johnson-Frey SH. Stimulation through simulation? Motor imagery and functional reorganization in hemiplegic stroke patients. Brain Cogn 2004;55:328–31. Kasai T, Kawai S, Kawanishi M, Yahagi S. Evidence for facilitation of motor evoked potentials (MEPs) induced by motor imagery. Brain Res 1997;744:147–50. Kiers L, Fernando B, Tomkins D. Facilitatory effect of thinking about movement on magnetic motor-evoked potentials. Electroencephalogr Clin Neurophysiol 1997;105:262–8. Liu KP, Chan CC, Lee TM, Hui-Chan CW. Mental imagery for promoting relearning for people after stroke: a randomized controlled trial. Arch Phys Med Rehabil 2004;85:1403–8. Marconi B, Koch G, Pecchioli C, Cavallari P, Caltagirone C. Breakdown of inhibitory effects induced by foot motor imagery on hand motor area in lower-limb amputees. Clin Neurophysiol 2007;118:2468–78.
Nair DG, Fuchs A, Burkart S, Steinberg FL, Kelso JA. Assessing recovery in middle cerebral artery stroke using functional MRI. Brain Inj 2005;19:1165–76. Porro CA, Francescato MP, Cettolo V, Diamond ME, Baraldi P, Zuiani C, et al. Primary motor and sensory cortex activation during motor performance and motor imagery: a functional magnetic resonance imaging study. J Neurosci 1996;16:7688–98. Rossi S, Pasqualetti P, Tecchio F, Pauri F, Rossini PM. Corticospinal excitability modulation during mental simulation of wrist movements in human subjects. Neurosci Lett 1998;243:147–51. Rossini PM, Rossi S, Pasqualetti P, Tecchio F. Corticospinal excitability modulation to hand muscles during movement imagery. Cereb Cortex 1999;9:161–7. Stephan KM, Fink GR, Passingham RE, Silbersweig D, CeballosBaumann AO, Frith CD, et al. Functional anatomy of the mental representation of upper extremity movements in healthy subjects. J Neurophysiol 1995;73:373–86. Stinear CM, Byblow WD. Motor imagery of phasic thumb abduction temporally and spatially modulates corticospinal excitability. Clin Neurophysiol 2003;114:909–14. Stinear CM, Byblow WD, Steyvers M, Levin O, Swinnen SP. Kinesthetic, but not visual, motor imagery modulates corticomotor excitability. Exp Brain Res 2006;168:157–64. Yahagi S, Kasai T. Facilitation of motor evoked potentials (MEPs) in first dorsal interosseous (FDI) muscle is dependent on different motor images. Electroencephalogr Clin Neurophysiol 1998;109:409–17. Yahagi S, Shimura K, Kasai T. An increase in cortical excitability with no change in spinal excitability during motor imagery. Percept Mot Skills 1996;83:288–90.
Kinesthetic motor imagery and spinal excitability: The effect of contraction intensity and spatial localization Patrick M. Cowley*, Brian C. Clark1, Lori L. Ploutz-Snyder Musculoskeletal Research Laboratory, Department of Exercise Science at Syracuse University, 201 Women’s Building, 820 Comstock Ave, Syracuse, NY 13244, USA Accepted 4 April 2008 Available online 16 May 2008
Abstract Objective: Data on whether motor imagery (MI) modulates spinal excitability are equivocal. The purpose of this study was to determine if imagined muscle contractions of the left plantar flexor (PF) alter spinal excitability, and if so, to determine whether this alteration is intensity dependent and/or localized to the target muscles. Our research questions required two experiments. Methods: In experiment 1, 16 healthy volunteers performed imagined muscle contractions using a kinesthetic approach with their left PF at 25% and 100% of imagined effort (IE). The soleus H-reflex was evoked during three conditions, which were separated by about 15 s: rest (preceding MI), during MI, and recovery (following the cessation of MI). In experiment 2, a subset of subjects from experiment 1 performed MI with their left PF at 100% of IE, while either the soleus or flexor carpi radialis (FCR) H-reflex was measured. Results: In experiment 1, we observed a facilitation of soleus H-wave amplitude during MI compared to the rest and recovery conditions (p < 0.05). Furthermore, the soleus H-wave amplitude was greater during 100% than 25% of IE (p < 0.05). In experiment 2, soleus and FCR H-wave amplitude increased during imagined muscle contractions of the left PF (p < 0.05). These changes were independent of voluntary muscle activity. Conclusions: These findings suggest MI can increase spinal excitability by the intensity of imagined effort, but this effect is not fully localized to the task specific muscle. Significance: These data provide evidence that MI can increase spinal excitability in healthy subjects, which suggests future studies are warranted to examine the clinical relevance of this effect. These studies are needed to help establish a therapeutic theory by which to advance motor function rehabilitation using MI. Ó 2008 International Federation of Clinical Neurophysiology All rights reserved. Keywords: Motor imagery; H-Reflex; Spinal excitability
1. Introduction Motor imagery (MI) training has recently intrigued researchers and clinicians for its potential role in mitigating the effects of motor deficits and strength loss following cerebrovascular trauma and periods of prolonged disuse (Butler and Page, 2006; Cicinelli et al., 2006; Clark et al., *
Corresponding author. Tel.: +1 315 443 1411; fax: +1 315 443 9375. E-mail address: [email protected] (P.M. Cowley). 1 Present address: Department of Biomedical Sciences, 211 Irvine Hall, Ohio University, Athens, OH 45701, USA.
