tonic spasticity, parkinson's

Journal of Neurology, Neurosurgery, and Psychiatry, 1972, 35, 477-486

Tonic vibration reflex in spasticity, Parkinson'sdisease, and normal subjects

DAVID BURKE', COLIN J. ANDREWS2, AND JAMES W. LANCE

From the Division ofNeurology, the Prince Henry Hospital, and the School ofMedicine, the University ofNewSouth Wales, Sydney, Australia

SUMMARY The tonic vibration reflex (TVR) has been studied in the quadriceps and triceps suraemuscles of 34 spastic, 15 Parkinsonism, and 10 normal subjects. The TVR of spasticity developsrapidly, reaching a plateau level within 2-4 sec of the onset of vibration. The tonic contraction wasoften preceded by a phasic spike which appeared to be a vibration-induced equivalent of the tendonjerk. The initial phasic spike was usually followed by a silent period, and induced clonus in somepatients. No correlation was found between the shape of the TVR and the site of the lesion in thecentral nervous system. The TVR of normal subjects and patients with Parkinsonism developedslowly, starting some seconds after the onset of vibration, and reaching a plateau level in 20-60 sec.A phasic spike was recorded occasionally in these subjects, but the subsequent tonic contractionfollowed the usual time course. Muscle stretch increased the quadriceps TVR of all subjects, includingthose with spasticity in whom the quadriceps stretch reflex decreased with increasing stretch. It issuggested that this difference between the tonic vibration reflex and the tonic stretch reflex arisesfrom the selective activation of spindle primary endings by vibration, while both the primary and thesecondary endings are responsive to muscle stretch. The TVR could be potentiated by reinforcementin some subjects. Potentiation outlasted the reinforcing manoeuvre, and was most apparent at shortmuscle lengths. As muscle stretch increased, thus producing a larger TVR, the degree of potentiationdecreased. It is therefore suggested that the effects of reinforcement result at least partially from theactivation of the fusimotor system. Since reinforcement potentiated the TVR of patients with spinalspasticity in whom a prominent clasp-knife phenomenon could be demonstrated, it is suggested thatthe effects of reinforcement are mediated by a descending pathway that traverses the anterior quadrantof the spinal cord.

The tonic reflex evoked by muscle vibration hasbeen studied extensively, both in normal man(Lance, 1965; Eklund and Hagbarth, 1965,1966, De Gail, Lance, and Neilson, 1966; Hag-barth and Eklund, 1966a; Lance, De Gail,and Neilson, 1966; Rushworth and Young,1966; Marsden, Meadows, and Hodgson, 1969)and in the cat (Matthews, 1966, 1969; Gillies,Burke, and Lance, 1971a, 1971b). Relativelylittle attention has been given to this 'tonicvibration reflex' (TVR) in spastic man.

It has been reported that the TVR is usuallydiminished on the affected side of hemipareticpatients (Hagbarth and Eklund, 1966b; Lance etal., 1966), and absent in patients with transec-tion of the spinal cord (De Gail et al., 1966).

1 Commonwealth Postgraduate Scholar and Adolph Basser ResearchFellow in Neurology.

2 Edwin and Daisy Street Research Fellow in Neurology.

From studies in the cat (Gillies et al., 1971a,1971b), it appears that the abolition of the TVRby spinal transection is due to severance of thevestibulospinal and pontine reticulospinal path-ways in the anterior columns of the spinal cord.Hagbarth and Eklund (1968) reported thatmuscle vibration often produces a rapidly de-veloping tonic contraction in spastic limbs similarto that found in the experimental animal, ratherthan the slowly increasing tonic contraction ofnormal man. To date there have been no reportsof systematic studies in man of the differencesbetween the normal TVR and the spastic TVR,and there have been no attempts to correlatechanges in the TVR with specific central nervoussystem lesions.

In the upper limbs of spastic patients, theTVR increases as the vibrated muscle is stretched(Hagbarth and Eklund, 1968). The tonic stretch

[77

David Burke, Colin J. Andrews, andJames W. Lance

reflexes of the biceps and triceps brachii musclesof spastic man also increase with increasingmuscle stretch (Ashby and Burke, 1971), sothat the behaviour of the TVR parallels that ofthe stretch reflex in the upper limb. However,vibration is a selective stimulus for the spindleprimary ending, and has little effect on the firingfrequency of spindle secondary endings (Brown,Engberg, and Matthews, 1967). In the lower limbof spastic patients the stretch reflexes of extensormuscles are inhibited by increasing musclelength, while flexor stretch reflexes are facilitated,effects arising from spindle secondary endings(Burke, Gillies, and Lance, 1970, 1971; Burke,Andrews, and Ashby, 1971). It has yet to beestablished whether the TVR resembles thestretch reflex in this respect.

