3rournal ofNeurology, Neurosurgery, and Psychiatry 1995;59:493-498

Changes in the balance between motor corticalexcitation and inhibition in focal, task specificdystonia

M C Ridding, G Sheean, J C Rothwell, R Inzelberg, T Kujirai

MRC HumanMovement andBalance Unit,Institute ofNeurology,Queen Square,London WC1N 3BG,UKM C RiddingJ C RothwellNational Hospital forNeurology, QueenSquare, LondonWCIN 3BG, UKG SheeanTel Aviv Universityand the WeizmannInstitute of Science,IsraelR InzelbergThe ThirdDepartment ofInternal Medicine,Yamagata UniversitySchool ofMedicine,2-2-2 lida-Nishi,Yamagata 990-23,JapanT KujiraiCorrespondence to:Dr Rothwell.Received 22 March 1995and in revised form29 June 1995Accepted 11 July 1995

AbstractTranscranial magnetic stimulation hasbeen used in a double pulse paradigm toinvestigate the excitability of intrinsicmotor cortical circuits in 15 patients withfocal task specific dystonia of the righthand and a group of eight age matchedcontrols. The left hemisphere was exam-

ined in five patients; in the remainder,both hemispheres were tested. There was

no significant difference in stimulationthreshold between patients and controlsnor between the left and right hemi-spheres in the patients. There was a sig-nificant decrease in early corticocorticalsuppression when comparing stimulation

of the left hemisphere in the patients andcontrols at interstimulus intervals of 1-15ms (P < 0.01). There was no difference inthe amount of suppression in the rightand left hemispheres of the patients. It isconcluded that in focal task specific dys-tonia there is shift in the balance betweenexcitation and inhibition in local circuitsof the motor cortex which leads to a netdecrease in the amount of short latencysuppression. These changes reflect dis-turbed basal ganglia input to the motorcortex. Reduced excitability of corticalinhibitory circuits may be one factorwhich contributes to the excessive andinappropriate muscle contraction whichoccurs during fine motor tasks in patientswith focal dystonia.

(7 Neurol Neurosurg Psychiatry 1995;59:493-498)

Keywords: dystonia; cortex; inhibition

Focal, task specific, dystonia, of whichwriter's cramp and musician's dystonia are

examples, is characterised by excessive mus-

cular activation during fine manipulativetasks. In the case of writing this excessiveactivity is usually seen in muscles of the fore-arm and hand. On the rare occasions when an

abnormality can be demonstrated it is usuallyconfined to the putamen, caudate, thalamus,and globus pallidus or their connecting path-ways.' Writer's cramp is therefore believed tobe a disease of the basal ganglia.

There have been several electrophysiologi-cal studies in patients with dystonia. Many ofthese have focused on spinal or brainstemabnormalities. For example, a reduction in

reciprocal inhibition between antagonist fore-arm muscles has been described in writer'scramp,23 and changes in the blink reflexrecovery cycle have been seen in patients withblepharospasm and cranial dystonia.G6 Morerecently, however, several reports have con-sidered the possibility that pathophysiologicalchanges may also occur within the cerebralcortex, particularly in the motor areas whichare a primary projection target of the basalganglia. The Bereitschaftspotential, which isthought to be generated in the supplementaryand primary motor areas, is decreased (in itslater phase) before self initiated movements.78During free choice joystick movements PETstudies have shown increased activity in therostral supplementary motor area anddecreased activity in the caudal supplemen-tary motor area and primary motor cortex.Reilly et a19 reported bilateral abnormalities ofthe N30 component of the somatosensoryevoked potential. Finally, Thompson et al'0reported that the duration of the silent periodafter transcranial magnetic stimulation was inthe low normal range.

Recently, we have reported a techniqueemploying double pulse magnetic stimulationwhich gives an indication of the balancebetween excitability of local intracorticalinhibitory and excitatory neuronal popula-tions."I With this technique it has been possibleto show that the relative excitability ofinhibitory circuits is reduced in Parkinson'sdisease,'2 in which there is known disease ofthe basal ganglia. In the present study we haveexamined intracortical inhibition in the motorcortex of a group of patients with focal taskspecific dystonia.

