OUP user menu

Abnormalities of sensorimotor integration in focal dystonia
A transcranial magnetic stimulation study

Giovanni Abbruzzese, Roberta Marchese, Alessandro Buccolieri, Bruno Gasparetto, Carlo Trompetto
DOI: http://dx.doi.org/10.1093/brain/124.3.537 537-545 First published online: 1 March 2001

Summary

It has been postulated that sensorimotor integration is abnormal in dystonia. We investigated changes in motor cortical excitability induced by peripheral stimulation in patients with focal hand dystonia (12 patients with hand cramps) and with cervical dystonia (nine with spasmodic torticollis) compared with 16 age-matched normal controls. Motor evoked potentials (MEP) to focal (figure-of-eight coil) transcranial magnetic stimulation of the hand area were recorded from the right abductor pollicis brevis (APB), first dorsal interosseus (FDI), flexor carpi radialis and extensor carpi radialis muscles. Changes of test MEP size following conditioning stimulation of the right median nerve (or of the index finger) at conditioning-test (C-T) intervals of 50, 200, 600 and 1000 ms were analysed. Peripheral stimulation significantly reduced test MEP size in the APB and FDI muscles of normal control and spasmodic torticollis patients. The inhibitory effect was larger upon median nerve stimulation and reached a maximum at the C-T interval of 200 ms. On the contrary, hand cramp patients showed a significant facilitation of test MEP size. This study suggests that MEP suppression following peripheral stimulation is defective in patients with focal hand dystonia. Central processing of sensory input is abnormal in dystonia and may contribute to increased motor cortical excitability.

  • cortical excitability
  • focal dystonia
  • peripheral stimulation
  • sensorimotor integration
  • transcranial magnetic stimulation
  • ANOVA = analysis of variance
  • APB = abductor pollicis brevis
  • C-T = conditioning-test
  • ECR = extensor carpi radialis
  • FCR = flexor carpi radialis
  • FDI = first dorsal interosseus
  • MEP = motor evoked potential
  • SEP = somatosensory evoked potential
  • TMS = transcranial magnetic stimulation

Introduction

Idiopathic focal dystonia is a disorder characterized by prolonged involuntary muscle contractions, causing abnormal postures of a single body part (blepharospasm, spasmodic torticollis, writer's cramp). The pathophysiology of primary focal dystonia is still uncertain. Reduced spinal cord and brainstem inhibition is likely to contribute to dystonia, but cannot be considered the only cause (Berardelli et al., 1998). On the other hand, studies with magnetic brain stimulation have suggested the occurrence of an increased gain of the input–output relationship in the motor cortex of dystonic subjects (Berardelli et al., 1998).

Although dystonia is regarded as a movement disorder and there is no primary deficit of sensation, several sensory phenomena occur in dystonia, suggesting that sensory processing may be impaired. The possible role of sensory input is particularly evident in focal dystonia. `Sensory tricks' can relieve or improve dystonic spasms, while manipulation of sensory inputs can trigger (muscle vibration) or relieve (muscle afferent block) dystonia (Kaji et al., 1995; Yoshida et al., 1998). In connection with this, it has been postulated that frequent use of a body part or trauma to a body part, often preceding dystonia (Defazio et al., 1998), may induce dystonic movements via a sensory mechanism (Hallett, 1995).

The activity of motor cortical areas can be investigated by recording motor evoked potentials (MEPs) to transcranial magnetic stimulation (TMS) of the brain. The excitability of the corticomotoneuronal system is modulated by peripheral afferent stimulation. In normal subjects, both facilitation (Deuschl et al., 1991; Rossini et al., 1991; Deletis et al., 1992; Komori et al., 1992; Maertens de Noordhout et al., 1992; Hirashima and Yokota, 1997) and inhibition (Mariorenzi et al., 1991; Maertens de Noordhout et al., 1992; Clouston et al., 1995; Inghilleri et al., 1995) of MEPs have been shown to take place in the first 100 ms following peripheral stimulation, while a decreased motor cortex excitability has been reported at longer intervals, >200 ms (Chen et al., 1999).

To investigate possible abnormalities of sensorimotor integration mechanisms in dystonia, we studied changes of motor cortical excitability (as tested with TMS) following a conditioning peripheral nerve stimulation in patients with focal dystonia compared with normal controls.