This study was undertaken to analyse furtherthe TVR of spastic man in an attempt to answerthese questions, and to contrast the results withthose obtained by vibrating muscles of normalsubjects and patients with Parkinsonism.

METHODS

The TVR was studied in 34 spastic patients, 15patients with Parkinson's disease, and 10 normalsubjects. Rigidity and bradykinesia were the mainmanifestations of Parkinson's disease, the patientshaving been selected for treatment with L-dopa.None of the patients with Parkinsonism had clinicalevidence of an upper motor neurone lesion.

All of the 34 spastic patients had typical velocity-dependent hypertonia, and the clasp-knife phenom-enon was demonstrable in the quadriceps stretchreflex. Of the 34 patients, four were hemiplegic dueto cortical or internal capsular lesions, and three hadbilateral spasticity due to brain-stem disease. Eightpatients suffered from multiple sclerosis and two fromfamilial spastic paraplegia. Five patients were para-paretic: one each from transverse myelitis, spinalcord angioma, spinal metastases, spinal osteomye-litis, and from arterial insufficiency of the spinalcord. The patient with the angioma was examined ona number of occasions, before and after the develop-ment of a complete cord lesion which was confirmedat laminectomy. Twelve patients were quadri-plegic after trauma to the cervical spinal cord, thelesion being clinically complete in five patients.

Usually the TVR of the quadriceps femoris musclewas studied, but in some patients the triceps suraemuscle was also vibrated. Vibration was applied tothe muscle belly or to the muscle tendon using eitherof two vibrators: an orbital physiotherapy vibrator(Kurt Stoll, Neidlingen: Tech., Germany) which wascapable of 50 Hz, or a small cylindrical vibrator (TVRVibrator V1A-F:A Keydon, Uppsala, Sweden)

which was capable of frequencies up to 200 Hz.The baseplate of the physiotherapy vibrator wasapplied directly to the muscle belly, or the edge of thebaseplate was applied to the muscle tendon. Thecylindrical vibrator was fixed to the limb over themuscle belly or muscle tendon by a rubber strap.When vibration was applied to the quadriceps

muscle, the patient reclined in a semi-supine positionwith both legs hanging over the edge of the examina-tion couch, flexed at the knee. The force exerted bythe TVR of the quadriceps muscle was estimated bya force transducer strapped to the leg immediatelyabove the ankle. The transducer consisted of a four-arm strain gauge bridge bonded to a rigid metal bar.The bridge was activated by the DC power supplyof an Offner Dynograph. For records of the TVRelicited under isometric circumstances, the leg wassupported forwards by heavy duty springs at theankle, and braced backwards by ties attached toeither end of the force transducer. Kneejoint positionwas altered by tightening or loosening these ties.When the quadriceps muscle was vibrated during astretching movement, the leg was moved by a handleattached to the force transducer. The TVR of thetriceps surae was recorded with the patient in theprone position, and with the leg to be examinedimmobilized in a rigid metal frame, fixed at both ankleand knee. The foot rested against a movable base-plate to which a second strain gauge bridge was bond-ed.The electromyogram (EMG) ofthe relevant muscle

was recorded by surface electrodes applied 5 cm apartover the belly of the muscle. The electrodes wereshielded from the vibrator by a lead earth 10 x 5 cmin order to reduce vibration-induced artefact. Theposition of the knee joint was measured by a gonic-meter. The EMG potentials were monitored on anoscilloscope to ensure that the records were artefact-free, and in some experiments the outputs of theforce transducer and the goniometer were alsomonitored. Tension, joint position and EMG wererecorded on a four-channel Offner Dynograph.Occasionally the EMG potentials were integrated(time constant 0-2 sec), and recorded in this formas well.The effects of reinforcement on the TVR were

studied using the Jendrassik manoeuvre. During suchstudies subjects were continually exhorted to maxi-mum effort in order to produce a constant andreproducible degree of reinforcement.