2006; Dickstein et al., 2004; Johnson-Frey, 2004; Liu et al., 2004; Nair et al., 2005). It has been proposed that the degree of imagined effort should correspond to the effort during actual performance, but at a lower magnitude (Decety et al., 1993, 1991; Guillot and Collet, 2005). Evidence from functional imaging studies support this viewpoint, as MI and motor task performance have been shown to activate similar areas of the brain (Ehrsson et al., 2003; Gerardin et al., 2000; Hanakawa et al., 2003; Porro et al., 1996; Stephan et al., 1995); the change in functional MRI signal intensity during MI is 30% of that during actual task performance (Porro et al., 1996). MI also
1388-2457/$34.00 Ó 2008 International Federation of Clinical Neurophysiology All rights reserved. doi:10.1016/j.clinph.2008.04.004
1850
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
acutely increases cortical excitability as measured via transcranial magnetic stimulation (Facchini et al., 2002; Hashimoto and Rothwell, 1999; Rossi et al., 1998; Stinear and Byblow, 2003; Stinear et al., 2006; Yahagi and Kasai, 1998; Yahagi et al., 1996). Collectively, these findings suggest MI results in activation of the motor cortex; however, data on whether MI modulates spinal excitability have been equivocal. While some authors have reported an increase in spinal excitability during MI (Bonnet et al., 1997; Gandevia et al., 1997; Hale et al., 2003; Kiers et al., 1997; Rossini et al., 1999), others have not (Facchini et al., 2002; Hashimoto and Rothwell, 1999; Kasai et al., 1997; Stinear and Byblow, 2003; Stinear et al., 2006; Yahagi et al., 1996). For example, using the H-reflex technique Kiers et al. and Hale et al. showed MI increases spinal excitability, but this effect was limited to 50% of subjects and was independent of imagined effort (Hale et al., 2003; Kiers et al., 1997). These findings suggest MI is not only a supraspinal process, but also facilitates spinal excitability. When data from the literature are synthesized it appears that certain methodological differences may account for the lack of consistent findings on the effect of MI on spinal excitability. For example, Hale et al. reported MI increases spinal excitability in the absence of voluntary muscle activity when employing a high-intensity MI task (Hale et al., 2003), whereas many other studies using a low-intensity, fine MI task frequently report no effect on spinal excitability (Bonnet et al., 1997; Hashimoto and Rothwell, 1999; Kasai et al., 1997; Stinear and Byblow, 2003). The primary aim of our study was to determine if MI contraction intensity effects spinal excitability using the H-reflex technique. A secondary aim of this study was to determine if an acute alteration in excitability during MI is localized to the target muscle. Our research questions were addressed through two separate experiments. For our first experiment we examined whether soleus H-wave amplitude was altered during kinesthetic MI at 100% and 25% of imagined effort (IE) using the left plantar flexor (PF) muscle group. We hypothesized that changes in soleus H-wave amplitude would be facilitated during the imagined effort, and that this facilitation would be greatest during the higher imagined contraction intensity. The findings from our first experiment led us to explore if the effect of MI is localized to the target muscle. For this experiment we examined changes in soleus and flexor carpi radialis (FCR) H-wave amplitude during kinesthetically imagined maximal PF muscle contractions. Here, we hypothesized that changes in H-wave amplitude would only be observed in the soleus muscle.
MI trials each subject performed 3 voluntary contractions at 100% and 25% of maximum voluntary contraction (MVC) using the left PF, so they could establish a reference point to relate back to during the MI trials. The time between the voluntary contractions and MI trials was 10 min. A total of 20 trials of MI were performed at the two different intensities (10 per intensity performed in a randomized order), and the soleus H-reflex was elicited at rest, during MI, and at rest following MI (recovery). Fifteen seconds separated H-reflex measurements. For the MI trial each subject was given a verbal command to imagine contracting their left PF at 100% or 25% of IE, which lasted approximately 15 s, and between the 8th and 15th second the soleus H-reflex was elicited. In order to ensure there was no voluntary activation of the left PF during the MI trial the interference electromyogram (EMG) signal was quantified (root mean squared EMG) 50 ms immediately before the H-reflex measurement. Our findings from the first experiment indicated MI can acutely increase soleus H-wave amplitude, and that this was contraction intensity dependent with a slightly higher facilitation during the maximal MI task (see Section 3 for complete details). Therefore, to further investigate these findings we conducted a follow-up experiment to determine if the change in spinal excitability was localized to the soleus. To do this we invited a subset of the subjects from the first experiment back to the laboratory. Subjects in whom soleus H-wave amplitude increased by at least the mean increase (5.1%) were invited to participate in the follow-up experiment (Fig. 1). Of these seven subjects, six volunteered to participate in the second experiment. The decision to use the mean increase of 5.1% in H-wave amplitude as a cut-off point to include subjects in experiment 2 was based on our research question, which was to examine whether the increase in H-wave amplitude was localized to the target muscle. Thus requiring subjects who showed an increase in H-wave in response to MI.