Material and methods

Subjects

Two groups of patients with focal dystonia were studied: 12 patients with hand cramps and nine with spasmodic torticollis. The controls were 16 age-matched healthy subjects with normal neurological examination and no history of neurological disorders. Informed consent was obtained from all subjects and the local ethics committee approved the experimental protocol. All patients and controls were right handed. Patients underwent a complete neurological examination, which revealed no additional abnormality. The results of biochemical and neuroimaging studies were normal so that dystonia was diagnosed to be idiopathic in all the patients (Fahn et al., 1998). The main clinical features of the patients are reported in Table 1.

View this table:
Table 1

Patient characteristics

SexAge (years)DiagnosisAssociated symptomsDuration (years)Treatment
F = female; M = male; R = right, L = left; AC = anticholinergics; BTX = botulinum toxin. *Patients under treatment with botulinum toxin were examined 3 months after the last injection.
Hand cramps
1M49Writer's cramp (R)10
2M28Musician's cramp (R) 4
3F54Writer's cramp (R) 7AC
4M27Writer's cramp (R)Spasmodic dysphonia 3AC
5M26Musician's cramp (R) 3AC–BTX*
6M61Upper limb dystonia (R)Blepharospasm27AC
7M43Writer's cramp (R) 5AC
8M49Writer's cramp (R) 1AC
9F50Writer's cramp (R) 2AC
10F37Writer's cramp (R) 8BTX*
11F34Writer's cramp (R) 2
12M62Writer's cramp (R) 6AC
Mean ± SD43.3 ± 12.8
Spasmodic torticollis
1M53Laterocollis (L)Head tremor15AC
2M59RetrocollisHead tremor 8AC
3M78Rotatocollis (R)Head tremor 5BTX*
4F72Rotatocollis (R) 2BTX*
5M74Rotatocollis (R)Blepharospasm 3BTX*
6M40Rotatocollis (R) 4AC
7M59Rotatocollis (L)Head tremor10AC
8M43Rotatocollis (L)Head tremor 9AC
9F33AnterocollisBlepharospasm 1
Mean ± SD56.8 ± 15.9
Controls6 F/10 M
Mean ± SD52.0 ± 19.0

EMG recordings

Subjects were seated in a comfortable reclining chair and EMG activity was recorded from the right abductor pollicis brevis (APB), first dorsal interosseus (FDI), flexor carpi radialis (FCR) and extensor carpi radialis (ECR) muscles using surface (Ag–AgCl) electrodes with a tendon-belly arrangement. The EMG signals were amplified (×2000), filtered (bandwidth 50–2000 Hz, –3 dB) using an EMG amplifier Type 16C01 (Dantec, Denmark), displayed on a screen, captured on a laboratory computer and converted by an analogue-to-digital interface at a sampling rate of 5 kHz for further analysis. Each recorded epoch lasted 800–1500 ms, of which 100 ms preceded the stimulus.

TMS

TMS was performed with a figure-of-eight coil (9.5 cm external diameter) powered by a Magstim 200 stimulator (Magstim Company Ltd, Whitland, UK). The coil was placed over the optimal position of the left scalp for evoking MEPs from the right APB muscle, with the handle of the coil pointing backwards. Stimulus intensity was adjusted to produce MEPs of ~1 mV peak-to-peak amplitude in the relaxed APB muscle. Stimulus intensity was expressed as a percentage of the maximal stimulator output. The subjects were relaxed throughout the experiments and muscle relaxation was checked by audio-visual EMG monitoring. The resting trials, in which background EMG activity was present in the pre-stimulus time period, were rejected off-line.

Experimental procedures

Experiment I

The effect of a conditioning stimulation of the right median nerve on MEP amplitudes was investigated in all subjects. The median nerve was stimulated at the wrist with standard surface electrodes (cathode proximal), using 0.2 ms square wave electrical pulses delivered by a Grass S88 stimulator (Grass Instruments Co., Quincy, Mass., USA) at an intensity just above the thumb twitch threshold. Median nerve stimulation was timed to precede TMS at four conditioning-test (C-T) intervals: 50, 200, 600 and 1000 ms. At least eight unconditioned (TMS alone) and eight conditioned trials (median nerve stimulation + TMS) were collected for each subject for each C-T interval. A pseudo-random order of presentation of the different intervals was used, and conditioned and unconditioned stimuli were randomly intermixed and given every 10 s.

