Movement Cueing and Motor Execution in Patients WithDystonia: A Kinematic Study

*†Antonio Curra, MD, *†Alfredo Berardelli, MD, *Rocco Agostino, MD, *Morena Giovannelli, MD,*Giacomo Koch, MD, and *†Mario Manfredi, MD

*Dipartimento di Scienze Neurologiche, Universita´ di Roma “La Sapienza,” Rome, Italy; and †Istituto NeurologicoMediterraneo Neuromed, Pozzilli, IS, Italy

Summary: To investigate whether the type of movement cue-ing influences motor performance in patients with dystonia, westudied externally triggered (ET) and self-initiated (SI) sequen-tial rapid arm movements in patients with generalized or focaldystonia and healthy control subjects. The ET task requiredsubjects to initiate movements in response to consecutive visualcues; the SI task allowed them to start at will. To determinewhether patients found sequential motor tasks more difficultthan single tasks, we also analyzed single ET movements. Con-trol subjects performed the SI task significantly faster than theET task. Their single ET movements and first ET sequentialsubmovements had similar speeds. Patients with generalizeddystonia were slow in performing the single movement, the ETand the SI sequential tasks, and they executed the SI sequencemore slowly than the ET. They made long pauses between SIsequential submovements, had longer reaction times during the

ET sequences, and performed the first ET submovement moreslowly than the single ET movement. Patients with focal dys-tonia had normal reaction times but they performed single andsequential tasks slowly, made long pauses during SI tasks, andalso executed the first ET submovement more slowly than thesingle ET movement. Our findings indicate that patients withdystonia have a general impairment of sequential movements.The more marked slowness in executing SI than ET movementsobserved in patients with generalized dystonia shows that dys-tonia impairs internal cueing more than external cueing mecha-nisms. Overall, these findings imply abnormal activation ofprimary and nonprimary motor areas during movement in dys-tonia. The greater impairment of SI tasks as well as the delayedmotor responses during ET task suggest predominant underac-tivity of the supplementary motor area.Key Words: Internalcueing—External cueing—Motor sequences—Dystonia.

Dystonia is a syndrome characterized by prolongedmuscle contractions leading to sustained and twistingmovements and causing abnormal postures of the af-fected body segments.1 Besides dystonic movements,clinical examination of patients with dystonia shows ab-normal execution of voluntary movements. In patientswith hand dystonia, neurophysiological investigationsdisclose reduced selectivity in performing independentfinger movements and overflow of muscle activity withco-contraction of agonist and antagonist muscle groups.2

Investigation of rapid single-joint movements shows thatdystonic patients perform elbow flexion movements atreduced speed and with prolonged electromyographic ac-tivity in agonists and antagonists. The velocity profiles

show accelerative and decelerative phases of similar du-ration indicating that motor programs governing the ex-ecution of rapid elbow movements are preserved in dys-tonia.3 Motor performance abnormalities have also beenreported during complex, horizontal planar and reachingmovements. In a study of self-initiated sequential planarmovements, Agostino et al.4 found that dystonic patientswere slower than control subjects in completing indi-vidual movements and took longer than control subjectsto switch from one submovement to the next, but did notshow progressive slowing with sequence completion. Inaddition, when performing a single movement as a sub-movement of a motor sequence, patients with dystoniabecame even slower (“extra-slowness”). Inzelberg et al.5

studied time-constrained arm-reaching movements ex-ecuted in a fully cued reaction time paradigm andshowed that patients with dystonia had normal reactiontimes, were slightly slower than control subjects, and hada prolonged deceleration phase. This latter abnormality

Received July 7, 1998; revision received December 2, 1998. Ac-cepted August 12, 1999.

Address correspondence and reprint requests to Alfredo Berardelli,MD, Dipartimento di Scienze Neurologiche, Universita´ di Roma “LaSapienza,” Viale dell’Universita´ 30, 00185 Rome, Italy.

Movement DisordersVol. 15, No. 1, 2000, pp. 103–112© 2000 Movement Disorder Society

103

worsened without visual feedback of the moving limb,suggesting defective integration of sensory feedback cor-rections in dystonia.

More information is needed to explain whether move-ment cueing influences motor performance. In normalsubjects, we recently investigated sequential movementsexecuted with and without advance knowledge of themotor pathway.6 When the sequence path was unknown,the subjects moved in response to consecutively appear-ing targets that specified the end position of each sub-movement; when the path was known, they started themovement at will. In this way, the unknown sequenceswere externally triggered, whereas the known sequenceswere self-initiated. We found that subjects moved fasterwhen they knew the pathway in advance, indicating thatinternal determination of movement influenced the speedof execution.

Although there is no simple dichotomy,7 external andinternal motor initiation are thought to be under the con-trol of distinct motor systems.8,9 The lateral premotorarea may be more concerned with external initiation10;the supplementary motor area (SMA) may be more con-cerned with internal initiation.11,12Although the cerebel-lum and basal ganglia project to both nonprimary motorareas,13,14 the lateral premotor area mainly receives thecerebellar outputs,15,16whereas the supplementary motorarea mainly receives the basal ganglia outputs.17,18

Hence, the basal ganglia-supplementary motor area cir-cuit (medial motor system) mainly controls SI move-ments, whereas the cerebellum-lateral premotor cortexcircuit (lateral motor system) preferentially mediates ETmovements.