RESULTS

TONIC VIBRATION REFLEX A TVR could be elicit-ed from the quadriceps or triceps surae musclesof all spastic patients except those with clinicallycomplete spinal cord lesions. As recordedisometrically, the TVR of spastic patients usuallydeveloped abruptly, reaching a plateau tension

478

Tonic vibration reflex in spasticity, Parkinson's disease, and normal subjects

within 2-4 sec of the onset of vibration (Fig. la).An initial phasic spike was often recorded afterthe onset of vibration, tension then subsidingmomentarily before rapidly assuming the plateaulevel (Fig. Ib). Occasionally a TVR of relativelyslow onset could be recorded in spastic limbs(Fig. 1c), but this could usually be altered by

a b

-ii

_-

C

FIG. 1. The TVR of spastic man. The three TVRs shownwere all obtainedfrom the same patient. In a, low fre-quency vibration was applied to the muscle tendon, inb highfrequency vibration was applied to the tendon, andin c low frequency vibration was applied to the musclebelly. Calibrations: vertical-4 Kg, I mV for a;2 m Vfor b and c; horizontal-J0 sec.

reinforcement, or by changing the frequencyof vibration or the site of application of thevibrator to the limb. The tension produced bythe TVR was generally well maintained duringcontinued vibration (up to five minutes of record-ing), and did not decline to less than 3000 of itsinitial value. The EMG response to the onset ofvibration usually showed an initial phasic burstof potentials after a latency of 50 to 100 msec,and this was often accompanied by a phasicspike of tension, as seen in Fig. lb. There wasthen a momentary pause in EMG activity for100 to 150 msec before the tonic EMG responsestarted. On cessation of vibration EMG andreflex tension decreased rapidly after a delay ofup to 0-5 sec.The TVR of patients with Parkinsonism did

not differ significantly from that of normal sub-jects in whom a recordable tonic contractionstarted some seconds after the onset of vibrationand then increased slowly over 20 to 60 sec(Fig. 2). The rising phase usually became morerapid if the frequency of vibration was increased.

FIG. 2. The TVR ofnormalman. The tonic contractiondevelops slowly, taking, in a, more than 60 sec to reacha plateau. In a, the onset of vibration evokes a phasicburst of EMG potentials and a phasic spike oftension. There is then a latent period of 10 sec beforethe tonic reflex response commences. The tonic con-

traction is potentiated by the simultaneous use of a

second vibrator, and also by reinforcement. In b,reinforcement doubles the size ofthe tonic contraction.Calibrations: vertical-5 kg and I mV; horizontal-10 sec.

The tonic contraction was usually quite wellmaintained, as in spastic patients, and subsidedfairly rapidly 0 5-2-0 sec after vibration ceased.In some tense normal subjects and some tenseParkinsonism patients the onset of vibration was

accompanied by a phasic spike of both EMG andtension, similar to that seen in spastic patients.However, reflex tension then subsided almostcompletely, and the usual incremental toniccontraction started, often after a prolongedlatent period (Fig. 2a). As in spastic patients,the latency of the initial spike was 50 to 100 msec.The TVR varied in amplitude from patient to

patient in any one group. In two spastic, twonormal, and three Parkinsonism patients a TVRcould not be recorded unless a Jendrassikmanoeuvre was performed and, in these in-stances, the TVR thus elicited often started witha phasic spike before assuming the pattern

aVibrator 1 Vibrator 2

Jendrassik

~--b

.Jendra-ssik

-

II

479

I..,

-1-- ---,


characteristic of its group. In general the un-reinforced TVR was larger in normal subjects(O to 15 kg) than in spastic patients (O to5 kg), but the overlap was considerable. TheTVR of patients with Parkinsonism was ofintermediate size.Although it is possible to drive spindle primary

endings maximally in the experimental animal(Brown, Engberg, and Matthews, 1967), thisdoes not appear to be possible in man, be henormal, spastic or with Parkinsonism. This wasdemonstrated by the use of both vibratorssimultaneously. During vibration of the patellartendon at 200 Hz, the resulting tonic contractioncould be increased by simultaneous vibrationof the muscle belly. Performance of a Jendrassikmanoeuvre during such vibration often increasedreflex tension even further (Fig. 2a). It thusappears that these vibrators, either individuallyor in combination, are inadequate to produce aTVR in man which is independent of fusimotordrive.

It was not possible to demonstrate any rela-tionship between variations in the TVR of spasticpatients and the type of spasticity or the site ofthe causative lesion. The exceptions were patientswith clinically complete spinal cord lesions, inwhom no TVR could be elicited. In hemiplegicspasticity, the TVR was essentially similar tothat recorded in patients with spinal lesions, buton the clinically normal side of such patients,the TVR appeared to be of normal configuration.