Experiment II

The conditioning effect of a pure cutaneous stimulation on MEP amplitudes was tested by stimulating the right index finger with ring electrodes (cathode distal to the metacarpal-phalangeal joint and anode distal to the proximal interphalangeal joint). Stimuli consisted of 0.2 ms square wave electrical pulses with an intensity set at three times the sensory threshold (i.e. the lowest stimulus intensity at which the subjects reported sensation of finger stimulation). Six normal controls (five males and one female) and five patients with focal hand dystonia (Cases 3, 7, 8, 9 and 12 with hand cramp) participated in this experiment. The same procedure and C-T intervals as in Experiment I were used.

Experiment III

To test the effects of median nerve stimulation on spinal excitability we recorded the F-wave in 10 normal subjects and in three patients with focal hand dystonia (Cases 6, 9 and 12), who also participated in Experiment I. F-waves were elicited in the right FDI muscle by supramaximal stimulation of the ulnar nerve at the wrist. Two conditions were tested: ulnar nerve stimulation alone and conditioning median nerve stimulation followed by ulnar nerve stimulation at a C-T interval of 200 ms. Ten trials for each condition were presented in a random order.

Somatosensory evoked potential (SEP) recording

To test possible changes in the excitability of primary somatosensory cortical areas, we recorded SEPs after stimulation of the right median nerve at the wrist (0.2 ms square waves delivered by a Grass S88 stimulator with a frequency of 0.5 Hz and an intensity just above the thumb twitch threshold). SEPs were recorded from the Erb's point, parietal (3 cm behind the vertex and 7 cm from the midline), and frontal (4 cm in front of the vertex and 5 cm from the midline) electrodes with an ipsilateral earlobe reference (Abbruzzese et al., 1997). At least three series of 256 responses were averaged over a 100 ms period (bandpass, 16–3000 Hz; sampling rate, 10 kHz per channel), with automatic rejection of samples with excessive EMG interference, and stored on a computer disk for further analysis. Latency and peak-to-peak amplitude of the Erb's potential (P9 component), parietal N20/P25 and frontal P22/N30 complexes were measured. SEP recordings were performed in 10 normal controls (seven males and three females; mean age (± standard deviation): 47.6 ± 16.1 years; height: 169 ± 7.9 cm), eight hand cramp patients (mean age: 41.4 ± 12.2 years; height: 169 ± 8.6 cm) and six spasmodic torticollis patients (mean age, 56.3 ± 10.6 years; height, 169.6 ± 8.9 cm).

Data analysis

The size (area of the rectified EMG signal, mV.ms) of single MEPs and the peak-to-peak F-wave amplitude were measured off-line. A preliminary analysis of the effect of conditioning stimulation of the median nerve (Experiment I) and of the index finger (Experiment II) on absolute MEP sizes was performed in normal controls by a single factor analysis of variance (ANOVA) for repeated measures (repeated measures ANOVA). Dunnett's multiple comparison test was used to compare values obtained at the different C-T intervals with control (unconditioned) values.

In order to reduce interindividual variability, the mean size of the conditioned MEPs at each C-T interval was expressed as a percentage of the mean size of the unconditioned MEPs (see Kujirai et al., 1993; Abbruzzese et al., 1999) and the results reported as mean ± standard errors. For each muscle, the time course of conditioning median nerve stimulation (at the various C-T intervals) was tested in the different groups (normal controls, hand cramps and spasmodic torticollis patients) with two-way ANOVA (first variable: groups; second variable: C-T intervals). Bonferroni's multiple comparison test was used for post tests. Within subjects comparison of median nerve and index finger conditioning stimulation was performed (in normal controls and hand cramp patients) by means of two-way ANOVA (first variable: stimulation type; second variable: C-T intervals). The effects of median nerve stimulation on F-wave amplitudes were tested with the paired t-test. SEP data were analysed with one-way ANOVA. Statistical analysis was performed by means of GraphPad Prism 3.0 software.

Results

Normal controls

Conditioning stimulation of the right median nerve significantly modified corticospinal excitability as measured by MEP size. Repeated measures ANOVA showed a significant reduction of the test MEP size in the APB [F(4,15) = 3.41, P = 0.014], FDI [F(4,15) = 3.18, P = 0.019] and FCR muscles [F(4,15) = 2.77, P = 0.034], but not in the ECR muscle [F(4,15) = 0.41, P = 0.797]. The reduction in test MEP amplitude was most evident for the APB muscle and reached the maximum at the C-T interval of 200 ms (APB: 67.1%, P < 0.01; FDI: 82.7%, P < 0.01; FCR: 82.5%, P < 0.05; ECR: 91.3%, not significant) The time course of corticospinal excitability changes, following right median nerve stimulation, in the normal control group is shown in Fig. 1.