Clinical and experimental evidence suggests a basalganglia abnormality in dystonia.19,20If the basal ganglia-SMA circuit preferentially controls self-initiated (SI)movements, then in patients with dystonia this taskshould be more impaired than that which is externallytriggered (ET). Prompted by this question, we comparedrapid sequential arm movements executed in a zig-zagpattern by patients with focal and generalized dystonia inresponse to internal cues (SI sequence) and to consecu-tive external visual triggers (ET condition). During SIbut not ET sequences, the internal cueing mechanism notonly activates submovement execution, but also controlsthe switching between submovements. Because this pro-cess is altered in dystonia, it might act as a confoundingfactor in the comparison of ET and SI sequences. Hence,we also investigated the execution of the same zig-zagpattern during a mixed-cued condition relying on bothexternal and internal cueing mechanisms. The executionof the whole sequence had to be initiated in response tothe simultaneous activation of all visual triggers, whereasthe switching between submovements was internallycued. SI sequences therefore differed from mixed-cuedsequences only in self-initiation.

Finally, we also analyzed the execution of a single ETmovement to investigate whether extra-slowness is evi-dent for externally cued motor acts.

METHODS

SubjectsNine patients with generalized dystonia (six men and

three women, mean age ± standard deviation, 30.6 ± 14.6yrs), six patients with focal, task-specific dystonia, thatis, writer’s cramp (five men and one woman, mean age49.8 ± 11.13 yrs) and two groups of age-matched healthysubjects (young control subjects, five men and fourwomen, mean age 31.5 ± 13 yrs for comparison withpatients with generalized dystonia; older control sub-jects, four men and one woman, mean age 50.8 ± 13 yrs,for comparison with patients with focal dystonia) par-ticipated in the study. Informed consent was obtainedfrom all the participants and the study was approved bythe local ethics committee. Clinical features of the pa-tients were assessed by the Fahn-Marsden EvaluationScale for Dystonia (FMESD, Table 1). All patients haddystonia of the right upper limb, and none were receivingneuroleptic medication. Patients with focal dystonia weretreated with botulinum toxin injections. All of them werestudied at least 4 months after the last treatment.

Apparatus and TasksMovements in the three-dimensional space were re-

corded by the ELITE motion analysis system (Bioengi-

TABLE 1. Clinical features of patients with dystonia

PatientsAge(yrs)

Dystoniamovement

scale

Dystoniadisability

scale

FMESDtotalscore

G1 28 35 5 40G2 26 92 8 100G3 24 56 6 62G4 24 41 8 49G5 16 83 13 96G6 19 27 7 34G7 32 56 10 66G8 46 25.5 7 32.5G9 61 11 6 17F1 65 1 3 4F2 55 8 2 10F3 45 2 2 4F4 54 3 1 4F5 32 1 1 2F6 48 2 0 2

G, patients with generalized dystonia; F, patients with focal dystonia.Dystonia movement scale and dystonia disability scale are subscales

of the FMESD (Fahn-Marsden Evaluation Scale for Dystonia).

A. CURRÁ ET AL.104

Movement Disorders, Vol. 15, No. 1, 2000

neering Technology and Systems [BTS, Milano, Italy]).Two infrared ray cameras recorded, at a 100-Hz sam-pling rate, the motion of a passive marker placed on thedistal phalanx of the second finger of the moving hand.A real-time TV-image converting processor connected tothe cameras digitalized the analogic data and recon-structed the coordinates of the marker motion. Coordi-nates were provided on the x, y, and z axes of a previ-ously calibrated 3-D reference system. Kinematic param-eters obtained from the mathematic arrangement ofspatial coordinates were displayed in graphic form.

As previously described,6 the motor tasks consisted ofjoining six visual targets by making five rapid zig-zagarm movements starting from the lower left and movingtoward the upper right target (Fig. 1). The subjects satcomfortably in front of a screen with their upper limbheld pronated at the shoulder, the forearm moderatelyflexed at the elbow, and the index finger extended. Thetargets were arranged in two parallel rows and the dis-tance between the target centers was kept constant so thatthe motor sequence consisted of a zig-zag pathway com-prising five segments of equal length (350 mm). Move-ment targets consisted of 35-mm-sided squares whichcould be displayed empty or solid depending on the typeof condition studied.

The subjects performed the motor sequence at theirmaximum speed, made five arm strokes, and stoppedinside the square targets “as accurately as possible.” Af-ter a few practice trials, blocks of 10 trials were recordedfor each condition. Conditions were presented in randomorder. To avoid fatigue, subjects were allowed a shortrest between trials and between blocks.

Experimental Paradigm

All sequences were visually guided, that is, the zig-zagpathway was displayed on the screen before the subjectsmoved so they could examine the sequence path freelybefore starting the task. The motor sequence was per-formed under self-initiated and externally triggered con-ditions. Self-initiated movements were started at the sub-jects’ will; externally triggered movements were startedin response to external stimuli.

