Skip to main content
PMC home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2010 Mar 1;588(Pt 8):1269–1280. doi: 10.1113/jphysiol.2009.183038

Illusory movements of a phantom hand grade with the duration and magnitude of motor commands

Lee D Walsh 1, Simon C Gandevia 1, Janet L Taylor 1
PMCID: PMC2872732  PMID: 20194129

Abstract

The senses of limb movement and position are critical for accurate control of movement. Recent studies show that central signals of motor command contribute to the sense of limb position but it is not clear whether such signals influence the distinctly different sense of limb movement. Nine subjects participated in two experiments in which we inflated a cuff around their upper arm to produce an ischaemic block, paralysing and anaesthetising the forearm, wrist and hand. This produces an experimental phantom wrist and hand. With their arm hidden from view subjects were asked to make voluntary efforts with their blocked wrist. In the first experiment, efforts were 20 and 40% of maximum and were 2 and 4 s in duration. The second experiment used 1 and 5 s efforts of 5 and 50% of maximum. Subjects signalled perceived movements of their phantom wrist using a pointer. All subjects reported clear perceptions of movement of their phantom hand for all levels and durations of effort. On average, subjects perceived their phantom wrist to move between 16.4 ± 3.3 deg (mean ± 95% confidence interval (CI)) and 30.2 ± 5.4 deg in the first experiment and between 10.3 ± 3.5 and 38.6 ± 6.7 deg in the second. The velocity of the movements and total displacement of the phantom graded with the level of effort, and the total displacement also graded with duration. Hence, we have shown that motor command signals have a novel proprioceptive role in the perception of movement of human joints.

Introduction

How efferent command signals contribute to the proprioceptive senses has long been controversial. Much of the focus has been on whether or not these command signals can create ‘sensations of movement’. Von Helmholtz (1867) proposed that to perceive the position of an object in our visual world it was necessary to know where the image formed on the retina and the position of the eyeball. Eyeball position could be signalled by ocular proprioceptors or by knowledge of the motor commands that moved the eye, termed a ‘sensation of innervation’. Later this sensation of innervation was termed a ‘corollary discharge’ (Sperry, 1950) or ‘efference copy’ (von Holst, 1954). Both terms refer to a signal derived from the motor command sent to the muscles. Such a system would minimise the reliance on delayed feedback from the periphery. The simple biomechanics of the eyeball are such that information from corollary discharges could be quite accurate. The current consensus is that the input from both ocular proprioceptors and centrally generated sources signal eyeball position (Donaldson, 2000).

Despite anecdotal reports from amputees that their phantom limbs moved with voluntary efforts (see Henderson & Smyth, 1948; McCloskey, 1978), it was thought that corollary discharges were not involved in sensations of limb movement (e.g. Goodwin et al. 1972; McCloskey & Torda, 1975; Gandevia & McCloskey, 1977). No movement or change in position was perceived when voluntary efforts were made to move paralysed limbs (for review see Matthews, 1982; McCloskey et al. 1983a; Gandevia, 1987). The only role given to corollary discharges in the sense of limb position and movement was a central subtraction whereby the corollary discharge corrected for muscle spindle input provided by fusimotor co-activation during a voluntary contraction (Goodwin et al. 1972; McCloskey et al. 1983b). However, recent studies have found that motor command signals are involved in the sense of limb position. By ‘motor command’ we mean a centrally generated signal that is monotonically related to motoneuronal output to the muscle. This signal may be related to drive reaching the motor cortex or it may arise in parallel to it. Voluntary efforts resulted in illusory changes in the position of a phantom hand which was created experimentally by ischaemic anaesthesia and paralysis of the arm (Gandevia et al. 2006). Similar illusions occurred when afferents were intact but the arm was paralysed, as well as when the arm was fully intact (Smith et al. 2009; Walsh et al. 2009b). In addition, other studies have shown an effect of fatigue on position sense that has been attributed to an influence of central command signals (Walsh et al. 2004; Allen & Proske, 2006; Walsh et al. 2006). Little has been done to look at the effect of central commands on the sense of limb movement.

In a recent study in which phantom hands were produced by ischaemia, some subjects gave a verbal report of movement (Gandevia et al. 2006). By contrast, Melzack & Bromage (1973) showed that experimental phantoms ‘move’ only if the block was incomplete and electromyographic activity persisted. However, they gave no indication of the level of effort produced by the subjects and thus the efforts may have been below the threshold to generate movement. They reported it was difficult to attempt to move paralysed limbs, and others have shown that it is difficult to sustain an effort to paralysed muscles (Stevens, 1978; Gandevia et al. 1993). From our studies that have assessed motor performance in the presence of deafferentation and paralysis (e.g. Hobbs & Gandevia, 1985; Gandevia et al. 1990, 2006; Smith et al. 2009), we have found that subjects may need encouragement to perform tasks that are physically impossible. In amputees who reported voluntary movements of phantoms, the movements were accompanied by twitching of muscles in the stump (Henderson & Smyth, 1948). It has been suggested that the muscles in the stump are differentially active for different phantom movements (Reilly et al. 2006) and that perception of movements depends on this activity through remapping of the hand representation onto remaining muscles. In contrast, transcranial magnetic stimulation over the hand area of the motor cortex in amputees can elicit phantom movements which do not depend on specific activation of stump muscles (Mercier et al. 2006). However, these movements are twitches, which can be hard to localise even in intact subjects, and it is difficult to know how specific the stimulus is for the motor hand area, especially if there has been cortical reorganisation.