Fig. 1

Time course of corticospinal excitability changes, following right median nerve stimulation at the wrist, in the normal control group. Each point corresponds to the mean (+ standard error) size of the conditioned MEP expressed as a percentage of the mean size of the unconditioned MEP. C-T intervals are reported in the abscissa. Repeated measures ANOVA showed a significant inhibition of the test MEP in the APB, FDI and FCR muscles, but not in the ECR muscle.

Conditioning stimulation of the right index finger also affected the corticospinal excitability, but the effects were considerably less evident than those with median nerve stimulation. The size of the test MEP was significantly reduced in the APB [repeated measures ANOVA: F(4,5) = 5.25, P = 0.0046], but not in the FDI [F(4,5) = 1.79, P = 0.171], FCR [F(4,5) = 1.52, P = 0.233] and ECR [F(4,5) = 0.41, P = 0.797]. However, within subjects comparison (two-way ANOVA in six normal controls) of the effects of conditioning median nerve and index finger stimulation on the test MEP size of APB and FDI muscles did not show significant differences [APB: stimulation type, F(1,40) = 0.74, P = 0.39; C-T intervals, F(3,40) = 0.49, P = 0.69; no interaction. FDI: stimulation type, F(1,40) = 0.15, P = 0.70; C-T intervals, F(3,40) = 1.12, P = 0.35; no interaction].

Conditioning stimulation of the median nerve produced little modification of the F-wave amplitudes at the C-T interval of 200 ms. No change was observed in five subjects, while a trend towards inhibition (three subjects) or facilitation (two subjects) was present in the others. Overall, the mean F-wave amplitude was not significantly affected (unconditioned: mean 0.31 ± 0.22 mV; conditioned: 0.27 ± 0.17 mV; paired t-test: P = 0.496, not significant).

Dystonic patients

The mean intensity of TMS used to evoke test MEPs in the target muscles was similar in normal controls and in the two groups of dystonic patients [normal controls: mean 84.0 ± 15.0%; hand cramp patients: 74.2 ± 14.0%; spasmodic torticollis patients: 87.0±15.0%; ANOVA: F(2,34) = 2.34, P = 0.119]. The mean size of the test MEPs for all the muscles was not significantly different among the three groups (P > 0.05).

The effect of conditioning stimulation of the median nerve varied significantly among groups for the APB [two-way ANOVA: groups, F(2,133) = 5.98, P = 0.0032; C-T intervals, F(3,133) = 1.25, P = 0.139; no interaction) and FDI muscles [groups, F(2,133) = 10.54, P < 0.0001; C-T intervals, F(3,133) = 0.72, P = 0.399; no interaction]. No significant difference was observed for the FCR [groups, F(2,133) = 3.0, P = 0.053] and ECR muscles [groups, F(2,133) = 1.45, P = 0.239]. Conditioning stimulation of the median nerve induced an inhibition of the test MEP in the APB and FDI muscles of normal controls and of patients with spasmodic torticollis, but not of those in patients with hand cramps, who showed a significant facilitation of the test MEP (Fig. 2). An example of the effect of conditioning stimulation of the median nerve (C-T interval 200 ms) in three representative subjects is shown in Fig. 3.

Fig. 2

Effect of conditioning stimulation of the median nerve on the test MEP amplitude in normal controls (NC), in patients with cervical dystonia (ST) and in patients with hand dystonia (HC). Each point corresponds to the mean (+ standard error) size of the conditioned MEP, expressed as a percentage of the mean size of the unconditioned MEP. The C-T intervals are reported in the abscissa. The test MEPs in the APB and FDI muscles are inhibited by conditioning stimulation of the median nerve at the wrist in NC and ST patients, while they are facilitated in the HC patients. No significant modification was observed for the test MEPs in the FCR and ECR muscles.

Fig. 3

Effect of conditioning stimulation of the median nerve in three representative subjects. Right traces: test MEPs recorded from the APB muscle; left traces: MEPs following conditioning stimulation (C-T interval 200 ms) of the median nerve at the wrist. Each trace is the average of eight trials. MEPs are inhibited by conditioning stimulation in the normal control (top) and in the patient with cervical dystonia (bottom), while they are significantly facilitated in the patient with focal hand dystonia (centre).