MethodsPATENTSWith the approval of the local ethics committeewe studied 15 patients with focal task specificdystonia (14 cases of writer's cramp and oneof simple musician's dystonia), mean age 47(SD 13) years), and eight neurologically nor-mal age matched controls (49 (SD 14) years).Patients and control subjects gave informedconsent. All patients and controls were righthanded. Four of the patients with writer'scramp were being treated with botulinumtoxin. The patients with writer's cramp weresubdivided into two groups depending ontheir symptoms. If their symptoms were spe-cific to writing and only involved musclesemployed for writing they were considered to

493 on 8 July 2018 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.59.5.493 on 1 N

ovember 1995. D

ownloaded from

Ridding, Sheean, Rothwell, Inzelberg, Kujirai

have simple writer's cramp. If their symptomswere specific to writing but also involved mus-cles not usually employed for writing (moregeneralised muscle involvement) then theywere considered to have dystonic writer'scramp. '3

All experiments were performed with a fig-ure of eight stimulating coil (external loopdiameters 9 cm) powered with two highpower Magstim 200 magnetic stimulators(Magstim, Dyfed, UK), which were linked viaa Bistim module (Magstim, Dyfed, UK). Thisallowed the delivery of two magnetic pulsesthrough the one coil at short interstimulusintervals. With this technique there is a reduc-tion in the power of the pulses due to theswitching characteristics of the Bistim mod-ule. This results in the power output beingabout 30% lower than that seen on the displayof the magnetic stimulators. Throughout thispaper the intensities quoted are for stimula-tion through the Bistim module and, there-fore, are some 30% higher than the trueoutput intensity of the stimulator. The coilwas held so that it induced electric current inthe motor cortex which flowed in a posteriorto anterior direction. Silver/silver chloride sur-face recording electrodes were used to recordevoked responses from the right first dorsalinterosseous muscle. The active electrode wasplaced over the muscle belly and the referenceelectrode was placed over the second metacar-pophalangeal joint. In 10 of the patientsresponses were also recorded from the leftinterosseous muscle after stimulation of theright motor cortex. Responses were recordedon to a PC using a 1401 laboratory interface(Cambridge Electronic Design, Cambridge,UK) for "off line" analysis.

THRESHOLDSThe threshold for evoking responses in thetarget muscle (first dorsal interosseous mus-cle) was assessed with single stimuli. Thestimuli were still delivered with the Bistimmodule connected and so the values quotedare about 30% higher than the values thatwould have been obtained with a single con-ventionally set up stimulator. Firstly, thethreshold for eliciting responses with the tar-get muscle relaxed was assessed. To help thesubject maintain relaxation audiovisual feed-back was given. The coil was placed on theoptimal scalp location for eliciting responsesin the target muscle and stimuli were applied.Initially the stimulus was increased in 5%steps until an estimation of threshold wasobtained. The stimulator output was thenadjusted in 1% steps until an accurate mea-sure of threshold was reached. Threshold wasdefined as being the stimulus intensity thatproduced clear EMG responses (of at least 50,uV peak to peak amplitude) in 50% of succes-sive trials. For the determination of activethreshold subjects were instructed to maintaina steady, minimal, background contraction(5-10% MVC) while the above procedurewas repeated. Subjects were again givenaudiovisual feedback to assist them in main-taining a steady contraction.

PAIRED PULSE SUPPRESSIONThis technique is described in detail else-where." Briefly, two magnetic stimuli weregiven through the same stimulating coil overthe motor cortex and the effect of the first(conditioning) stimulus on the second (test)stimulus was investigated. The conditioningstimulus was set at an intensity of 5% (ofstimulator output) below active threshold.The second, test, shock intensity was adjustedto evoke a muscle response in the relaxed firstdorsal interosseous muscle with an amplitudeof about 1 mV peak to peak. The timing ofthe conditioning shock was altered in relationto the test shock. Interstimulus intervalsbetween 1 and 15 ms were investigated.Three blocks of 40 trials were recorded foreach subject. Each block consisted of four dif-ferent conditions; test alone and test + condi-tioning at three different interstimulusintervals. The order of the presentation wasgenerated pseudorandomly by means of a1401 laboratory interface (CambridgeElectronic Design, Cambridge, England). Forthese recordings muscle relaxation is veryimportant and subjects were given audiovisualfeedback at high gain to assist in maintainingcomplete relaxation. If EMG activity becameapparent during data collection responseswere rejected. Measurements were made onindividual responses and the area of the con-ditioned response, at each interstimulus inter-val, was expressed as a percentage of the areaof the test response alone.

