Unilateral hand movements are accompanied by a transient decrease in corticospinal (CS) excitability of muscles in the opposite hand. However, the rules that govern this phenomenon are not completely understood. We measured the amplitude of motor evoked potentials (MEP) in the left first dorsal interosseus (FDI) elicited by transcranial magnetic stimulation (TMS) of the primary motor cortex in order to assess CS excitability changes that preceded eight possible combinations of unilateral and bilateral index finger movements with different right hand positions. Left FDI MEP amplitude (MEPLeft FDI) increased when this muscle acted as an agonist and tended to decrease when it was an antagonist. Additionally, MEPLeft FDI decreased substantially before right index finger abduction (a movement mediated by the right FDI) when both hands were lying flat (a movement mirroring left index finger abduction) but not when the right hand was turned at 90° or flat with the palm up. Therefore, CS excitability of the resting FDI was differentially modulated depending on the direction of the opposite index finger movement, regardless of muscles engaged in the task. These results indicate that inhibitory interactions preceding unilateral finger movements are determined by movement kinematics possibly to counteract the default production of mirror motions.
Coupling the activity of both hands in symmetrical mirror motions is a natural tendency that has been repetitively shown while performing bimanual movements (Serrien et al., 1999; Swinnen, 2002). This feature of motor behaviour is prominent in infants, in whom execution of strictly unilateral hand movements is particularly difficult (Mayston et al., 1999). Unilateral movements become easier later in life and are performed individually by the end of the first decade, possibly as a consequence of the functional maturation of the corpus callosum (e.g. Pujol et al., 1993; Rajapakse et al., 1996; Giedd et al., 1999; Banich and Brown, 2000).
In adult healthy volunteers, performance of unilateral finger movements is accompanied by a transient decrease in the excitability of CS neurons innervating muscles in the resting homonymous finger (Leocani et al., 2000; Liepert et al., 2001; Sohn et al., 2003; Weiss et al., 2003). It has been proposed that this mechanism permits to overcome the tendency to perform mirror movements when generating unilateral finger movements. However, whether this interhemispheric inhibition during movement preparation is selectively directed towards the muscles responsible for the mirror movement or directed towards the entire homonymous finger representation remains unknown. To address this issue, we measured CS excitability of the left FDI, a muscle responsible for index finger abduction, immediately before eight different combinations of unilateral and bilateral index finger movements. In addition, to study the relative weight of right index finger movement direction and right index finger muscle involvement on CS excitability of the left FDI, the right hand was placed in three different positions (‘flat palm down’, ‘90 degrees thumb up’, and ‘flat palm up’) in three separate experiments while the left hand was maintained ‘flat palm down’. We hypothesized that CS excitability of the left FDI would be selectively down-regulated by right index finger voluntary movements mirroring motions mediated by left FDI action.
Materials and Methods
Subjects and Experimental Condition
We studied 13 right-handed healthy volunteers (Oldfield, 1971) (6 women and 7 men; 32 ± 2.4 years old). All participants gave written informed consent. The NINDS Institutional Review Board and the ethic committee of the University of Louvain (UCL) approved the study protocol. Subjects sat on a comfortable chair in front of a computer screen. During the main experiment (n = 8) both hands laid flat, palm down with the arms semiflexed. In two additional experiments, the right hand was placed in two different positions; it was turned at 90° thumb up (n = 7) and positioned flat palm up (n = 5).
We evaluated changes in CS excitability of the left FDI immediately before eight different combinations of unilateral and bilateral index finger movements in a choice reaction time (cRT) paradigm (Fig. 1). Each trial started with a warning signal (a cross) displayed on a monitor, followed, after a delay of 1.4 s, by a visual cue that instructed the subject to perform a given combination of index finger movements.
Subjects practiced the task for a few min and then performed two blocks of 56 randomized trials (8 movement conditions × 7 trials per condition) to determine their mean RT for each movement condition in the absence of TMS. RT was defined as the time between the visual cue and onset of the voluntary EMG activity in the left FDI. In conditions with the left FDI at rest (Fig. 1A,B), cRT was measured from the right FDI EMG. The RT was then used to determine the timing of TMS application following the visual cue presentation, so that, for each subject and for each condition, TMS occurred 70 ms before the mean time of EMG onset estimated in unstimulated trials. Trials where TMS fell outside a ±30 ms window centred around 70 ms before EMG onset were discarded. This time window was chosen because previous studies have shown that corticospinal excitability changes start 100 ms before movement onset (Rossini et al., 1988; Chen and Hallett, 1999; Reynolds and Ashby, 1999; Leocani et al., 2000). Once the timing for TMS was determined, subjects participated in the core experiment, which consisted of six blocks of 64 movement trials (8 movement conditions × 8 trials each) and five randomly interleaved baseline trials, in which TMS was applied 1.4 s after the warning signal. Trials with background or mirror EMG activity were discarded. In this main experiment, both hands laid flat, palm down, so that right index finger abduction motions mirrored the action of the left FDI. To study the relative weight of right index finger movement direction and right index finger muscle involvement on MEPLeft FDI, we evaluated the influence of placing the right hand in two additional positions (‘90° thumb up’ in seven subjects and ‘flat palm up’ in five). In this way, whereas right index finger abduction mirrored the action of left FDI when the right hand was ‘flat palm down’, it was not the case in the two other hand positions.