In the ET condition (Fig. 1), subjects moved in re-sponse to the consecutive activation of visual targets.Target activation was signalled by changing the targetfrom white to black squares and occurred in a zig-zagorder (from the upper left to the upper right target). Asoftware program controlled activation to vary the inter-vals between each submovement in the sequence andeach sequence in the collection block. In this way, thesubjects could never predict the time to start submove-ments as they all confirmed at the end of the experimen-

tal session. The intervals were long enough to allow theongoing submovement to be terminated before the pro-gram activated the next target.

In the SI condition (Fig. 1), all six targets appeared onthe screen as solid black squares before the subjectsstarted the movement. After examining the pathway, the

FIG. 1. Sequential pathways for the single movement and the se-quence in the three conditions studied. Under the externally triggeredcondition, subjects assumed the starting position at the target indicatedby the arrow. The activation (from white to black) of a target repre-sented the signal for the subjects to start moving and perform thesubmovement as fast as possible stopping inside the activated target.Panels 1 to 6 show the consecutive activation of the visual targets.Under the mixed-cued condition, after assuming the starting position,subjects executed the whole sequence in response to the simultaneousactivation of all the targets, moving as fast as possible and stoppinginside the targets as briefly as possible. The single movement wastested only under the ET condition. Under the self-initiated condition,after assuming the starting position, the subject started the execution attheir will, moving as fast as possible and stopping inside the targets asbriefly as possible.

MOVEMENT CUEING IN DYSTONIA 105

Movement Disorders, Vol. 15, No. 1, 2000

subjects initiated the motor sequence in their own time,moving in a zig-zag pattern from the lower left to theupper right target as fast as possible and stopping insidethe targets as briefly as possible.

ET and SI sequences were studied in normal subjectsand in all patients with generalized and focal dystonia. Insix of nine patients with generalized dystonia and six ofnine control subjects we also investigated a mixed-cuedcondition (Fig. 1). From the initial position (lower lefttarget), subjects had to start moving as fast as possible inresponse to the simultaneous activation of all the visualtargets and to complete the zig-zag pattern toward theupper right target. The mixed-cued sequences differedfrom the SI sequences only for initial external cueing. Inthe mixed-cued condition, the intervals preceding the vi-sual go-signal were the same as those used to trigger thefirst submovement of the ET sequences.

In all subjects we also analyzed performance of asingle ET movement (Fig. 1), identical in length, orien-tation, and spatial position to the first sequential sub-movement and similar to the intervals preceding the gosignal.

Analysis of Kinematic Variables

Three-plane (vertical, sagittal, and horizontal) refer-enced kinematics were reconstructed by the ELITE soft-ware for each trial. Paths of submovement trajectorieswere plotted on each plane and analyzed by visual in-spection with special focus on the vertical plane onwhich movements primarily occurred.

Movement onset and termination were determined onthe velocity and displacement profiles using the arbitraryvalue of 50 mm/sec as a threshold for hand motion. Thestart of each submovement was considered when the ve-locity values reached and stayed over 50 mm/sec and theend of the submovement when values fell below thisvalue. Coordinates of hand position obtained from thedisplacement profile at movement onset and terminationprovided submovement amplitudes. In addition, if thesubject corrected the final hand position, the correctionswere excluded from submovement duration by inspec-tion of the displacement profile. Pauses between sub-movements executed during the SI condition were cal-culated as the time elapsing between the end of onesubmovement and the onset of the next. The reactiontime was calculated as the time elapsing between thetrigger and the onset of the movement.

Statistical Analysis

All data are expressed as mean ± standard deviationobtained from 10 trials for each condition. Data frompairs of age-matched groups, that is, patients with gen-

eralized dystonia/young control subjects and patientswith focal dystonia/older control subjects, were analyzedseparately using multivariate analyses of variance(MANOVA).

Amplitude and movement time of each submovementwere subjected to a three-way ANOVA with the factorsgroup (normal versus patient), condition (ET versus SI),and position (of the submovement in the sequence, 1-2-3-4-5). A two-way ANOVA with the factors group andposition was used to analyze reaction times during theET sequences. Pauses between submovements during theSI and the mixed-cued sequences were subjected to sepa-rate MANOVAs with the factors group and position (ofthe pause in the sequence, 1-2-3-4) and also with factorcondition for comparisons between SI and mixed-cuedsequences.

In addition, we compared the execution of singlemovement with first sequential ET submovement. Forreaction times and movement times, we used a two-wayANOVA with the factors group and condition (singlemovement versus first ET submovement). p values lessthan 0.05 were considered to indicate statistical signifi-cance.

RESULTS

After a brief practice period, all normal subjects anddystonic patients were able to perform the motor se-quences correctly.