STATISTICAL ANALYSISIn the analysis of the experiments we consid-ered the following comparisons: (a) patients'left hemisphere and controls' left hemisphere,(b) patients' right hemisphere and controls'left hemisphere (c) the left and right hemi-spheres of individual patients (d) thosepatients who were on botulinum toxin andthose on no medication, and (e) the level ofinhibition in patients with simple writer'scramp and those patients with dystonicwriter's cramp. For the main set of experi-ments with paired pulses the effect of group(patients/controls), interstimulus intervals,and the interaction between GROUP x inter-stimulus intervals were analysed using multi-variate analysis of variance (MANOVA). Thecomparison of the inhibition of the right andleft hemisphere in the patients, the effect ofbotulinum toxin, and the comparison of inhi-bition in the two subgroups of patients withwriter's cramp (simple and dystonic) was per-formed with MANOVA. For comparison ofthreshold data in the patients and controls a ttest was used. For comparison of right and lefthemispheric thresholds in the patients thepaired Student's t test was used.

ResultsTHRESHOLDSIn the relaxed condition, after stimulation ofthe left hemisphere, the mean threshold forthe controls was 56 (SD 9)% and 57 (SD10)% for the patients. The threshold after

494 on 8 July 2018 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.59.5.493 on 1 N

ovember 1995. D

ownloaded from

Changes in the balance between motor cortical excitation and inhibition in focal, task specific dystonia

AControl ISI 2 ms

Conditioning ialone I

Test alone

Test+Iconditioning -

Patient

Test alone

j 0.5 mV

Test+ XO1msconditioning

Patients PatientsA Controls o Left hemisphere A Right hemisphere

stimulation stimulation

200

a 150c0

uz 100

0

B

0 2 4 6 8 10 12 14

Interstimulus interval (ms)16 18

Figure 1 Time course ofpaired pulse effects. (A) Raw data traces from a control subject (top set of three traces) and arepresentative patient with writer's cramp (bottom set of two traces). These data were obtained after left hemisphericstimulation and responses recorded in the relaxed right first dorsal interosseous muscle. In both the control subject and thepatient the subthreshold conditioning stimulus evoked no response in the target muscle (see top trace), whereas the teststimulus evoked a clear EMG response of approximately lmV (peak to peak amplitude). In the control subject when theconditioning stimulus was given 2 ms before the test stimulus there was clear suppression of the response (bottom trace forthe control subject). In the case of the patient there was much less suppression of the test response when conditioned at aninterstimulus interval of 2 ms (see bottom trace). (B) Data obtained across all interstimulus intervals in controls andpatients (stimulation of both both left and right hemisphere).

Figure 2 (A) Averagelevel of inhibition acrossinterstimulus intervals of1-6 ms in controls andpatients (stimulation ofboth left and righthemisphere). They-axis isthe area of the averageconditioned responseexpressed in terms of thearea of the average responseto the test stimulus alone.There is significantly lessinhibition in the patientsafter either left(MANOVA,P < 0O005***) or righthemisphere stimulation(MANOVA,P <0O-05**). (B) Averagefacilitation of the testresponse when preceded bya conditioning stimulus atinterstimulus intervals of7-15 ms. There is nosignificant differencebetween the level offacilitation in controls andpatients (stimulation ofeither left or righthemisphere) (MANOVA,P > 0O1).

stimulation of the right hemisphere in thepatients was 55 (SD 9)%. When active, afterleft hemispheric stimulation, the threshold forthe controls was 40 (SD 9)% and for thepatients 42 (SD 1 1)%. After right hemi-spheric stimulation the threshold in thepatients was 40 (SD 8)%. There was no sig-nificant difference in threshold in either therelaxed or active conditions when comparingleft hemispheric stimulation in the patientsand controls (relaxed state: t test, P > 0.5;active state: P > 0 5). Also, no significant dif-ference was seen in the thresholds for stimula-tion when comparing the right and lefthemispheres in the patients (relaxed condi-tion: paired t test, P > 0-5; active condition:P > 0-1).

