doi:10.1093/brain/awh291 Brain (2004), 127, 2764–2778

Neck proprioception and spatial orientation incervical dystonia

Marco Bove,1 Giampaolo Brichetto,2 Giovanni Abbruzzese,2 Roberta Marchese2 and Marco Schieppati3

Correspondence to: Professor Marco Schieppati, Centro

Studi Attivita Motorie (CSAM), Fondazione Salvatore

Maugeri (IRCCS), Via Ferrata 8, 27100 Pavia, Italy

E-mail: mschieppati@fsm.it and marco.schieppati@unipv.it

1Department of Experimental Medicine, Section of Human

Physiology, 2Department of Neurological Sciences,

Movement Disorder Unit, University of Genoa and3Department of Experimental Medicine, Section of Human

Physiology, University of Pavia, and Human Movement

Laboratory (CSAM), Fondazione Salvatore Maugeri

(IRCCS), Scientific Institute of Pavia, Italy

SummaryNeck muscle vibration is known to influence body orienta-

tion and posture during locomotion and stance in normal

subjects. To verify the hypothesis that neck proprioceptive

input can be misinterpreted in patients with cervical dys-

tonia (CD), lateral continuous vibration was applied to the

sternocleidomastoid muscle during both stepping-in-place

and quiet stance, with eyes closed. The orienting responses

of CD patients were compared with those of normal sub-jects. Vibration effects on body orientation during step-

ping and stance were apparently different from normal,

since no effects were seen when all patients’ data collapsed

were analysed. However, while some patients did not

respond to vibratory stimuli regardless of the vibrated

side, others had a ‘good’ side, the stimulation of which

produced effects on body orientation similar to those

observed in normal subjects. Homogeneous groupswithin the patient population were identified, based on

the vibration-induced responses under stepping condi-

tions. The different orienting or postural responses

observed in CD patients were correlated with disease-

related features such as spontaneous head position, max-

imum range of voluntary head yaw, presence or absence of

a botulinum toxin treatment and disease duration. Our

data suggest that, in CD patients, the reference system

used in the control of body orientation in space is eitherrefractory to the lateralized proprioceptive neck input or

modified such that the input from both sides produces an

orientation shift in the same sense. This would depend on

the pathogenesis of the disease or on an adaptive process

connected to the head abnormal posture. It seems that this

refractoriness spreads to both sides of the neck with the

advancement of the disease, thereby possibly entraining a

progressive shift from a reference system based on thehead to a more reliable egocentric reference.

Keywords: cervical dystonia; neck muscle vibration; neck proprioception; quiet stance; stepping-in-place;

spatial orientation

Abbreviations: BTX = botulinum toxin; C = control condition; CD = cervical dystonia; CFP = centre of foot pressure;

CCW = counter-clockwise; CW = clockwise; SCM = sternocleidomastoid; Vl = vibration to the left side of the neck;

Vr = vibration to the right side of the neck.

Received February 23, 2004. Revised July 23, 2004. Accepted July 25, 2004. Advanced Access publication September 8, 2004

IntroductionProprioceptive input from the neck muscles plays an import-

ant role in the definition of the reference system for the control

of posture and locomotion. Cohen (1961) pointed to the

weight of neck proprioceptive inputs in the control of body

orientation and motor coordination in monkeys. In humans,

several studies led to the notion that neck afferent input plays

a significant role in the control of posture (Brandt, 1996) and

that disturbances to the neck musculature can provoke dizzi-

ness or unsteadiness, the so-called ‘cervico-genic dizziness’

(Brandt and Bronstein, 2001). Hlavacka and Njiokiktjien

(1985) and Fransson et al. (2000) showed an influence of

active and passive neck torsion on the direction of galvani-

cally induced postural responses, an effect expected on the

basis of the convergence of neck and labyrinth input onto

Brain Vol. 127 No. 12 # Guarantors of Brain 2004; all rights reserved

brainstem nuclei (Pompeiano et al., 1991) and vestibular neur-

ons (Anastasopoulos and Mergner, 1982). Recently, a pro-

longed contraction of dorsal neck muscles has been shown

to alter balance control through a mechanism connected to

fatigue-induced afferent inflow (Schieppati et al., 2003).

Proprioception is often studied by means of muscle

vibration, since this is an adequate stimulus for selectively

activating the spindle receptor, thereby inducing a vibration

frequency-entrained excitation of the primary endings and a

train of action potentials in the large diameter afferent fibres

(Burke et al., 1976a; Roll and Vedel, 1982; Matthews, 1988).

Lund (1980) showed that neck muscle vibration had obvious

effects on standing posture. Vibration has also been applied to

various body muscles under various behavioural conditions,

including postural and locomotor tasks (Courtine et al.,

2001). Lateral neck muscle vibration, in the absence of

vision, produces undershoot of target and deviation of the

trajectory during gait (Bove et al., 2001) and body rotation

during stepping (Bove et al., 2002); under both conditions,

the deviation is opposite to the vibration side. Ivanenko et al.

(1999, 2000) showed that when the head is horizontally

turned or the eyes are laterally rotated, vibration of dorsal

neck muscles during stepping-in-place causes stepping in

the direction of the naso-occipital axis or of the gaze,

respectively. Ledin et al. (2003) found that head tilt on

the sagittal plane, but not head rotation, produces changes

in the postural sway induced by leg muscle vibration in

standing subjects, thus confirming previous findings (Kogler

et al., 2000) on the absence of major effects of head torsion

on postural control.

The above effects can hardly be interpreted in terms of

simple reflex effects. Goodwin et al. (1972) originally

described proprioceptive illusions induced by muscle

vibration. It was shown later that neck muscle vibration elicits

apparent motion of a stationary visual target and deviation of

the perceived ‘straight ahead’, possibly by producing a change

in the egocentric body-centred coordinate system (Biguer

et al., 1988; Taylor and McCloskey, 1991; Karnath et al.,

1994). In standing subjects, neck muscle vibration induces

body tilt and increased sway, suggesting that posture is

organized with respect to a ‘body schema’, to the construction

of which neck input contributes together with eye and skeletal

muscle (Eklund, 1972; Roll et al., 1989; Gurfinkel et al., 1995;

Ivanenko et al., 1999; Kavounoudias et al., 1999). In ‘neglect’

patients, neck muscle vibration can compensate the horizontal

displacement of the sagittal midplane (Karnath et al., 1993;

Karnath, 1994). Bottini et al. (2001) have shown recently that

signals from the neck muscles and from the labyrinth

converge onto several areas of the cortex, including the

temporo-parietal junction, the insular and retroinsular cortex

and the secondary somatosensory area, and proposed the

notion that these areas contribute to the egocentric

representation of space.

Idiopathic cervical dystonia (CD) is the most common

adult-onset focal dystonia and is defined as head rotation

caused by an abnormal involuntary neck muscle contraction

(Nutt et al., 1988; Jankovic et al., 1991). Recently, the

existence of a pathogenetic role for sensory feedback in

dystonia has been discussed (Hallet, 1995; Abbruzzese

and Berardelli, 2003; Tinazzi et al., 2003). Sensory input

may be abnormal and trigger focal dystonia, or its defective

‘gating’ may cause an input–output mismatch in specific

motor programmes. Several observations strongly support

the idea that sensorimotor integration is impaired in focal

dystonia (Abbruzzese et al., 2001; Munchau et al., 2001;

Siggelkow et al., 2002). Moreover, dorsal neck input in the

context of whole-body postural control and spatial

orientation seems to be relatively ignored in CD patients

(Lekhel et al., 1997). These notions are consistent

with results obtained in these patients by means of psycho-

physical studies of spatial orientation (Anastasopoulus et al.,

1997a,b).

However, the matter is still controversial. A population of

dystonic patients has been shown to exhibit altered postural

control, and to improve their stance control after botulinum

toxin (BTX) injection that normalized neck muscle activity

(Wober et al., 1999). Other studies described a great variab-

ility in the capacity of CD patients to estimate their head and

trunk direction; nonetheless, these patients appeared to estim-

ate head and trunk displacements correctly in ego-centric and

space-centric spatial orientation tasks (Anastasopoulos et al.,

2003). The aim of the present study was to shed further light

on the effect of neck muscle proprioception in CD patients, by

using an experimental protocol recently developed, where

neck muscle vibration is used to modify the reference system

for spatial orientation during a dynamic, locomotor task (Bove

et al., 2001). In particular, we report here the effects of lateral

neck muscle vibration on body orientation during an extended

stepping-in-place task without visual information. Such sti-

mulation induces in normal subjects a clear-cut body rotation

to the side opposite to the vibrated side (Bove et al., 2002). We

argued that lateral neck vibration during stepping would be

a sensible enough protocol for answering the question of

whether the proprioceptive neck input is ignored in CD

patients in the context of spatial orientation, in which case

no effect of vibration should be observed, or whether the neck

input is centrally integrated in an abnormal way, in which case

anomalous responses would be obtained. Stepping-in-place

was used rather than linear locomotion, since stepping exhi-

bits several key characteristics of locomotion (Lyon and

Day, 1997), in addition to allowing prolonged periods of

locomotor-like activity to be recorded by means of a

movement analysis system. The stepping-in-place paradigm

has also been used extensively in the evaluation of patients

with vestibular disorders (Norre et al., 1989; Bonanni and

Newton, 1998), since the early observations by Fukuda

(1959) of body rotation during stepping in normal subjects

and vestibular patients.