2. Methods 2.1. General overview of experimental design and procedure For the first experiment we examined the effect of imagined PF muscle contractions at 100% and 25% of IE on soleus H-wave amplitude in sixteen subjects. Before the
Fig. 1. Percent change in soleus H20: Mmax from rest to motor imagery for each subject in experiment 1. The line represents the mean increase in soleus H20: Mmax for all subjects (mean increase of 5.1%). Subjects that had greater than a 5.1% increase in soleus H20: Mmax are represented by closed bars, and those that exhibited less than a 5.1% increase are represented by open bars. Subject’s 1, 3, 4, 5, 6, and 7 participated in experiment 2.
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
During this experiment we examined the effect of maximal imagined muscle contractions of the left PF on soleus H-wave amplitude, as well as on FCR (a forearm muscle) H-wave amplitude. The MI trials were preceded by three MVC’s using the left PF to establish a reference point for subjects to relate back to during the MI trials. The MVC’s and MI trials were separated by 10 min. Twelve trials of PF MI were performed with the soleus or FCR H-reflex being randomly assessed (6 per muscle performed in a randomized order) at rest before MI, during MI, and at rest after MI (recovery). Fewer MI trials were used in experiment 2 compared to experiment 1 because the results from experiment 1 indicated H-wave amplitude did not increase with the number of MI trials performed (analysis not shown) as has been previously shown (Hale et al., 2003). Fifteen seconds separated H-reflex measurements. In order to ensure there was no voluntary activation of the left PF or FCR before the H-reflex measurements the interference EMG activity was quantified (RMS EMG) for the 50 ms before the H-reflex measurements. For this experiment, the interference EMG activity of the left soleus is expressed relative to the interference EMG activity recorded during the actual MVC; referred to herein as normalized interference EMG.
1851
lowing the MVC trials in experiment 1 each subject performed three submaximal isometric contractions at 25% MVC. They were asked to match a target force displayed as a horizontal line on the computer monitor and maintain the contraction for 8 s. During the performance of these tasks subjects were asked to remember this feeling and use it during the MI trials. 2.4. Electrical recordings Electrical signals were recorded from both the soleus and FCR muscles using bipolar surface electrodes (Ag– AgCl, Soleus Electrodes: 36 mm diameter, Kendall MediTrace 200; FCR Electrodes: 15 mm diameter, Danlee Medical Products, Inc.). Before electrode placement, the subject’s skin was shaved and then abraded and cleaned with an alcohol pad to minimize skin impedance. Electrodes were placed longitudinally over the muscles with an interelectrode distance of 2 cm. For the soleus recordings a reference electrode was placed on the upper lateral aspect of the calf. For the FCR recordings a reference electrode was placed over the radial styloid. The EMG signals were amplified 500, band-pass filtered between 10 and 500 Hz, and sampled at 2500 Hz (MP150, BioPac Systems Inc., Goleta, CA).
2.2. Subjects 2.5. Electrical stimulation and spinal excitability Sixteen individuals participated in experiment 1 (mean ± standard deviation of 29 yrs ± 8, range of 20–48 yrs; 10 men and 6 women). Six individuals from experiment 1 (mean ± standard deviation of 27.3 yrs ± 4.3, range of 22–33 yrs; 4 men and 2 women) participated in experiment 2. The Institutional Review Board of Syracuse University approved the procedures used in this study, and all subjects gave written informed consent prior to participation. 2.3. Mechanical recordings and voluntary contraction trials Left PF force was assessed while the subject sat in a custom-modified PF dynamometer (Parabody 826, LifeFitness, Schiller Park, IL) equipped with a force transducer (TSD121C, BioPac Systems, Goleta, CA). The left leg of the subject was positioned so that the hip, knee, and ankle joints were at 90°, and a 43-cm computer monitor was positioned 1-m in front of the subject for visual feedback. Prior to the subjects performing the MI tasks, subjects performed voluntary PF contractions. For experiment 1, the subjects performed 3 voluntary contractions with their left PF at 100% and 25% of their MVC with 1–2 min between trials. For experiment 2, the three voluntary contractions were performed using the left PF at 100% MVC. The subjects did not perform a voluntary contraction using their left forearm flexors. For the MVC each subject was asked to gradually increase force and then perform a maximal effort for 2–3 s. The highest of the three was considered the MVC force. During the MVC’s the investigators provided strong verbal encouragement. Fol-
To measure spinal excitability we used the H-reflex technique. The H-reflexes were elicited using procedures previously outlined by our laboratory (Clark et al., 2006, 2007), which we have reported to have high reliability (intraclass correlation coefficient = 0.93) (Clark et al., 2007). In brief, subjects lay prone on an examination table with the ankle joint fixed at 45° of plantarflexion by a cylindrical cushion placed under the talus. The head was down in medial alignment with the torso, and rested in a face rest so that the neck was in a neutral position. The arms and hands were pronated, and placed along the side of the body. To elicit the soleus H-reflex a cathode (Ag–AgCl, 36 mm diameter; Kendal Medi-Trace 200) was placed over the tibial nerve in the popliteal fossa. The cathode site was determined using a hand-held stimulation probe. An anode was placed above the cathode on the posterior thigh (Ag–AgCl, 48 mm diameter; Kendal Medi-Trace 530). The electrical stimulus consisted of a 1 ms square pulse (Grass S88 stimulator coupled with a Grass SIU5 stimulus isolation unit, Astro-Med Inc., West Warwich, RI). The stimulus intensity was continuously increased to obtain a maximal Mwave (Mmax), and then adjusted and continuously monitored to evoke an M-wave equal to 20 ± 2.5% of the Mmax. The corresponding peak-to-peak amplitude of the H-wave was then calculated and expressed relative to the Mmax (H20: Mmax). To elicit the FCR H-reflex a cathode (Ag–AgCl, 15 mm diameter; Kendal Medi-Trace 200) was placed over the median nerve in the bicipital groove after the optimum site
1852
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
contracting, and not a visual approach where one visualizes oneself performing the task. The same investigator gave the instructions to each subject.