Submovement Amplitudes

Data for submovement amplitudes in SI and ET (Fig.2) sequences showed that all groups executed shorter ETthan SI submovements (F for factor condition in patientswith generalized dystonia/young control subjects419.8; in patients with focal dystonia/older control sub-jects F4 70.1; for both values p <0.01). ET submove-ments were 18 mm shorter in patients with generalizeddystonia/young control subjects and 27 mm shorter inpatients with focal dystonia/older control subjects. Pa-tients with generalized dystonia executed shorter move-ments than normal subjects (11.8 mm, F for factor group4 8.1; p <0.01). Because the submovement end pointsalways came within target area, note that differences inmovement amplitude observed between groups and con-ditions do not indicate inaccuracy. In addition, the slightincrease in movement amplitude observed during the SIcondition was probably related to the higher movementvelocity during the SI than during the ET condition. Asa measure of variability, the standard deviation of move-ment amplitudes was similar in the various groups for allsubmovements and conditions.

A. CURRÁ ET AL.106

Movement Disorders, Vol. 15, No. 1, 2000

Submovement Times

Submovement times for SI and ET sequences (Fig. 3)showed that all patients moved slower than normal sub-jects (F for factor group in patients with generalizeddystonia/young control subjects4 205.4; in patientswith focal dystonia/older control subjects F4 121.2; forboth values p <0.01). In the pair patients with focal dys-tonia/older control subjects, the SI sequences were ex-ecuted faster than the ET (F for factor condition4 43.1;p <0.01). In the pair patients with generalized dystonia/young control subjects, the patients executed the SI andET sequences at similar speeds, whereas normal subjectsexecuted SI sequences faster than ET (F for two-wayinteraction of group by condition4 4.3; p <0.03).

Submovement Reaction Times During theET Sequence

Patients with generalized dystonia had longer sub-movement reaction times than control subjects (F forfactor group4 15.4; p <0.01). Patients with focal dys-tonia had normal reaction times (Fig. 4).

Duration of Pauses in the SI Condition

All patients paused longer than control subjects duringthe SI condition (F for factor group in patients with gen-eralized dystonia/young control subjects4 26.3; in pa-tients with focal dystonia/older control subjects: F44.6; p <0.03; Fig. 5).

FIG. 2. Submovement amplitudes forexternally triggered and self-initiatedsequences. Left panel: data from thepaired patients with generalizeddystonia/young control subjects; rightpanel: data from the paired patientswith focal dystonia/older control sub-jects. Bars represent standard devia-tion.

FIG. 3. Submovement times for ex-ternally triggered and self-initiated se-quences. Left panel: data from thepaired patients with generalizeddystonia/young control subjects; rightpanel: data from the paired patientswith focal dystonia/older control sub-jects. Bars represent standard devia-tion.

MOVEMENT CUEING IN DYSTONIA 107

Movement Disorders, Vol. 15, No. 1, 2000

Comparison of the Execution of Single Movementsand the First Sequential ET Movements

Patients had normal reaction times for single move-ments and for the first submovements of the ET sequence(F for factor group4 1.4, F for factor condition4 2.0;for both values p <0.2). They were slower than controlsubjects in executing the single ET movement and thefirst ET submovement (F for factor group in patientswith generalized dystonia/young control subjects445.3; in patients with focal dystonia/older control sub-jects, F4 12.2; for both values, p <0.01) and executedthe first ET submovement significantly more slowly thansingle ET movement (F for factor condition in patientswith generalized dystonia/young control subjects4 7.5,p <0.01; in patients with focal dystonia/older controlsubjects F4 8.5, p <0.02; Fig. 6).

Comparison of the Execution of Mixed-CuedSequences and SI Tasks in Subgroups of Patients

With Generalized Dystonia/Young Control Subjects

Patients with generalized dystonia were slower thancontrol subjects (F for factor group4 203.4; p <0.01)and had longer pauses (F for factor group4 20.0;p <0.01). Both groups executed faster movements (F forfactor condition4 32.1; p <0.01) and pauses (F forfactor condition4 12.1; p <0.01) during the mixed-cuedthan the SI sequence (Fig. 7). Patients with generalizeddystonia reduced the movement time from the SI to themixed-cued condition more than control subjects (24.5%

versus 17.7%; F for two-way interaction of group bycondition4 6.9; p <0.01). They also shortened the timespent at the targets between submovements (F for two-way interaction of group by condition4 7.1; p <0.01).The smaller reduction in movement time and pause du-rations observed in normal subjects was clearly the resultof a floor effect.

Comparison of the reaction times for the mixed-cuedsequence, for the single movement, and for the first sub-movement of the ET sequence yielded no statisticallysignificant difference between groups and conditions.

DISCUSSION

To find out whether internal cueing, preferentiallymodulated by the basal ganglia-SMA circuit, is moreimpaired than external cueing in patients with dystonia,we compared the execution of self-initiated (SI), exter-nally triggered (ET), and mixed-cued (SI and ET) se-quential rapid arm movements in patients with focal andgeneralized dystonia. Our results show that patients withdystonia have a general impairment of movement se-quencing. They also specify that dystonia impairs inter-nally cued motor tasks more than externally cued tasks.