In this work, the neck vibration was also administered to the

same CD patients during quiet stance.

The orienting and postural responses of CD patients and

normal subjects were therefore analysed and compared, in

Proprioception and orientation in dystonia 2765

order to understand whether, in CD patients: (i) a long-term

neck vibration affected body orientation during stepping as in

normal subjects; and (ii) vibration administered during quiet

stance produced effects coherent to those observed during

stepping. The responses were correlated with disease-related

features such as spontaneous head position, maximum range

of voluntary head yaw, presence or absence of a BTX treat-

ment and disease duration.

Material and methodsSubjectsTwelve CD patients [four men and eight women, age range

33–82 years, mean 59 6 15.1 (SD); mean disease duration

9.4 6 5.5 years] and 12 normal subjects (seven men and five

women, age range 31–76 years, mean 51 6 15.5) volunteered for

these experiments. All subjects gave their informed consent for the

study, which was approved by the local ethics committee, and the

study conformed to the Declaration of Helsinki. Normal subjects had

no history of neurological diseases or signs or symptoms of cervical

diseases or trauma. On clinical examination, subjects and patients

showed no vestibular deficits or discrepancies between right and left

lower limbs that can affect veering of locomotion (Boyadjian et al.,

1999). In CD patients, both the amplitude of head deviation from the

primary position (spontaneous head position) and the maximum

range of the voluntary head rotation (yaw) were measured during

quiet upright stance. These and other features are summarized in

Table 1. Of the 12 patients studied, seven had received type A

botulinum toxin (BTX-A) injections. The BTX-A treatment ranged

between 10 and 40 BTX-A injections [mean 19 6 11.3 (SD)] admi-

nistered in several sessions. In most cases, the injected neck muscles

were the sternocleidomastoid (SCM) controlateral to the head rota-

tion, and one of the ipsilateral dorsal muscles. During the BTX-A

treatment, the injections scheme was modified in some patients as a

function of the head posture changes. Two patients had received

BTX-A injections in both SCM muscles, and one had only the

left dorsal muscles injected. All these patients were evaluated at

least 3 months after the last injections. The remaining five patients,

candidate for receiving for the first time BTX treatment, were eval-

uated before the injection.

ProcedureWhen they first came to the laboratory, subjects and patients prac-

tised stepping-in-place, with eyes open and closed, for �30 s before

recording. For the experimental session, they were instructed to start

stepping-in-place at their own preferred pace, on a verbal go signal

until they were told to stop after 60 s. This was done under both

control (C, no vibration) and vibration (V) conditions, in which case

they were told not to react to the applied stimulation; in particular,

there was no instruction to keep any pre-set facing direction or any

definite head or body position. The CD patients were free to maintain

their spontaneous head position during stepping. In different trials,

vibration was delivered randomly to the right (Vr) and left side (Vl)

of the neck. All subjects and patients easily succeeded in the

stepping-in-place task in all trials and, within each trial, for the entire

epoch of acquisition (1 min). The room was dim lit and the subjects

kept their eyes closed during stepping. Under all conditions, the onset

of acquisition of the kinematics signal started �3 s after the verbal go

signal for stepping onset. By this time, subjects and patients had

already performed at least one or two complete stepping cycles.

For V, the vibrator was set on concurrently with the onset of the

acquisition. The vibrator and the acquisition were shut off 60 s later,

at the end of the trial. Two trials were performed randomly for each

C, Vr and Vl condition (six trials in total). The entire session lasted

�60 min, including rest periods. After the session, subjects were

asked specifically about head tilt or rotation illusions: under no

circumstances did normal subjects or CD patients report explicit

sensations of head turning or tilting on the trunk in response to

the vibration. This is not contradictory to the finding reported in

the study of Taylor and McCloskey (1991), where dorsal neck vibra-

tion rarely produced illusions of head movement. We did not check,

Table 1 Clinical features of the CD patients

Patientno.

Sex Age(years)

Duration(years)

Type ofcervicaldystonia

Associatedsigns

Treatment andSCM muscleinjected

Head deviation(�) ( + , R; �, L)

Max head voluntaryrotation (�)

to R to L Total

1 F 45 11 Rotatocollis (R) Head tremor BTX-A, R/L 8.5 32.6 59.0 91.62 M 82 13 Rotatocollis (R) – BTX-A, L 7.5 41.3 57.7 993 F 62 4 Rotatocollis (L) – BTX-A, R �25.8 22.3 37.6 59.94 F 33 11 Rotatocollis (L) – – �24.7 78.3 33.5 111.85 F 55 3 Rotatocollis (L) – – �6.5 22.0 27.4 49.46 M 47 13 Rotatocollis (L) – BTX-A, R; AC �14.4 50.0 53.1 103.17 M 63 6 Rotatocollis (L) Head tremor;

right arm dystoniaBTX-A, R/L; AC �23.2 71.1 11.9 83

8 F 42 2 Rotato- andlaterocollis (L)

– – �15.2 72.0 51.1 123.1

9 F 76 6 Rotatocollis (R) – BTX-A, L 44.5 5.2 90.7 95.910 F 70 9 Rotatocollis (L) Head tremor;

upper limbs tremor– �19.8 52.3 1.1 53.4

11 F 74 21 Laterocollis (R) Head tremor – 29.6 16.3 96.9 113.212 M 55 14 Retro- and

rotatocollis (R)– BTX-A 8.3 39.9 47.8 87.7

R = right; L = left; BTX-A = botulinum toxin type A; AC = anticholinergics; SCM = sternocleidomastoid.

2766 M. Bove et al.

however, whether vibration produced errors in the execution of

movements pointing at a specific spot in the head.

In a separate session, three of the normal subjects performed addi-

tional stepping-in-place trials, in C and V conditions, with the head

deliberately kept rotated �20�, first on the right and then on the left

side. Five other naıve subjects (two men and three women, age range

24–40 years, mean 29.6 6 6.2), who did not participate in the main

experiment, were also recruited for this experiment. This was done to

check, first, any effects of static head yaw on body rotation during

stepping under control condition and, secondly, possible differences

in the effects of neck muscle vibration on stepping orientation

when the head was in the primary position versus rotated to an

extent comparable with the spontaneous position of the head in

many CD patients. These subjects repeated three trials for each

condition.

All normal subjects and CD patients who participated in the

stepping-in-place sessions also stood upright on a dynamometric plat-

form, on the same day, after a rest period of at least 10 min (see below).

StimulationA DC motor with an eccentric on the shaft, embedded in a plastic

tube (Dynatronic, France), produced 5 N peak-to-peak vibrations at

90 Hz. The vibrator was applied on each side of the neck (left and

right sides were vibrated separately) by means of an elastic strap that

passed around the neck. The vibrated spot was the belly of the SCM

muscle, �40% of its length below its insertion on the mastoid

(�5–6 cm from the mastoid apex), with the cylinder axis about

normal to the direction of muscle. The vibrator was put in place

at the beginning of the experiment and kept in place throughout the

session. In order to evaluate the propagation of the vibration to

the temporal bone, which could possibly induce activation of the

labyrinthine receptors, in three normal subjects we used two identical

accelerometers (TSD109, BIOPAC Systems Inc., Santa Barbara,

CA), one fixed to the vibrator located on the muscle, the other placed

on the ipsilateral mastoid bone. The amplitude of the main peak in the

power spectrum of the vibration signal recorded at the mastoid bone

was, on average, <1% (0.66 6 0.03%) of the signal measured at the

muscle. This finding indicates a negligible propagation of the

mechanical wave to the labyrinth, and strongly lessens the likelihood

of vestibulum-mediated vibration effects.

KinematicsBody movements were recorded by an optoelectronic motion

analysis system (ProReflex, Qualisys AB, Sweden). Four infrared

cameras were located around the experimental field, identifying a

space of �2.5 3 2.5 3 2.5 m. The reflective markers were attached

to the skin on the forehead, vertex of the head, left and right acromion

and on the right lateral femur epicondyle. Two markers on the floor

identified a fixed reference axis. The markers’ position was sampled

at a frequency of 100 Hz. The following parameters were computed

for each trial and under each condition (C, Vr and Vl). (i) Step cycle

frequency. The recording of the vertical displacement of the knee

versus time was subjected to a fast Fourier analysis and the value

of the main peak was taken as the mean value of step frequency.