200 A Average inhibition across ISI 1-6 ms

100 -

CD

uz 200-C:

c1

100

B Average facilitation across ISI 7-15 ms

O-

Corntrols Patients Patientsleft hemisphere; right hemisphereE

PAIRED PULSE SUPPRESSIONWhen the whole time course (interstimulusintervals 1-15 ms) was examined there was asignificant difference between the patients'left hemisphere and the control subjects' lefthemisphere (MANOVA, P < 0-01). Also,there was a significant difference when com-paring the patients' right hemisphere with thecontrol subjects' left hemisphere (MANOVA,P < 0 05). In both of these cases there was nosignificant interaction between group andinterstimulus intervals (P > 0-05). Figure 1shows the time course of the inhibition. Weknow from previous studies and this studythat in normal subjects the time course can bedivided into an "inhibitory phase" (interstimu-lus intervals of 1-6 ms) and a "facilitatoryphase" (interstimulus intervals of 7-15 ms).On the basis of this division further analysiswas performed on the data. When we aver-aged the data across interstimulus intervals of1-6 ms-the "inhibitory phase"-there wassignificantly less inhibition both when com-paring the patients' left hemisphere with thecontrol subjects' left hemisphere (MANOVA,P < 0-001) and also when comparing thepatients' right hemisphere with the controlsubjects' left hemisphere (MANOVA,P < 0-01). When we averaged the data acrossinterstimulus intervals of 7-15 ms-the "facil-itatory phase"-there was no significant dif-ference comparing either the patients' lefthemisphere with the controls' left hemisphere(MANOVA, P > 0 2) or the patients' righthemisphere with the controls' left hemisphere(P > 0 5). Figure 2 shows the average level ofinhibition and facilitation across these blocksof interstimulus intervals. In the controls,conditioned responses were suppressed to anaverage of 50 (SD 15)% of the test responsealone across interstimulus intervals of 1-6 ms

495

g 0_

on 8 July 2018 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.5.493 on 1 Novem

ber 1995. Dow

nloaded from

Ridding, Sheean, Rothwell, Inzelberg, Kujirai

after left hemispheric stimulation. In thepatients, responses were suppressed to only80 (SD 17)% after left hemispheric stimula-tion and 76 (SD 20)% after right hemisphericstimulation. In the control subjects responseswere facilitated to an average of 12 (SD 35)%of the test response across interstimulus inter-vals of 7-15 ms, whereas in the patients,responses were facilitated to 144 (SD 44)%after stimulation of the left hemisphere and127 (SD 23)% after stimulation of the righthemisphere. Botulinum toxin had no signifi-cant effect on the level of inhibition observed inthe left hemisphere of patients (MANOVA,P > 0 05). There was no significant differencein the level of inhibition when comparingpatients with simple writer's cramp and thosewith dystonic writer's cramp (MANOVA,P > 0-5).

DiscussionWriter's cramp and musician's dystonia areexamples of task specific focal dystonias. As aresult they are believed to be due to dysfunc-tion of the basal ganglia or their connections.The physiological basis of these conditionshas been debated for a long time. As outlinedin the introduction, there is evidence of disor-dered spinal cord and brainstem reflexes inseveral types of focal dystonia. As no abnor-mality has ever been found at either of thesesites, however, it is usually assumed that thechanges are due to an alteration in descendinginputs from other, higher, structures.Unfortunately, it is unclear which particularstructures might be important in relayingthese effects. For example, changes in theblink reflex recovery could be due to descend-ing connections between basal ganglia orbrainstem, or due to changes in projections tobrainstem from cerebral cortex deprived of itsnormal input from basal ganglia. Similarly,spinal mechanisms of reciprocal inhibitionreceive input from many supraspinal struc-tures including both the cortex and brain-stem.The purpose of the present paper was to