Here we aimed to determine if subjects could perceive continuous movements of a paralysed and anaesthetised wrist when they made voluntary efforts. The alternative is that they perceive an instantaneous change in position without any movement. The aim was to quantify any illusory movements that were experienced and to determine if there was a dependence of the size and velocity of these movements on the direction, magnitude and duration of effort. An important control was to check that subjects could accurately indicate unexpected movements of their real hand in the absence of visual feedback. It was also important to establish that subjects could produce at least two different efforts without this feedback. Our hypothesis was that subjects would perceive the velocity of their phantom wrist during voluntary efforts and that these movements would move the joint further and faster for larger efforts and further for longer efforts.

Methods

Nine healthy subjects (4 male) aged 23–51 participated in the study. Eight participated in the first experiment and five of those plus the ninth participated in the second experiment. All subjects gave informed consent and the experimental procedures were carried out in accordance with the Declaration of Helsinki. The University of New South Wales Human Research Ethics Committee approved the study. All subjects were informed about the experimental procedures but were kept unaware of the experimental hypothesis. They were informed that their right arm would be paralysed and anaesthetised and that they would be asked to make voluntary efforts and then signal any movement with their left hand using a pointer.

Experimental setup

Subjects had their right hand strapped into a manipulandum mounted on a table that held the hand in a semi-pronated position with the fingers extended (Fig. 1). The manipulandum restricted movement of the wrist, allowing only wrist flexion and extension. The manipulandum was fitted with a load cell (250N XTran, Applied Measurement, Australia) to measure forces generated by the wrist. Electromyographic activity (EMG) was measured from over the flexor and extensor carpi radialis muscles using Ag–AgCl surface electrodes (band-pass filtered 16–1000 Hz, CED 1902 amplifier). The purposes of measuring EMG were to ensure subject compliance during the control experiment and to ensure EMG was absent during the block. The subject's arm was covered below the elbow. A pointer, the axis of which was co-linear with the axis of the wrist, sat above the wrist and could be moved by the subject's left hand to indicate position and movement about the right wrist. Potentiometers mounted in the wrist manipulandum and the pointer were calibrated to measure actual wrist angle and pointer angle (i.e. perceived wrist angle), respectively. The manipulandum could be locked so that subjects could perform isometric contractions with the wrist. When unlocked, the manipulandum acted against an adjustable motorcycle steering damper, which applied a viscous load to the wrist. The damper had seven discrete settings allowing us to choose viscosities for each subject based on their strength. The reason for using this type of load was that we had an anecdotal report from a previous study that perceived movements of a phantom wrist had the quality ‘as if moving through treacle’ (Gandevia et al. 2006, p. 707). Using this load also allowed us to change the resistance so that the subjects could not predict the movement that would occur for a particular effort. The force, angle and EMG data were all sampled through a CED Power1401 data acquisition system (Cambridge Electronic Design, Cambridge, UK) and stored on a computer for analysis.

Figure 1. The experimental setup.

Figure 1

Open in a new tab

Subjects sat with their forearm strapped to the table and hand clamped in a manipulandum that allowed movement around the axis of the wrist joint into both flexion and extension. The rotating platform could be locked into position for isometric trials otherwise it was free to move against a viscous load, which was applied by a motorcycle steering damper. The forearm was hidden from subjects for the whole experiment. A pointer with an axis colinear with the axis of the wrist joint was moved by the subject's left hand to indicate the position and movement of the right wrist joint throughout the experiment. The cuff was placed around the upper arm. This figure is used with permission from Smith et al. (2009).

Experiment 1

Subjects (n= 8) initially performed maximal voluntary contractions (MVCs) of the wrist flexor and wrist extensor muscles to determine maximal flexion and extension forces which were used to set submaximal target forces. Subjects were then trained to perform efforts of 20 and 40% of maximum for 2 or 4 s. At first, visual feedback of force was provided on an oscilloscope, and then it was removed and verbal feedback was given by the experimenter. Once subjects could repeatably make 20 and 40% efforts with their wrist flexors and wrist extensors without feedback, they performed a series of control trials designed to test their ability to indicate the position and velocity of their wrist during these efforts. There were trials in which subjects missed the target force by a large margin and these trials were discarded and repeated. This only occurred in about 1 in 10 trials.

Efforts with the arm intact

In each trial, subjects performed the same effort twice. The first time, the manipulandum was locked in position so that the effort resulted in an isometric contraction. The second time, the manipulandum was unlocked so that the subject pushed against an unknown viscous load. Here it was possible for the subject's wrist to move during the effort but the distance and velocity of any movement depended upon the voluntary force exerted. To provide a variety of unpredictable movements required efforts at two intensities (20 and 40% of maximum), two durations (2 and 4 s) and two viscous loads (big and small). We used four conditions. The first condition comprised a 20% effort for 2 s against the big load. The second was a 20% effort for 2 s against the small load. The third was a 40% effort for 2 s against the big load and the fourth was a 20% effort for 4 s against the big load. Subjects performed each condition four times, in both directions of wrist flexion and extension (total of 32 trials) and the order of trials was randomised.

Subjects were not told which viscous load they would be pushing against or the duration of the effort, but they were told to maintain the effort until the experimenter said ‘relax.’ The order of events was as follows. The subject was told the level and direction of effort (e.g. ‘get ready to make a 40% effort into flexion’). Then the subject was asked to push and the experimenter counted them though the effort before telling the subject to relax (e.g. for a trial that was 4 s long, ‘Push, 2, 3, 4, relax’). Next, the subject was instructed ‘show me where your wrist is’. The subject then used the pointer to indicate the current position of their wrist. The manipulandum was then unlocked and the subject was instructed to make the same effort with the same instruction as previously (when wrist movements could occur). After the subject relaxed, the experimenter sometimes moved the manipulandum in the direction opposite to the subject's active movement, so that the subject experienced a combination of active and passive movements. The subject was then instructed, ‘show me what your wrist did’. The subject then moved the pointer through the perceived path taken by their right hand during the just completed effort. Instructions to the subjects emphasised that all parts of each movement should be tracked, including any pauses in movement, so that the record of the pointer signal should duplicate that of the wrist in both time and displacement. The left panel of Fig. 2 shows an example of one of these trials and the data derived from it.