In five spastic patients clonus could be inducedby muscle vibration, and in two normal subjectsa phenomenon similar to clonus could be record-ed. Clonus only appeared when a phasic spikefollowed the onset of vibration, and it could beavoided by changing the site or amplitudeof vibration, or by increasing the frequency ofvibration gradually, so that a phasic spike was notproduced. Flexor spasms could be induced byvibration in patients with complete spinal lesionsand in four patients with incomplete lesions if theknee were flexed. Anaesthetizing the skin under-lying the vibrator with procaine hydrochloride1% was sufficient to prevent such spasms in theone patient in whom this was done. In threespastic patients with predominantly extensorhypertonicity vibration provoked extensorspasms when the knee was extended to 30 45°.In two patients with spinal spasticity and grosslyexaggerated flexor tone, vibration of the patellartendon set up a small TVR in the quadricepsmuscle at first, but this was interrupted after

2 to 5 sec by a powerful sustained tonic contrac-tion of the hamstrings. This indirect TVR of thehamstrings presumably resulted from spread ofvibration to excite the spindle receptors of thehamstrings muscles.

In normal subjects, and to a somewhat lesserextent in patients with Parkinsonism, the TVRcould be abolished voluntarily. If vibration wascontinued and the subject's attention distracted,the TVR resumed its previous level of activity.The ability to suppress the TVR voluntarilyappeared to be impaired in spastic patients,particularly those with spinal lesions. Vibrationofa paretic muscle potentiated the force producedby voluntary contraction of that muscle, pre-sumably due to a synergistic effect of reflexand voluntary activation of motor units. Theseresults confirm those of Hagbarth and Eklund(1968). This potentiation of voluntary power didnot significantly outlast the duration of vibration,although some patients reported that their limbsfelt more relaxed and free of spasms for one totwo hours after the cessation of vibration, andthey were therefore able to make better use oftheir paretic muscles. By contrast, vibration didnot significantly alter the force ofmaximal volun-tary contraction in normal subjects. The durationof vibration was limited in all subjects by thegeneration of heat.

EFFECT OF MUSCLE STRETCH ON TVR The effectof muscle stretch on the quadriceps stretch reflexof spastic patients is seen in Fig. 3. The reflexconsists of a dynamic response to the stretchingmovement, there being little or no response tomaintained static stretch. If the amplitude of thestretching movement is divided into equal se-quential steps performed at the same velocityof stretch, reflex EMG is maximal in the firststep of the movement, but decreases in subse-quent steps as muscle length increases. In con-trast, the quadriceps TVR of spastic patients wasfound to increase as muscle stretch increases,using similar stepwise stretching movements(Fig. 4). Again the dynamic response to thestretching movement was prominent.The EMG and resistance produced by the TVR

during slow stretching movements becamemaximal at the position of greatest stretch(Fig. 5a). If the rate of stretching was increased,peak EMG and peak resistance sometimes ledthe position of greatest stretch (Fig. 5b), butthis did not produce the clinical sensation of aclasp-knife phenomenon. Since the firing rate

480


Time * . .

Velocity

Angle

Integrated EMG

EMG

seconds

] 2500/sec

e- 00

f g0o

to isolate the effects of changes in muscle lengthuncontaminated by such velocity-dependent fac-tors, the TVR was elicited isometrically at differ-ent positions of the knee joint. Again, musclestretch was found to increase the TVR (Fig. 6).In spastic patients a TVR usually could not beobtained until the knee joint was flexed to 30 to450. With further flexion the TVR increased,but the response to increasing length tended toreach a plateau at 90° flexion.To demonstrate that this facilitatory effect was

an autogenic phenomenon, due to stretch of thevibrated muscle rather than to shortening of theantagonist or activation of joint receptors, the

] I mV

FIG. 3. The effect of muscle stretch on the quadricepsstretch reflex ofspastic man. The stretching movementhas been divided into three steps of equal amplitudeperformed at the sanme velocity ofstretch. Reflex EMGis greatest during the first step, and subsequent stepsevoke only a small EMG response. ReflexEMG occursonily in response to the stretching movement, and thereis no maintained response to static stretch.

a

Angle

IntegratedEMGO

b

1:°f J 9o

4kg

JNJdKJV-EMG V=2Z

5#C'

FIG. 5. The effect of the velocity of the stretchingmovement on the TVR of spastic man. In a, with aslow stretching movemenit, the reflex response becomesmaximal at the position ofgreatest stretch. In b withfaster stretching movements, reflex EMG and tenisionpeak at a position slightly in advance ofthe position ofgreatest stretch. The patellar tendon is being vibratedthrolughout each record.