Lesion studies in animals suggest that movement cue-ing is under the control of nonprimary motor areas. Ac-cordingly, bilateral cooling of SMA impairs tasks requir-ing internal preparatory processes21 but not visually trig-gered simple movements.22 On the other hand, bilateralcooling of the lateral premotor area disrupts cued-reaction time tasks23 but does not disrupt the perfor-mance of a fixed sequence of fore-limb movements.24

The functional selectivity of nonprimary motor areasemerging from lesion studies has been only partially con-firmed by cell recordings. Although most SMA neuronsfiring before self-initiated movements are not activatedbefore externally triggered movements,25–27 other evi-dence failed to show overall differences in SMA activityfor visually triggered and self-paced movements.28 De-spite the absence of a strict functional dichotomy,7 it istherefore accepted that the supplementary motor area ismore related to SI movements, whereas the lateral pre-motor cortex preferentially mediates ET movements. Inpeople this view is partially supported by PET scan-ning,29–34functional magnetic resonance imaging,35 andstudies of movement-related potentials.32,36,37Althoughthe SMA and the lateral premotor area are both activatedduring ET and SI movements,7 the reciprocal degree oftheir involvement during these tasks differs.31,32,34,38

The SI condition in our experimental paradigm prefer-entially activates the SMA, whereas the ET conditionmainly activates the lateral premotor cortex. The kine-

FIG. 4. Reaction times for externally triggered sequence. Upper panel:data from the patients with generalized dystonia/young control sub-jects; lower panel: data from the patients with focal dystonia/oldercontrol subjects. Bars represent standard deviation.

A. CURRÁ ET AL.108

Movement Disorders, Vol. 15, No. 1, 2000

matic features of ET and SI tasks may therefore differaccording to the degree of medial or lateral motor systemactivation that the movement demands.

Normal subjects performed SI sequences faster thanET sequences. We specifically designed our experimentsso that certain motor control variables were similar in theET and SI tasks whereas others differed. Under bothconditions, subjects received advance information on thedirection and amplitude of each submovement of thesequence. They also moved at their maximum speed.Hence, differences in the performance of SI and ET tasksunquestionably arose from mechanisms other than pro-gramming of movement direction, amplitude, and speed.Interestingly, animal studies have demonstrated that neu-rons precisely for these variables are commonly found inprimary motor cortex, nonprimary motor areas,39,40 andalso in some cortical areas in the parietal lobe.41

The motor variable that differed in the ET and SI taskswas the timing of submovement execution. In the SIcondition, internal cueing allowed subjects to predict thetiming of each submovement, thus the automatic execu-tion of the sequence. In the ET condition, the subjectswere unable to predict submovement timing. Convincingevidence shows that the timing of movement initiationcomes under the control of the SMA.12,42This is true forsequences of predictable steps.43 However, it is not true

for tasks imposing unpredictable timing cues, such as theET condition studied here.37,44,45 Therefore, when theinitiation of a submovement has to fit into a precisetiming plan, the sequencing relies mainly on the basalganglia-SMA circuit. Hence, the shorter movement time(15%) during the SI than during the ET sequence weobserved in normal subjects presumably reflects prefer-ential activation of the medial motor system. This inter-pretation was confirmed by the further improvementin movement speed observed in normal subjects duringthe mixed-cued condition: 30.2% less than ET and17.7% less than SI movement time. In the mixed-cuedcondition, once subjects started to move in response tothe external cue, for switching between submovementsthey had to rely on an internal cueing mechanism. Themixed cueing of submovement execution probably leadsto a facilitation of the lateral and the medial motor systems.

The finding that ET, and to an even greater extent SI,motor tasks are impaired in patients with dystonia is inagreement with studies showing that basal ganglia proj-ect to both lateral and medial nonprimary motor areas,but predominantly to the SMA.14,46 The greater impair-ment of basal ganglia SMA circuit is also supported byother neurophysiological evidence. For example, themovement-related cortical potentials are depressed inamplitude in patients with primary,47 secondary,48 and

FIG. 5. Pause durations for self-initiated sequence. Upper panel: datafrom the paired patients with general-ized dystonia/young control subjects;lower panel: data from the paired pa-tients with focal dystonia/older controlsubjects. Bars represent standard devia-tion.

MOVEMENT CUEING IN DYSTONIA 109

Movement Disorders, Vol. 15, No. 1, 2000

focal49 dystonia. The movement-related cortical poten-tials recorded from the scalp before the onset of a self-paced voluntary movement reflect the activity of the pri-mary motor cortex, but predominantly of the SMA.50

The decreased activation of cortical motor areas alsoagrees with the observation that paired transcranial mag-netic stimulation delivered at long interstimulus intervalselicit more inhibited muscle responses in patients with

dystonia than in control subjects.51 PET scan studies inpatients with dystonia performing freely chosen pacedjoystick movements have shown overactivity of the pre-motor areas and reduced activation of the sensorimotorcortex.52,53 The underactivity of the caudal SMA re-ported in patients with focal or segmental dystonia52 hasnot been replicated in patients with familial generalizeddystonia.53 Why functional imaging and electrophysi-ological investigations disclose opposite patterns of brainactivity remains unexplained. One explanation is that ex-citatory and inhibitory synaptic neuronal activity canboth increase rCBF, a low temporal resolution might notdiscriminate properly events occurring before or after themovement.