for stimulation was located by a hand-held stimulation probe. The anode was placed above the cathode (Ag–AgCl, 36 mm diameter; Kendal Medi-Trace 530). The electrical stimulus consisted of a 1 ms square pulse, and the intensity was continuously increased to obtain a maximal M-wave (Mmax). The stimulus intensity was continuously adjusted and monitored to evoke an M-wave equal to 5 ± 2.5% of the Mmax. The corresponding peak-to-peak amplitude of the H-wave was then calculated and expressed relative to the Mmax(H5: Mmax). To compare and interpret changes in evoked H-reflex responses during the experimental conditions, it was vital that the effective stimulus strength remain similar between the recording sessions, because the size of the H-reflex is heavily modulated by changes in stimulus intensity. As such it has been suggested that when repeated H-reflex measurements are performed, it is advantageous to use a stimulation intensity that produces similar M-wave responses to ensure that the same number of motor axons are recruited in each trial, which indicates that stimulus intensity to the efferent nerve is also kept constant (Aagaard et al., 2002). Based on this logic, we set limits on the evoked M-wave (M-wave equal to 20 ± 2.5% of the Mmax for soleus and 5 ± 2.5% of the Mmax for FCR) to make sure any effect observed was limited to motor imagery and not stimulus intensity. For the H-reflex recordings, if an MI trial was out of range (i.e., greater than or less than the established limits relative to the Mmax) it was not included in the analysis. A complete trial included an H-reflex measurement at rest, during MI, and recovery. Trials that were out of range were not repeated because it has been shown that spinal excitability is modulated by the number of trials performed (i.e., practice) (Hale et al., 2003). Data from one subject was excluded from the analysis for experiment 1 because we were unable to obtain a reasonable number of trials within the given range. For the subjects included in experiment 1, 85% of their trials were complete, and 81% of soleus trials and 92% of FCR trials for experiment 2 were complete.
For experiment 1, to examine the effect of MI on soleus H-wave amplitude a repeated measure analysis of variance was performed with condition (H-reflex at rest, during MI, and recovery) and intensity (100% and 25% of IE) as the within subject factors. To account for the possibility that voluntary muscle activity influenced H-wave amplitude during the MI condition we performed an additional analysis. Separate repeated measure analysis of covariance was performed for the 100% and 25% of IE trials with condition (H-reflex at rest, during MI, and recovery) as the within subject factor and the interference EMG (RMS EMG 50 ms immediately before the MI H-reflex measurement) for the respective intensity as a covariate. For experiment 2, to examine the effect of MI on soleus and FCR H-wave amplitude separate repeated measure analysis of variances was used with condition (H-reflex at rest, during MI, and recovery) as the within subject factors. We also examined the effect of MI on soleus normalized interference EMG and FCR interference EMG (RMS EMG for 50 ms immediately before the H-reflex measurements) by performing separate repeated measure analysis of variance with condition (EMG at rest before MI, during MI, and recovery) as the within subject factor. When a significant main effect or interaction was observed, a Sidak post hoc test was used to determine where the main effect or interaction was significant. The assumption of normality was examined using the Kolmogorov–Smimov Test of Normality and normal probability plots. Data are presented as means ± SEM, unless otherwise stated. The alpha was set at p 6 0.05, and the partial Eta squared effect size statistic (g2) and power of main effects and interactions are reported to further aid in data interpretation. The SPSS (version 13.0, Chicago, IL) statistical package was used for data analysis.