FIG. 4. The effect of muscle stretch on the TVR ofspastic man. Vibration of the patellar tendon continuesthroughout the entire record. Stretch of the quadricepsmuscle produced by stepwise flexion movements at theknee joint results in progressive increase in size of thereflex response. There is a prominlent dynamic response

to the stretching movement.

of the spindle primary ending will increase inresponse to the velocity of the stretch, thesestudies do not examine exclusively the effects ofchanges in muscle length on the TVR. In order

quadriceps muscle was stretched by pressure

applied to the patella, knee joint position remain-ing constant. Vibration was applied to the musclebelly above the patella, and the effect on the TVRwas recorded by changes in EMG activity.Again, vibration-induced EMG activity increasedas the quadriceps muscle was stretched.

Essentially similar results were found when theeffect of muscle stretch on the TVR of normaland subjects with Parkinsonism was studied.Usually the TVR increased throughout a stretch-ing movement, but sometimes, especially atfaster velocities of stretch, reflex EMG diminishedand resistance collapsed as the position of great-est stretch was approached. When the effectof muscle stretch on the TVR was studied under

e. 00

Angle

,'egaf9d0

IntegratedEMG L L tEMG -L _ Si- --l. , -l-A11s-.

$-Wo

1- .--I 1.

-17-7. -, "-1

481

David Burke, Coliii J. Andrews, andJames W. Lance

450 60°0

Ll~K

750

K

900

il

- --1-

4 kg1 mV

10 sec

FIG. 6. The effect ofmuscle stretch on the TVR ofspastic man. Increasing muscle stretch increases the size oftheTVR. The effect tends to plateau between 75° and 90°.

isometric circumstances, the TVR increased withincreasing muscle stretch (Figs 7, 8a), indicatingthat the apparent inhibition was related to thevelocity of the stretching movement rather thanto the length of muscle, and was possibly of

5.0

> 255

0

00 300 60

JOINT POSITION

FIG. 7. The effect of muscle stretch on the TVR ofpatients with Parkinsonism. For each of threepatients, increasing muscle stretch increases the sizeof the resulting TVR, as recorded isometrically. Atleast three TVRs were obtained for each patient ateach joint position, the range being indicated by thevertical lines.

IOU 30 Boo goo,.

a

b I< , ,-,-.-

* ^ ^~~~~~-AFIG. 8. The effect of reinforcement on the responseof the TVR to muscle stretch. In a, the TVR increasesas muscle stretch increases. In b, vibration was appliedduring reinforcement. A large TVR is produced at theposition of least stretch and this changes little as themuscle is stretched. Calibrations: vertical-tensionrecords, 5 kg for a and first two traces of b; 10 kg forlast two tracesof b. EMG, ImVfora and b; horizontal-20 sec.

volitional or supraspinal origin. The effect ofmuscle stretch on the TVR of normal or Parkin-sonism subjects was thus much the same as inspastic patients, except that a TVR could some-times be obtained in normal subjects and thosewith Parkinsonism when the knee joint wasflexed to less than 1°O (Fig. 7).

EFFECT OF REINFORCEMENT ON TVR The effects ofreinforcement varied from patient to patient.In some normal and some Parkinsonism patientsthe effect of a Jendrassik manoeuvre was small

iLk

482

,,_


and inconstant, but in others a reproduciblefacilitatory action could be demonstrated whenreinforcement was performed before, during orthroughout vibration (Fig. 2). Similar resultswere obtained in spastic patients, reinforcementhaving no significant effect in some patients,but consistently increasing the force generated bythe TVR in others (Fig. 9a). In normal, Parkin-sonism, and spastic patients the effects of theJendrassik manoeuvre significantly outlastedthe duration of the manoeuvre (Fig. 9a). Theduration of this facilitation was studied byperforming the manoeuvre at variable intervalsbefore the onset of vibration. The greatestfacilitation occurred if vibration started within0 5 sec of the cessation of the manoeuvre, butsome potentiation persisted for as long as 10sec (Fig. 9b).

a b

J

251

2-

I--

I

of interaction of these facilitatory influences.The amount by which the TVR could be po-tentiated by a Jendrassik manoeuvre was greatestwhen muscle stretch was least, and was leastwhen muscle stretch was greatest (Fig. 8).When evoked during reinforcement, increasingmuscle stretch had less effect on the TVR, sothat the TVR varied little from one joint positionto the next (Fig. 8b). It thus appeared that atless stretched positions reinforcement wascapable of compensating for the degree ofmuscle stretch.