(ii) Body displacement. The first and last 100 samples (=1 s) of the

vertex marker at the beginning and the end of each trial were aver-

aged in order to obtain a mean position of the head during the first and

last stepping cycles and minimize the error connected with head

oscillation. The averaged initial position was then subtracted from

the averaged final position to obtain the final body displacement. (iii)

Body rotation. This was represented by the changes in the angle

defined by the shoulder axis and the fixed reference axis. The first

and last 100 samples of this angle value (recorded at the beginning

and the end of each trial) were averaged. The initial mean position

was then subtracted from the final mean position. In normal subjects,

the body rotation was �0 in the control condition, and could become

negative or positive when the neck was vibrated on the right or left

side of the neck, respectively (Bove et al., 2002). Negative values

indicate a counter-clockwise (CCW), and positive values a clockwise

(CW) rotation. (iv) Head–shoulder angle (yaw). This was the mean

angle between head antero-posterior axis and shoulder medio-lateral

axis evaluated during stepping-in-place.

Stabilometric recordingThe sampling frequency of the force exerted on three strain gauges of

a dynamometric platform (Medicapteurs, Toulouse, France) was set

at 10 Hz. The subjects were asked to stand still with their eyes closed,

their bare feet on a patterned surface at an angle of 30�, with their heels

separated by 10 cm, and their arms kept at their sides. Each trial lasted

51.2 s, regardless of the condition tested. The vibration was applied

to the same spot and side as in the stepping-in-place sessions. The

antero-posterior and medio-lateral displacement of the centre of foot

pressure (CFP) was measured: (i) during the stance trials without

vibration (C); (ii) during Vr; and (iii) during Vl. The trials

were spaced by at least 3 min, during which subjects were allowed

to move across the laboratory in order to allow recovery from

possible long-lasting effects of vibration (Wierzbicka et al., 1998)

and from the effects of repetition of stance trials (Tarantola

et al., 1997).

Statistical analysisThe effects of vibration on stepping frequency, body displacement,

body rotation, head–shoulder angle and displacement of CFP were

assessed by means of an analysis of variance (ANOVA) for each

subject in each experimental condition. Homogeneity of data var-

iances was evaluated by means of the Bartlett’s test. Comparisons

within groups or between groups (normal subjects or CD patients as

independent variables) were performed by means of a one-way or

two-way repeated measure ANOVA, respectively, where the depen-

dent variables were the three different experimental conditions (C,

Vr and Vl). When repeated measure ANOVA gave a significant (P <

0.05) result, the post hoc Newman–Keuls test was employed to assess

differences among C and V conditions for both the stepping-in-place

and quiet stance protocol.

Linear regression analysis was used to evaluate, in CD patients, the

possible relationship between head yaw angle and CFP mean position

in the sagittal and frontal plane during quiet stance, or between head

yaw angle and body rotation angle during stepping-in-place.

ResultsStepping-in-place

Stepping frequencyIn CD patients, a lower and marginally significant stepping

frequency with respect to normal subjects was observed

Proprioception and orientation in dystonia 2767

within all the experimental conditions (normal subjects, C,

0.93 6 0.02 Hz; Vr, 0.9 6 0.019 Hz; Vl, 0.9 6 0.018 Hz; CD

patients,C,0.7560.07Hz;Vr,0.7660.08Hz;Vl,0.7760.08Hz;

means6SE)[two-wayrepeatedmeasureANOVA,F(2,44) = 4.1;

P= 0.05]. Neck muscle vibration (V) administered during step-

ping did not modify the stepping frequency with respect to C, in

either normal subjects or CD patients [two-way repeated

measure ANOVA, F(2,22) = 0.275; P = 0.76].

Body displacement

The effect of the vibration on the capacity of subjects and

patients to step on the spot was evaluated by the distance

between their final and initial body positions. The mean dis-

tance from the starting position is indicated in Fig. 1A (normal

subjects) and B (CD patients) by the length of the dashed lines

radiating from the origin of the polar plots.

In normal subjects, under C, Vr and Vl conditions, there was

an average body displacement of 58.26 37.6, 59.36 29.4 and

52.5 6 22.8 cm (SD), respectively. This body displacement

was usually oriented forward under the control condition. The

displacement was instead directed to the left or to the right

when the vibration was applied to the right and left side of the

neck, respectively (Fig. 1A). Regardless of the actual direction

of the stepping, the final distance attained during vibration did

not show significant changes with respect to the control

condition (Vr versus C, P = 0.88; Vl versus C, P = 0.45).

In CD patients, the average body displacement was 94.9 6

51.9, 83 6 42.8 and 74.2 6 44.7 cm (SD), in C, Vr and Vl

conditions, respectively (Fig. 1B). The difference in the body

displacement between normal subjects and patients, all con-

ditions pooled, was only marginally significant [two-way

repeated measure ANOVA, F(1,22) = 4.27; P = 0.051].

A marginally significant difference across conditions was

present for the CD patients [two-way repeated measure

ANOVA, F(2,44) = 3.16; P = 0.052]. In fact, the body dis-

placement in both V conditions was lower than that observed

in the control condition, but only in the Vl condition was the

difference significant (P = 0.02).

Body rotation

Normal subjects and patients could rotate slightly on

their vertical axis during stepping-in-place under control

conditions. In the normal subjects, a slow yet continuous

body rotation was the main and consistent effect of the lateral

Fig. 1 (A and B) Polar graph representations of the effects of lateral neck vibration on body rotation and displacement duringstepping-in-place in normal subjects and CD patients. Dashed lines represent the mean angular value of body rotation; solid lines representthe 95% confidence interval (CI) of the mean value, evaluated across all subjects under the different experimental conditions. The lengthof the lines represents the main final distance of body displacement. Normal subjects exhibited only a slight body rotation under the control(C) condition. During vibration (V), there was a systematic rotation, in all subjects, in the direction opposite to the vibrated side. CDpatients showed an ample scatter of non-systematic body rotations under the control condition, leading to a mean value not different fromthat found in the normal subjects. During vibration, the mean body rotations were smaller in the patients than in the normal subjects,but the 95% CIs were about twice those obtained for the normal subjects. The final displacements were larger in the patients than in thenormal subjects. (C and D) The histogram bars represent the mean head yaw value (6SE) during 60 s of stepping-in-place in the threeexperimental conditions across all subjects. On average, in both normal subjects and CD patients, no significant vibration-induced headrotation was observed. However, the variability of the head yaw was much higher in the CD patients than in normal subjects.

2768 M. Bove et al.

neck vibration. In the patients, the rotation during stepping

was less consistent (see below). The mean rotation values

attained by the populations of normal subjects and of CD

patients at the end of the stepping trial are represented by

the angular coordinate of the dashed lines radiating from the

origin of the polar plots in Fig. 1A (normal subjects) and B

(CD patients).

Normal subjects exhibited no major rotation under control

condition, as indicated by the mean value, very close to 0�

(1.2� CCW, dashed black line), and by the small data variance,

as indicated by the two 95% confidence limits (continuous

black lines) (see Fig. 1A). Under vibration conditions, there

was a systematic rotation, in the direction opposite to the

vibrated side. In the figure, the scatter of the rotations induced

by right and left vibrations are indicated by the sectors

between the continuous dark grey and light grey lines, respect-

ively, which correspond to the two confidence limits. On

average, the rotation was �53� CCW and 68� CW with

respect to the initial body position, for vibration applied to

the right and left neck side, respectively. These values were

significantly different from the negligible rotations observed

under the control condition [two-way repeated measure

ANOVA, F(2,44) = 29.06; P < 0.005 for Vr, and P <

0.0001 for Vl]. The distribution of the rotation data around

the population mean value was similar across the two vibra-

tion conditions (Bartlett’s test: P = 0.34).

CD patients exhibited a much smaller consistency in body

rotation than normal subjects under control conditions, in

spite of their average orientation being similar to that of

normal subjects (Fig. 1B). The average final position of

the population was rotated by only a few degrees (7.8�

CW) with respect to the initial position. However, the scatter

of the rotation data was much larger across the patients’

population, as indicated by the broad sector comprised

between the 95% confidence intervals, and the variance of

the data obtained under control conditions proved to be

different between normal subjects and CD patients (Bartlett’s

test: P < 0.001).

CD patients as a population showed no significant differ-

ence in body rotation between vibration and control condi-

tions (Vr versus C, P = 0.09; Vl versus C, P = 0.56). The

amplitude of the confidence interval was similar under control

and vibration conditions, but was about twice that observed

for the normal subjects (under any conditions). The variance

of the data recorded under the C and V conditions proved to be

different between normal subjects and patients (Bartlett’s test,

C, P < 0.01; Vr, P < 0.01; Vl, P < 0.01). This scatter, however,

was more of a population result rather than the expression of

inconsistencies within patients’ performances. To show this,

the absolute differences between the rotations of the two trials

performed by each patient under each condition were aver-

aged and compared in the two populations: no significant

difference between normal subjects and CD patients was

observed in this value [two-way repeated measure

ANOVA, F(2,44) = 1.17; P = 0.32], indicating a similar

variability in each of the subjects of the two groups.