Patients with dystonia also exhibit an abnormalswitching process between two successive submove-ments, as shown by the long pauses during the SI con-dition found here and in a previous study.4 Because thisprocess is implemented during the SI but not during theET condition, it might, in theory, cause the predominantimpairment in the SI sequence. Yet our finding that pa-tients with generalized dystonia improved their perfor-mance during the mixed-cued condition suggests other-wise. When the sequence requiring automatic execution(SI) was externally cued (mixed-cued condition), pa-tients shortened their movement time by 16.7% com-pared with ET and by 24.5% compared with SI se-quences. In addition, external cueing of the whole se-quence reduced not only the movement time, but also thepause durations. Under our experimental conditions, theprominent bradykinesia seen in patients with generalizeddystonia during the SI condition must therefore depend

FIG. 6. Reaction times and movement times for the single-movementand the first submovement of the externally triggered sequence. Upperpanel: reaction times; lower panel: movement times. Bars representstandard deviation.

FIG. 7. Submovement times andpause durations for self-initiated andmixed-cued sequence. Upper panel:duration of pauses; lower panel: dura-tion of submovements. Bars representstandard deviation.

A. CURRÁ ET AL.110

Movement Disorders, Vol. 15, No. 1, 2000

less on the abnormal switching process per se and moreon movement cueing.

In patients with motor impairment resulting from basalganglia disorders, bradykinesia during motor sequencesis the result of slowness observed during the execution ofthe single movement plus an “extra-slowness” that ap-pears when the same movement forms part of a se-quence.54,55 In a previous paper we described extra-slowness in patients with dystonia performing SI sequen-tial arm movements.4 In this article we now describe thesame change during ET sequences. In addition, we con-firm the presence of extra-slowness in patients with gen-eralized and focal dystonia. Hence, we conclude that inpatients with dystonia, extra-slowness is a motor-sequence feature that does not depend on movement cue-ing or on the body distribution of dystonia.

Previous investigators showed that patients with dys-tonia have normal reaction times for single move-ments.5,56 Our results confirm this finding and alsospecify normal reaction times for sequential movements.However, when the motor task required sequential motorresponses, patients with generalized dystonia had slowerreaction times for each ET submovement. Hence, thesepatients probably have an impaired motor set. The motorset is the function that enables subjects to execute con-secutive ET submovements as a sequence. In corticalneurons, set-related activity occurs throughout the delayperiod before experimental subjects receive the stimulusto move.57 Animal studies have shown that set-relatedneurons are common in premotor cortex and SMA39;sequence-specific neurons are common only in SMA.10

The abnormalities of movement execution describedhere are similar to those observed in patients with Par-kinson’s disease. In an earlier paper we showed that dys-tonic and parkinsonian patients are slow in executingsimple and sequential movements.4 Further experiments,similar to those used here, designed to test the influenceof internal and external cues on bradykinesia, haveshown that like dystonic patients, parkinsonian patientsalso find internally cued motor sequences harder thanexternally cued sequences.6 These abnormalities suggestthat a common defect of basal ganglia disorders consistof an inability to perform internally cued sequentialmovements efficiently.

In conclusion, in this study we found that patients withdystonia perform single and sequential free arm move-ments more slowly than control subjects, they show ex-tra-slowness in ET sequences, and they pause longer be-tween submovements in SI sequences. In addition, pa-tients with generalized dystonia have longer reactiontimes during the ET condition. They are also more bra-dykinetic during the SI than the ET sequences, owing to

their particular difficulty when engaged in SI motortasks. Overall, the slowness of single and sequential mo-tor acts independently from the cueing mechanism sug-gests a generalized disorder of movement execution indystonia. The more severe impairment of movement timein SI than ET sequences indicates that in dystonia, inter-nal cues are less efficiently processed than external cues.In addition, we suggest that the marked slowness in ex-ecuting self-initiated movements and the delayed motorresponses during externally triggered sequences are theresult of underactivity of the basal ganglia-SMA circuitin dystonia.

Acknowledgment: AC was supported by an IPSEN grant.

REFERENCES

1. Fahn S, Marsden CD, Calne DB. Classification and investigationof dystonia. In: Marsden CD, Fahn S, eds.Movement Disorders 2.London: Butterworth, 1987:332–358.

2. Cohen G, Hallett M. Hand cramp: clinical features and electro-myographic patterns in a focal dystonia.Neurology1998;38:1005–1012.

3. van der Kamp W, Berardelli A, Rothwell JC, Thompson PD, DayBL, Marsden CD. Rapid elbow movements in patients with torsiondystonia.J Neurol Neurosurg Psychiatry1989;52:1043–1049.

4. Agostino R, Berardelli A, Formica A, Accornero N, Manfredi M.Sequential arm movements in patients with Parkinson’s disease,Huntington’s disease and dystonia.Brain 1992;115:1481–1495.