Conditioning stimulation of the right index finger did not affect corticospinal excitability in the five hand cramp patients examined. In particular, single factor, repeated measures ANOVA showed that the size of the test MEPs was not significantly modified in the APB [F(84,4) = 0.596, P = 0.67] or the FDI [F(4,4) = 1.93, P = 0.154] muscles. Within-subjects comparison of the effects of conditioning stimulation of the median nerve and index finger did not show significant differences [APB: stimulation type, F(1,32) = 0.38, P = 0.54; C-T intervals, F(3,32) = 1.11, P = 0.35; no interaction. FDI: stimulation type, F(1,32) = 0.07, P = 0.79; C-T intervals, F(3,32) = 2.34, P = 0.09].

No apparent change of the F-wave amplitude was observed upon conditioning stimulation of the median nerve in the three hand cramp patients examined. Overall, the mean F-wave amplitude was not significantly affected (unconditioned: mean 0.36 ± 0.42 mV; conditioned: 0.37 ± 0.11 mV; paired t-test: P = 0.781, not significant).

SEP recordings

Both latency and peak-to-peak amplitude of SEP early components were not significantly different in dystonic patients and controls (Table 2) [ANOVA: Erb's P9 amplitude, F(2,21) = 0.338, P = 0.705; parietal N20/P25 amplitude, F(2,21) = 0.537, P = 0.570; frontal P22/N30 amplitude, F(2,21) = 0.06, P = 0.938].

View this table:
Table 2

Latency and amplitude of SEP components in patients and controls

Normal controls (n = 10)Hand cramps (n = 8)Spasmodic torticollis (n = 6)
Results shown are expressed as mean ± SD. ANOVA: P > 0.05, not significant.
Latency (ms)
Erb P98.73 ± 0.61 8.95 ± 0.7 9.07 ± 0.76
Parietal N2020.3 ± 1.54 19.7 ± 0.7219.83 ± 1.7
Frontal N3029.4 ± 0.8230.03 ± 2.8528.31 ± 2.5
Amplitude (V)
Erb P913.5 ± 6.0 10.7 ± 7.7 12.5 ± 7.9
Parietal N20/P254.05 ± 1.3 3.27 ± 0.5 4.42 ± 3.4
Frontal P22/N304.25 ± 2.1 4.08 ± 2.8 4.55 ± 2.6

Discussion

In line with Chen et al. (1999), we have shown that in normal subjects, motor cortical excitability (as tested with TMS) is significantly reduced to 200–1000 ms after conditioning stimulation of the contralateral median nerve. Such inhibition was also effective in the case of muscles not innervated by the median nerve (i.e. FDI muscle). A similar inhibitory effect was also present following digit stimulation, but, at variance with the data reported by Chen et al. (1999), we observed a more extensive inhibitory effect upon stimulation of the median nerve (activating both muscle and cutaneous afferents) than upon digit stimulation (activating mainly cutaneous afferents).

Since H-reflexes cannot be elicited from relaxed hand muscles, we used the F-wave to test possible changes of spinal excitability. The F-wave has been shown to be sensitive to changes in spinal excitability induced by TMS (Mercuri et al., 1996) or peripheral stimulation (Walk and Fisher, 1993). Although we cannot rule out subtle changes, not assessed by this method, the lack of modification of the F-wave amplitudes (at the C-T interval of 200 ms) suggests that MEP inhibition induced by conditioning stimulation of the median nerve was not due to changes of spinal motoneuron excitability, but was likely to reflect changes at the cortical level.

The modulation of motor cortical excitability induced by peripheral afferent activation was evidently impaired in the patients with focal dystonia (`occupational cramps'), affecting the stimulated hand. In these patients, MEPs of the intrinsic hand muscles were not suppressed by conditioning stimulation; indeed an opposite trend of facilitation was consistently observed. In contrast, patients with cervical dystonia (spasmodic torticollis) did not behave differently from normal controls.

The results of this study suggest, therefore, that suppressive mechanisms induced by median nerve stimulation are defective in patients with focal hand dystonia. This finding is in agreement with several studies suggesting that intracortical inhibition is reduced in primary dystonia (Ridding et al., 1995; Filipovic et al., 1997; Rona et al., 1998; Siebner et al., 1999). In this regard, it should be pointed out that preliminary observations showed that intracortical inhibition induced by paired TMS and MEP inhibition following peripheral activation are positively correlated in normal subjects (Trompetto et al., 2000).

Different pathways and cortical sites may be involved in this defective inhibitory mechanism. Previous studies on the effects of peripheral stimulation on corticomotoneuronal excitability of normal subjects mainly reported a facilitatory effect in the first 100 ms after conditioning stimulation (Deuschl et al., 1991; Rossini et al., 1991; Deletis et al., 1992; Komori et al., 1992; Maertens de Noordhout et al., 1992; Hirashima and Yokota, 1997). This effect was interpreted as both spinal and supraspinal in origin, possibly involving a transcortical loop through the primary (SI) sensory cortex. On the other hand, it has been suggested (Chen et al., 1999) that changes of motor cortical excitability induced by peripheral afferent stimulation at C-T intervals >100 ms are probably due to corticocortical connections from sensory cortices. At such latencies both the primary (SI) and secondary (SII) somatosensory cortices, which project directly to the motor cortex with a somatotopic arrangement, are activated (Jones, 1983, 1986; Gosh et al., 1987). Such organization is consistent with our observation that MEP size modulation following peripheral afferent stimulation at the wrist was more pronounced in the hand muscles.

Indeed, cerebral blood flow studies in patients with focal dystonia have shown underactivation of the sensorimotor cortex in response to vibration (Tempel and Perlmutter, 1993; Feiwell et al., 1999) or during movement (Ceballos-Baumann et al., 1995; Ibanez et al., 1999). Such metabolic underactivity might indicate a reduced effectiveness of intracortical inhibitory circuits. Conflicting results have been previously reported on the behaviour of SEPs in response to upper limb stimulation. In particular, the amplitude of the frontal N30 component was found to be reduced (Mazzini et al., 1994; Grissom et al., 1995) or enhanced (Reilly et al., 1992; Kañovský et al., 1998). These discrepancies are likely to depend on methodological differences as well as on the clinical features of the various populations of dystonic patients examined. Overall, it appears that SEP amplitudes are not consistently modified in resting conditions (Tinazzi et al., 2000), and we found no difference between patients and normal controls in the amplitude of the parietal (N20) or of the frontal (N30) SEP component. This might indicate that the excitability of sensory areas per se is not modified in focal dystonia, whereas sensorimotor processing following peripheral activation is different in patients and controls. Indeed, recent SEP studies (Tinazzi et al., 2000) have shown that the integration of afferent inputs coming from adjacent body parts is abnormal in dystonia.

Several studies have suggested that central processing of sensory input might be abnormal in dystonia. Perception of motion is impaired in idiopathic focal dystonia because of a dysfunction in Ia afferent signal elaboration (Grünewald et al., 1997; Rome and Grünewald, 1999). Kaji and colleagues showed that dystonia can be modified by modulation of afferents from muscle spindles (Kaji et al., 1996). A reduced reciprocal inhibition of H reflexes has been shown in dystonic patients (Nakashima et al., 1989), possibly due to defective presynaptic inhibition, and botulinum toxin injections can restore defective reciprocal inhibition (Priori et al., 1995) as well as defective intracortical inhibition (Gilio et al., 2000).

It may be postulated, therefore, that in genetically susceptible individuals, overuse or repetitive trauma may modify the characteristics of sensory inputs from a specific body area leading to an abnormal sensorimotor integration or even to plastic cortical reorganization (Bara-Jimenez et al., 1998). An increased synaptic connectivity might be responsible for the increased input/output relationship in dystonics (Ikoma et al., 1996). In particular, sensory inputs might be abnormally processed in the brain of dystonic patients with a defective activation of local cortical inhibitory systems. This would in turn increase the excitability of the motor cortical areas leading to an inappropriate output and more widespread muscle activation.

Abnormalities of spinal cord inhibitory mechanisms (reciprocal inhibition) have also been observed outside the clinically involved territory (Deuschl et al., 1992; Chen et al., 1995). In this study, on the other hand, the impairment of sensory motor integration following peripheral nerve stimulation at the wrist was present only in patients with focal hand dystonia, but not in those with cervical dystonia. Such apparent loco-regional specificity could be tested in the latter group of patients by a protocol investigating the activation of peripheral afferents from the cervical region.

Acknowledgments

We wish to thank Professor Marco Schieppati for discussing the preliminary version of the manuscript and Mrs Rosemary Allpress for assistance in reviewing the English text. This study was supported by M.U.R.S.T. (Cofinanziamento 1999) and by C.N.R.

References

View Abstract