Therefore, the responses to vibration of the individual CD

patients were reproducible, as were those in the normal sub-

jects, in spite of their responses being sometimes very differ-

ent in absolute terms from those of normal subjects. In fact, in

the normal subjects, the confidence interval of the responses

to each vibration condition covered only one quadrant of the

polar diagram. Conversely, in the CD patients, the range

between the confidence intervals of the rotations under

each vibration condition partly overlapped around zero.

This is the outcome of a non-systematic response pattern to

vibration across the whole group of CD patients: CCW and

CW responses alike could occur in fact for the same neck

vibration side in different patients.

The mean position of the head on the vertical axis with

respect to the body sagittal plane was evaluated in normal

subjects and CD patients during the stepping-in-place under

all conditions. Under control conditions, normal subjects had

a mean head yaw angle of �0.08 6 0.9� (SE) (Fig. 1C), while

in the patients this mean angle reached �3.8 6 3.9� (Fig. 1D).

Interestingly enough, the mean head yaw angle during the

stepping trials underwent no significant vibration-induced

changes across any condition. This was true for both the

normal subject group [one-way ANOVA, F(2,22) = 0.006;

P = 0.99] (Fig. 1C) and the patient group [one-way ANOVA,

F(2,22) = 1.14; P = 0.34] (Fig. 1D).

Patient groups according to their responsesTo better describe the responses to neck vibration in CD

patients, data were analysed in each single patient, in order

to extract potential features common to other patients of the

same population, and possibly identify homogeneous sub-

groups. The mean body rotations under vibration conditions

are shown in Fig. 2A, for each patient. Figure 2B depicts the

same data for the normal subjects population, and allows a

comparison of the vibration effects on body rotation in the

two groups. Dashed lines are the limits of the 95% confidence

interval of the body rotations evaluated within all the CD

patients (Fig. 2A) and normal subjects (Fig. 2B) under control

conditions. The interval thus identified is considered, for each

group, as the zone where the neck vibration effects on the body

rotation are non-significant. The bars are the mean rotation

values calculated from the two trials accomplished during

vibration of the same side. Across normal subjects, the ampli-

tude and sign of the vibration-induced rotations were very con-

sistent: all subjects rotated CW on Vr and CCW on Vl. Across

the patients, vibration induced two main patterns. (i) The

responses to vibration applied to both the right and the left

side lay inside the confidence interval. Five of the 12 patients

(subjects2,3,4,6and11)had thiskind of response. Henceforth,

these patients are indicated asbelonging toagroup ‘insensitive’

to vibration stimuli (INS group). (ii) The body rotations in

response to vibration were significantly different from those

observed under control conditions (and congruent with the

responses observed in the normal subjects) when one side of

the neck was vibrated, but they were either non-significant

Proprioception and orientation in dystonia 2769

or were incongruent when the other side of the neck was

vibrated, since the body rotated in the ‘wrong’ direction

compared with normal subjects’ rotation. Hence, there was

in this group one ‘sensitive’ side and one ‘insensitive’ or

‘wrong’ side (note, however, that the wrong rotations never

exceeded the correct response on contralateral stimulation).

Seven patients showed this pattern, and were divided in turn

into two subgroups related to the neck side, the vibration of

which induced the ‘correct’ response (the RIGHT group con-

sisted of subjects 8, 9 and 12; the LEFT group consisted of

subjects 1, 5, 7 and 10).

Figure 3A, B and C summarizes the mean responses to

lateral neck vibration of INS, RIGHT and LEFT groups,

respectively. As expected, the INS group (Fig. 3A) showed

no significant rotation under vibration or control conditions

[one-way repeated measure ANOVA, F(2,8) = 1.95; P = 0.2],

while the RIGHT (Fig. 3B) and LEFT (Fig. 3C) groups were

selectively and significantly sensitive to vibration [one-way

repeated measure ANOVA: RIGHT group, F(2,4)= 10.44; P <

0.05; LEFT group, F(2,6) = 6.13; P < 0.05]. The mean RIGHT

group response to right vibration was a CCW rotation of

�111� (Vr versus C, P < 0.05), while left vibration induced

no significant effects. Conversely, the LEFT group rotated

CW �83� for vibration to the left side (Vl versus C, P <

0.05), and no significant rotation was induced by right

side vibration. Interestingly, the variability of the vibration-

induced body rotations within each of these groups, in

spite of the small number of patients, was of the same

magnitude as that in the normal subjects population.

Head–shoulder angle (yaw) versus bodyrotation angleThe spontaneous head yaw mean angles ranged between 3 and

50� in the CD patients. We tested the hypothesis that the head

positioncouldaffect thevibration-inducedbodyrotation. Since

it is known (Leis et al., 1992; Lekhel et al., 1997; Karnath et al.,

2000) that neck muscle vibration can produce sizeable effects

on head yaw in CD patients, we recorded head yaw relative to

trunk during vibration. The plot in Fig. 4 shows, for each pati-

ent, the mean head yaw during Vr and Vl (ordinate) versus the

spontaneous head yaw. Across the patients, there were no

major vibration-induced changes, as indicated by the data

points clustering close to the identity line. There were non-

negligible effects in some patients, but the effects were appar-

ently not side specific in this population. Moreover, there was

apparently no difference in the vibration-induced effects on

head yaw among the three patient groups, as defined above

(INS, RIGHT and LEFT), since the data of the patients belong-

ing to these groups formed no clear clusters in the scatterplot.

The possible effect of the head posture on the body orienta-

tion during stepping-in-place was then analysed across indi-

vidual patients stepping under both control and vibration

conditions. All individual trials from all the patients were

segregated in Fig. 5, which shows the scatter graphs of

body rotation angles against head yaws, for C (Fig. 5A),

Vr (Fig. 5B) and Vl (Fig. 5C) conditions. No significant rela-

tionships between head–shoulder angle and body rotation

angle were found either in C or V conditions (C, P = 0.26;

Vr, P = 0.12; Vl, P = 0.095). Similar results were found in

normal subjects (C, P = 0.61; Vr, P = 0.59; Vl, P = 0.59), as

already shown in Bove et al. (2002).

Head deliberately rotated in normal subjectsEight normal subjects stepped with eyes closed with the head

rotated to either side to an extent similar to that found in the

Fig. 2 Effects of vibration on body rotation for each CD patient(A) and normal subject (B). The bars are the mean values of thetwo trials performed by each patient and subject. The rotationeffects are separated for the vibration side. Dashed lines representthe upper and lower limits of the 95% confidence interval of themean evaluated across all subjects in the control condition. Theinterval thus identified is considered as the zone where the neckvibration effects on the body rotation are non-significant. In (A),the circles at the top of the bar pairs indicate CD patientsbelonging to the RIGHT (open symbols) or LEFT group (filledsymbols). The other CD patients belong to the INS group.

2770 M. Bove et al.

patients (Fig. 6A). Under the control condition, no or non-

systematic body rotations were observed, across subjects or

trials (Fig. 6B). The mean head–shoulder yaw angles were

(mean 6 SE): head in primary position, 0.68 6 1.17�; left

rotation, �28.74 6 3.11�; right rotation, 22.83 6 2.64�. These

head positions were also broadly maintained during vibration.

The vibrations, on either the right or the left side, induced

significant and systematic CCW and CW body rotations,

respectively (Fig. 6B) [two-way repeated measure

ANOVA, F(2,42) = 91.28; P < 0.001]. The analysis also

confirmed that head position did not affect the vibration-

induced body rotation [two-way repeated measure

ANOVA, interaction effect, F(4,42) = 0.96; P = 0.44]. The

deliberate head rotation did not significantly affect the abso-

lute value of the responses to vibration compared with those

observed with the head in the primary position (post hoc, left

rotation in both Vr and Vl, P > 0.1; right rotation in both Vr

and Vl, P > 0.1). To assess whether the variability of the body

rotation angles during stepping was affected by the head pos-

ture (possibly producing a scatter in the data comparable with

that seen in the CD patients), the data obtained in these trials

with the head deliberately rotated were compared with those

obtained in the population of normal subjects stepping with

the head in the primary position, illustrated in Fig. 1A. The

homogeneity of variances between these two sets of data was

confirmed by the Bartlett’s test (P = 0.19).

Upright stance

Antero-posterior CFP positionUnder control conditions, CD patients were inclined forward,

on average, somewhat more than normal subjects (Fig. 7).