5. Inzelberg R, Flash T, Schechtman E, Korczyn AD. Kinematicproperties of upper limb trajectories in idiopathic torsion dystonia.J Neurol Neurosurg Psychiatry1995;58:312–319.

6. CurraA, Berardelli A, Agostino R, et al. Performance of sequentialarm movements with and without advance knowledge of motorpathways in Parkinson’s disease.Mov Disord1997;12:646–654.

7. Tanji J. New concepts of the supplementary motor area.Curr OpinNeurobiol1996;6:782–787.

8. Goldberg G. Supplementary motor area structure and function:review and hypotheses.Behav Brain Sci1985;8:567–616.

9. Passingham RE. Two cortical systems for directing movements[Review]. Ciba Found Symp1987;132:151–164.

10. Mushiake H, Inase M, Tanji J. Neuronal activity in the primatepremotor, supplementary and precentral motor cortex during visu-ally guided and internally determined sequential movements.JNeurophysiol1991;66:705–718.

11. Tanji J, Shima K. Role of supplementary motor area cells in plan-ning several movements ahead.Nature1994;371:413–416.

12. Gerloff C, Corwell B, Chen R, Hallett M, Cohen L. Stimulationover the human supplementary motor area interferes with the or-ganization of future elements in complex motor sequences.Brain1997;120:1587–1602.

13. Wiesendanger R, Wiesendanger M. Cerebello-cortical linkage inthe monkey as revealed by transcellular labelling with the lectinwheat germ agglutinin conjugated to the marker horse-radish per-oxidase.Exp Brain Res1985;59:105–117.

14. Wiesendanger R, Wiesendanger M. The thalamic connections withmedial area 6 (supplementary motor cortex) in the monkey (ma-caca fascicularis).Exp Brain Res1985b;59:91–104.

15. Schell GR, Strick PL. The origin of thalamic inputs to the arcuatepremotor and supplementary motor areas.J Neurosci1984;4:539–560.

16. Matelli M, Luppino G, Fogassi L, Rizzolati G. Thalamic input toinferior area 6 and area 4 in the macaque monkey.J Comp Neurol1989;280:468–488.

MOVEMENT CUEING IN DYSTONIA 111

Movement Disorders, Vol. 15, No. 1, 2000

17. Hoover JE, Strick PL. Multiple output channels in the basal gan-glia. Science1993;259:819–821.

18. Matelli M, Luppino G, Rizzolati G. Architecture of superior andmesial area 6 and adjacent cingulate cortex in the macaque mon-key. J Comp Neurol1991;311:445–462.

19. Bhatia KP, Marsden CD. The behavioural and motor consequencesof focal lesions of the basal ganglia in man.Brain 1994;117:859–876.

20. Berardelli A, Hallett M, Rothwell JC, Thompson PD, Manfredi M,Marsden CD. The pathophysiology of primary dystonia.Brain1998;121:1195–1212.

21. Tanji J, Kurata K, Okano K. The effects of cooling supplementarymotor area and adjacent cortical areas.Exp Brain Res1985;60:423–426.

22. Schmidt EM, Porter R, McIntosh JS. The effects of cooling supple-mentary motor area and midline cerebral cortex on neuronal re-sponses in area 4 of monkeys.Electroencephalogr Clin Neuro-physiol1992;85:61–71.

23. Sasaki K, Gemba H. Effects of premotor cortex cooling uponvisually initiated hand movements in the monkey.Brain Res1986;374:278–286.

24. Halsband U, Passingham RE. Premotor cortex and the conditionfor movement in monkeys (macaca fascicularis).Behav Brain Res1985;18:269–277.

25. Mushiake H, Inase M, Tanji J. Selective coding of motor sequencein the supplementary motor area of the monkey cerebral cortex.Exp Brain Res1990;82:208–210.

26. Romo R, Schultz W. Role of primate basal ganglia and frontalcortex in the internal generation of movements. III: neuronal ac-tivity in the supplementary motor area.Exp Brain Res1992;92:396–407.

27. Halsband U, Matsuzaka Y, Tanji J. Neuronal activity in the pri-mate supplementary, pre-supplementary and premotor cortex dur-ing externally and internally instructed sequential movements.Neurosci Res1994;20:149–155.

28. Okano K, Tanji J. Neuronal activities in primate motor fields of theagranular frontal cortex preceding visually triggered and self-pacedmovement.Exp Brain Res1987;66:155–166.

29. Deiber M-P, Passingham RE, Colebatch JG, Friston KJ, Nixon PD,Frackowiak RSJ. Cortical areas and the selection of movement: astudy with positron emission tomography.Exp Brain Res1991;84:333–342.

30. Playford ED, Jenkins IH, Passingham RE, Nutt J, Frackowiack RS,Brooks DJ. Impaired mesial frontal and putamen activation in Par-kinson’s disease: a positron emission tomography study.Ann Neu-rol 1992;32:151–161.

31. Remy P, Zilbovicius M, Leroy-Willig A, Syrota A, Samson Y.Movement- and task-related activations of motor cortical areas: apositron emission tomographic study.Ann Neurol1994;36:19–26.