2.6. Kinesthetic motor imagery
3. Results
The verbal directions used during the MI trials are similar to those reported by our laboratory (Clark et al., 2006). Briefly, on a verbal signal to begin, the subject was instructed to ‘‘concentrate on your left calf muscles, and imagine your foot pushing at 100% (or 25%) of your maximum against the footpad. Feel yourself pushing up at 100% (or 25%) of your maximum with your left calf muscles. Keep pushing up at 100% (or 25%) of your maximum with your left calf muscles. Keep imagining you are pushing at 100% (or 25%) of your maximum, keep pushing, keep pushing. . .now stop.” The duration of each imagined contraction was 15 s and the subjects were asked to adopt a kinesthetic imagery approach, in which they were urged to create an internal representation and feeling of their muscle
3.1. Experiment 1
2.7. Statistical analysis
For experiment 1, we observed a significant main effect of condition, and intensity by condition interaction. First, soleus H-wave amplitude increased during MI at 100% and 25% of IE (F(2,28) = 11.1, p 6 0.05, g2 = 0.44, power = 0.93). Further, there was a significant intensity by condition interaction (F(2,28) = 4.1, p 6 0.05, g2 = 0.23, power = 0.56). Subsequent analysis revealed soleus H-wave amplitude was greater during MI at 100% (50.9 ± 3.4 H20: Mmax) than 25% of IE (49.9 ± 3.5 H20: Mmax) compared to the rest (F(1,14) = 5.7, p 6 0.05) and recovery conditions (F(1,14) = 4.8, p 6 0.05) (Fig. 2). An example of the change in soleus H-wave amplitude during MI compared to the
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
rest and recovery conditions is presented in Fig. 3. The interference EMG immediately before the soleus H-reflex measurement during the 100% and 25% of IE trials was 0.002 mV ± 0.00 and 0.002 mV ± 0.00, respectively (means ± standard deviation). The separate repeated measure analysis of covariance performed for the 100% and 25% IE trials using the interference EMG as a covariate showed that voluntary muscle activity did not affect H-wave amplitude (F(2,26) = 4.2, p 6 0.05 and F(2,26) = 3.5, p 6 0.05 for the 100% and 25% IE trials, respectively). Thus, H-wave amplitude increased during MI and was greater at 100% than 25% of IE, and this was not influenced by voluntary muscle activity, as our analysis was significant when we did and did not account for any potential changes in the interference EMG during MI.
1853
Fig. 3. An example of the soleus H-wave amplitude during the rest, motor imagery, and recovery conditions. The stimulus intensity was adjusted to elicit a motor response equal to 20 ± 2.5% of maximum, and the corresponding H-wave amplitude was normalized to the maximum motor response (H20: Mmax).
3.2. Experiment 2 For experiment 2, we observed a significant main effect of condition for the soleus muscle (F(2,10) = 13.0, p 6 0.05, g2 = 0.72, power=0.98) and the FCR muscle (F(2,10) = 8.1, p 6 0.05, g2 = 0.62, power = 0.88). Subsequent analysis revealed soleus H-wave amplitude was higher during MI than during the rest condition (55.3 ± 4.2 vs. 50.3 ± 5.2 H20: Mmax; p 6 0.05), but that there was no difference in soleus H-wave amplitude between the MI and recovery conditions (55.3 ± 4.2 vs. 52.1 ± 4.9 H20: Mmax; p > 0.05). As expected there was no difference in H-wave amplitude between the rest and recovery conditions (p > 0.05). Based on the fact that soleus H-wave amplitude increased during MI compared to rest, with no difference between MI and recovery, and rest and recovery conditions an additional analysis was performed to aid in the interpretation of these data. The values for the rest and recovery conditions were averaged and compared to the value obtained during MI using a
Fig. 2. Soleus H-wave amplitude during the rest, motor imagery (MI), and recovery conditions. Soleus H-wave amplitude increased during MI compared to the rest and recovery conditions, however it increased more during MI at 100% than 25% of imagined effort (p 6 0.05). *Significant increase in soleus H-wave amplitude during MI compared to the rest and recovery conditions. §Significantly greater increase in soleus H-wave amplitude during MI at 100% than 25% of imagined effort.
paired t-test; this analysis was significant (t(1,5) = 3.8, p 6 0.05; 55.3 ± 4.2 vs. 51.2 ± 5.1 H20: Mmax), indicating soleus H-wave amplitude increased during the MI condition compared to the non-MI conditions (Fig. 4). There was no difference in the normalized interference EMG immediately before the soleus H-reflex measurements (F(2,10) = 0.6, p > 0.05; 0.40 ± 0.12% vs. 0.44 ± 0.14% vs. 0.67 ± 0.29% for the rest, MI, and recovery conditions, respectively). Follow-up analysis revealed FCR H-wave amplitude was higher during MI than the recovery condition (22.3 ± 5.0 vs. 20.3 ± 4.7 H5: Mmax; p 6 0.05), but that there was no difference in FCR H-wave amplitude between the MI and rest conditions (22.3 ± 5.0 vs. 20.5 ± 4.7 H5: Mmax; p > 0.05). As expected there was no difference in H-wave amplitude between the rest and recovery conditions (p > 0.05). Based on the fact that FCR H-wave amplitude increased during MI compared to recovery, with no
Fig. 4. Soleus and flexor carpi radialis (FCR) H-wave amplitude during the non-motor imagery conditions (average of the rest and recovery) compared to motor imagery (MI). The amplitude of the soleus and FCR H-waves was expressed relative to Mmax (H20: Mmax for soleus and H5: Mmax for FCR). *Both soleus and FCR H-wave amplitude significantly increased during MI compared to the non-MI conditions.