DISCUSSION

Significant differences have been demonstratedbetween the TVR of spastic patients and that ofnormal or Parkinsonism subjects. Although

control

.* 0

0--

If the Jendrassik manoeuvre started beforethe onset of vibration a phasic contraction couldbe induced in some normal and some Parkinson-ism subjects, and also in some spastic patientsin whom a phasic spike had not previously beenrecorded. In normal and Parkinsonism subjectstension and EMG subsided after the phasicspike and the usual incremental TVR thenensued. In spastic patients tension and EMG de-creased only slightly before rapidly climbingto the plateau level.The degree of potentiation of the TVR pro-

duced by reinforcement at differentjoint positionswas studied in order to determine the degree

5LATENCY seconds

FIG. 9. The duration of thepotentiation of the TVR pro-duced by reinforcement. Ina, a Jendrassik manoeuvre(J),performed during musclevibration, potentiates theTVR. The potentiation out-lasts the manoeuvre. In b,the effect ofa Jendrassikmanoeuvre conditioning asubsequent TVR is shown.The effect is maximal im-mediately after the cessationof the manoeuvre, but somepotentiation persists for aslong as 10 sec. Calibrations

o for a-vertical 2 kg and1 m V; time marker,I sec/division.

variations in the speed of onset of the TVR wererecorded in each group, it is apparent that therapidly developing TVR is characteristic ofspasticity, while that of normal man or patientswith Parkinsonism develops relatively slowly.In some normal subjects the onset of the TVRbecame more rapid if the frequency of vibrationwas increased, and in some spastic patients aslowly increasing tonic contraction could beevoked by reducing the frequency or changingthe site of vibration. As pointed out by Hagbarthand Eklund (1968), the TVR of spasticity usuallyresembles that found in the decerebrate oranaesthetized cat (Matthews, 1966; Gillies et al.,

7--

483

.


1971a, 1971b). When low frequency vibrationwas applied to the non-denervated limb ofanaesthetized cats, Gillies et al. (1971a) wereable to record the slowly progressive increasein EMG and reflex tension characteristic of theTVR of normal man. They concluded that thedifference in time course of the TVR of thedecerebrate animal and that of normal man wasrelated to the innervation of the limb other thanthat of the muscle vibrated. However, this con-clusion does not explain the differences in theTVR ofnormal and spastic man and it is probablethat, given a comparable vibration stimulus, thespastic TVR differs significantly from the normalTVR.At the outset of this study it was hoped that

consistent differences in the TVR might be foundin spasticity of different aetiologies, such that agradation between the normal TVR and thespastic TVR could be demonstrated. No suchdifference could be found. The variation encoun-tered in any one subject was so great that nosingle pattern could be regarded as specific forany particular form of spasticity. The similarityof the normal and Parkinsonism TVR and theiridentical response to muscle stretch suggest thatspinal reflex mechanisms are normal in Parkin-son's disease. This conclusion supports thepostulate of Landau (1969) that rigidity ofParkinsonism arises from increased supraspinaldrive of essentially normal segmental mecha-nisms.An initial burst of EMG resulting in a phasic

rise in tension followed the onset of vibrationin many spastic patients, and in some normalor Parkinsonism subjects during reinforcement.The latency of this burst of EMG was such thatit can be attributed only to a spinal reflex. Theheight of the resulting phasic spike of tension(cf Figs 1, 6, 9a) suggest synchronous firing ofmotoneurones, and it is probable that this phasicresponse is a vibration-induced tendon jerk.That the latency of the response could be aslong as 100 msec probably arises from theinertia of the vibrator when first applied to thelimb, and temporal summation at motoneuronelevel of the afferent activity from the initialstrokes of the vibrator.The phasic spike was followed by a brief

pause in EMG activity of 100 to 150 msec. Thissilent period may be accounted for by unloadingof the muscle spindle by the initial phasic con-traction and by autogenic inhibition from Golgitendon organs activated by it. If most moto-

neurones were activated in the phasic contractionthere would be a tendency for these motoneuronesto fire synchronously when activity resumed afterthe silent period. It is thus not surprising thatvibration may initiate clonus in spastic patientsand in some normal subjects. Since the silentperiod is probably determined by the occurrenceof a phasic response to the onset of vibration,it might be expected that the presence of clonuswould be correlated with the presence of aphasic contraction, and that efforts to avoid thisphasic contraction would also prevent clonus.