32. Jahanshahi M, Jenkins IH, Brown RG, Marsden CD, PassinghamRE, Brooks DJ. Self-initiated versus externally triggered move-ments. An investigation using measurement of regional cerebralblood flow with PET and movement-related potentials in normaland Parkinson’s disease subjects.Brain 1995;118:913–933.

33. Larsson J, Guylas B, Roland PE. Cortical representation of self-paced finger movement.Neuroreport1996;7:463–468.

34. Roland PE, Zilles K. Functions and structures of the motor corticesin humans.Curr Opin Neurobiol1996;6:773–781.

35. Boecker H, Kleinschmidt A, Requardt M, Hanichi W, MerboldtKD, Frahm J. Functional cooperativity of human cortical motorareas during self-paced simple finger movements: a high resolutionMRI study.Brain 1994;117:1231–1239.

36. Papa SM, Artieda J, Obeso JA. Cortical activity preceding self-initiated and externally-triggered voluntary movement.Mov Dis-ord 1991;6:217–224.

37. Cunnington R, Iansek R, Bradshaw JL, Phillips JG. Movement-related potentials in Parkinson’s disease. Presence and predictabil-ity of temporal and spatial cues.Brain 1995;118:935–950.

38. Grafton ST, Fagg AH, Woods RP, Arbib MA. Functional anatomyof pointing and grasping in humans.Cereb Cortex1996;6:226–237.

39. Kurata K, Wise SP. Premotor and supplementary motor cortex inrhesus monkeys: set-related activity during externally and inter-nally-instructed motor tasks.Exp Brain Res1988b;72:237–248.

40. Alexander GE, Crutcher MD. Movement related neuronal activityselectively coding either direction or muscle pattern in three motorareas of the monkey.J Neurophysiol1990;64:151–163.

41. Wise SP, Boussaoud D, Johnson PB, Caminiti R. Premotor andparietal cortex: corticocortical connectivity and combinatorialcomputations.Annu Rev Neurosci1997;20:25–42.

42. Deecke L, Kornhuber HH, Lang W, Lang M, Schreiber H. Timingfunctions of the frontal cortex in sequential motor and learningtasks.Human Neurobiology1985;4:143–154.

43. Marsden CD, Obeso JA. The functions of the basal ganglia and theparadox of stereotaxic surgery in Parkinson’s disease.Brain 1994;117:877–897.

44. Lang W, Obrig H, Lindinger G, Cheyne D, Deecke L. Supplemen-tary motor area activation while tapping bimanually differentrhythms in musicians.Exp Brain Res1990;79:99–104.

45. Halsband U, Ito N, Tanji J, Freund H-J. The role of premotorcortex and the supplementary motor area in the temporal control ofmovement in man.Brain 1993;116:243–266.

46. Matelli M, Luppino G. Thalamic input to mesial and superior area6 in the monkey.J Comp Neurol1996;372:59–87.

47. van der Kamp W, Rothwell JC, Thompson PD, Day BL, MarsdenCD. The movement related cortical potential is abnormal in pa-tients with idiopathic torsion dystonia.Mov Disord 1995;5:630–633.

48. Feve A, Bathien N, Rondot P. Abnormal movements related po-tentials in patients with lesions of the basal ganglia and anteriorthalamus.J Neurol Neurosurg Psychiatry1994;57:100–104.

49. Deuschl G, Toro C, Matsumoto J, Hallett M. Movement relatedcortical potential in writer’s cramp.Ann Neurol1995;38:862–868.

50. Ikeda A, Luders HO, Burgess RC, Shibasaki H. Movement-relatedpotentials recorded from supplementary motor area and primarymotor area.Brain 1992;115:1017–1043.

51. Rona S, Berardelli A, Inghilleri M, Vacca L, Manfredi M. Excit-ability of motor cortex in patients with dystonia.Mov Disord1998;13:118–124.

52. Ceballos-Baumann AO, Passingham RE, Warner T, Playford ED,Marsden CD, Brooks DJ. Overactive prefrontal and underactivemotor cortical areas in idiopathic dystonia.Ann Neurol1995;37:363–372.

53. Playford ED, Passingham RE, Marsden CD, Brooks DJ. Increasedactivation of frontal areas during arm movement in idiopathic tor-sion dystonia.Mov Disord1998;13:309–318.

54. Benecke R, Rothwell JC, Day BL, Mardsen CD. Disturbance ofsequential movements in patients with Parkinson’s disease.Brain1987;110:361–379.

55. Thompson PD, Berardelli A, Rothwell JC, et al. The coexistence ofbradykinesia and chorea in Huntington’s disease and its implica-tions for theories of basal ganglia control of movement.Brain1988;111:223–244.

56. Kaji R, Ikeda A, Ikeda T, et al. Physiological study of cervicaldystonia. Task-specific abnormality in contingent negative varia-tion. Brain 1995;118:511–522.

57. Kurata K, Wise SP. Premotor cortex in rhesus monkeys: set-relatedactivity during two conditional motor tasks.Exp Brain Res1988a;69:327–343.

A. CURRÁ ET AL.112

Movement Disorders, Vol. 15, No. 1, 2000