Figure 2. Typical trials from one subject during the ‘intact’ and ‘blocked’ states.

Figure 2

Open in a new tab

A shows one trial during the control experiment when the arm was intact. The downward arrow indicates when the subject was told to ‘show me where your wrist is’, at which point the subject moved the pointer from its starting position to where they perceived their wrist. The upward arrow indicates when the subject was told to ‘show me what your wrist did’. The upper row of boxes below the indicator trace mark when the wrist was locked, so that when the subject made the effort their wrist was either ‘isometric’ or ‘free to move’. The lower row of boxes shows when the subject made a voluntary ‘effort’. The subject was told a level and direction of effort. For example ‘get ready to make a 20% effort into flexion.’ The box labelled EM indicates experimenter movement, when the experimenter moved the wrist back towards the starting position. After the experimenter-imposed movement, the subject was told to ‘show me what your wrist did.’ The horizontal dashed lines in the top of A indicate where the change in position was measured and the dotted line indicates where the average velocity was measured. B is a trial from the same subject during the block. For these trials the subject was first instructed to ‘show me where your wrist is’ (at the downward arrow) and moved the pointer from its starting position at either full extension (shown here) or full flexion to their perceived wrist position. Then they were told the level of effort and the direction of effort to make and after the effort was complete the subject was told to ‘show me what your wrist did’ (at the upward arrow).

Efforts with the arm blocked

After subjects had completed the control trials, an ischaemic block of the right forearm and hand was performed. A two-chamber cuff placed around the upper arm was inflated to 250 mmHg and subjects remained relaxed until light touch sensation was abolished below the elbow. This typically took 35–40 min. At this time, the forearm and hand were paralysed and anaesthetised (Gandevia et al. 2006). However, all subjects continued to have a perception of wrist and hand position although this did not necessarily correspond to the actual position of their wrist and hand. That is, subjects developed a ‘phantom’ of their forearm and hand (Walsh et al. 2009a).

Once the block was complete, the subjects were instructed to perform efforts that were 20 or 40% of maximum into flexion or extension at the wrist for durations of 2 or 4 s. Each trial was repeated three times (total 24) and the order was randomised. Again, subjects were not told in advance the duration of each effort. Before each effort, subjects were told to ‘show me where your wrist is’ and they did so with the pointer. Next the experimenter told them the size and direction of the effort to make. After they made the effort with the experimenter counting them through it and relaxed on instruction, the subjects were told to ‘show me what your wrist did.’ As in the control trials, they moved the pointer to track the path of any perceived movement of their phantom that was associated with the preceding effort. An example of the data from these trials is shown in the right panel of Fig. 2.

After the experiment the cuff was deflated and subjects recovered from the block within 5–10 min. Subjects were then asked about any additional perceptions that they could not indicate with the pointer during the experiment. In addition, a structured interview was conducted to gain additional information about each subject's perception of any wrist movements during the block. The experimenter transcribed the responses.

Experiment 2

In experiment 1, subjects perceived wrist movements when they made voluntary efforts during an ischaemic block, but the differences between conditions were small (see Results). Therefore, the experiment was repeated with greater differences between levels of effort (5 and 50%) and durations of effort (1 and 5 s). As in experiment 1, subjects (n= 7) initially made maximal voluntary flexion and extension contractions about the right wrist and were then trained to make 5 and 50% maximal flexion and extension efforts without feedback. One subject, who did not participate in experiment 1, completed a series of control trials with the arm intact to demonstrate an ability to reproduce unexpected wrist movements using the pointer. All subjects then underwent an ischaemic block of the right forearm and hand. When light touch was lost below the elbow, 6 of the 7 subjects reported a phantom wrist and hand. One subject reported phantom fingers but no wrist and was excluded from the study. With the forearm and hand paralysed and anaesthetised, subjects (n= 6) performed trials as described in experiment 1 above. Subjects tracked perceived wrist movement after making 5 or 50% efforts into flexion or extension at the wrist for durations of 1 or 5 s.

Data and statistical analysis

Data processing was done using Spike 2 version 6 (Cambridge Electronic Design) and Igor Pro version 6 (Wavemetrics, Lake Oswego, OR, USA). Statistical testing was done using SPSS ver. 17 (SPSS Inc., Chicago, IL, USA). Wrist angle was measured immediately prior to each effort and at peak displacement. Pointer angle was measured immediately prior to being moved and at peak displacement. Changes in angle were calculated for both the wrist and the pointer. Average velocity was taken as the average slope of the position trace between the start and end angle for both the wrist and the pointer. Peak velocity was also measured, but the results were qualitatively similar to those for the average velocity and are not mentioned further. Data are presented as a mean ± 95% confidence interval, except data about the timing of movements, which are presented as a range with a median. The data from the control experiment (Fig. 3) were analysed with regression analyses between perceived wrist position and the actual wrist position, and perceived velocity and actual velocity. The changes in angle and velocities were tested using a repeated measures general linear model with the factors of direction (flexion vs. extension), level of effort (20 vs. 40% or 5 vs. 50%) and duration (2 vs. 4 s or 1 vs. 5 s). Threshold for significance for all statistical tests was set as P < 0.05.

Figure 3. Subjects’ performance during the matching of wrist position and velocity in the control study before the arm was paralysed and anaesthetised.