Their CFP lay significantly ahead with respect to that of

normal subjects [two-way repeated measure ANOVA,

F(1,22) = 12.96; P = 0.0015]. Across the patients, there

was no relationship between the spontaneous head yaw

angle and the CFP position in the sagittal plane (F = 0.33;

P = 0.58). Vibration exerted no significant effect on the

antero-posterior position of the CFP in either normal subjects

Fig. 3 Effects of vibration on body rotation. Average values of thethree patient groups. (A) The INS group showed no significanteffects on body rotation during vibration on either neck side.(B) The RIGHT group was sensitive to vibration administered tothe right side of the neck. Left side vibration induced a small andnon-significant response toward the ‘wrong’ side with respect tothat observed for the same condition in normal subjects. (C) TheLEFT group was sensitive to vibration when vibration wasadministered to the left side of the neck, while right side vibrationinduced a response similar to that observed in the controlcondition.

Fig. 4 Mean head yaw during vibration as a function ofspontaneous head position during stance. Each patient enters thegraph with two data points for each value of the abscissa. Thesetwo data points represent the mean head yaw during vibration(white for Vr condition, grey for Vl condition). In turn, each valueon the ordinate results from the mean of two equal-conditiontrials. In general, there were no major vibration-induced effects onhead position across the patients, as indicated by the data pointsclustering close to the identity line. There were non-negligibleeffects in some patients, but these effects were apparently not sidespecific. The patient groups are identified with different symbols.There is apparently no difference in the vibration-inducedeffects on head yaw among the three groups.

Proprioception and orientation in dystonia 2771

(squares) or in CD patients (circles) [two-way repeated

measure ANOVA, F(2,44) = 2.91; P = 0.07].

Medio-lateral CFP positionUnder control conditions, the medio-lateral CFP position

was not different between normal subjects and CD patients

(Fig. 7). As for the antero-posterior plane, there was no

significant regression between spontaneous head yaw

angle and CFP position in the frontal plane (F = 0.085;

P = 0.78).

Lateral neck muscle vibration induced in the normal sub-

jects a significant CFP displacement in the frontal plane with

respect to C (Fig. 7), the amplitude of which was in turn

compatible with published data (Bove et al., 2001). In particu-

lar, Vr induced a shift to the left and Vl to the right. In the

CD patients, vibration had a much weaker effect on medio-

lateral displacement of CFP. Two-way repeated measure

ANOVA showed an effect of group (normal subjects versus

CD) and an interaction between groups and conditions.

In particular, significant differences in the medio-lateral

CFP displacements induced by vibration were observed

between normal subjects and CD patients [F(2,44) = 11.71;

P < 0.001]. Post hoc test showed a significant side-related

effect of vibration in normal subjects (P < 0.001), but no effect

in CD (P = 0.11). Therefore, on average, the patients had a

smaller sensitivity to vibratory stimuli than normal subjects

under quiet stance as well as under stepping-in-place

condition.

Effects of vibration during stance in thepatients’ groupsThe patients’ CFP was re-analysed based on the separation in

the three groups defined on the basis of the findings collected

with the stepping-in-place protocol, and the postural

responses to neck vibration of the three groups compared

separately. Figure 8A, B and C shows the mean postural

responses to lateral neck vibration of the INS, RIGHT and

LEFT groups, respectively. The INS group did not show

significant differences in medio-lateral CFP displacement

during either Vl or Vr conditions [one-way repeated measure

ANOVA, F(2,8) = 2.58; P = 0.137]. However, during V

conditions, a significant backward displacement of the CFP

along the sagittal plane with respect to C was observed [one-

way repeated measure ANOVA, F(2,8) = 4.79; P = 0.043]. In

the RIGHT group, neither Vl nor Vr evoked significant

antero-posterior displacements during vibration (one-way

repeated measure ANOVA, F(2,4) = 1.79; P = 0.28]. How-

ever, vibration produced a significant medio-lateral displace-

ment with respect to the control condition (one-way repeated

measure ANOVA, F(2,4) = 7.61; P = 0.043]. This

displacement (a mean shift to the left on the frontal plane

of�12 mm) was significant only for Vr (P < 0.05). In contrast,

Vl induced only a slight and non-significant medio-lateral

displacement with respect to C (P = 0.093) in the ‘wrong’

direction (to the left), compared with the response observed in

the normal subjects group under the same condition (Fig. 7).

In the LEFT group, no significant vibration effects on CFP

displacement on the sagittal plane were observed [one-way

repeated measure ANOVA, F(2,6) = 0.69; P = 0.54]. How-

ever, in this group, the CFP position in the frontal plane was

significantly sensitive to Vl [one-way repeated measure

ANOVA, F(2,6) = 4.62; P = 0.046]: a displacement to the

Fig. 5 Head yaw angles versus body rotation angles under C (A),Vr (B) and Vl (C) conditions. All trials of all CD patients arecollapsed. Negative head yaw angles correspond to head deviationto the left, positive correspond to head deviation to the right.Negative body rotation angles refer to counter-clockwise bodyrotation, while positive angles refer to clockwise rotation. Therewas no significant relationship between head–shoulder angle andbody rotation in any of the three experimental conditions.

2772 M. Bove et al.

right of �15 mm was induced by Vl, while Vr did not exert

any influence on CFP medio-lateral displacement (P = 0.99).

DiscussionStepping-in-placeLateralized neck muscle vibration, applied during stepping-

in-place with the eyes closed, is known to induce clear-cut

body rotation and a moderate displacement from the starting

position in normal young subjects. The rotation is invariably

toward the side opposite to the vibrated site, i.e. CW on left

and CCW on right side vibration, respectively, and the dis-

placement is in the forward direction (Bove et al., 2002). The

normal subjects recruited in the present study, who repres-

ented a population age-matched to the CD patients, repro-

duced the same patterns of body displacement observed in

the young subjects under control conditions and in response

to vibration.

Forward displacementThe CD patients showed significant differences from normal

subjects in their body progression during stepping-in-place,

under control conditions. Forward body displacement was

almost twice as large as that observed in normal subjects.

The body therefore moved with a relatively higher velocity

than in normal subjects, since the trial duration was constant.

It is worth noting that the stepping frequency of CD patients

was significantly lower than that of normal subjects, thus

patients take ‘longer steps’ than normals. The significantly

lower stepping frequency in CD patients would be in keeping

with the occurrence in these patients of ‘parkinsonian symp-

toms’ such as bradikynesia (Couch, 1976). The increased step

length should not be taken as a contradiction, since we do not

deal here with true locomotion, and the ‘high’ velocities were

in the order of 1 m/min. These incongruities may rather

Fig. 6 Neck vibration effects on the body rotation, in eight normalsubjects, during stepping-in-place with the head positioned in theprimary position or voluntarily rotated to either side. Thehistogram bars are separated into three groups, according to thecondition (C, control; Vr, vibration right; Vl, vibration left).(A) The left bar of each group refers to the head in the primaryposition, the middle bar to the head rotated to the right, and theright to the head rotated to the left. In (B), the body rotation anglesare shown for each of the head positions corresponding to thevalues reported in A. Under control conditions, there were no ornon-systematic body rotations across subjects and trials. Therewere significant body rotations when the neck vibration wasapplied to the right (Vr) or left (Vl) side. Systematic counter-clockwise and clockwise body rotations for vibration to the rightand left side, respectively, were observed. For all conditions, nosignificant effects of head position on rotatory responses tovibration were observed.

Fig. 7 Postural responses to lateral neck muscle vibration in thenormal subjects (NS, squares) and in all the CD patients (circles)are represented by the medio-lateral and antero-posterior shift ofthe centre of foot pressure (CFP) during quiet stance. Grandaverages of all subjects and trials are shown, separately for eachcondition (C, Vr and Vl). Normal subjects showed significantmedio-lateral displacements of the CFP towards the side oppositeto vibration. The position of the CFP in the CD patients wassignificantly advanced with respect to that of normal subjects(P < 0.01), but showed no significant change on right or leftvibration in either the frontal or sagittal plane.

Proprioception and orientation in dystonia 2773

indicate for CD patients a greater difficulty than normal sub-

jects to step ‘in place’ without visual information.

Vibration of the SCM muscle did not modify forward

displacement in normal stepping subjects, but in CD patients

it reduced the distance reached. Ivanenko et al. (2000) showed

that, during stepping-in-place, dorsal neck vibration produced

in normal subjects an involuntary forward stepping, which has

been explained as the consequence of an illusion of forward

movement of the head or forward sway (Lund, 1980; Lekhel

et al., 1997) to which the appropriate response is to step

forward. In our case, the vibration was not acting on dorsal

muscles; indeed, normal subjects did not increase their pro-

gression rate. It is curious nonetheless that CD patients

reduced their progression with vibration, while a forward

progression might perhaps have been expected as an effect

of diffusion of the vibration to the dorsal muscles, not least

because of the abnormal head position.

Body rotationThe CD patients exhibited a much smaller consistency in body

rotation than normal subjects under control conditions, as

proved by a significant difference between the data variances,

in spite of their average orientation being similar to that of

normal subjects. The effects of the neck muscle vibration on

body rotation during stepping were also different from nor-

mal, since when the collapsed data from all patients were

analysed, there was no effect of vibration on body rotation.