EMG activity was recorded from surface electrodes placed over the right and left FDI muscles for 1500 ms (500 ms before and 1000 ms after TMS application). The EMG signal was amplified and bandpass filtered (50–2000 Hz; Counterpoint Electromyograph: Dantec Electronics, Skovlunde, Denmark); then it was digitized at 5 kHz and stored on a personal computer for off-line analysis.
Transcranial Magnetic Stimulation
Subjects wore a tightly fitting EEG cap with pre-marked coordinates in a 1 cm grid pattern. TMS was applied using a figure-of-eight magnetic coil (diameter of wings 70 mm) connected to a Magstim 200 magnetic stimulator (Magstim, Whitland, Dyfed, UK). The magnetic coil was placed tangentially on the scalp, over the right primary motor cortex (M1), with the handle pointing backward and laterally at a 45° angle away from the midline, approximately perpendicular to the central sulcus. The hot spot was defined as the optimal position to elicit MEPs in the left FDI. We chose to record MEPs from a left hand muscle (left FDI) because it is known that interhemispheric inhibition is stronger from the dominant to the non-dominant motor cortex (Leocani et al., 2000). The resting motor threshold (rMT) was defined, at the hot spot, as the minimal TMS intensity needed to evoke MEPLeft FDI larger than 50 μV peak-to-peak in the relaxed FDI in 5 out of 10 consecutive trials. The intensity of TMS was then set at 10% above the resting motor threshold during the whole experiment.
Consistent H reflexes can be recorded from intrinsic hand muscles in rare occasions (Mazzocchio et al., 1995). We screened every participant but were able to obtain H reflex from left FDI in only one subject. Electrical stimulation was applied to the left ulnar nerve through bipolar electrodes placed just above the wrist (1 ms rectangular pulses) at a stimulus intensity set to elicit consistent H reflexes. M waves were measured to monitor the stability of the nerve stimulation. We tested the influence of the four unilateral movement conditions (Fig. 1A,B,C,F) on H reflexLeft FDI and MEPLeft FDI amplitudes when TMS and electrical stimulation were randomly intermixed. At the end of the experimental session, the maximum M wave (M max) was recorded in order to express MEPLeft FDI and H reflexLeft FDI amplitudes relative to this value.
Log transformations were applied to the amplitudes of MEPLeft FDI (log MEPLeft FDI amplitude) for each of the nine conditions (the eight movement conditions and the baseline) after testing for normality (Kolmogorov–Smirnov tests; failed, P < 0.03). The log MEPLeft FDI measurements for the nine Movement conditions were analyzed in a general linear model (repeated measures ANOVA) with Subjects, Left FDI action(agonist, antagonist, rest) and Right FDI action(agonist, antagonist, rest) as factors. In order to explore further the interaction identified in this initial analysis, we performed a separate repeated measure (rm) ANOVA evaluating the effects of factor Movement condition on log MEPLeft FDI amplitude. Post-hoc pairwise comparisons for both the repeated measures and subjects were implemented using Sidak and Scheffe adjustments for multiple comparisons, respectively. In addition, a two-way factorial ANOVA with factors Right hand position(flat palm down, 90 degrees thumb up, flat palm up) and Right index finger movement direction(rest, mirror, non-mirror) was used to study the relative influence of unilateral right index finger movement direction and muscle involvement on MEPLeft FDI expressed in percentage of baseline. Post-hoc pairwise comparisons were implemented using Tukey Tests adjusted for multiple comparisons. Reaction times were analyzed using a two-way repeated measures ANOVA with Movement condition and Stimulation(TMS, no TMS) as factors. Data are expressed as mean ± SE.