Figure 3

Open in a new tab

A shows the change in position that subjects indicated with the pointer versus the actual change in the wrist position. B shows the wrist velocity that subjects indicated versus the actual wrist velocity. The continuous lines are the lines of best fit and the dashed lines show their 95% confidence intervals. The dotted lines are the lines of identity and positive angles indicate a displacement or velocity into extension of the wrist. Subjects overestimated changes in the position and the velocity of their wrist movements, but their judgments have a strong linear correlation to the true movements of their wrist (P < 0.001).

Results

We measured subjects’ ability to indicate, with a pointer, position and movement about the wrist joint during efforts against viscous loads. Subjects then used the same pointer to indicate any perceived movement of the wrist during efforts made while the forearm and hand were anaesthetised and paralysed by an ischaemic block. After the initial training period, subjects were able to reproduce target efforts repeatably without feedback. Efforts varied about the target by 5.3 ± 0.5% (mean ± 95% CI) of their maximum.

Experiment 1

Efforts with the arm intact

The purpose of this condition was to ensure that subjects could indicate unexpected movements of their wrist. With the arm intact, subjects tended to overestimate both the change in position and the velocity of any wrist movements (Fig. 3). However, there were significant linear relationships between movements of the pointer and actual wrist movements (P < 0.001). These correlations were strong with R2 values of 0.83 and 0.88 for changes in position and velocity, respectively. Thus, subjects were able to signal different magnitudes and velocities of unexpected movements of the wrist joint shortly after the event using a pointer.

Efforts with the arm blocked

With the arm blocked (i.e. paralysed and anaesthetised), subjects made efforts of 20 or 40% maximum for 2 or 4 s. On average, subjects perceived their wrist to move between 16.4 ± 3.3 and 30.2 ± 5.4 deg with all combinations of level and duration of effort. However, movement was not always perceived to continue for the entire duration of the effort. In many trials the initial movement of the phantom was fast at the start of the effort with the phantom then pausing at the final position until the subject ceased making an effort (Fig. 4). The duration of the perceived movement varied from 0.6 to 2.4 s (median 1.4 s) for efforts that were 2 s in duration and 0.4 to 3.9 s (median 2.3 s) for efforts of 4 s duration. The duration of the pauses at the final position was 0 to 1.6 s (median 0.47 s) and 0 to 3.3 s (median 0.6 s) for 2 and 4 s efforts, respectively. Larger displacements of the wrist were reported for 40% efforts compared to 20% efforts (F1,7= 13.0, P < 0.01) as well as for efforts of 4 s duration compared to efforts of 2 s duration (F1,7= 8.95, P < 0.05; Fig. 5). However, these differences in displacement were small (2 to 9 deg). Perceived velocities of wrist movement were between 10.9 ± 2.2 and 16.9 ± 4.1 deg s−1 in the different conditions. There were no significant effects of level or duration of effort on the perceived velocity. After the efforts were stopped, subjects signalled that their hand returned, on average, to within ∼0.1 deg of their pre-effort position. While this number was significantly different from zero, it is too small to be physiologically relevant.

Figure 4. Examples of perceived movements during 4 s efforts.

Figure 4

Open in a new tab

Traces of perceived wrist position from two subjects when they moved the pointer to signal perceived movement during a preceding 4 s effort. The thick line is a trial in which the subject's phantom wrist moved throughout the effort. The dashed line shows a trial in which the subject perceived a quicker movement at the start of the effort followed by a slower movement and a pause at the final position for the remainder of the effort. In this example the mean velocity (10 deg s−1) does not match the velocity calculated (3 deg s−1) from the final displacement (14 deg) and the duration of the effort (5 s). In each trial, the return of the phantom towards its original position occurred with the end of effort and has been truncated for the illustration.

Figure 5. Perceived change in phantom wrist position and wrist velocity during 20 and 40% efforts for individual subjects and the group.

Figure 5

Open in a new tab

A shows the changes in wrist position perceived during efforts of 20 and 40% of maximum and with durations of 2 and 4 s. B shows the velocity of the perceived movements. The thick lines and circles indicate the group mean ± 95% CI (8 subjects) and the thin lines represent the mean data from each subject. Data for efforts into wrist extension (continuous lines, filled circles) and flexion (dashed lines, open circles) are shown. On average, subjects reported velocities of greater than zero for all efforts. * indicate significant differences in the change in wrist position perceived during a 20% effort vs. a 40% effort, and between a 2 s effort and 4 s effort (P < 0.05).

Experiment 2

Subjects made efforts of 5 or 50% maximum for 1 or 5 s with the arm blocked. As in the first experiment, subjects perceived movements of their phantom wrist but sometimes the movement did not last the entire duration of the effort. For 1 s efforts the duration of the movement was 0.7 to 0.8 s (median 0.7 s) and the pause at the final wrist position was 0 to 1.4 s (median 0.6 s). The duration of the movement for 5 s efforts varied from 0.1 to 4.4 s (median 1.3 s) and the pause at the final position varied from 0 to 4.1 s (median 1.0 s). This range is large because one subject indicated very rapid movements and reported that their phantom changed position instantaneously. On average, subjects perceived the wrist to move between 10.3 ± 3.5 and 38.6 ± 6.7 deg with all combinations of level and duration of effort (Fig. 6). There was an effect of both level of effort (F1,5= 50.9, P < 0.005) and duration (F1,5= 10.4, P < 0.05) on the perceived change of wrist angle and these effects are larger (15 to 30 deg) than in experiment 1 (2 to 9 deg). The longer duration (5 s) and the larger effort (50% maximum) produced the larger perceived movements. The velocity of perceived movements was 13.0 ± 3.6 to 37.0 ± 7.9 deg s−1 with a significant effect of the level of effort (F1,5= 14.8, P < 0.05). However, there was no significant effect of effort duration on the velocity of perceived movements. Interestingly, the trend was towards slower movements with a longer effort, which is the same as the first experiment (Fig. 5). As with the first block, on average subjects indicated that after an effort their hand returned very near (within 0.1 deg) to the position that they had indicated their wrist occupied before the effort.

