INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Intracortical inhibitory interneurons within the primary motor cortex (M1) inhibit the corticospinal neurons, found in layer V (Di Lazzaro et al. 1998). They form a horizontal network of connections within the cortex and seem to play an important role in shaping the output from M1 (Donoghue et al. 1996). Animal research has demonstrated that pharmacological blockade of intracortical inhibition with a GABA_A antagonist (bicuculline methobromide) degrades the spatial selectivity of the output from M1 (Jacobs and Donoghue 1991; Matsumura et al. 1991, 1992). This is thought to result from the disinhibition of horizontal intracortical connections between areas of M1, which promotes the recruitment of additional muscle or movement representations.

The role of intracortical inhibition in the functional coupling between areas of M1 has recently been explored by Schneider et al. (2002). These authors used intracortical microstimulation techniques to identify two motor cortical points within M1 that activated separate muscles in anesthetized cats. They found that when a GABA_A antagonist (bicuculline methobromide) was administered to one motor cortical point, stimulation of the other motor cortical point produced combined activation of the muscles previously activated by stimulation of each site separately. This did not appear to be due to diffusion of the bicuculline methobromide or current spread from the stimulation site. Furthermore, depolarization of the corticospinal neurons at the site of drug administration was not sufficient to allow their recruitment by stimulation of the other motor cortical point. These authors conclude that disinhibition is a vital step in the functional coupling of motor cortical points and the formation of muscle synergies during natural movement (Schneider et al. 2002). This study also provides evidence of a critical role for intracortical inhibitory processes in the prevention of co-activation of separate motor cortical points.

In humans, paired-pulse transcranial magnetic stimulation (TMS) provides a noninvasive means of studying intracortical inhibitory function. When a suprathreshold magnetic test stimulus is preceded by a subthreshold magnetic conditioning stimulus, the resulting motor evoked potential (MEP) may be inhibited or facilitated, depending on the interstimulus interval (ISI) (Kujirai et al. 1993). Generally, short ISIs (1-5 ms) produce inhibition of the test MEP, while longer ISIs (10-15 ms) produce facilitation of the test MEP (Kujirai et al. 1993; Chen et al. 1998). The conditioning stimulus is thought to excite GABA-ergic intracortical inhibitory interneurons, which inhibit corticospinal cells with a latency of between 1 and 5 ms (Ziemann et al. 1996a,b). Other authors have shown that even minimal levels of voluntary activation of the target muscle significantly reduce the degree of inhibition produced by subthreshold conditioning stimuli delivered at ISIs between 1 and 6 ms (Hanajima et al. 1998; Ridding et al. 1995). This is probably due to reduced excitability of the inhibitory interneurons that project to the corticospinal neurons responsible for activation of the target muscle (Ridding et al. 1995).

The potential role of intracortical inhibition in the prevention of unwanted muscle activation has recently been explored using paired-pulse TMS. Liepert et al. (1998) measured the degree of intracortical inhibition (ICI) acting on the M1 representations of abductor pollicis brevis (APB) and fourth dorsal interosseous (4DIO) before and after performing repetitive thumb abductions for 5 min, paced at 1 Hz. Responses to single and paired TMS stimuli were recorded while the subjects were at rest and before and after each of four blocks of thumb abductions. Over the course of the blocks of training, they found a decrease in the degree of ICI acting on the M1 representations of both APB and 4DIO, which were both activated during task performance. However, when subjects were provided with EMG feedback from 4DIO and instructed to perform the thumb movements while keeping 4DIO relaxed, they found a significant increase in the ICI of responses recorded from this muscle. The authors conclude that ICI acting on the M1 representations of different hand muscles can be differentially modulated and volitionally increased (Liepert et al. 1998). However, it is worth noting that ICI was tested while subjects were at rest. The potentially differential modulation of ICI, and its contribution to selective muscle activation, during task performance was not examined.