investigate the physiology of motor corticalareas as these are known to be one of the pri-mary output targets of the basal ganglia. Todo this we used the technique of transcranialmagnetic stimulation over the motor cortex.Like others we found that the threshold forevoking an EMG response in active or relaxedmuscle was the same as in normal subjects.'0By contrast, major changes in excitabilitybecame evident when we used double pulsetesting at short interstimulus intervals. Withthis technique, a subthreshold conditioningshock is given before a suprathreshold testshock. In control subjects this produces a pro-nounced suppression of the EMG response tothe test shock at interstimulus intervals of 1-6ms. We have argued previously"I that this is ofa cortical nature and suggest that it reflectsactivity of local intracortical inhibitory (proba-bly GABAergic) interneurons. The main rea-sons for this are (1) the conditioning stimulusis of such a low intensity (5% or more below

the threshold for evoking EMG responses inpreactivated muscle) that it probably does notproduce any descending volley in the pyrami-dal tract; (2) a conditioning stimulus pro-duces greater suppression of responses evokedby a magnetic than an anodal electrical stimu-lus to the motor cortex." Because both formsof stimulation are thought to activate the samedescending corticospinal pathways, the condi-tioning shock is unlikely to be acting at a sub-cortical level. Instead, the differential effect isthought to arise because of differences in theway electric and magnetic stimuli activate thecorticospinal system. Anodal electrical stimuliover the scalp tend to activate the pyramidaltract neurons directly whereas magnetic stimulitend to activate the same neurons transynapti-cally. '4 Thus the size of responses evoked bymagnetic stimulation should be more affectedby changes in cortical excitability than thoseevoked by anodal stimuli. If the conditioningstimulus affects cortical excitability then thisexplains why there is a greater suppression ofresponses to magnetic than electrical stimula-tion.

PAIRED PULSE SUPPRESSION IN DYSTONIAThe time course of paired pulse testing iscomplex, beginning with a period of pro-nounced suppression followed by less promi-nent facilitation. Many circuits, withoverlapping time courses could contribute tothe effect. The curve shows only that inhibi-tion dominates the early timing whereas facili-tation dominates later. Less suppression earlyin the time course in the patients with focaldystonia indicates that that there is a relativedecrease in the excitability of corticalinhibitory systems.The question is whether this reduction has

any relation to the clinical picture of dystonia.It is interesting to note that in primates injec-tion of the GABA antagonist bicuculline intothe motor cortex results in loss of directionalspecificity of cortical cells'" and inappropriatecontraction of antagonist muscles during reac-tion time wrist movements.'6 We thereforesuggest that the reduction in excitability ofintracortical inhibitory circuits which we havefound in patients with focal dystonia (theactivity of which, we suggest, is measured bythis paired pulse technique) may be one factorwhich contributes to the inappropriate andexcessive muscle activity which occurs whenthey attempt fine manipulations. Reducedactivity in inhibitory circuits may also explainwhy PET studies of patients with idiopathicdystonia show a decreased blood flow in theprimary motor cortex during voluntary joy-stick movements.'7One criticism of the proposed role of cortical

inhibition in producing dystonia is that thereduction in paired pulse suppression is seenboth in the motor cortex contralateral to theaffected hand and in the other hemisphere(although the second comparison was madebetween the patients' right hemisphere andthe controls' left hemisphere we have notexperienced any major variation in theamount of inhibition in the two hemispheres

496

on 8 July 2018 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.59.5.493 on 1 Novem

ber 1995. Dow

nloaded from

Changes in the balance between motor cortical excitation and inhibition in focal, task specific dystonia

of normal subjects). Such bilateral abnormali-ties in cases of unilateral dystonia9 18 are not an

uncommon result. Interestingly, when some

patients with writer's cramp learn to writewith their "unaffected" hand, because of thedifficulties with their dominant hand, as manyas 25% of them go on to develop symptoms inthe other hand.13 This suggests that subclini-cal abnormalities may have beenpresent bilaterally from the beginning of thedisease.