Figure 6. Perceived change in phantom wrist position and wrist velocity during 5 and 50% efforts for individual subjects and the group.

Figure 6

Open in a new tab

A shows the changes in wrist position perceived during efforts of 5 and 50% of maximum and with durations of 1 and 5 s. B shows the velocity of the perceived movements. The thick lines and circles indicate the group mean ± 95% CI (6 subjects) and the thin lines represent the mean data from each subject. Data for efforts into wrist extension (continuous lines, filled circles) and flexion (dashed lines, open circles) are shown. The mean data show that on average movements were perceived by subjects during all conditions except the 1 s long 5% efforts (the CI for the mean velocity includes zero). * indicates significant differences in the change in perceived wrist position during a 5% effort vs. a 50% effort within durations, and between the change in wrist position with a 1 s effort vs. a 5 s effort. For perceived velocity of movement, there is a significant difference between a 5% effort and a 50% effort.

Subject reports after the block

All subjects reported that they had made two distinct levels of effort while paralysed and anaesthetised, and that they perceived definite movement of their phantom hand during efforts. However, one subject who participated in both experiments reported perception of movement only on the first occasion. On the second occasion, the subject said that the phantom hand simply jumped from position to position with no perceivable movement. Three subjects reported that perception was the same in both flexion and extension, while the others reported that it was easier to make an effort in flexion or extension. Five subjects said that the movements started immediately after they started an effort and four reported that their phantom always moved for the whole duration of their effort. The majority of subjects (5 out of 9 subjects) felt that the phantom movements were like pushing through a viscous substance. The substances suggested by the subjects were oatmeal, honey, wet concrete, water or glue.

Discussion

We have shown that motor command signals can generate graded sensations of continuous movement in the absence of sensory input. Subjects made no real wrist movements, had no EMG in the wrist flexors or extensors, and had no changes in afferent input related to movement. Yet, subjects not only tracked perceived wrist movements with a pointer but they later reported, in all but one case, that their efforts evoked a distinct continuous movement of the phantom rather than an instantaneous change in its position. Subjects perceived their phantom wrist to move faster if they made bigger efforts and further if they made longer efforts.

In the control experiment, subjects made wrist flexion and extension efforts against two loads that were not known in advance. Thus, their voluntary efforts generated movements of unpredictable displacement and velocity. Under these conditions, subjects indicated reliably the wrist movements with a pointer above their hand. While their performance was not perfect (Fig. 3), it was consistent, as shown by the strong linear relation between the actual and perceived wrist position. Subjects tended to overestimate both the change in position and the velocity of the movements. Figure 3 shows that the gradient of the relation is ∼1.3 for perceived position and ∼1.7 for velocity. It is unlikely that use of the pointer to indicate movements causes this overestimation as subjects accurately indicate wrist position with a pointer when the wrist is moved passively (Gandevia et al. 2006) and indicate similar elbow positions with an arm or a pointer (Gritsenko et al. 2007). Thus, the overestimation most probably represents perception of larger, faster movements than those that occurred. This could be due to a contribution of motor command signals to limb position sense (Smith et al. 2009; Walsh et al. 2009b), and our current results show that there may be a similar effect on perception of limb velocity. However, overestimation of remembered limb position has previously been reported early in passive as well as active elbow movements (Gritsenko et al. 2007) so that overestimation of movement extent and velocity when subjects were intact may be independent of the presence of voluntary effort.

These results on altered position are consistent with recent reports that voluntary efforts can alter the perceived position of a phantom wrist during ischaemic block of the arm and can also influence perceived wrist position during local curarisation (Gandevia et al. 2006; Smith et al. 2009). In contrast, previous studies, in which subjects attempted to move paralysed (McCloskey & Torda, 1975) or paralysed and anaesthetised limbs or digits (Goodwin et al. 1972), reported that no limb movements were perceived. In these studies, efforts were brief and may not have allowed time to generate illusory movements. In the current study, brief efforts lasting 1 s produced perceptions of wrist movement in all subjects, but not on all attempts. In addition, the current study promoted uncertainty about possible wrist movements through prior exposure to trials in which efforts against unexpected loads produced movements that varied from 0 to 70 deg. The role of motor commands in the ability to track velocities has also been examined by altering the relationship between muscle recruitment and arm velocity through fatigue of the muscle (Allen & Proske, 2006). Passive movements of one arm were tracked by voluntary movements of the other. This ability was altered by muscle vibration but not by fatigue, and it was concluded that motor commands did not contribute. However, the task was not ideal as subjects controlled the speed of their arm with eccentric voluntary contractions. Thus, the speed was determined by how quickly their muscles were relaxed voluntarily. Furthermore, comparison of active and passive movements did not allow comparison of motor command between arms.

When subjects made efforts during the ischaemic block, the perception that the wrist moved in the direction of the effort was robust. On average, the size of the movements, that is the change in position with effort, graded with both the level and the duration of voluntary effort. Figure 7A shows the data for position from both experiments plotted together (with the data from flexion and extension pooled). However, the grading was not a one-to-one ratio. An effort for five times as long (1 s efforts vs. 5 s efforts) did not displace the phantom five times as far. In addition, multiplying the level of effort by ten (5% line vs. 50% line) only increased the size of the phantom movement by 2.5–3 times. The curves are plotted as linear, because there are only two data points, but below 1 s the data are non-linear, as the curves must pass through the origin. Perceived velocity of the phantom also increased with an increased level of effort (Fig. 7B) and graded within each experiment. The poor grading between experiments is probably due to two different subject groups being used. Perhaps the subjects in the second experiment paid closer attention to the velocity of the phantom movements they perceived. Again, a tenfold increase in effort (5 to 50% effort) did not increase perceived speed tenfold, rather the increase was about twofold. In contrast to movement extent, perceived velocity did not increase with increased duration of effort.