1854
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
difference between rest and MI, and rest and recovery conditions an additional analysis was performed to aid in the interpretation of these data. The values for the rest and recovery conditions were averaged and compared to the value obtained during MI using a paired t-test; this analysis was significant (t(1,5) = 3.5, p 6 0.05; 22.3 ± 5.0 vs. 20.4 ± 4.8 H20: Mmax), indicating FCR H-wave amplitude increased during the MI condition compared to the nonMI conditions (Fig. 4). There was no difference in the interference EMG immediately before the FCR H-reflex measurements (F(2,10) = 2.1, p > 0.05; 0.0007 ± 0.00 mV vs. 0.0007 ± 0.00 mV vs. 0.0018 ± 0.00 mV for the rest, MI, and recovery conditions, respectively). 4. Discussion Our first experiment examined the effect of imagined PF muscle contractions at 100% and 25% of IE on soleus Hwave amplitude. We hypothesized that changes in soleus H-wave amplitude would occur during MI, and that these changes would be higher during MI at 100% than 25% of IE. Our findings are consistent with this hypothesis, as soleus H-wave amplitude increased during both MI tasks when compared to the rest and recovery conditions; increasing to a greater extent during the maximal MI task. The changes in H-wave amplitude during MI were independent of voluntary muscle activity, as when we covaried for the degree of interference EMG during the MI task no effect was observed. This observation is consistent with other studies (Hale et al., 2003; Kiers et al., 1997). In experiment 2 we examined the effect of imagined maximal PF muscle contractions on soleus and FCR Hwave amplitude. We hypothesized that changes in H-wave amplitude would increase in the soleus, but not in the FCR indicating a localized facilitation and not a systemic response. Soleus and FCR H-wave amplitude increased during MI when compared to the average of the non-MI conditions (rest and recovery). This is in contrast to our hypothesis, as we did not expect there to be any facilitation in the FCR. While it is somewhat difficult to determine the magnitude of each of these responses some insight can be gleaned from the reported effect sizes (partial eta-squared), which represents the proportion of total variation attributable to the factor, partialing out other factors from the total non-error variation. As such, the slightly larger effect size observed for the MI task in the soleus (g2 = 0.72) in comparison to the FCR (g2 = 0.62) suggests that the soleus H-wave amplitude increased slightly more than the FCR’s, and that the increase in spinal excitability during MI is only partially localized to the target muscle. It is important to interpret these data with caution, as the sample size was small (n = 6). However, these findings warrant additional investigations into this effect. In experiment 1 we observed a significant increase in Hwave amplitude during MI, which was primarily due to the large change in H-wave amplitude from 7 of the 15 subjects (illustrated in Fig. 1). A potential reason why some sub-
jects’ spinal excitability did not increase to as great an extent of those in Fig. 1 may be because PF force and the H-reflex measurements were obtained in different positions (sitting for PF force vs. lying prone for the H-reflex measurement). However, this observation is consistent with that of Kiers et al. (Kiers et al. (1997)), as only 4 of the 8 subjects in their study increased spinal excitability during MI. For the 6 subjects that completed both experimental paradigms, the change in soleus H20: Mmax during MI compared to the rest condition was 11% during experiment 1 and 10% during experiment 2. Collectively, these findings suggest MI is mainly a supraspinal process; however, it can also affect excitability at the level of the spinal cord. Modulation of H-wave amplitude can result from a number of factors. These factors can include: (1) presynaptic inhibition of Ia fibers, (2) variation in the amount of Ia neurotransmitter release, and (3) changes in excitability of the motorneuron due to changes in excitatory or inhibitory inputs (Clark et al., 2006). We cannot explain which of these factors effects H-wave amplitude during MI; thus, we will refer to these changes herein as changes in spinal excitability. 4.1. MI can increase soleus spinal excitability by the intensity of imagined effort It has previously been shown that MI shares many of the same neural substrates as actual performance (Ehrsson et al., 2003; Gerardin et al., 2000; Hanakawa et al., 2003; Porro et al., 1996; Stephan et al., 1995); therefore, it has been proposed that the degree of imagined effort should correspond to the effort during actual performance, but at a lower magnitude (Decety et al., 1991, 1993; Guillot and Collet, 2005). These data are consistent with this suggestion, as MI can acutely increase spinal excitability in an intensity dependent fashion, with a maximal imagined contraction being greater than an imagined contraction at an intensity equal to 25% of maximum. We are aware of only two studies that have examined the effect of imagined effort on soleus spinal excitability. Bonnet et al. assessed spinal excitability using the H- and tendon-reflex (T-reflex) techniques by having subjects imagine either ‘‘weak” (2% of IE) or ‘‘strong” (10% of IE) PF isometric contractions (Bonnet et al., 1997). They found the change in T-reflex amplitude during MI, but not H-reflex amplitude intensity dependent. We cannot compare our results to those of Bonnet et al. as interference EMG activity increased during their MI condition (which we did not observe). In addition, the contraction intensities used by Bonnet et al. during the MI condition were extremely low (2% vs. 10%). Thus, it is possible that subtle differences in spinal excitability during MI at very low contraction intensities cannot be discerned using the H-reflex technique. More recently, Hale et al. assessed soleus H-reflex excitability during MI by having subjects perform imagined PF contractions at 40%, 60%, 80%, and 100% of IE (Hale et al., 2003). They showed spinal excitability increased with
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
1855
the number of MI trials performed, but not by the intensity of imagined effort. Differences in the intensity of the imagined contraction may explain the differing results. We obtained our measurements at 25% and 100% of IE, while the lowest measured intensity by Hale and colleagues was 40%. Therefore, because we used a lower contraction intensity this may have allowed us to better delineate the effect of MI intensity on spinal excitability (25% vs. 100% compared to 40% vs. 100%). Our data suggest MI can modulate spinal excitability in an intensity dependent fashion, independent of voluntary muscle activity.