A more recent study by Sohn et al. (2002) partially addressed this issue. These authors had subjects perform a precued Go/NoGo reaction time task, where they were asked to extend their index finger in response to the Go stimulus and keep their hand relaxed in response to the NoGo stimulus. Paired-pulse TMS was used to evaluate ICI of responses recorded from extensor indicis proprius (EIP) and abductor digiti minimi (ADM). Stimuli were delivered at rest and after the Go and NoGo stimuli. They observed a significant increase in the ICI of responses recorded from both muscles after the NoGo stimulus compared with rest. The authors interpreted this finding as evidence of a globalized increase in ICI during volitional inhibition of motor activity (Sohn et al. 2002). This suggests that ICI may be increased to prevent unwanted muscle activation and is consistent with the findings of Liepert et al. (1998). However, given that subjects were required to keep their whole hand at rest in response to the NoGo stimulus, it is perhaps unsurprising that the authors found a globalized, rather than muscle-specific, increase in ICI. The muscle specificity of the differential modulation of ICI that may occur during the performance of a focal motor task therefore remains unknown.

The aim of this study was to determine if the ICI acting on the M1 representations of intrinsic hand muscles was temporally modulated in a muscle-specific way during the performance of a phasic manual task. This was achieved by recording MEPs in response to single- and paired-pulse TMS while subjects were at rest and performing a phasic index finger flexion task paced at 1 Hz. Responses were recorded from first dorsal interosseous (FDI) as it is activated during index finger flexion, and APB as a control muscle. During task performance, stimuli were delivered during finger flexion (on phase) and between finger flexions (off phase). Two hypotheses tested are the following: 1) there is a decrease in the ICI acting on the M1 representation of FDI during task performance, and this decrease is greater during the on phase than the off phase of the task; and 2) there is an increase in the ICI acting on the M1 representation of APB during task performance, and this increase is greater during the on phase than the off phase of the task.

By testing these hypotheses, this study was designed to explore the potential role of differential modulation of ICI in the selective activation of hand muscles during the performance of a precise manual task.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Fifteen neurologically normal subjects participated in this experiment (6 women and 9 men; mean age, 28 yr; range, 20-56 yr). Using a handedness questionnaire (Oldfield 1971), 12 were deemed right-handed [mean laterality quotient, 85.7 ± 27.0% (SD)], and 3 were deemed left-handed (mean laterality quotient, -89.7 ± 8.5%). The Auckland Ethics Committee approved the procedure in accordance with the Declaration of Helsinki, and informed consent was obtained from all participants.

Preparation

Subjects were comfortably seated next to a table on which they rested their dominant forearm and hand in a pronated position. Their dominant hand rested on a computer mouse, with the middle and distal phalanges of their index finger resting on the mouse button. Their dominant thumb was supported by a foam block in a slightly abducted position, so that it did not contact the mouse. Their dominant wrist and forearm were supported by a large piece of foam so that the wrist joint was maintained in a neutral position. Subjects were instructed to keep their hand and forearm muscles as relaxed as possible and apply downward pressure to the mouse button using only their index finger.

Depression of the mouse button activated a sound transducer, which emitted a high-pitched tone while the mouse button switch was closed. Subjects practiced depressing the switch and creating a tone in time with an auditory metronome that was emitting 800 Hz tones of 200 ms duration at a rate of 1 Hz. The output from the mouse button was collected at a 200-Hz sampling rate with a National Instruments 16 bit A/D converter (PCI-MIO-16XE-50) and PC LabView program and stored to disk for subsequent analysis. Electromyography (EMG) data were collected from the FDI and APB muscles of the dominant hand via a pair of 12 mm diam surface Ag-AgCl Hydrospot electrodes (Physiometrix, N. Billerica, MA), using standard techniques. Signals were amplified by two Grass P511AC EMG amplifiers (Grass Instrument Division, W. Warwick, RI). The EMG data were band-pass filtered at 30-1,000 Hz, sampled at 4 kHz with a 12-bit MacLab A/D acquisition system and software, and stored to disk for subsequent analysis.

This experiment was conducted in two separate sessions, separated by 24 h. In the first session, the TMS variables were optimized for stimulation of APB, and in the second session, the TMS variables were optimized for stimulation of FDI, or vice versa. The same procedure was followed for both sessions. A pair of MagStim 200 magnetic stimulators (maximum output intensity 2.0 T; MagStim, Dyfed, UK) connected by a BiStim unit was used to stimulate the motor cortex via a figure-of-eight coil (7 cm coil diam). The coil was held tangentially to the scalp and perpendicular to the central sulcus, so that the induced current flow was in a posterior to anterior direction. The optimal location for eliciting MEPs in the muscle of interest (APB or FDI) was determined by stimulating at sites over the contralateral motor cortex in a 1-cm grid pattern with the subject at rest. The location that produced the greatest peak-to-peak amplitudes in the muscle of interest was considered optimal. This was found to be 2-4 cm lateral to the vertex for all subjects.

Rest threshold for the muscle of interest was then established at the optimal locations by altering the stimulator output intensity initially in 5% and then in 1% increments. Rest threshold (RTh) was defined as the stimulator output intensity that produced MEPs with a peak-to-peak amplitude of 50 µV in four of eight consecutive trials. For FDI, active threshold (ATh) was also established by determining the stimulator output intensity that produced MEPs with a peak-to-peak amplitude of 100 µV in four of eight consecutive trials while the subject held the mouse button down with their index finger. Test stimulus intensity was set to 120% RTh. For both muscles, the conditioning stimulus intensity was initially set to 80% RTh. The interstimulus interval (ISI) was then altered in 0.1 ms steps to determine the ISI that produced maximal inhibition of the test response. This ISI was used for all subsequent paired-pulse trials, and was found to be between 2.3 and 2.8 ms for all subjects. For FDI, conditioning stimulus intensity was then set to 90% ATh. For APB, the conditioning stimulus intensity was then reduced to a level that produced approximately 30% inhibition of the test response (CS₃₀). This level was chosen to avoid the "floor effect" (Fisher et al. 2002) and allow the detection of any increases in the level of inhibition acting on the APB representation during task performance. The conditioning stimulus intensities for both FDI and APB most likely intersect the linear region of the stimulus-response curve for the intracortical inhibitory interneurons (Fisher et al. 2002), thus allowing the detection of changes in ICI that occur with changes in the slope of this linear region during task performance (Devanne et al. 1997).

Experimental protocol

Two types of trials were conducted: active task performance and rest. During active trials, an auditory metronome produced 800-Hz tones, of 200 ms duration, at a rate of 1 Hz. For active trials subjects were instructed to use their dominant index finger to depress the mouse button in time with the auditory metronome, while keeping the rest of their hand as relaxed as possible. Depressing the mouse button activated a sound transducer that provided auditory feedback to subjects. They were instructed to match the onset and offset of the tones they generated by depressing the mouse button with those of the auditory metronome. During rest trials, subjects were instructed to relax their dominant hand on the computer mouse. Single and paired TMS pulses were delivered during active and rest trials in blocks of eight stimuli, at a frequency of 0.2 Hz. During active trials, stimuli were delivered either 50 or 600 ms prior to the metronome beat. This meant that stimuli were delivered either during the FDI EMG burst (50 ms offset, on phase) or while the FDI was at rest between EMG bursts (600 ms offset, off phase; Fig. 1). Test stimulus intensity was initially maintained at 120% RTh during task performance, enabling the assessment of the modulation of corticospinal excitability by task performance. Test stimulus intensity was then adjusted to match nonconditioned MEP amplitude during the on and off phases with the nonconditioned MEP amplitude recorded at rest. Prior to data collection, subjects completed active practice trials, and their FDI EMG activation and timing checked. Blocks of eight TMS stimuli were delivered in a pseudorandomized order such that a total of 16 MEPs were collected in response to TMS for each condition.

View larger version (10K): [in this window] [in a new window]   Fig. 1. Example first dorsal interosseous (FDI) and abductor pollicis brevis (APB) EMG traces from a typical group 1 subject, during a practice trial where no stimulation was given. Key is the signal from the mouse button, with downward deflection representing depression of the button. Met is the trigger to the auditory metronome, with upward deflection representing the activation of the metronome. Stim is the trigger to the magnetic stimulator, which was turned off during this practice trial. Calibration bar: 0.5 mV and 200 ms.

Data analysis

The output from the mouse button was transformed and visually displayed so that the timing of mouse button depression with respect to the auditory metronome beep could be visually inspected. Trials were accepted if the onset of the mouse button depression occurred within a 200-ms window, extending from 150 ms before to 50 ms after the onset of the auditory metronome beep, and the release of the mouse button occurred within the 300 ms following the onset of the auditory metronome beep. If <60% of a subject's data were accepted on the basis of these criteria, their entire data set was discarded.

The trials recorded under each combination of experimental conditions for each subject and muscle were ranked in ascending order of peak-to-peak MEP amplitude. The top 20% and bottom 20% of the ranked trials were then discarded. A trimmed mean was then calculated from the remaining trials, which gives an accurate representation of centrality (Wilcox 2001). For the retained trials, the pretrigger level of EMG activity was determined by calculating the root mean square (r.m.s., mV) value for a 20-ms period ending 5 ms prior to the stimulus artifact.

For each subject, muscle and condition, the average nonconditioned MEP amplitude in response to the unadjusted test stimulus (120% RTh) was normalized to the maximum MEP amplitude recorded from that muscle. The degree of MEP facilitation produced by task performance was then calculated by subtracting the normalized MEP amplitude recorded at rest from the normalized MEP amplitudes recorded during the on and off phases of the task, in response to the unadjusted test stimulus (120% RTh).

A posteriori, subjects were divided into two groups, according to the relative degree of APB MEP facilitation during task performance. Group 1 was comprised of seven subjects whose APB MEP facilitation was greatest during the on phase of the task. Group 2 was made up of the remaining eight subjects whose APB MEP facilitation was greatest during the off phase of the task (Fig. 2). This division allowed any related differences in the patterns of ICI and pretrigger EMG modulation to be detected.

View larger version (17K): [in this window] [in a new window]   Fig. 2. Example APB EMG traces from a typical subject from each group. Two EMG traces per panel. A: group 1 subject. B: group 2 subject. Unmatched is the response to 120% RTh stimulus intensity. Nonconditioned is the response to the adjusted test stimulus intensity, to match MEP amplitude across task conditions. Conditioned is the response to the adjusted test stimulus intensity when it is preceded by CS30. Calibration bars (A) 1 mV, 25 ms; (B) 0.5 mV, 25 ms.

The degree of inhibition for each muscle under each condition was then calculated using the following formula
where NC = average matched nonconditioned MEP amplitude and C = average conditioned MEP amplitude. This results in a number that represents the degree of inhibition as a percentage, which at least partly compensates for interindividual variability in MEP amplitudes. The modulation of ICI produced by task performance was then calculated by subtracting the percentage ICI calculated at rest from the percentage ICI calculated during the on and off phases of the task.

The pretrigger EMG data were normalized using a log₁₀ transformation. The normality and homoscedasticity of the data were confirmed prior to statistical analysis (Bruning and Kintz 1987). For NC MEP amplitude and ICI data, a mixed repeated measures ANOVA was performed, with muscle (FDI, APB) and phase (on, off) as the within-group factors. For pretrigger EMG data, a mixed repeated measures ANOVA was performed for each muscle, with session (FDI session, APB session) and phase (rest, on, off) as the within-group factors. Planned contrasts to test for temporal modulation of NC MEP amplitude and ICI were then carried out, using paired t-tests.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject variables

There were no differences between the two groups of subjects in their age, handedness, or temporal accuracy of task performance (Table 1). The three left-handed subjects were in group 1. Both groups depressed the mouse button switch around 23 ms prior to the metronome beep (on average) and held the switch closed for around 190 ms. The percentage of trials discarded due to temporal inaccuracy was between 4% and 7% (on average) for both groups during both experimental sessions. The rest thresholds, active thresholds, and CS₃₀ intensities were comparable. The degree of ICI produced at rest by the 90% ATh conditioning stimulus for FDI and by the CS₃₀ conditioning stimulus for APB were comparable between groups. The nonconditioned test MEP amplitudes during task performance were adequately matched to those recorded at rest for each muscle (Table 1).

Corticospinal excitability

The modulation of corticospinal excitability by task performance was examined by comparing MEP amplitude facilitation during the on and off phases in response to the 120% RTh test stimulus. For APB, it was apparent that MEP amplitude facilitation was greatest during the on phase for seven subjects (group 1), and during the off phase for the remaining eight subjects (group 2; Fig. 3, A and B). As expected, FDI MEP amplitude facilitation was greater during the on phase than the off phase of the task (Fig. 3B).

View larger version (25K): [in this window] [in a new window]   Fig. 3. Modulation of corticospinal excitability by task performance. MEP amplitudes were first normalized to the maximum motor evoked potential (MEP) amplitude recorded from each muscle for each subject. The degree of facilitation was then calculated by subtracting the normalized MEP amplitude recorded at rest from the normalized MEP amplitudes recorded during the on and off phases of the task in response to the unadjusted test stimulus (120% RTh). A: individual data for facilitation of APB MEP amplitude during the on and off phases of task performance (n = 15). B: group data for facilitation of FDI and APB MEP amplitude. Error bars = SE. ***P < 0.001; **P < 0.01; *P < 0.05.

The subsequent mixed repeated measures ANOVA revealed main effects of muscle and phase. Mean MEP amplitude facilitation was significantly higher for FDI than APB (FDI 39.4%, APB 24.0%, F = 11.7, P < 0.01) and during the on phase than the off phase (on 39.1%, off 29.0%, F = 7.7, P < 0.05). A significant interaction between group, muscle, and phase was also found (F = 14.6, P < 0.01). This was due to the degree of APB MEP facilitation during the on phase being significantly higher for group 1 than group 2 (group 1 40.0%, group 2 16.5%, P < 0.001) and the degree of APB MEP facilitation during the off phase being significantly higher for group 2 than group 1 (group 2 41.6%, group 1 16.6%, P < 0.01; Fig. 3B). Planned contrasts revealed temporal modulation of the MEP amplitudes recorded from FDI in group 1 (on 54.9%, off 25.0%, P < 0.001) and group 2 (on 47.0%, off 30.8%, P = 0.057), and APB in group 1 (on 40.0%, off 16.6%, P < 0.001) and group 2 (on 41.6%, off 16.6%, P < 0.001).

ICI

The modulation of ICI by task performance was examined by comparing the change in the degree of inhibition from rest levels during the on and off phases of the task. The mixed repeated measures ANOVA revealed main effects of muscle and phase. The reduction in ICI during task performance was significantly greater for FDI than APB (FDI -35.3%, APB -7.9%, F = 6.3, P < 0.05), and during the on phase than the off phase (on -30.7%, off -12.5%, F = 10.5, P < 0.01). A significant interaction between group, muscle, and phase was also revealed (F = 5.2, P < 0.05). This was due to the modulation of inhibition of APB responses during the on phase being significantly different and in opposite directions between the groups (group 2 24.7%, group 1 -40.3%, P < 0.001; Fig. 4).

View larger version (11K): [in this window] [in a new window]   Fig. 4. Modulation of intracortical inhibition by task performance. The degree of intracortical inhibition (ICI) was calculated by subtracting the percentage ICI calculated at rest from the percentage ICI calculated during the on and off phases of the task. Positive values reflect an increase in ICI above that observed at rest, while negative values reflect a decrease in ICI below that observed at rest. Error bars = SE. ***P < 0.001; **P < 0.01; * P < 0.05.

Planned contrasts revealed temporal modulation of the inhibition of responses recorded from FDI in group 1 (on -54.9%, off -10.3%, P < 0.001) and group 2 (on -57.4%, off -18.8%, P < 0.01), and APB in group 2 (on 24.7%, off -4.5%, P < 0.05). There was no statistically significant temporal modulation of inhibition of responses recorded from APB in group 1 (P = 0.09).