Figure 7. Comparison of the changes in perceived position and velocity from both experiments.

Figure 7

Open in a new tab

Data for flexion and extension trials have been pooled and the data from one subject who reported perceiving instantaneous changes in position without movements were excluded. A, the mean changes in perceived wrist position are plotted against the duration of effort for the four levels of effort (5, 20, 40 and 50% maximum). The size of perceived movement of the phantom wrist scales with the level of effort and with the duration of effort. However, the relation is not 1:1. Thus, a tenfold increase in voluntary effort does not produce a movement that is ten times bigger, nor does an effort lasting five times longer produce a movement five times bigger. The shape of the curve below 1 s is unknown (dotted lines) but must approach the origin. B, the perceived wrist velocity plotted against duration of effort for the four levels of effort. The perceived wrist velocity scales with the level of effort within experiments. The velocity scales inversely with duration of effort. As for position, neither relation is 1:1 and the dotted parts of the curve represent unknown data that must intersect the origin.

Although all subjects perceived movements of their phantom wrist in the direction of effort, they did not have identical experiences. These differences were revealed both by the movements indicated with the pointer and by subjects’ reports after the block. While some subjects perceived uniform movement throughout the efforts, others reported movements at the start of the effort followed by a pause before the return movement on relaxation. Thus, it was not the duration of the effort that was misperceived but rather that movement was not perceived throughout the effort. The cessation of movement during an effort was most obvious with the longest efforts but also occurred with the brief efforts of 1 s duration. This suggests that the phantom wrist movements are not simply reaching the end of their range of motion for longer efforts. There was not a set time limit to the perceived movement (e.g. the first 0.5 s of effort), because longer duration efforts produced larger movements. In addition, some subjects reported that there was a delay between when they started the effort and when the phantom began to move, while others reported that the phantom movements started immediately. Those subjects who reported a delay between the beginning of the effort and the phantom movement described a feeling of nothing happening initially, and then their hand began to move.

We argue that subjects perceive movements of the phantom wrist due to a contribution of motor command signals to the sense of limb movement. However, other factors could contribute. As mentioned in the Introduction, a signal of motor command in this context refers to a central signal that is monotonically related to the motoneuronal output. Studies in experimental psychology have considered other perceptions such as will, agency and intention that might also be related to motor command (Jeannerod, 1999). For example, awareness of the intention to act is believed to begin at a time after the selection of a movement but during the preparatory phase and to arise in areas of the brain that are ‘higher’ than the motor cortex (see Haggard, 2005). However, it seems unlikely that awareness of the intention to act would be sufficient to generate the perceptions of wrist movement in the current experiment. Subjects were asked to make different levels of effort with their wrist flexors and extensors and so they did not plan movements. Thus, awareness of intention might signal timing and level of effort but should not give information on the resultant unplanned wrist movement. Visual input can also contribute to perceptions of limb movement. Amputees experience phantom movements of their amputated limb when a mirror is used to duplicate a movement of the intact limb in the space occupied by the phantom (Ramachandran & Rogers-Ramachandran, 2000). Our study specifically excluded vision of the ischaemically blocked arm. Although subjects may have visualised movements of their phantom wrist when asked to make voluntary efforts, this is unlikely as subjects were not given any instruction in this regard and were never told to expect their voluntary efforts to produce phantom movements and they did not report visualisation of any movements. Finally, activation of proximal muscles has been linked to the perception of movement of phantom limbs. In the current study, once the ischaemic block was complete, the arm distal to the cuff was paralysed and anaesthetised. However, activity in muscles proximal to the cuff was not excluded and proximal muscles were probably performing their normal postural roles during voluntary efforts with the phantom. Reilly et al. (2006) have shown in amputees that proximal muscles activate differentially for different phantom movements and that ischaemia of muscles in the stump diminishes perceived movement. The interpretation of this and its comparison to acute experimental phantoms is complicated by the fact that amputees probably have undergone cortical remapping (Kew et al. 1994; Schieber & Deuel, 1997; Mercier et al. 2006). It is difficult to see why, during an acute ischaemic block in otherwise normal subjects, activation of a proximal muscle should be responsible for the perception of movement of a distal joint. None of the muscles that were intact or partially intact in our experiments are involved in control of the wrist and afferents activated by proximal contractions would be expected to signal unambiguously forces, changes in position or movements at proximal joints. While it is possible that information from proximal muscles may contribute to the perceptions of phantom movement, it is more likely that motor command signals related to the wrist are the dominant contributor to these perceptions of movement.

What information can be derived from a motor command signal? First, it can signal both the start and end of a voluntary action (McCloskey et al. 1983a). Second, the direction of the voluntary movement can be determined as this depends upon which muscles are activated (Gandevia & Rothwell, 1987; Gandevia et al. 1990). Finally, the motor command indicates how much of the muscle is activated, that is, how much drive is sent to the muscle to perform the action. Further information about the actions of motor command on the unloaded limbs could be learned from experience. For example, with the assistance of afferent feedback, experience could teach the brain how much force is produced by a muscle for a given motor command. Then, provided the muscle is not fatigued, the brain could control voluntary force based only on the level of motor command. Furthermore, experience could provide information about the weight and inertia of the unloaded limbs and the effects of gravity. This information, along with the relation between muscle force and motor command output, could be stored in a model of the kinematics of the unloaded body. This would indicate how the unloaded limbs behave and move with motor commands, including how fast they move for a given command. Such a model could control the unloaded body using only information derived from motor command signals, although perhaps not perfectly (Balslev et al. 2007). The remaining information required to control body are the properties of any external loads and the state of the muscles, including the level of muscle fatigue. This information can be derived with the use of afferent information (e.g. Gandevia & McCloskey, 1978).