Repeated measures ANOVA showed a significant effect of Movement condition (F = 7.25; P < 0.001) but not Stimulation (F = 1.43; P = 0.28) on reaction times; their interaction was not significant (F = 0.97; P = 0.46, Table 1). Post-hoc testing showed slightly shorter RTs in the two conditions involving a unilateral adduction of the left or right index finger (< −, − >) than in movements involving abduction of the left index finger (> −, > >, > <) and in the mirror condition (< >) (P < 0.05).
|No TMS||TMS||P value|
|No TMS||TMS||P value|
TMS = transcranial magnetic stimulation; NS = not significant (P > 0.05); − < = right index finger abduction; − > = right index finger adduction; > − = left index finger abduction; > > = right bimanual parallel movement; > < = inward bimanual mirror movement; < − = left index finger adduction; < < = left bimanual parallel movement; < > = outwards bimanual mirror movement; SE = standard error.
Premovement MEPLeft FDI
In baseline trials, the amplitude of MEPLeft FDI was 1.2 ± 0.43 mV (n = 8). Repeated measures ANOVA showed a significant effect of Left FDI action(agonist, antagonist, rest), Right FDI action(agonist, antagonist, rest) and their interaction on log MEPLeft FDI (F = 157.7, P < 0.0001, F = 6.2, P = 0.002 and F = 3.8, P = 0.004 respectively). Log MEPLeft FDI was significantly larger when left FDI acted as an agonist (P < 0.0001). It was smaller when right FDI acted as agonist than when it acted as antagonist (P < 0.001). Rm ANOVA showed a significant effect of Movement conditions on log MEPLeft FDI amplitude (F = 20.248; P < 0.001, Fig. 2a,b). In trials with the left FDI acting as an agonist to the intended movements (Fig. 1C–E), the amplitude of MEPLeft FDI was significantly larger than in the baseline condition, irrespective of the unilateral or bilateral character of the task (P < 0.001; Fig. 2C–E). In contrast, in trials with the left FDI acting as an antagonist to the intended movement (Fig. 1F–H), MEPLeft FDI amplitude tended to be smaller than in baseline trials although it was not significant (Fig. 2F–H).
In movement trials where left FDI remained at rest (Fig. 1A,B), MEPLeft FDI amplitude was modulated depending on the right index finger movement direction. Specifically, in trials with right index finger abduction (Fig. 1A) MEPLeft FDI amplitude was smaller (P = 0.005, Fig. 2A: − <) but remained unchanged in trials with right index finger adduction (Fig. 2B: − >). There was a significant difference between MEPLeft FDI amplitudes in trials with right index finger abduction and those with adduction (P = 0.02).
A factorial ANOVA showed a significant effect of Right index finger movement direction(mirror, non-mirror) but not Right hand position(flat palm down, flat palm up, 90 degrees thumb up) nor their interaction (F = 9.876, P < 0.001; F = 0.0734, ns and F = 0.924, ns) on normalized MEPLeft FDI amplitudes. Therefore, for each right hand position, MEPLeft FDI were smaller with movements performed towards the midline (mirroring the action of L FDI) than at baseline (P = 0.03). Additionally, right index finger movements towards the midline (mirroring the action of left FDI) resulted in MEPLeft FDI of smaller amplitude than non-mirror movements, regardless of the muscles engaged in the task (P < 0.001; see Fig. 3).
Premovement H ReflexLeft FDI
We tested simultaneously MEPLeft FDI and H reflexLeft FDI amplitudes in the only subject in whom we could elicit H reflexes from left FDI. Consistent with the group results, MEPLeft FDI amplitude increased when this muscle acted as an agonist (Fig. 4aC), did not change when it acted as an antagonist (Fig. 4aD) and decreased in trials with right index finger abduction (Fig. 4aA–B). In contrast, we observed that H reflexLeft FDI amplitude increased to a similar extent before any movement of the left index finger, regardless of the involvement of the FDI as an agonist or antagonist (Fig. 4c,dC–D, > −, < −); H reflexLeft FDI amplitude was not modulated before movements of the right index finger (Fig. 4c,dA–B, − <, − >).
In the present study, we found that, during movement preparation, CS excitability of the resting index finger abductor muscle was differentially modulated depending on the direction of the movement performed by the opposite index finger, regardless of the muscles engaged in the task.
The purpose of these experiments was to study changes in CS excitability of one finger muscle representation during preparation of unilateral and bilateral finger movements in different directions. Measure of MEP amplitudes immediately before movement onset allowed us to identify inhibitory and excitatory influences modulated by the motor set, without contamination by the actual movement (Starr et al., 1988; Hoshiyama et al., 1996; Chen et al., 1998). In our experimental paradigm, subjects could not predict the movement to make until the visual cue appeared, and therefore TMS was always applied at a comparable stage in the process of generation of various unilateral and bilateral movements. It is of note that we did not find TMS-induced RT increases most likely because TMS was applied at ∼70 ms before the mean RT in unstimulated trials for each condition and at relatively low (110% rMT) stimulus intensities (Ziemann et al., 1997; Leocani et al., 2000; see Table 1).
Previous studies have shown that preparation to move one finger is associated with a transient inhibition of muscles controlling movements of the homonymous finger in the resting hand (Leocani et al., 2000), probably an effect of cortical origin (Weiss et al., 2003). Our results now indicate that these inhibitory interactions are much more specific than previously thought. First, we found that generation of a unilateral index finger abduction (i.e. right) resulted in inhibition exclusively directed towards the abductor muscle in the resting index finger (i.e. left). This inhibition occurred when both hands were symmetrically placed flat on the board (‘flat palm down’), and therefore when right index finger abduction mirrored the function of the left FDI in the resting hand. Second, performance of the same right index finger abduction movement with the hand placed in another position than ‘flat palm down’ (e.g. ‘90° thumb up’ or ‘flat palm up’) failed to modulate activity of the left FDI in the resting hand. The main difference between these two conditions and ‘flat palm down’ is that abduction of the right index finger occurred in directions other than mirror to the pulling direction of the left FDI. These results are consistent with the view that inhibitory interactions before unilateral finger movements are determined by the movement direction defined in an extrinsic (external) coordinate frame (Kakei et al., 2001, 2003) and are therefore heavily influenced by the dynamics of the motor set.
During childhood, performance of unilateral finger movements is associated with mirroring (Mayston et al., 1999). Over time, maturation of the CNS leads to a progressive decrease in intensity and occurrence of mirror activity, more prominently during the first decade of life (Wolff et al., 1983; Heinen et al., 1998). In normal adults, mirror movements are usually absent except when movements are performed against high resistance (Armatas et al., 1994), under conditions of muscle fatigue (Liederman and Foley, 1987; Aranyi and Rosler, 2002) or in the setting of cortical lesions like stroke (Cohen et al., 1991; Nelles et al., 1998; Kim et al., 2003). Interestingly, bimanual voluntary movements are easy to perform in a mirror direction whereas the same movements performed unilaterally or asymmetrically require some practice to be successfully implemented without mirroring (Swinnen et al., 1993; Wenderoth et al., 2003). These findings are consistent with the view that bilateral mirror- or symmetrical-motor activity may represent the default mode of operation of the CNS (Swinnen, 2002). The ability to perform unilateral movements relies therefore on the adequate inhibition of such a default operating mode (Chiarello and Maxfield, 1996; Schnitzler et al., 1996).
Our finding of a lack of correspondence between changes in CS excitability and H-reflexes (see Fig. 4) is consistent with the hypothesis that premovement interactions occur at a cortical level, most probably through interhemispheric connections. Anatomical and physiological evidence for interhemispheric interactions between motor representations are numerous in both animals (Rouiller et al., 1994; Aboitiz and Montiel, 2003) and humans (Ferbert et al., 1992; Thut et al., 1997; Meyer et al., 1998); they have been shown to be predominantly inhibitory (Ferbert et al., 1992; Gerloff et al., 1998) and operate in the process of generation of voluntary movements (Murase et al., 2004). The precise site of these cortical inhibitory interactions remains to be determined. Possible candidates include the supplementary motor area, largely connected with its homonymous brain region through the corpus callosum (Rouiller et al., 1994), and the primary and premotor cortices, contributing to unimanual independence and bimanual coordination (Brinkman, 1984; Sadato et al., 1997; Donchin et al., 1998; Immisch et al., 2001; Meyer-Lindenberg et al., 2002). However, the present finding that movement kinematics influence interhemispheric interactions suggests that they occur between cortical regions where movements are represented in extrinsic coordinates. The issue of movement reference frames in motor areas is still very controversial (Scott and Kalaska, 1997; Kakei et al., 1999, 2001; Todorov, 2000), and therefore it remains to be determined which cortical regions contribute to these inhibitory interhemispheric interactions.
In conclusion, the present results argue for a high degree of specificity in interhemispheric inhibition that is influenced by the kinematics of the intended movements, possibly to counteract the default production of mirror movements.
J.D. is a research assistant at the Belgian National Funds for Scientific Research (FNRS). We are thankful to Friedhelm Hummel for his critical comments on an earlier draft of this manuscript.
1Human Cortical Physiology Section, National Institute of Neurological Disorders and Stroke, NIH, Bethesda, MD, USA, 2Laboratoire de Neurophysiologie, Université de Louvain, Brussels, Belgium, 3Sezione di Neurofisiologia Clinica, Dipartimento di Scienze Neurologiche e del Comportamento, Universita' di Siena, Italy and 4Department of Neurology, Tokushima University Faculty of Medicine, Tokushima, Japan