In fact, no significant difference in the average body rotation

during vibration, either on the left or on the right side, was

detected with respect to their average rotation under control

conditions. This could have been the outcome of a smaller

sensitivity to vibration in CD patients or of a ‘wrong’, abnor-

mal response to lateral neck vibration, or both. However, one

of the reasons for the lack of a systematic vibration-induced

body rotation in CD patients proved to be the ample variability

in the body rotation across subjects, under both vibration and

control conditions. This was in fact reflected in a confidence

interval of the population data that was about twice as large as

and significantly different from that observed for the normal

subjects, under any condition.

CD patients’ groupsThis analysis of the responses of each patient gave the pos-

sibility to extract potential common features and to identify

homogeneous groups within the population of CD patients. In

fact, it soon appeared that there were patients who did not

rotate, regardless of the vibrated side, or patients who had a

‘good’ side, the stimulation of which produced effects on

body rotation similar to those observed in normal subjects.

It also appeared that no CD patient exhibited a fully normal

response (CW on Vl and CCW on Vr); rather, patients were

refractory to vibration on either or both sides of the neck. In

those with one good side, the contralateral side could be

refractory or exhibit the same response observed on the vibra-

tion of the good side, as if the CNS could produce one and the

same type of rotatory response on receiving the propriocep-

tive input from either side of the neck. CD patients could then

Fig. 8 CD patients under quiet stance condition. Postural responsesto lateral neck muscle vibration of the INS, RIGHT and LEFTgroups. (A) In the INS group, vibration induced no effect on bodydisplacement in the frontal plane, while it induced a significantbackward displacement in the sagittal plane with respect to control.(B) The RIGHT group patients were sensitive to vibrationadministered to the right side of the neck. Left side vibration induceda slight and non-significant medio-lateral displacement with respectto control in the ‘wrong’ direction, compared with the responseobserved in the INS group under the same condition (see Fig. 5).(C) The LEFT group was sensitive to vibration administered to theleft side of the neck. Postural responses similar to those observed inthe control condition were obtained when vibration wasadministered on the right side of the neck.

2774 M. Bove et al.

be classified into three groups, according to their complete

(INS) or partial (RIGHT or LEFT) sensitivity.

Head abnormal posture is unrelated toabnormal body orientation response to vibrationduring stepping-in-placeA concern was represented by the possibility that the spontan-

eous head abnormal posture of CD patients would ‘deter-

mine’ the body rotation induced by the lateralized neck

vibration, either because of the ongoing asymmetrical ‘nat-

ural’ input possibly being able to modify the orientation of

stepping or because during vibration it could counteract in

some way the effect of the asymmetrical neck muscle pro-

prioceptive input. A further concern was represented by the

possibility that lateral neck muscle vibration would produce

head-on-shoulder rotation in both normal subjects and

patients, in turn possibly affecting the orientation of the

body during stepping. Neck muscle vibration in fact has

been reported to induce involuntary movements of the head

in some patients (Lekhel et al., 1997; Karnath et al., 2000). A

still further concern was that the changes in muscle length or

contraction level or in passive or active stiffness of the

vibrated muscles, due to the head posture, could modify

the mechanical effect of the vibratory stimulus (Burke

et al., 1976b).

However, during the stepping trials, no significant changes

in the head primary position were observed on lateralized

neck vibration in normal subjects. In the control experiments

on normal subjects, with the head deliberately turned, with or

without vibration, no significant relationship between head

yaw angle and body rotation angle was observed across sub-

jects and conditions, as already reported in Bove et al. (2002).

The spontaneous head position (yaw) did show some

vibration-induced changes in a few CD patients; however,

no significant relationships between head yaw and body rotation

angle during stepping-in-place across all the patients and

conditions (control and vibration) were observed. Further,

it might be noted that these relationships were characterized

by a similar slope, to underscore again a general lack of

dependence of spatial orientation effects on head postural

response to vibration of CD patients.

In spite of the variability across subjects, the absence of

significant body rotation when the head was deliberately

turned, in the normal subjects under both no vibration and

vibration conditions, points to the irrelevance of the natural

asymmetric neck input on the body rotation during the step-

ping trial. This result would be taken as evidence against the

possible influence of the muscle stiffness on the vibration-

induced afferent volley, because if stiffness or contraction had

conditioned the efficacy of the vibration, the spindle input

from the vibrated muscle would be expected to change

depending on the muscle being active and shortened or

being passive and lengthened during the head rotations.

In the treated patients studied here, the single dose of the

BTX was within the range commonly used in CD treatment

(between 15 and 100 IU). Further, the behaviour of the groups

did not differ with respect to the spontaneous head position,

BTX treatment and maximum range of voluntary head yaw

(Fig. 9). Interestingly, the spontaneous angular head deviation

with respect to the primary position (Stell et al., 1988; Tsui

et al., 1996) was not correlated with the degree of maximal

voluntary rotation of the head in any of the three groups. All

these features suggest that the absent or limited insensitivity to

vibration was not directly related to the disease effects on the

neck torsion (see below). In addition, CD patients who had

never received the BTX treatment showed the same responses

as the other patients. Moreover, treated and untreated patients

were distributed in the different CD groups and they shared

the same mean features for both spontaneous head position

and maximum head rotation angle of the other patients. This

would exclude the possibility that the observed ‘insensitivity’

was caused by the BTX injections, through its effects on

spindle afferent activity as seen in rat (Filippi et al., 1993;

Rosales, 1996) and humans (Priori et al., 1995; Naumann and

Reiners, 1997; Gilio et al., 2000). We do not know whether

larger doses or a longer period of BTX-A treatment could

affect the spindle’s transduction properties, or whether long-

term effects of the BTX could induce any effect on spindle

sensitivity. However, we observed that the two patients who

received several BTX-A injections on both SCM muscles

showed only a partial insensitivity to vibratory stimuli,

while other patients treated only on one neck side fall within

the INS group.

Quiet stanceUnder this condition, lateral neck vibration produces a medio-

lateral body displacement, opposite to the side of the vibration

Fig. 9 Maximum voluntary head rotation angle in the horizontalplane plotted against mean spontaneous head position in CDpatients. Patients were identified with different symbols for eachgroup. Black filled symbols correspond to the fiveBTX-A-untreated patients.

Proprioception and orientation in dystonia 2775

(Eklund, 1972; Bove et al., 2001, 2002) and it did so in our

normal subjects tested during stance. In CD patients, during

quiet upright stance, the effects of neck muscle vibration were

different from normal, since when all the patients’ data

collapsed were analysed, there was no effect of vibration

on posture. The absence or reduction of postural responses

was also seen by Lekhel et al. (1997) when the dorsal neck

muscles were vibrated.

Interestingly enough, during quiet upright stance, patients

belonging to the RIGHT and LEFT group revealed behaviours

reminiscent of that observed under the stepping-in-place con-

dition. In fact, these patients correctly responded only to

vibration of the same side of the neck that produced rotation

during stepping-in-place. It should be noted that in both

stepping-in-place and upright stance conditions, the ‘correct’

responses of CD patients implied body rotations or medio-

lateral displacements of similar amplitude to those observed

in normal subjects. The INS group showed no significant

medio-lateral displacements induced by vibration during

stance, and the same insensitivity to vibration was observed

during stepping-in-place. However, a small backward dis-

placement occurred, which was significantly different with

respect to the control condition when either side of the

neck was vibrated. This effect might be explained as a

postural response to a vibratory input interpreted by the

patients as if vibration were applied in front of them, rather

than laterally. This would correspond to a wrong central inter-

pretation of the input signal. In these otherwise ‘insensitive’

patients, therefore, vibration appears to have some postural

effects on the sagittal plane. Since these patients were not

those with the head turned most, the wrong interpretation

would not depend on the mechanisms responsible for the

abnormal head posture, as shown for other motor functions

(Anastasopoulos et al., 1998). In passing, all CD patients had

on average a posture slightly but significantly inclined

forward (body pitch) with respect to normals under all the

experimental conditions, but more so without vibration,

perhaps in the search for a ‘safer’ posture, as occurs in normals

on closing the eyes (Schieppati and Nardone, 1991;

Schieppati et al., 1994) or in spastic patients (Nardone

et al., 2001).

Reference frame for spatial orientationThe vibration effect on body orientation would be related to

the capacity of the neck proprioceptive input (in this case an

asymmetric lateralized input) to modify coherently the ego-

centric body-centred coordinate system that allows us to estim-

ate our body position with respect to the environment. The

vibration-induced input from one SCM muscle would mimic

head rotation toward the vibrated side, since during head

turning to one side it is the ipsilateral SCM which is being

lengthened passively, while the muscle opposite to the rota-

tion direction contracts and shortens (Mazzini and Schieppati,

1992). The tonic neck afferent signals thus elicited would

converge onto central networks and give an estimation of

the head and trunk posture. In fact, the imbalance of neck

proprioceptive inputs can modify the ‘representation’ of the

spatial orientation scheme in the context of whole body

control, inducing postural, visual and oculomotor responses

(Popov et al., 1999; Bove et al., 2001, 2002).