(measured using the H-reflex technique) in select subjects performing imagined MVC’s of their left PF. The clinical significance of MI training enhancing spinal excitability is uncertain at this time. However, these data provide a necessary first step in examining this effect in healthy subjects. Future studies examining this effect in clinical populations to determine any effect on clinical outcomes are warranted. These studies are needed to help establish a therapeutic theory by which to approach motor function rehabilitation following cerebrovascular trauma or during periods of prolonged disuse (e.g., bed rest, casting, spaceflight).
4.2. Changes in spinal excitability are not localized to the target muscle
References
Our data indicate MI increases soleus spinal excitability during imagined PF contractions, but based on our observation of a concurrent facilitation in the FCR spinal excitability it appears that the effect is not fully localized to the target (soleus) muscle. Based on the reported effect sizes (g2 = 0.72 for soleus, and g2 = 0.62 for FCR), however, it appears the larger facilitation occurred at the soleus. This is in contrast to our hypothesis, as we did not expect there to be any facilitation in the FCR. These findings are consistent with that of Marconi et al., who found imagined PF muscle contractions increased cortical excitability of the resting FCR in lower limb amputees and control subjects without amputation (Marconi et al., 2007). We extend the findings of Marconi et al. by showing that spinal excitability increases in the FCR during imagined PF contractions as well. This effect was first observed during voluntary muscle contractions of the foot, which caused reciprocal changes in the excitability of the resting wrist muscles (Baldissera et al., 2002; Borroni et al., 2004). Recent evidence indicates something of central origin modulates this effect (Borroni et al., 2004; Byblow et al., 2007; Cerri et al., 2003). Byblow et al. recently showed neural networks involving secondary motor areas, specifically the dorsal premotor cortex, and the primary motor cortex facilitate these reciprocal changes in excitability of the resting wrist muscles during voluntary contraction of the foot (Byblow et al., 2007). The clinical relevance of this finding remains to be determined. However, Byblow et al. suggested that voluntary movement of the leg may induce effects on upper limb primary motor cortex excitability, and this may be of considerable value given the importance of the dorsal premotor cortex during recovery from stroke (Byblow et al., 2007). 5. Conclusion We examined the effect of MI on spinal excitability. Our data indicate that MI can increase soleus spinal excitability in an intensity dependent fashion when performing a highintensity MI plantarflexion task in select subjects, but the entire effect is not fully localized to the soleus muscle. Collectively, these findings suggest kinesthetic MI is not solely a supraspinal process, but modulates spinal excitability
Aagaard P, Simonsen EB, Andersen JL, Magnusson P, Dyhre-Poulsen P. Neural adaptation to resistance training: changes in evoked V-wave and H-reflex responses. J Appl Physiol 2002;92:2309–18. Baldissera F, Borroni P, Cavallari P, Cerri G. Excitability changes in human corticospinal projections to forearm muscles during voluntary movement of ipsilateral foot. J Physiol 2002;539:903–11. Bonnet M, Decety J, Jeannerod M, Requin J. Mental simulation of an action modulates the excitability of spinal reflex pathways in man. Brain Res Cogn Brain Res 1997;5:221–8. Borroni P, Cerri G, Baldissera F. Excitability changes in resting forearm muscles during voluntary foot movements depend on hand position: a neural substrate for hand–foot isodirectional coupling. Brain Res 2004;1022:117–25. Butler AJ, Page SJ. Mental practice with motor imagery: evidence for motor recovery and cortical reorganization after stroke. Arch Phys Med Rehabil 2006;87:S2–S11. Byblow WD, Coxon JP, Stinear CM, Fleming MK, Williams G, Muller JF, et al. Functional connectivity between secondary and primary motor areas underlying hand–foot coordination. J Neurophysiol 2007;98:414–22. Cerri G, Borroni P, Baldissera F. Cyclic H-reflex modulation in resting forearm related to contractions of foot movers, not to foot movement. J Neurophysiol 2003;90:81–8. Cicinelli P, Marconi B, Zaccagnini M, Pasqualetti P, Filippi MM, Rossini PM. Imagery-induced cortical excitability changes in stroke: a transcranial magnetic stimulation study. Cereb Cortex 2006;16:247–53. Clark BC, Cook SB, Ploutz-Snyder LL. Reliability of techniques to assess human neuromuscular function in vivo. J Electromyogr Kinesiol 2007;17:90–101. Clark BC, Manini TM, Bolanowski SJ, Ploutz-Snyder LL. Adaptations in human neuromuscular function following prolonged unweighting: II. Neurological properties and motor imagery efficacy. J Appl Physiol 2006;101:264–72. Decety J, Jeannerod M, Durozard D, Baverel G. Central activation of autonomic effectors during mental simulation of motor actions in man. J Physiol 1993;461:549–63. Decety J, Jeannerod M, Germain M, Pastene J. Vegetative response during imagined movement is proportional to mental effort. Behav Brain Res 1991;42:1–5. Dickstein R, Dunsky A, Marcovitz E. Motor imagery for gait rehabilitation in post-stroke hemiparesis. Phys Ther 2004;84:1167–77. Ehrsson HH, Geyer S, Naito E. Imagery of voluntary movement of fingers, toes, and tongue activates corresponding body-part-specific motor representations. J Neurophysiol 2003;90:3304–16. Facchini S, Muellbacher W, Battaglia F, Boroojerdi B, Hallett M. Focal enhancement of motor cortex excitability during motor imagery: a transcranial magnetic stimulation study. Acta Neurol Scand 2002;105:146–51. Gandevia SC, Wilson LR, Inglis JT, Burke D. Mental rehearsal of motor tasks recruits alpha-motoneurones but fails to recruit human fusimotor neurones selectively. J Physiol 1997;505(Pt 1):259–66.