Muscle stretch enhanced the TVR in normal,Parkinsonism, and spastic man. Presumably thisarises from the facilitatory effect of increasingmuscle length on the response of the spindleprimary ending to vibration (Brown et al., 1967),since recordings of afferent fibre activity in manhave shown that muscle stretch increases thegroup Ia response to vibration (Hagbarth andVallbo, 1967, 1968). In the cat a TVR cannot beelicited in a slack muscle, and does not becomemaximal until the vibrated muscle is stretched toits maximal physiological length, producing apassive tension of up to 200 g (Matthews, 1966;Gillies et al., 1971a). A similar situation appearsto exist in spastic man in whom the quadricepsTVR usually cannot be elicited until the kneejoint is flexed to 30-45'. In normal man andpatients with Parkinsonism the threshold lengthat which a TVR can be elicited appears to beless, but otherwise the response to increasingmuscle length is qualitatively the same.

In spastic man the quadriceps stretch reflexdecreases as muscle length increases, an effectprobably due to autogenic inhibition from thespindle secondary ending (Burke et al., 1970).However the secondary ending is insensitiveto muscle vibration (Brown et a!., 1967). It isprobable then that in spasticity the quadricepsTVR increases as muscle length increases becausestretch enhances the primary ending's responseto vibration but does not alter the slight responseof the secondary endings to vibration. The in-hibitory effects of the secondary ending in-crease in proportion to increasing muscle length,but presumably this is masked by the greaterresponse of the primary ending to vibration.It is apparent then that the TVR cannot beequated with the tonic stretch reflex, at least inspasticity.

Potentiation of the TVR by reinforcement wasnot found in all subjects, but when present itwas usually reproducible. In all spastic patients,

484

Tonic vibration reflex in spasticity, Parkinson's disease, and niormal subjects

including those spinal patients in whom reinforce-ment potentiated the TVR, a prominent clasp-knife phenomenon could be elicited in the quadri-ceps stretch reflex. It has been suggested that theclasp-knife phenomenon in cases of spinalspasticity results from release of the group IIafferent pathway due to a lesion of the postero-lateral funiculus of the spinal cord (Burke,Knowles, Andrews, and Ashby, 1972). In thesecircumstances it is therefore likely that thedescending pathways mediating the facilitatoryeffects of reinforcement traverse the anteriorquadrant of the spinal cord.The duration of the potentiation induced by

reinforcement outlasted the reinforcing man-oeuvre by up to 10 sec. In contrast to the effectson alpha motoneurones, reflex or supraspinalactivation of gamma motoneurones evokes highfrequency repetitive firing, thus producing effectswhich may long outlast the stimulus (Hunt andPaintal, 1958). The long duration of the potentia-tion induced by reinforcement in man is thereforeconsistent with activation ofthe fusimotor system,in addition to any direct action on the alpha moto-neurone. The suggestion that reinforcement actsat least partially through the fusimotor systemis supported by the finding of interaction betweenthe effects of the Jendrassik manoeuvre andthose of muscle stretch. Reinforcement was mostconspicuous when the muscle was in the shortenedposition, and became proportionately less asstretch increased. This suggests that stretch andreinforcement use a common path to producetheir effects on the TVR. Thus reinforcementprobably acts at least in part through the fusi-motor system, setting the contraction of intra-fusal fibres at a level which compensates for thelength of extrafusal fibres. The demonstrationof an effect on the gamma efferent system doesnot exclude a parallel effect on the alpha moto-neurone in addition. In man, Hagbarth andVallbo (1968, 1969) and Hagbarth, Hongell,and Wallin (1970) have demonstrated coactiva-tion of alpha and gamma motoneurones duringvoluntary contraction of muscle, and it is there-fore to be expected that reinforcement wouldalso activate both types of motoneurone.

The authors would like to thank Dr. J. D. Gilliesand Miss Lyndsay Knowles for assistance withsome of the experiments reported here. Mr K.Norcross and Mr N. Skuse have given valuabletechnical aid. The financial assistance of the NationalHealth and Medical Research Council of Australia,

the Adolph Basser Trust, and Mr and Mrs EdwinStreet is gratefully acknowledged. Illustrations werephotographed by the Department of Medical Illus-tration, University of New South Wales.

REFERENCES

Ashby, P., and Burke, D. (1971). Stretch reflexes in the upperlimb of spastic man. Jouirnal of Neurology, Neurosuirgery,and Psychiatry, 34, 765-771.

Brown, M. C., Engberg, I., and Matthews, P. B. C. (1967).The relative sensitivity to vibration of muscle receptors ofthe cat. Jouirnal ofPhysiology, 192, 773-800.

Burke, D., Andrews, C., and Ashby, P. (1971). Autogeniceffects of static muscle stretch in spastic man. Archives ofNeurology. 25, 367-372.

Burke, D., Gillies, J. D., and Lance, J. W. (1970). Thequadriceps stretch reflex in human spasticity. Journal ofNeurosurgery, and Psychiatry, 33, 216-223.