Normal subjects, who voluntarily rotated the head, showed

no significant modification in the body orientation under con-

trol or vibration conditions. From this finding, it seems that the

tonic afferent input connected with the abnormal head posture

is not enough to change the reference frame for orientation, at

least during stepping-in-place. Obviously, in the normal sub-

jects, voluntary head turning, as opposed to lateralized vibra-

tion, is accompanied by the brain activity connected to the

volition to turn the head (or the ‘efference copy’; von Holst

and Mittelstaedt, 1950), which can succeed in adapting

the reference frame during deliberate maintenance of

neck torsion.

In the CD patients, the abnormal spontaneous head posture

was ineffective, as was the deliberate rotation in the normal

subjects. In the patients, the trunk and the body do not appear

to match the vibration-induced sensory input from the neck, or

to re-align the trunk to the head, at least during stepping-

in-place. Perhaps as a consequence of their abnormal head

posture, they simply become less sensitive to neck muscle

vibration. Possibly, they have learned to rely less on the

proprioceptive sensory input from the SCM (or other neck

muscles) for their orientation during their navigation, and

rely more for this task on the input from other muscle

groups not influenced by the head posture, e.g. the trunk

paraspinal muscles (Anastasopoulos et al., 1998; De Nunzio

et al., 2003).

These findings do not allow the establishment of whether

proprioceptive abnormalities are a cause of the abnormal neck

muscle contractions or a consequence of an imbalance in other

systems controlling neck movements, e.g. the vestibular sys-

tem (Bronstein and Rudge, 1986; Mazzini and Schieppati,

1994; Colebatch et al., 1995; Munchau and Bronstein,

2001). On the other hand, neck input shapes the output of

the vestibular nucleus and its contribution to posture, gaze and

perception (Gdowski and McCrea, 1999, 2000, in the

primate). Admittedly, we are not in the position to interpret

the ‘insensitivity’ to vibration in these patients in the light

of any possible subclinical vestibular deficit or of the abnor-

mal interaction between proprioceptive neck and (normal)

vestibular input. Perhaps, on the basis of a recent suggestion

of Anastasopoulos et al. (2003), the abnormal response to

lateral neck vibration in CD patients can be explained as

the consequence of an offset of a non-sensory set point signal

in the neck proprioceptive loop for head-on-trunk control. It is

also difficult to draw conclusions from one snapshot of a

population of patients with a different natural history of dis-

ease. We can only note that the mean disease duration of the

INS group (12.4 6 6.1 years) was longer than that in the

RIGHT and LEFT groups (RIGHT, 7.3 6 6.1 years;

LEFT, 7.3 6 3.5 years), while the mean age was not different

in the three groups.

2776 M. Bove et al.

ConclusionsOne may hypothesize that the reference system used in the

control of body orientation in space by the CD patients during

a locomotor task is refractory to proprioceptive lateral input

from the neck or that the input has been re-addressed or

biased, whereby the vibration to both sides produces an

orientation shift in the same sense. This refractoriness or

distortion would be either primitive, connected to the patho-

genesis of the disease, or the result of an adaptive process,

whereby the proprioceptive input from muscles having an

abnormal length or tone is being progressively cancelled. If

anything, it seems that this refractoriness increases and occu-

pies both sides with the progress of the disease. This phenom-

enon might also entrain a slow shift from a reference system

based on the head position to a more reliable reference based

on another part of the body, such as the trunk; in this process,

head posture control is obviously lost.

AcknowledgementsThis work was supported by a grant from the Italian Ministero

Istruzione Universita Ricerca (MIUR), FIRB 2001 and

PRIN 2003.

References

Abbruzzese G, Berardelli A. Sensorimotor integration in movement disorders.

[Review]. Mov Disord 2003; 18: 231–40.

Abbruzzese G, Marchese R, Buccolieri A, Gasparetto B, Trompetto C.

Abnormalities of sensorimotor integration in focal dystonia: a transcranial

magnetic stimulation study. Brain 2001; 124: 537–45.

Anastasopoulos D, Mergner T. Canal–neck interaction in vestibular nuclear

neurons of the cat. Exp Brain Res 1982; 46: 269–80.

Anastasopoulos D, Bhatia K, Bisdorff A, Bronstein AM, Gresty MA, Marsden

CD. Perception of spatial orientation in spasmodic torticollis. Part I: the

postural vertical. Mov Disord 1997a; 12: 561–9.

Anastasopoulos D, Bhatia K, Bronstein AM, Marsden CD, Gresty MA.

Perception of spatial orientation in spasmodic torticollis. Part 2: the visual

vertical. Mov Disord 1997b; 12: 709–14.

Anastasopoulos D, Nasios G, Psilas K, Mergner T, Maurer C, Lucking CH.

What is straight ahead to a patient with torticollis? Brain 1998; 121:

91–101.

Anastasopoulos D, Nasios G, Mergner T, Maurer C. Idiopathic spasmodic

torticollis is not associated with abnormal kinesthetic perception from neck

proprioceptive and vestibular afferences. J Neurol 2003; 250: 546–55.

Biguer B, Donaldson IM, Hein A, Jeannerod M. Neck muscle vibration

modifies the representation of visual motion and direction in man. Brain

1988; 111: 1405–24.

Bonanni M, Newton R. Test–retest reliability of the Fukuda Stepping Test.

Physiother Res Int 1998; 3: 58–68.

Bottini G, Karnath HO, Vallar G, Sterzi R, Frith CD, Frackowiak RS, Paulesu

E. Cerebral representations for egocentric space: functional–anatomical

evidence from caloric vestibular stimulation and neck vibration. Brain

2001; 124: 1182–96.

Bove M, Diverio M, Pozzo T, Schieppati M. Neck muscle vibration disrupts

steering of locomotion. J Appl Physiol 2001; 91: 581–8.

Bove M, Courtine G, Schieppati M. Neck muscle vibration and spatial orien-

tation during stepping in place in humans. J Neurophysiol 2002; 88:

2232–41.

Boyadjian A, Marin L, Danion F. Veering in human locomotion: the role of

the effectors. Neurosci Lett 1999; 265: 21–4.

Brandt T. Cervical vertigo—reality or fiction? [Review]. Audiol Neurootol

1996; 1: 187–96.

Brandt T, Bronstein AM. Cervical vertigo. J Neurol Neurosurg Psychiatry

2001; 71: 8–12.

Bronstein AM, Rudge P. Vestibular involvement in spasmodic torticollis.

J Neurol Neurosurg Psychiatry 1986; 49: 290–5.

Burke D, Hagbarth KE, Lofstedt L, Wallin BG. The responses of human

muscle spindle endings to vibration during isometric contraction. J Physiol

(Lond) 1976a; 261: 673–93.

Burke D, Hagbarth KE, Lofstedt L, Wallin BG. The responses of human

muscle spindle endings to vibration of non-contracting muscles. J Physiol

(Lond) 1976b; 261: 695–711.

Cohen LA. Role of eye and neck proprioceptive mechanisms in body orienta-

tion and motor coordination. J Neurophysiol 1961; 24: 1–11.

Colebatch JG, Di Lazzaro V, Quartarone A, Rothwell JC, Gresty M. Click-

evoked vestibulocollic reflexes in torticollis. Mov Disord 1995; 10: 455–9.

Couch JR. Dystonia and tremor in spasmodic torticollis. Adv Neurol 1976; 14:

245–58.

Courtine G, Pozzo T, Lucas B, Schieppati M. Continuous, bilateral Achilles’

tendon vibration is not detrimental to human walking. Brain Res Bull 2001;

55: 107–15.

De Nunzio A, Stapley P, Schmid M, Courtine G, Schieppati M. Trunk muscle

vibration disrupts the steering of locomotion. Proceedings of the 3rd

International Posture Symposium, Smolenice, Slovakia; 2003.

Eklund G. General features of vibration-induced effects on balance. Uppsala J

Med Sci 1972; 77: 112–24.

Filippi GM, Errico P, Santarelli R, Bagolini B, Manni E. Botulinum A toxin

effects on rat jaw muscle spindles. Acta Otolaryngol 1993; 113: 400–4.

Fransson PA, Karlberg M, Sterner T, Magnusson M. Direction of

galvanically-induced vestibulo-postural responses during active and

passive neck torsion. Acta Otolaryngol 2000; 120: 500–3.

Fukuda T. The stepping test. Two phases of the labyrinthine reflex. Acta

Otolaryngol 1959; 50: 95–108.

Gilio F, Curra A, Lorenzano C, Modugno N, Manfredi M, Berardelli A.

Effects of botulinum toxin type A on intracortical inhibition in patients

with dystonia. Ann Neurol 2000; 48: 20–6.

Gdowski GT, McCrea RA. Integration of vestibular and head movement

signals in the vestibular nuclei during whole body rotation. J Neurophysiol

1999; 82: 436–49.