Another apparent problem with the presentfindings is that there was no significant differ-ence in the level of inhibition when we com-

pared patients with simple writer's cramp withthose with dystonic writer's cramp, even

though the dystonic spasms were worse in thesecond group. This may have been becausethe muscle investigated (first dorsalinterosseous muscle) was equally likely to beaffected in either group of patients and henceshowed a similar lack of cortical inhibition inthe two groups. It may be that ifwe had inves-tigated muscles that were affected in patientswith dystonic writer's cramp but not in thesimple group there may have been a differencein the level of inhibition. Unfortunately, it ismore difficult to perform these measurementson proximal muscles as the threshold for stim-ulation is usually much higher than in distalmuscles. Also it is important to rememberthat these recordings were obtained when thesubjects were at rest, and had no dystonicsymptoms. It is possible that if we had investi-gated the patients when active a different pic-ture may have emerged.We conclude that, although it is impossible

to be certain, it is likely that these alterationsin the excitability of inhibitory systems withinthe motor cortex play a part in the abnormalmovements seen in dystonia.

Four of the patients studied were beingtreated with botulinum toxin. Botulinumtoxin acts principally at the neuromuscularregion but may also have an effect on centralstructures, either by retrograde transmissionup motor axons, or by secondary changes inneural circuitry consequent on altered periph-eral input caused by muscle weakness. Forexample, it has been reported that reciprocalinhibition in patients with writer's cramp isnormalised after botulinum toxin treatment.'9At the level of the cortex, however, recentPET studies have failed to show any signifi-cant alterations in the level of cortical bloodflow after botulinum toxin treatment.20Similarly, our results showed no significantdifference in the level of corticocortical inhibi-tion in those patients who were being treatedwith botulinum toxin and those who were not.Although only a few patients were studied,this suggests that botulinum toxin has no

effect on the activity of local corticalinhibitory circuits.

COMPARISON WITH PREVIOUS RESULTS INPATIENTS WITH PARKINSON'S DISEASEWe have shown previously that, as in the pre-

sent results, paired pulse suppression isreduced in patients with Parkinson's disease.'2

Close inspection of the data in the two groupssuggests that there may be subtle differencesin the form of the suppression in Parkinson'sdisease and dystonia. The latter seem to havereduced suppression at all intervals, whereasthe former show more reduction at specifictimings (2, 4, and 5 ms). It is, however, prob-ably inapropriate to speculate on such differ-ences until larger numbers of patients arestudied. It is interesting to note, though, thatalthough the lack of suppression occurs inboth these instances of basal ganglia disease,suppression is normal in patients with cerebel-lar deficits.21

Finally, the question arises as to why thedecreased cortical inhibition produces exces-sive muscular activity in patients with dysto-nia, but not in patients with Parkinson'sdisease. One possibility is that the movementcommand itself is reduced in Parkinson's dis-ease (resulting in bradykinesia) and the effectsof reduced cortical inhibition are only appar-ent in the presence of a relatively normalmovement command. Hence the effects oncortical inhibition are masked in untreatedParkinson's disease. When the movementcommand is restored, as during drug induceddyskinesiae, the resulting excess muscle activitybecomes clear.

In conclusion, we have shown abnormali-ties of inhibition in the motor cortex ofpatients with two manifestations of task spe-cific focal dystonia-namely, writer's crampand musician's dystonia. These abnormalitiesare seen bilaterally and are not confined to themotor cortex projecting to the affected limb.Disease in the basal ganglia may affect inhibi-tion in the bilateral areas of motor cortex butonly leads to symptoms with the repeated per-formance of skilled manipulative tasks.

MCR was supported by Action Research. RI was supportedby the British Council and the Clore Foundation.

1 Rothwell JC, Obeso JA. The anatomical and physiologicalbasis of torsion dystonia. In: Movement disorders 2.London: Butterworths, 1987:313-31.