Pretrigger EMG

For FDI, the mixed repeated measures ANOVA revealed a main effect of group, with higher mean levels of pretrigger EMG for group 1 than group 2 (group 1 0.010 mV, group 2 0.005 mV, F = 8.6, P < 0.05). There was also a main effect of phase (F = 70.4, P < 0.001). The mean pretrigger EMG during the on phase (0.015 mV) was found to be significantly greater than during the rest and off phases (rest 0.003 mV, P < 0.001; off 0.005 mV, P < 0.001). The mean pretrigger EMG during the off phase was also found to be significantly greater than during rest (P < 0.05). There was no effect of session and no significant interactions (Fig. 5).

View larger version (11K): [in this window] [in a new window]   Fig. 5. Modulation of pretrigger EMG activity by task performance. Error bars = SE. *P < 0.05.

For APB, there was a main effect of phase (F = 37.2, P < 0.001) and a significant interaction between phase and group (F = 6.8, P < 0.01). There was no effect of session. The mean pretrigger EMG during the on phase (0.005 mV) was found to be significantly greater than during the rest and off phases (rest 0.003 mV, P < 0.001; off 0.003 mV, P < 0.001). There was no significant difference between mean pretrigger EMG levels during the rest and off phases. The significant interaction was found to arise from the mean pretrigger EMG during the on phase being greater in group 1 than in group 2 (group 1 0.007 mV, Group 2 0.003 mV, P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The findings from this study suggest that the level of intracortical inhibition acting on the M1 representation of a muscle not involved in a precise phasic task can be increased above rest levels during the on phase activation of a muscle that is involved in task performance. This appears to decrease the excitability of the corticospinal projection to the uninvolved muscle and prevent its unintended activation during task performance. This is a novel finding that extends the work of others (Liepert et al. 1998; Sohn et al. 2002) by demonstrating muscle-specific temporal modulation of intracortical inhibition that appears to play a role in preventing unintended muscle activation during the performance of a phasic task. This study also lends support to the hypothesis that intracortical inhibitory processes play an important role in shaping the temporal and spatial characteristics of the output from M1.

As expected, the excitability of the corticospinal projection to FDI was significantly increased during task performance compared with rest. This was associated with a significant decrease in the level of ICI acting on the FDI representation and a significant increase in the level of FDI activation. These changes were temporally modulated, such that corticospinal excitability and FDI activation were significantly increased and ICI significantly decreased during the on phase of the task compared with the off phase. These findings are in agreement with those from previous studies (e.g., Devanne et al. 1997; Flament et al. 1993; Ridding et al. 1995) and support the hypothesis that there is a temporal pattern of ICI and corticospinal modulation during phasic muscle activation.

For APB, the group as a whole demonstrated no clear temporal modulation of corticospinal excitability. This was because some subjects had greater MEP facilitation during the on phase, while others had greater MEP facilitation during the off phase (Fig. 3A). Subsequent analysis demonstrated that group 1 displayed the same pattern of temporal modulation of corticospinal excitability, ICI, and muscle activation for both FDI and APB during task performance. The temporal modulation of MEP amplitude facilitation was significant for both FDI and APB, as was the temporal modulation of disinhibition for FDI. This was associated with the level of pretrigger EMG being significantly higher during the on and off phases compared with rest and during the on phase compared with the off phase for both FDI and APB. However, there was no significant difference in the degree of disinhibition of APB during the on and off phases.

In contrast, group 2 displayed differential patterns of temporal modulation of corticospinal excitability and ICI for FDI and APB during task performance (Figs. 3B and 4). FDI MEP amplitude facilitation appeared greater during the on phase than the off phase of the task, as for group 1, although this effect did not reach the conventional level of significance (P = 0.057). This was associated with a reduction in ICI and increase in pretrigger EMG that was temporally modulated and statistically significant, as for group 1. The overall level of FDI activation during task performance was lower for than for group 1 (P < 0.05). However, for APB, there was less MEP amplitude facilitation during the on phase than the off phase, and this was associated with an increase in ICI during the on phase. This represents a differential modulation of corticospinal excitability and ICI for these muscles in group 2. During the on phase of the task, the FDI representation was released from inhibition and its excitability increased, while the APB representation was more inhibited and its excitability decreased. The increase in ICI acting on the APB representation may have helped to prevent the unwanted activation of this muscle, enhancing the selectivity of muscle activation during task performance. This is evidenced by the significantly lower level of APB activation during the on phase of the task observed in group 2 compared with group 1 (Fig. 5).

One of the methodological limitations of paired-pulse TMS studies is that the MEP responses are influenced by changes in the slopes of the sigmoidal stimulus-response curves for both the corticospinal and inhibitory interneurons (Capaday 1997; Devanne et al. 1997). In this experiment, the slopes of the stimulus-response curves for these types of neuron may have been altered by task phase, which hampers the comparison of the effects of conditioning under different task conditions (Capaday 1997). This cannot be compensated for by matching the amplitudes of the nonconditioned MEPs (Capaday 1997). Despite this, the differential modulation of nonconditioned APB MEP amplitude in response to the unmatched test stimulus demonstrates differential changes in the corticospinal excitability for the two groups of subjects. The decrease in APB corticospinal excitability during the on phase observed in group 2 subjects is likely to have occurred due to an increase in inhibition acting on its M1 representation.

It is apparent that subjects in group 2 were able to perform the task using lower overall levels of FDI activation. This may be because other finger flexors (such as flexor digitorum profundus) made a greater relative contribution to task performance. However, this remains unknown as the relative contribution to task performance by other finger flexors was not assessed. The higher levels of FDI activation by subjects in group 1 is associated with disinhibition of the APB representation and its unintended activation. In contrast, group 2 subjects achieved the task using lower levels of FDI activation, associated with increased inhibition of the APB representation that may have helped prevent its unintended activation.

Individuated finger movements often require the stabilizing action of many intrinsic and extrinsic hand muscles (Li et al. 2000; Schieber 1995; Slobounov et al. 2002). As the level of force production by a single digit increases, the recruitment of stabilizing hand muscles and involuntary movement of other digits also increases (Slobounov et al. 2002). The index finger and thumb, however, are the digits most capable of individuated movement, and their movement requires relatively low levels of stabilizing activity (Häger-Ross and Schieber 2000; Slobounov et al. 2002). The need for stabilizing activation by other hand muscles was minimized in this experiment by having subjects perform the task with their index finger and ensuring that the entire forearm and hand, and all digits, were fully supported. The index finger used to perform the task remained in contact with the mouse button switch at all times, and the distance moved by the switch surface when it was depressed was approximately 1 mm. These combined factors minimized the degree of mechanical stabilization required from other hand muscles. However, it is possible that subjects in group 1 involuntarily recruited APB during the on phase of the task as part of a stabilization strategy. Unfortunately, it is not possible to distinguish between stabilizing activation and involuntary activation that serves no stabilizing function in the context of this experiment. However, even if group 1 subjects were performing the task with a strategy that included the stabilizing activation of APB, this does not detract from the finding that group 2 subjects performed the task with significantly less activation of APB, and that this seemed to be achieved by decreasing the excitability and increasing the inhibition of its M1 representation during the on phase of the task.

Overall, these results suggest that the selective activation of intrinsic hand muscles during a precision task requiring high levels of temporal and spatial accuracy is enhanced by an increase in ICI acting on the M1 representations of those muscles not involved in the task. The spatially and temporally differential modulation of ICI observed in this experiment may function to enhance motor contrast, by increasing the excitability of the FDI representation and decreasing the excitability of the APB representation during the on phase of the task.

In conclusion, this study supports the hypothesis that intracortical inhibitory processes play an important role in shaping the temporal and spatial characteristics of the output from M1. Intracortical inhibition may help to prevent the co-activation of cortical muscle representations, thus enhancing motor contrast and the selectivity of muscle activation during the performance of precise manual tasks.