In the present study, afferent information from below the cuff was blocked from reaching the spinal cord, and the motor commands were blocked from reaching the agonist muscles. However, any central model about the effect of motor commands on the limbs should remain. When subjects attempted voluntary efforts with their paralysed and anaesthetised wrist muscles, they perceived both movement and displacement of their phantom wrist. This suggests that the motor command signals were still processed by the model of body kinematics and that the output of the model was the perception that the phantom hand moved with a velocity and became displaced from its starting position. Why then did the phantom return to its starting position after subjects stopped making efforts? For the phantom wrist, afferent signals do not change with voluntary efforts or passive movement but the brain presumably still interprets this unchanging information. With the cessation of the motor command signal, the phantom may revert to a hand position which is the brain's interpretation of the unchanging afferent information. This raises the question of why the phantom moves at all if the afferents continuously signal a static position. Presumably the low unchanging firing rates of the afferent signals are given little weight by the model, so that the motor command information dominates during the voluntary efforts, but afterwards the low-level unchanging afferent signal is all that is left.

Why do subjects not report that the limb moved as if it were unloaded? They describe a feeling of pushing through a viscous substance. It is possible that perceptions were influenced by the control experiment which involved pushing isometrically or against a viscous load. However, a subject volunteered a similar report of pushing through a viscous substance in a previous experiment in which no ‘training’ was given (Gandevia et al. 2006). This perception may relate to the motor commands signalling movement and a change of position, while the afferent signals were unchanging. Perhaps these competing signals evoke the sensation of a slower more difficult movement.

In summary, we present two novel findings. First, subjects can perceive the velocity of a phantom limb during voluntary efforts and the velocity of these movements depends on the level of effort, with larger efforts generating larger velocities. Second, the extent of movements of a phantom limb is larger if subjects make voluntary efforts for longer. These results show that motor command signals have an additional role in proprioception: as well as their recently established role in the sense of limb position, they can generate a sense of limb movement.

Acknowledgments

This work was supported by the National Health and Medical Research Council of Australia. We are grateful to Dr Lorimer Moseley and Professor Uwe Proske for their comments on the manuscript.

Author contributions

All authors contributed to all aspects of the study and all approved the final version. All experiments were performed at the Prince of Wales Medical Research Institute in Sydney, Australia.