2 Panizza ME, Hallett M, Nilsson J. Reciprocal inhibition inpatients with hand cramps. Neurology 1989;39:85-9.

3 Nakashima K, Rothwell JC, Day BL, Thompson PD,Shannon K, Marsden CD. Reciprocal inhibitionbetween forearm muscles in patients with writer's crampand other occupational cramps, symptomatic hemidys-tonia and hemiparesis due to stroke. Brain 1989;112:681-97.

4 Beradelli A, Rothwell JC, Day BL, Marsden CD.Pathophysiology of blepharospasm and oromandibulardystonia. Brain 1985;108:593-608.

5 Tolosa E, Montserrat L, Bayes A. Blink reflex studies infocal dystonias: enhanced excitability of brainsteminterneurones in cranial dystonia and spasmodic torti-collis. Mov Dis 1988;3:61-9.

6 Nakashima K, Rothwell JC, Thompson PD, et al. Theblink reflex in patients with idiopathic torsion dystonia.Arch Neurol 1990;47:413-6.

7 Feve A, Bathien N, Rondot P. Abnormal movementrelated potentials in patients with lesions of the basalganglia and anterior thalamus. J Neurol NeurosurgPsychiaty 1994;57:100-4.

8 Van Der Kamp W, Rothwell JC, Thompson PD, Day BL,Marsden CD. The movement related potential is abnor-mal in patients with idiopathic torsion dystonia. Mov Dis1995 (in press).

9 Reilly JA, Hallett M, Cohen LG, Tarkka IM, Dang N.The N30 component of somatosensory evoked poten-tials in patients with dystonia. Electroencephalogr ClinNeurophysiol 1992;84:243-47.

10 Thompson ML, Wilson SA, Thickbroom GW, Laing BA,Stell R, Mastaglia FL. Cortical excitability and topogra-phy of excitatory and inhibitory areas in focal dystonia.Mov Dis 1994;9(suppl 1):204.

497 on 8 July 2018 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.59.5.493 on 1 N

ovember 1995. D

ownloaded from

Ridding, Sheean, Rothwell, Inzelberg, Kujirai

11 Kujirai T, Caramia MD, Rothwell JC, et al. Cortico-cortical inhibition in the human motor cortex. Physiol1993;471:501-19.

12 Ridding MC, Inzelberg R, Rothwell JC. Changes inexcitability of motor cortical circuitry in patients withParkinson's disease. Ann Neurol 1995;37:181-8.

13 Sheehy MP, Rothwell JC, Marsden CD. Writer's cramp.Adv Neurol 1988;50:457-72.

14 Day BL, Dressler D, Maertens De Noordhout A, et al.Electric and magnetic stimulation of human motor cor-tex: surface EMG and single motor unit responses. J7Physiol 1989;412:449-73.

15 Matsumura M, Sawaguchi T, Kubota K. GABAergic inhi-bition of neuronal activity in the primate motor and pre-motor cortex during voluntary movement. Neurophysiol1992;68:692-702.

16 Matsumura M, Sawaguchi T, Oishi T, Ueki K, Kubota K.Behavioral deficits induced by local injection of

bicuculline and muscimol into the primate motorand premotor cortex. Neurophysiol 199 1;65:1542-53.17 Brooks DJ. Functional imaging studies of movement.Move Dis 1994;9(suppl 1):M36.

18 Tempel LW, Perlmutter JS. Abnormal cortical responses inpatients with writer's cramp. Neurology 1993;43:2252-7.

19 Priori A, Berardelli A, Mercuri B, Manfredi M.Physiological effects produced by botulinum toxin treat-ment of upper limb dystonia: changes in reciprocal inhi-bition between forearm muscles. Brain 1995;118:801-8.

20 Ceballos-Baumann AO, Sheean G, Passingham RE,Marsden CD, Brooks DJ. Cerebral activation withstereotyped writing in patients with writer's crampbefore and after botulinum toxin treatment: a PETstudy. Neurology 1995;45(suppl 4):A393.

21 Ugawa Y, Hanajima R, Kanazawa I. Motor cortex inhibi-tion in patients with ataxia. Electroencephalogr ClinNeurophysiol 1994;93:225-9.

498 on 8 July 2018 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.59.5.493 on 1 N

ovember 1995. D

ownloaded from