1856
P.M. Cowley et al. / Clinical Neurophysiology 119 (2008) 1849–1856
Gerardin E, Sirigu A, Lehericy S, Poline JB, Gaymard B, Marsault C, et al. Partially overlapping neural networks for real and imagined hand movements. Cereb Cortex 2000;10:1093–104. Guillot A, Collet C. Contribution from neurophysiological and psychological methods to the study of motor imagery. Brain Res Brain Res Rev 2005;50:387–97. Hale BS, Raglin JS, Koceja DM. Effect of mental imagery of a motor task on the Hoffmann reflex. Behav Brain Res 2003;142:81–7. Hanakawa T, Immisch I, Toma K, Dimyan MA, Van Gelderen P, Hallett M. Functional properties of brain areas associated with motor execution and imagery. J Neurophysiol 2003;89:989–1002. Hashimoto R, Rothwell JC. Dynamic changes in corticospinal excitability during motor imagery. Exp Brain Res 1999;125:75–81. Johnson-Frey SH. Stimulation through simulation? Motor imagery and functional reorganization in hemiplegic stroke patients. Brain Cogn 2004;55:328–31. Kasai T, Kawai S, Kawanishi M, Yahagi S. Evidence for facilitation of motor evoked potentials (MEPs) induced by motor imagery. Brain Res 1997;744:147–50. Kiers L, Fernando B, Tomkins D. Facilitatory effect of thinking about movement on magnetic motor-evoked potentials. Electroencephalogr Clin Neurophysiol 1997;105:262–8. Liu KP, Chan CC, Lee TM, Hui-Chan CW. Mental imagery for promoting relearning for people after stroke: a randomized controlled trial. Arch Phys Med Rehabil 2004;85:1403–8. Marconi B, Koch G, Pecchioli C, Cavallari P, Caltagirone C. Breakdown of inhibitory effects induced by foot motor imagery on hand motor area in lower-limb amputees. Clin Neurophysiol 2007;118:2468–78.
Nair DG, Fuchs A, Burkart S, Steinberg FL, Kelso JA. Assessing recovery in middle cerebral artery stroke using functional MRI. Brain Inj 2005;19:1165–76. Porro CA, Francescato MP, Cettolo V, Diamond ME, Baraldi P, Zuiani C, et al. Primary motor and sensory cortex activation during motor performance and motor imagery: a functional magnetic resonance imaging study. J Neurosci 1996;16:7688–98. Rossi S, Pasqualetti P, Tecchio F, Pauri F, Rossini PM. Corticospinal excitability modulation during mental simulation of wrist movements in human subjects. Neurosci Lett 1998;243:147–51. Rossini PM, Rossi S, Pasqualetti P, Tecchio F. Corticospinal excitability modulation to hand muscles during movement imagery. Cereb Cortex 1999;9:161–7. Stephan KM, Fink GR, Passingham RE, Silbersweig D, CeballosBaumann AO, Frith CD, et al. Functional anatomy of the mental representation of upper extremity movements in healthy subjects. J Neurophysiol 1995;73:373–86. Stinear CM, Byblow WD. Motor imagery of phasic thumb abduction temporally and spatially modulates corticospinal excitability. Clin Neurophysiol 2003;114:909–14. Stinear CM, Byblow WD, Steyvers M, Levin O, Swinnen SP. Kinesthetic, but not visual, motor imagery modulates corticomotor excitability. Exp Brain Res 2006;168:157–64. Yahagi S, Kasai T. Facilitation of motor evoked potentials (MEPs) in first dorsal interosseous (FDI) muscle is dependent on different motor images. Electroencephalogr Clin Neurophysiol 1998;109:409–17. Yahagi S, Shimura K, Kasai T. An increase in cortical excitability with no change in spinal excitability during motor imagery. Percept Mot Skills 1996;83:288–90.