References

  1. Allen TJ, Proske U. Effect of muscle fatigue on the sense of limb position and movement. Exp Brain Res. 2006;170:30–38. doi: 10.1007/s00221-005-0174-z. [DOI] [PubMed] [Google Scholar]
  2. Balslev D, Cole J, Miall RC. Proprioception contributes to the sense of agency during visual observation of hand movements: evidence from temporal judgments of action. J Cognit Neurosci. 2007;19:1535–1541. doi: 10.1162/jocn.2007.19.9.1535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Donaldson IM. The functions of the proprioceptors of the eye muscles. Philos Trans R Soc Lond B Biol Sci. 2000;355:1685–1754. doi: 10.1098/rstb.2000.0732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Gandevia SC. Roles for perceived voluntary motor commands in motor control. Trends Neurosci. 1987;10:81–85. [Google Scholar]
  5. Gandevia SC, Killian K, McKenzie DK, Crawford M, Allen GM, Gorman RB, Hales JP. Respiratory sensations, cardiovascular control, kinaesthesia and transcranial stimulation during paralysis in humans. J Physiol. 1993;470:85–107. doi: 10.1113/jphysiol.1993.sp019849. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Gandevia SC, McCloskey DI. Sensations of heaviness. Brain. 1977;100:345–354. doi: 10.1093/brain/100.2.345. [DOI] [PubMed] [Google Scholar]
  7. Gandevia SC, McCloskey DI. Interpretation of perceived motor command by reference to afferent signals. J Physiol. 1978;283:493–499. [PMC free article] [PubMed] [Google Scholar]
  8. Gandevia SC, Macefield G, Burke D, McKenzie DK. Voluntary activation of human motor axons in the absence of muscle afferent feedback. The control of the deafferented hand. Brain. 1990;113:1563–1581. doi: 10.1093/brain/113.5.1563. [DOI] [PubMed] [Google Scholar]
  9. Gandevia SC, Rothwell JC. Knowledge of motor commands and the recruitment of human motoneurons. Brain. 1987;110:1117–1130. doi: 10.1093/brain/110.5.1117. [DOI] [PubMed] [Google Scholar]
  10. Gandevia SC, Smith JL, Crawford M, Proske U, Taylor JL. Motor commands contribute to human position sense. J Physiol. 2006;571:703–710. doi: 10.1113/jphysiol.2005.103093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Goodwin GM, McCloskey DI, Matthews PBC. The contribution of muscle afferents to kinaesthesia shown by vibration induced illusion of movement and by the effects of paralysing joint afferents. Brain. 1972;95:705–748. doi: 10.1093/brain/95.4.705. [DOI] [PubMed] [Google Scholar]
  12. Gritsenko V, Krouchev NI, Kalaska JF. Afferent input, efference copy, signal noise, and biases in perception of joint angle during active versus passive elbow movements. J Neurophysiol. 2007;98:1140–1154. doi: 10.1152/jn.00162.2007. [DOI] [PubMed] [Google Scholar]
  13. Haggard P. Conscious intention and motor cognition. Trends Cogn Sci. 2005;9:290–295. doi: 10.1016/j.tics.2005.04.012. [DOI] [PubMed] [Google Scholar]
  14. Henderson WR, Smyth GE. Phantom limbs. J Neurol Neurosurg Psychiatry. 1948;11:88–112. doi: 10.1136/jnnp.11.2.88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hobbs SF, Gandevia SC. Cardiovascular responses and the sense of effort during attempts to contract paralysed muscles: Role of the spinal cord. Neurosci Lett. 1985;57:85–90. doi: 10.1016/0304-3940(85)90044-8. [DOI] [PubMed] [Google Scholar]
  16. Jeannerod M. The 25th Bartlett Lecture. To act or not to act: perspectives on the representation of actions. Q J Exp Psychol A. 1999;52:1–29. doi: 10.1080/713755803. [DOI] [PubMed] [Google Scholar]
  17. Kew JJ, Ridding MC, Rothwell JC, Passingham RE, Leigh PN, Sooriakumaran S, Frackowiak RS, Brooks DJ. Reorganization of cortical blood flow and transcranial magnetic stimulation maps in human subjects after upper limb amputation. J Neurophysiol. 1994;72:2517–2524. doi: 10.1152/jn.1994.72.5.2517. [DOI] [PubMed] [Google Scholar]
  18. McCloskey DI. Kinaesthetic sensibility. Physiol Rev. 1978;58:763–820. doi: 10.1152/physrev.1978.58.4.763. [DOI] [PubMed] [Google Scholar]
  19. McCloskey DI, Colebatch JG, Potter EK, Burke D. Judgements about onset of rapid voluntary movements in man. J Neurophysiol. 1983a;49:851–863. doi: 10.1152/jn.1983.49.4.851. [DOI] [PubMed] [Google Scholar]
  20. McCloskey DI, Gandevia S, Potter EK, Colebatch JG. Muscle sense and effort: motor commands and judgments about muscular contractions. Adv Neurol. 1983b;39:151–167. [PubMed] [Google Scholar]
  21. McCloskey DI, Torda TA. Corollary motor discharges and kinaesthesia. Brain Res. 1975;100:467–470. doi: 10.1016/0006-8993(75)90503-x. [DOI] [PubMed] [Google Scholar]
  22. Matthews PBC. Where does Sherrington's ‘muscular sense’ originate? Muscles, joints, corollary discharges. Annu Rev Neurosci. 1982:189–218. doi: 10.1146/annurev.ne.05.030182.001201. [DOI] [PubMed] [Google Scholar]
  23. Melzack R, Bromage PR. Experimental phantom limbs. Exp Neurol. 1973;39:261–269. doi: 10.1016/0014-4886(73)90228-8. [DOI] [PubMed] [Google Scholar]
  24. Mercier C, Reilly KT, Vargas CD, Aballea A, Sirigu A. Mapping phantom movement representations in the motor cortex of amputees. Brain. 2006;129:2202–2210. doi: 10.1093/brain/awl180. [DOI] [PubMed] [Google Scholar]
  25. Ramachandran VS, Rogers-Ramachandran D. Phantom limbs and neural plasticity. Arch Neurol. 2000;57:317–320. doi: 10.1001/archneur.57.3.317. [DOI] [PubMed] [Google Scholar]
  26. Reilly KT, Mercier C, Schieber MH, Sirigu A. Persistent hand motor commands in the amputees’ brain. Brain. 2006;129:2211–2223. doi: 10.1093/brain/awl154. [DOI] [PubMed] [Google Scholar]
  27. Schieber MH, Deuel RK. Primary motor cortex reorganization in a long-term monkey amputee. Somatosens Motor Res. 1997;14:157–167. doi: 10.1080/08990229771024. [DOI] [PubMed] [Google Scholar]
  28. Smith JL, Crawford M, Proske U, Taylor JL, Gandevia SC. Signals of motor command bias joint position sense in the presence of feedback from proprioceptors. J Appl Physiol. 2009;106:950–958. doi: 10.1152/japplphysiol.91365.2008. [DOI] [PubMed] [Google Scholar]
  29. Sperry RW. Neural basis of the spontaneous optokinetic response produced by visual inversion. J Comp Physiol Psychol. 1950;43:482–489. doi: 10.1037/h0055479. [DOI] [PubMed] [Google Scholar]
  30. Stevens J. The corollary discharge: is it a sense of position or a sense of space? Behav Brain Sci. 1978;1:163–165. [Google Scholar]
  31. von Helmholtz H. Helmholtz's Treatise on Physiological Optics. Vol. 3. Menasha, Wisconsin: Optical Society of America; 1867. [Google Scholar]
  32. von Holst E. Relations between the central nervous system and the peripheral organs. Br J Anim Beh. 1954;2:89–94. [Google Scholar]
  33. Walsh LD, Allen TJ, Gandevia SC, Proske U. The effect of eccentric exercise on position sense at the human forearm in different postures. J Appl Physiol. 2006;100:1109–1116. doi: 10.1152/japplphysiol.01303.2005. [DOI] [PubMed] [Google Scholar]
  34. Walsh LD, Hesse CW, Morgan DL, Proske U. Human forearm position sense after fatigue of elbow flexor muscles. J Physiol. 2004;558:705–715. doi: 10.1113/jphysiol.2004.062703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Walsh LD, Inui N, Taylor JL, Gandevia SC. Australian Neuroscience Society 29th Annual Meeting. Canberra, Australia: 2009a. Acute changes in the body schema: development of a phantom hand. [Google Scholar]
  36. Walsh LD, Smith JL, Gandevia SC, Taylor JL. The combined effect of muscle contraction history and motor commands on human position sense. Exp Brain Res. 2009b;195:603–610. doi: 10.1007/s00221-009-1832-3. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES