Motor evoked potential depression following repetitive central motor initiation

Abstract

Prior reports have described a transient and focal decline in transcranial magnetic stimulation (TMS)-induced motor evoked potential (MEP) amplitude following fatiguing motor tasks. However, the neurophysiological causes of this change in MEP amplitude are unknown. The aim of this study was to determine whether post-task depression of MEPs is associated with repetitive central motor initiation. We hypothesized that MEP depression is related to repeated central initiation of motor commands in task-related cortex independent of motor fatigue. Twenty healthy adults had MEPs measured from the dominant first dorsal interosseous (FDI) muscle before and after six different tasks: rest (no activity), contralateral fatiguing hand-grip, ipsilateral fatiguing hand-grip, contralateral finger tapping, ipsilateral finger tapping, and imagined hand-grip (motor imagery). Changes in MEPs from baseline were assessed for each task immediately following the task and at 2-min intervals until MEPs returned to a stable baseline. Measures of subjective effort and FDI maximum voluntary contractions (MVC) were also recorded following each task. A statistically significant drop in MEP amplitude was noted only with contralateral finger tapping and imagined grip. Changes in MEP amplitude did not correlate with subjective fatigue or effort. There was no significant change in FDI MVCs following hand-grip or finger-tapping tasks. This study extends our knowledge of the observed decline in MEP amplitude following certain tasks. Our results suggest that central initiation of motor programs may induce a change in MEP amplitude, even in the absence of objective fatigue.

Introduction

In 1993, Brasil-Neto and colleagues reported the first use of transcranial magnetic stimulation (TMS) to investigate central fatigue (Brasil-Neto et al. 1993). In this paper, the authors reported a transient depression of motor evoked potentials (MEPs) elicited at 0.3 Hz beginning 30 s following the completion of a fatiguing task and lasting several minutes. The authors hypothesized that the drop in MEP amplitude reflected a depletion of neurotransmitters specifically with repetitive 0.3-Hz firing of motor neurons. Although the role of neurotransmitter depletion and need for 0.3-Hz stimulation has not been validated (Taylor and Gandevia 2001), this general paradigm of studying post-task MEP depression has been widely replicated and used in research on central fatigue (Gandevia 2001). Numerous investigators have replicated and extended this measure of central fatigue, including use in a variety of muscles (Takahashi et al. 2011), task types (Taylor and Gandevia 2008), and patient populations (Lou et al. 2003; Liepert et al. 2005). This finding has been shown to be specific to the muscle fatigued by a task (Zanette et al. 1995) and at least in part reflects the duration of the fatiguing task (Samii et al. 1997).

Although this post-task depression of MEP has been purported to measure central fatigue, the significance of post-task MEP depression remains open to debate. Samii et al. (1997) found MEP depression to be related to task duration, with a minimal task duration of 90 s needed to see MEP depression. Of note, MEP depression was not seen with shorter task durations, even if objective fatigue had already occurred. Perhaps most puzzling is that the time course of the observed MEP depression does not correlate with changes in force output. In fact, at the time of maximum performance fatigue (immediately following task failure), the MEP is frequently facilitated (Samii et al. 1996). The MEP also may remain depressed well beyond the point of recovery of maximal voluntary contraction (MVC) force (Avanzino et al. 2011). While several authors have claimed that MEP depression may be a corollary for feelings of effort or subjective fatigue, this has also yet to be convincingly demonstrated. Regarding fatigue in pathological conditions, Liepert et al. (2005) found a correlation between duration of MEP depression and subjective fatigue ratings in multiple sclerosis patients, whereas Lou et al. (2003) found no such correlation in patients with Parkinson’s disease.

If the finding of MEP depression is reflective of central motor changes, then it might be expected to be modulated by tasks known to effect central motor cortex excitability. Triggs et al. (1998) have previously shown that repetitive tasks may be more likely to facilitate central excitability than sustained contractions. They hypothesized that primary motor cortex may be primarily important in initiating motor movements, while other motor pattern generators may be more involved in the subsequent maintenance of force. Similarly, numerous investigators have demonstrated that motor imagery may facilitate MEPs over task-relevant motor cortex without significant changes at spinal or peripheral levels (Hashimoto and Rothwell 1999).

The aims of this study were thus to investigate whether tasks involving repeated or continuous central motor initiation might induce MEP depression, even in the absence of performance fatigue. We further sought to determine whether either sense of effort or subjective fatigue would correlate with these MEP changes in healthy subjects.

Methods

Subjects

Twenty healthy subjects (mean age 45.4 ± 14 years, 11 women) participated in these experiments. All subjects gave informed written consent in accordance with the Colorado Multiple Institutional Review Board. All subjects were strongly right-handed per the Edinburgh Handedness Inventory and free of significant neurological or psychiatric disease. All testing sessions were performed at 9 a.m., and subjects were asked to refrain from drinking caffeine on the morning of testing.

Behavioral tasks

Subjects were asked to perform six separate tasks as described below.

1. No activity: Subjects rested with eyes open for a period of 2 min.

2. Contralateral sustained hand-grip: Prior to the start of this task, subjects were asked to give three maximum voluntary contractions (MVC) with their dominant hand using a hydraulic grip dynamometer (Lafayette Instrument Company, Lafayette, IN). This dynamometer was modified to allow for connection to a force transducer (Measurement Specialties Inc., Hampton, VA) and display of force information on a laptop computer to give subjects visual feedback in terms of percent MVC of their performance. Subjects were then asked to maintain a force above 75% MVC for as long as possible and then an MVC thereafter until they were no longer able to maintain a force of 40% MVC or greater with verbal encouragement.

3. Ipsilateral sustained hand-grip: This task was identical to the contralateral sustained hand-grip except that subjects used their non-dominant hand.

4. Contralateral repetitive finger tapping: Subjects were fitted with metallic tape over their dominant thumb and index finger, and each digit was connected to wires coming from a custom device to count the number of taps (contacts). These data were presented to subjects every 5 s as visual feedback of their performance. After the first minute, subjects were asked to maintain finger tapping as fast as possible for either a 10-min period or until they were no longer able to maintain a pace greater than 40% of their best 5-s period for 10 s.

5. Ipsilateral repetitive finger tapping: This was identical to the contralateral repetitive finger-tapping task except that subjects used their non-dominant hand.

6. Imagined contralateral sustained hand-grip: Subjects were asked to imagine squeezing something with their dominant hand as hard as possible for a 2-min period. Subjects were encouraged to use mental imagery, including imagining what they were squeezing and what it would feel like. Subjects were asked to not activate their hand and EMG was used to monitor for covert muscle activation.

Subjects always performed the no activity task first and the imagined grip task last. The remaining four tasks were performed in a pseudorandomized fashion balanced across subjects. Subjects rested between tasks until their MEP returned to baseline and they felt subjectively rested or a minimum of 12 min had elapsed.

Subjective measures of effort and fatigue

Immediately following the completion of each task, subjects were asked to rate their effort on the task using Borg’s Rating of Perceived Exertion (RPE) scale (Borg 1998). Subjects were also asked to separately rate their overall fatigue and local muscular hand fatigue using a 1–10 visual analogue scale.

TMS and M-wave protocol

All TMS stimulation was performed over the scalp vertex in the optimal coil orientation for right FDI activation using a 2-tesla 9-cm-diameter circular coil and Magstim 200 magnetic stimulator unit (Magstim Company, Dyfed, UK). M-waves were elicited using a single supramaximal pulse from a Nicolet bipolar stimulator probe Electromyograph (Nicolet Biomedical, Madison, WI) oriented orthodromically over the ulnar nerve just proximal to the wrist. Surface electromyography (EMG) was recorded using disposable electrodes (Nicolet) overlying the first dorsal interosseous (FDI) muscle of the right hand. EMG signals were band-pass filtered (50 Hz–2 kHz), digitized at a rate of 5 kHz, and stored for off-line analysis using a Viking II Electromyograph (Nicolet).

Resting motor threshold (MT) was defined as the minimal stimulator output required to elicit four of eight motor evoked potentials (MEPs) of at least 50 microV. For each behavioral condition described below, we elicited eight MEPs at 0.3 Hz and 120% MT as well as three M-waves also at 0.3 Hz prior to each task. We repeated this cycle of MEP and M-wave collection immediately following each task and every 2 min thereafter until the subject’s MEP returned to within 20% of their baseline value on two consecutive trials or a minimum of 12 min. If subjects took longer than 10 min to return to baseline, MT testing was repeated to ensure that changes in MT or subject positioning were not responsible for this change. MEPs and M-waves for each task were normalized to a baseline value measured just before the commencement of each task. To control for potential peripheral changes on MEP amplitude(Vagg et al. 1998), we also analyzed the changes in MEP normalized to the M-wave at each time point(Kalmar and Carafelli 2004).

Test of muscle fatigue in first dorsal interosseous

In separate independent samples of 10 subjects each, we sought to determine whether our sustained grip task or finger-tapping task induced measurable fatigue in the dominant FDI muscle. Subjects had FDI abduction force measured by isometric lateral abduction of their index finger against a fixed Jamar dynamometer Sammons Preston Patterson Medical Division, Bolingbrook, IL), with their hand and arm otherwise fixed using straps and plates. Subjects performed 3 MVCs with visual feedback and then performed either a 2-min maximal hand-grip (30.9 ± 8 years, eight women) or 10-min finger-tapping task (29.1 ± 7 years, six women) with visual and auditory feedback to maintain maximal performance followed immediately by an additional three FDI MVCs.

Statistical analysis

Statistical analysis was performed using SAS version 9.2 (SAS Inc., Cary, NC). All data were pre-processed for completeness, outliers, and distributional forms. Wilcoxon rank-sum test was used for comparison of sample means, and Spearman’s r was used to test bivariate correlations between continuous variables. These tests were chosen given their robustness in the face of relatively small sample sizes and non-parametric data. We grouped our analyses to answer three primary experimental questions. First, do continuous contractions result in MEP decrements in the absence of motor fatigue (comparisons: no activity, contralateral grip, and ipsilateral grip)? Second, do repetitive contractions result in decrements in MEP in the absence of motor fatigue (comparisons: no activity, contralateral tapping, and ipsilateral tapping)? Third, does continuous motor imagery result in decrements in MEP (comparisons: no activity and motor imagery)? P values less than 0.05 were considered significant.

Results

Behavioral data

Subjects showed significant behavioral change (drop in force output) for both dominant and non-dominant continuous grips with a mean time to task failure of 116 (SD 46) and 110 (SD 30) s, respectively. In contrast, subjects showed minimal change in tapping speed (0.0 ± 24% and 3.7 ± 21% decrease in dominant and non-dominant hands, respectively) and all subjects completed the full 10 min without dropping below our prespecified performance endpoint of 40% maximal speed. Subjects reported high effort and fatigue with all tasks except the baseline task. See Table 1 for a summary of these results.

Table 1.

Summary of Behavioral outcomes by task [mean (SD)]

Task	Mean performance change as % of maximal performance	Duration of task in duration of task	Borg’s rating of perceived exertion (6–20 scale)	Rating of overall general fatigue (1–10 VAS)	Rating of hand focal hand fatigue (1–10 VAS)
No activity	N/A	2 min	6.38 (0.74)	1.29 (0.85)	1.19 (0.54)
Dominant sustained grip	40%	116.6 (46) s	16.86 (2.26)	3.74 (2.47)	6.56 (2.45)
Non-dominant sustained grip	40%	110.1 (30) s	16.64 (2.46)	3.44 (1.78)	6.18 (2.01)
Dominant finger tapping	0.0 (24)	10 min	15.09 (3.53)	3.65 (2.67)	5.26 (2.76)
Non-dominant finger tapping	−3.7 (21)	10 min	15.52 (3.39)	3.76 (2.80)	5.73 (2.56)
Imagined sustained grip	N/A	2 min	13.43 (3.77)	3.52 (2.45)	3.78 (2.55)

MEP and M-wave data

To confirm FDI involvement, EMG activity was seen over FDI muscle for both grip and tapping tasks in all patients. Table 2 displays MEP and M-wave changes for all tasks, and Fig. 1 illustrates these changes graphically. We did not find significant facilitation immediately following any of the tasks. Regarding the question of whether continuous contractions cause MEP decrements in the absence of fatigue, significant changes from baseline were not found for either ipsilateral or contralateral sustained grip tasks. Regarding the question of whether repetitive contractions can alter MEPs in the absence of fatigue, significant decreases 2 min following task completion were seen only for MEPs in the contralateral hand-tapping (24 ± 36%, P = 0.01) but not for the ipsilateral hand-tapping tasks. Regarding the question of whether motor imagery can alter MEPs, there was a marginally significant decrease in MEPs 2 min following the imagined grip task (16 ± 28%, P = 0.03) but not for the no activity task. There were no significant changes in M-wave amplitude for any of the tasks. By 4 min, all MEPs and M-wave amplitudes were not significantly different from baseline. For MEP changes normalized to the M-wave, dominant hand-tapping (30 ± 36%, P = 0.0046) and imagined grip tasks (20 ± 33, P = 0.04) again were the only tasks that had significant decrements at 2 min. MEP changes were not associated with RPE, general fatigue, or hand fatigue for either the dominant tapping or imagined grip task.

Table 2.

Motor evoked potential and m-wave data

Task	Percent change in MEP immediately following task	Percent change in MEP 2 min following task	Percent change in M-wave immediately following task	Percent change in M-wave 2 min following task
No activity	0.03 (25)	0.04 (31)	−0.0 (3.2)	−2.1 (5.2)
Dominant sustained grip	5.8 (63)	−5.7 (51)	3.7 (14)	3.0 (15)
Non-dominant sustained grip	32.6 (63)	13.6 (60)	2.8 (5.7)	1.4 (11)
Dominant finger tapping	−8.1 (51)	−24.3 (36)*	6.0 (14)	4.9 (9.4)
Non-dominant finger tapping	0.3 (41)	0.0 (38)	−0.5 (7.2)	0.04 (7.7)
Imagined sustained grip	−1.9 (58)	−16.1 (28)*	−3.9 (11)	1.9 (16)

^*Significant change from baseline with P value <0.05 using Wilcoxon signed rank test

Fig. 1.

MEP and M-wave data normalized to pre-task values for all tasks and all time points. Data points and standard errors presented are from pre-task, immediately post-task (time 0), and every 2 min thereafter for the contralateral grip (a), ipsilateral grip (b), contralateral tapping (c), ipsilateral tapping (d), motor imagery (e), and no activity (f) conditions

Effect of tasks on first dorsal interosseous force

Neither the 2-min maximal hand-grip (14.6 ± 34% increase in MVC, P = 0.28) nor 10-min maximal finger-tapping task (28.6 ± 42% increase in MVC, P = 0.05) induced any significant reduction in FDI MVC.

Discussion

Although MEP depression following task completion has been used as a measure of central fatigue, we here demonstrate that MEP depression may be found after two tasks without obvious motor fatigue in the muscle measured. We interpret these results as evidence that repeated central initiation of movement (e.g., tapping) or sustained central activation (e.g., motor imagery) may be associated with subsequent MEP depression, even in the absence of fatigue. These changes appear to be specific to our task and muscle of interest in that neither ipsilateral tapping nor fatiguing grip tasks (presumably involving similar changes in effort and arousal) induced similar MEP changes. We propose that factors other than central motor fatigue may contribute to the MEP depression seen after prolonged motor tasks.

Avanzino et al. (2011) recently published similar findings of MEP depression following a fatiguing skilled finger movement sequence. Although they reported some deterioration in the temporal and sequence accuracy using a sensor-engineered glove, they noted that there was no deterioration in force which they interpreted as evidence against peripheral fatigue. While they found objective behavioral deterioration, they also questioned the relation of MEP depression to performance in that the MEP depression did not correlate with the time course of behavioral recovery. Similar to our study, these investigators also found no change in peripheral (M-wave) or spinal (H-reflex) measures. This is consistent with a large body of literature, including intradural recordings, suggesting that post-task MEP decrements are of central origin (Di Lazzaro et al. 2003).

It is possible that if we had used more precise measures of finger kinematics, we may have noted changes in either temporal variability or movement amplitude, despite the lack of observed change in MVC or rate. Rodrigues et al. (2009) looked at behavioral and EMG correlates of a maximum voluntary rate (MVR) finger movement task and reported slowing over time as well as increased co-contraction of antagonist muscles on EMG, also without a loss of force or single movement speed. They concluded that MVR fatigue was thus due to central rather than peripheral factors. Interestingly, subjects in their task had a 73% decrease in rate over 20 s. This suggests that our subjects may have consciously or unconsciously paced their tapping knowing it could be a 10-min task. Alternatively, our task of tapping index finger and thumb together in an unrestrained fashion may have been more of an over learned movement and thus less fatiguing compared with the task used by Rodrigues and colleagues.

Prior research has shown that motor imagery can influence cortical activity and excitability similar to actual motor actions (Munzert et al. 2009). For example, Cesari et al. (2011) found that imagining grips of different shaped objects elicited muscle-specific changes in TMS excitability in a pattern similar to that expected with actual grip of these objects. What was perhaps unexpected about our results was that in our experiment motor imagery induced greater changes in cortical excitability for our muscle of interest than the equivalent motor task. We speculate that maintaining the motor image required more sustained involvement of central motor generators than the actual physical sustained grip task. In a different paradigm in which grip was used as a fatiguing task but motor imagery was performed after motor fatigue was established, Pitcher et al. (2005) found that motor imagery still increased MEP amplitude following the sustained fatiguing grip contraction. This suggests that cortical motor areas involved in imagery are independent, and possibly upstream, of those responsible for MEP amplitude decrements in a purely motor fatiguing task. We are not aware of any other studies that have used fMRI, TMS, or other modalities to compare the central activation of sustained versus repeated motor imagery or to compare motor imagery to sustained contractions. It is possible that as motor imagery is often less precise than actual movement, the FDI representation was more involved in the imagined task. It is unclear whether this motor imagery finding may be linked to a fatigue or habituation of mental imagery in that there is no direct way to measure the behavioral output. Interestingly, subjects reported subjective fatigue following the motor imagery task in general and in their hand intermediate to the values reported with no task and the sustained grip task.

It should be noted that the MEP amplitude decrements we observed were more variable, smaller, and of shorter duration than those typically described in protocols designed to elicit fatigue in the muscle tested (Gandevia 2001). This suggests that repeated central initiation is but one factor in these MEP changes. The degree of central activation may also be higher in fatiguing protocols, and thus, the difference may be more quantitative than qualitative. Further studies are needed to clarify these differences, including more detailed fatiguing and non-fatiguing tasks within the same motor group. More detailed TMS measures can also better elucidate neurophysiological changes. For example, Vucic et al. (2011) elegantly showed differential decrements and roles of short intracortical inhibition (SICI) induced at 1- and 3-ms interstimulus intervals during and following a fatiguing contraction and suggest that these changes in SICI may be important in the maintenance of force and cortical motor output.

In conclusion, we suggest that repeated initiation of central commands or sustained central activation can produce post-task decrements in cortical excitability without motor fatigue. Further studies are needed to determine the physiological origins and behavioral significance of these changes.