Burke, D., Gillies, J. D., and Lance, J. W. (1971). Hamstringsstretch reflex in human spasticity. Journal of Neurology,Neurosurgery, and Psychiatry, 34, 231-235.

Burke, D., Knowles, L., Andrews, C., and Ashby, P. (1972).Spasticity, decerebrate rigidity and the clasp-knife pheno-menon: an experimental study in the cat. Brain. (In press.)

De Gail, P., Lance, J. W., and Neilson, P. D. (1966). Differen-tial effects on tonic and phasic reflex mechanisms producedby vibration of muscles in man. Journal of Neurology,Neuirosurgery, and Psychiatry, 29, 1-11.

Eklund, G., and Hagbarth, K.-E. (1965). Motor effects ofvibratory muscle stimuli in man. (Abstract.) Electro-encephalography and Clinical Neurophysiology, 19, 619.

Eklund, G., and Hagbarth, K.-E. (1966). Normal variabilityof tonic vibration reflexes in man. Experimental Neurology,16, 80-92.

Gillies, J. D., Burke, D. J., and Lance, J. W. (1971a). Tonicvibration reflex in the cat. Journal of Neurophysiology, 34,252-262.

Gillies, J. D., Burke, D. J., and Lance, J. W. (1971b). Supra-spinal control of tonic vibration reflex. Jouirnal of Neuiro-physiology, 34, 302-309.

Hagbarth, K.-E., and Eklund, G. (1966a). Motor effects ofvibratory muscle stimuli in man. In Muscular Afferents andMotor Control. Proceedings of the First Nobel SymposiuimHeld in Jlne 1965 at Sodergarn on the Island of Lidingo inthe County of Stockholm. pp. 177-186. Edited by R. Granit.Almqvist and Wiksell: Stockholm.

Hagbarth, K.-E., and Eklund, G. (1966b). Tonic vibrationreflexes (TVR) in spasticity. Brain Research, 2, 201-203.

Hagbarth, K-E., and Eklund, G. (1968). The effects of musclevibration in spasticity, rigidity, and cerebellar disorders.Journal of Neuirology, Neurosuirgery, and Psychiatry, 31,207-213.

Hagbarth, K.-E., Hongell, A., and Wallin, B. G. (1970). Theeffect of gamma fibre block on afferent muscle nerve activityduring voluntary contractions. Acta Physiologica Scandi-navica, 79, 27A-28A.

Hagbarth, K.-E., and Vallbo, A. B. (1967). Afferent responseto mechanical stimulation of muscle receptors in man.Acta Societatis Medicoruim Upsaliensis, 72, 102-104.

Hagbarth, K.-E., and Vallbo, A. B. (1968). Discharge charac-teristics of human muscle afferents during muscle stretchand contraction. Experimental Neurology, 22, 674-694.

Hagbarth, K.-E., and Vallbo, A. B. (1969). Single unit record-ings from muscle nerves in human subjects. Acta Physiologi-ca Scandinavica, 76, 321-334.

Hunt, C. C., and Paintal, A. S. (1958). Spinal reflex regula-tion of fusimotor neurones. Journal of Physiology, 143,195-212.

Lance, J. W. (1965). The mechanism of reflex irradiation.Proceedings of the Auistr-alian Association of Neurologists, 3,77-81.

485


Lance, J. W., De Gail, P., and Neilson, P. D. (1966). Tonicand phasic spinal cord mechanisms in man. Journal ofNeurology, Neurosurgery, and Psychiatry, 29, 535-544.

Landau, W. M. (1969). Spasticity and rigidity. In RecentAdvances in Neurology, pp. 1-32. Edited by F. Plum.Davis: Philadelphia.

Marsden, C. D., Meadows, J. C., and Hodgson, H. J. F.(1969). Observations on the reflex response to musclevibration in man and its voluntary control. Brain, 92,829-846.

Matthews, P. B. C. (1966). The reflex excitation of the soleusmuscle of the decerebrate cat caused by vibration appliedto its tendon. Journal ofPhysiology, 184, 450-472.

Matthews, P. B. C. (1969). Evidence that the secondary aswell as the primary endings of the muscle spindles may beresponsible for the tonic stretch reflex of the decerebratecat. Journal ofPhysiology, 204, 365-393.

Rushworth, G., and Young, R. R. (1966). The effect of vibra-tion on tonic and phasic reflexes in man. Journal ofPhysiol-ogy, 185, 63P-64P.

486

tonic spasticity, parkinson's

Documents