Gdowski GT, McCrea RA. Neck proprioceptive inputs to primate vestibular

nucleus neurons. Exp Brain Res 2000; 135: 511–26.

Goodwin GM, McCloskey DI, Matthews PB. The contribution of muscle

afferents to kinaesthesia shown by vibration induced illusions of

movement and by the effects of paralysing joint afferents. Brain 1972;

95: 705–48.

Gurfinkel VS, Ivanenko YP, Levik YS, Babakova IA. Kinaesthetic reference

for human orthograde posture. Neuroscience 1995; 68: 229–43.

Hallett M. Is dystonia a sensory disorder? Ann Neurol 1995; 38: 139–40.

Hlavacka F, Njiokiktjien C. Postural responses evoked by sinusoidal galvanic

stimulation of the labyrinth. Influence of head position. Acta Otolaryngol

1985; 99: 107–12.

Ivanenko YP, Grasso R Lacquaniti F. Effect of gaze on postural responses to

neck proprioceptive and vestibular stimulation in humans. J Physiol

(London) 1999; 519: 301–14.

Ivanenko YP, Grasso R, Lacquaniti F. Neck muscle vibration makes walking

humans accelerate in the direction of gaze. J Physiol (Lond) 2000; 525:

803–14.

Jankovic J, Leder S, Warner D, Schwartz K. Cervical dystonia: clinical

findings and associated movement disorders. Neurology 1991; 41:

1088–91.

Karnath HO. Subjective body orientation in neglect and the interactive

contribution of neck muscle proprioception and vestibular stimulation.

Brain 1994; 117: 1001–12.

Karnath HO, Christ K, Hartje W. Decrease of controlateral neglect by neck

muscle vibration and spatial orientation of trunk midline. Brain 1993; 116:

383–96.

Karnath HO, Sievering D, Fetter M. The interactive contribution of neck

muscle proprioception and vestibular stimulation to subjective ‘straight

ahead’ orientation in man. Exp Brain Res 1994; 101: 140–6.

Proprioception and orientation in dystonia 2777

Karnath HO, Konczak J, Dichgans J. Effect of prolonged neck muscle

vibration on lateral head tilt in severe spasmodic torticollis. J Neurol

Neurosurg Psychiatry 2000; 69: 658–60.

Kavounoudias A, Gilhodes JC, Roll R, Roll JP. From balance regulation to

body orientation: two goals for muscle proprioceptive information

processing? Exp Brain Res 1999; 124: 80–8.

Kogler A, Lindfors J, Odkvist LM, Ledin T. Postural stability using different

neck positions in normal subjects and patients with neck trauma. Acta

Otolaryngol 2000; 120: 151–5.

Ledin T, Hafstrom A, Fransson PA, Magnusson M. Influence of neck

proprioception on vibration-induced postural sway. Acta Otolaryngol

2003; 123: 594–9.

Leis AA, Dimitrijevic MR, Delapasse JS, Sharkey PC. Modification of

cervical dystonia by selective sensory stimulation. J Neurol Sci 1992;

110: 79–89.

Lekhel H, Popov K, Anastasopoulos D, Bronstein A, Bhatia K, Marsden CD,

et al. Postural responses to vibration of neck muscles in patients with

idiopathic torticollis. Brain 1997; 120: 583–91.

Lund S. Postural effects of neck muscle vibration in man. Experientia 1980;

36: 1398.

Lyon IN, Day BL. Control of frontal plane body motion in human stepping.

Exp Brain Res 1997; 115: 345–56.

Matthews PB. Proprioceptors and their contribution to somatosensory

mapping: complex messages require complex processing. [Review]. Can

J Physiol Pharmacol 1988; 66: 430–8.

Mazzini L, Schieppati M. Preferential activation of the sternocleidomastoid

muscles by the ipsilateral motor cortex during voluntary rapid head rota-

tions in humans. In: Berthoz A, Vidal PP, Graf W, editors. The head–neck

sensory motor system, Oxford: Oxford University Press; 1992. p. 597–600.

Mazzini L, Schieppati M. Short-latency neck muscle responses to vertical

body tilt in normal subjects and in patients with spasmodic torticollis.

Electroencephalogr Clin Neurophysiol 1994; 93: 265–75.

Munchau A, Bronstein AM. Role of the vestibular system in the pathophy-

siology of spasmodic torticollis. [Review]. J Neurol Neurosurg Psychiatry

2001; 71: 285–8.

Munchau A, Corna S, Gresty MA, Bhatia KP, Palmer JD, Dressler D, et al.

Abnormal interaction between vestibular and voluntary head control in

patients with spasmodic torticollis. Brain 2001; 124: 47–59.

Nardone A, Galante M, Lucas B, Schieppati M. Stance control is not affected

by paresis and reflex hyperexcitability: the case of spastic patients. J Neurol

Neurosurg Psychiatry 2001; 70: 635–43.

Naumann M, Reiners K. Long-latency reflexes of hand muscles in idiopathic

focal dystonia and their modification by botulinum toxin. Brain 1997; 120:

409–16.

Norre ME, Forrez G, Beckers A. Vestibulospinal findings in two syndromes

with spontaneous vertigo attacks. Ann Otol Rhinol Laryngol 1989; 98:

191–5.

Nutt JG, Muenter MD, Aronson A, Kurland LT, Melton LJ. Epidemiology of

focal and generalized dystonia in Rochester, Minnesota. Mov Disord 1988;

3: 188–94.

Pompeiano O, Manzoni D, Barnes CD. Responses of locus coeruleus neurons

to labyrinth and neck stimulation. [Review]. Prog Brain Res 1991; 88:

411–34.

Popov KE, Lekhel H, Faldon M, Bronstein AM, Gresty MA. Visual

oculomotor responses induced by neck vibration in normal subjects and

labyrinthine-defective patients. Exp Brain Res 1999; 128: 343–52.

Priori A, Berardelli A, Mercuri B, Manfredi M. Physiological effects

produced by botulinum toxin treatment of upper limb dystonia.

Changes in reciprocal inhibition between forearm muscles. Brain 1995;

118: 801–7.

Roll JP, Vedel JP. Kinaesthetic role of muscle afferents in man, studied by

tendon vibration and microneurography. Exp Brain Res 1982; 47: 177–90.

Roll JP, Vedel JP, Roll R. Eye, head and skeletal muscle spindle feedback in

the elaboration of body references. Prog Brain Res 1989; 80: 113–23.

Rosales RL, Arimura K, Takenaga S, Osame M. Extrafusal and intrafusal

muscle effects in experimental botulinum toxin-A injection. Muscle Nerve

1999; 19: 488–96.

Schieppati M, Nardone A. Free and supported stance in Parkinson’s

disease. The effect of posture and ‘postural set’ on leg muscle responses

to perturbation, and its relation to the severity of the disease. Brain 1991;

114: 1227–44.

Schieppati M, Hugon M, Grasso M, Nardone A, Galante M. The limits of

equilibrium in young and elderly normal subjects and in parkinsonians.

Electroencephalogr Clin Neurophysiol 1994; 93: 286–98.

Schieppati M, Nardone A, Schmid M. Neck muscle fatigue affects postural

control in man. Neuroscience 2003; 121: 277–85.

Siggelkow S, Kossev A, Moll C, Dauper J, Dengler R, Rollnik JD. Impaired

sensorimotor integration in cervical dystonia: a study using transcranial

magnetic stimulation and muscle vibration. J Clin Neurophysiol 2002; 19:

232–9.

Stell R, Thompson PD, Marsden CD. Botulinum toxin in spasmodic

torticollis. J Neurol Neurosurg Psychiatry 1988; 51: 920–3.

Tarantola J, Nardone A, Tacchini E, Schieppati M. Human stance stability

improves with the repetition of the task: effect of feet position and visual

condition. Neurosci Lett 1997; 228: 75–8.

Taylor J, McCloskey DI. Illusions of head and visual target displacement

induced by vibration of neck muscles. Brain 1991; 114: 755–9.

Tinazzi M, Rosso T, Fiaschi A. Role of the somatosensory system in primary

dystonia. [Review]. Mov Disord 2003; 18: 605–22.

Tsui JK. Botulinum toxin as a therapeutic agent. [Review]. Pharmacol Ther

1996; 72: 13–24.

von Holst E, Mittelstaedt H. Das Reafferenzprinzip. Weschselwirkung

zwischen Zentralnervensystem und Peripherie. Naturwissenschaften

1950; 37: 464–76.

Wierzbicka MM, Gilhodes JC, Roll JP. Vibration-induced postural

posteffects. J Neurophysiol 1998; 79: 143–50.

Wober C, Schnider P, Steinhoff N, Trattnig S, Zebenholzer K, Auff E.

Posturographic findings in patients with idiopathic cervical dystonia before

and after local injections with botulinum toxin. Eur Neurol 1999;

41: 194–200.

2778 M. Bove et al.