J. Neurol. Neurosurg. Psychiat., 1963, 26, 39

Normal plantar response: integration of flexorand extensor reflex components

LENNART GRIMBY

From the Department of Neurology, Karolinska Institute, Serafimerlasarettet, Stockholm, Sweden

The reflexes elicited by painful stimulation of theplantar surface of the foot have been studiedextensively for a long time and the relation betweenthe reflexes obtained in normal and in pathologicalcases has been the subject of considerable debate. Anexcellent survey of previous investigations is to befound in the review by Walshe (1956). As in moststudies of human reflexes, the technique commonlyused has, however, not permitted an exact deter-mination of the latency values of the reflexes, andit has thus not been possible to judge with certaintyto what extent the movements studied have beenpurely spinal and to what extent of cerebral origin.By means of brief electric stimuli and an electro-

myographic recording technique these latency valuescan, however, be exactly determined, and in this waya clear distinction can be made between purely spinalreflexes and movements of a more uncertain origin.Using this technique, the spinal skin reflexes of thetrunk (Kugelberg and Hagbarth, 1958; Hagbarthand Kugelberg, 1958), the leg (Hagbarth, 1960;Kugelberg, Eklund, and Grimby, 1960), and thefoot (Eklund, Grimby, and Kugelberg, 1959;Kugelberg et al., 1960) have previously beeninvestigated, and these studies have established theexistence of an extensive system of spinal skinreflexes representing a highly purposeful defencemechanism for appropriate withdrawal reactions.

In investigations of the spinal pain reflexes nor-mally elicited from the plantar surface of the footKugelberg et al. (1960) found that stimulation of theball and hollow of the foot evokes plantar flexionof the toe, whereas, conversely, stimulation of theball of the great toe produces dorsiflexion of thetoe. It was further established that in extreme patho-logical cases (with pronounced Babinski signs)painful stimulation of the plantar surface of the footelicits a stereotyped flexor reflex with dorsiflexion ofthe great toe, independently of the stimulus site, andthat the main difference between this pathologicalreflex and that normally evoked by hallux stimu-lation lies in the extent of the receptive field. Theconclusion was drawn that the two reflexes -areessentially identical but that in normal cases, due to

the suprasegmental control of the reflex centres, thereceptive field of the reflex is limited to the skin areawhere it is adequate for protective purposes, viz.,the ball of the great toe.

Previous investigations (Eklund et al., 1959;Kugelberg et al., 1960) have shown that the maindifference between the electromyographic pattern ofa flexor plantar response and that of an extensorplantar response is that the reflex plantar flexion ofthe great toe is associated with activity in the shorthallux flexor and reciprocal inhibition of thevoluntary activity in the short hallux extensor,whereas, conversely, reflex dorsiflexion of the greattoe is accompanied by activity in the short halluxextensor and reciprocal inhibition of the voluntaryactivity in the short hallux flexor. Landau and Clare(1959) have, however, put forward the view that theflexor plantar response differs from the extensorplantar response in that the long hallux extensor isengaged in th- latter but not in the former reflex type.The controversial results thus obtained in the twoinvestigations will be discussed below.On the basis of the results reached by Kugelberg

et al. (1960), on painful stimulation of the plantarsurface of the foot, investigations were started onpathological cases showing few pronounced oruncertain Babinski signs. In the course of the experi-ments it was, however, soon apparent that beforebeing able to judge of the results obtained inabnormal cases a closer study had to be undertakenof the receptive fields and the variability normallyexisting; the present work gives an account of theresults obtained in this investigation of normal cases.

In the present study, the latency values of therecorded bursts of activity have been determined(sect. 1). Only reflexes of shorter latencies than200 msec. have been included. The interest has beenfocused on, and all conclusions drawn from,responses recordable within 100 msec. The receptivefields of the various foot and toe reflexes which couldbe elicited by stimulation of the plantar surface ofthe foot have been examined in detail under varyingexperimental conditions (sect. 2). Special attentionhas been given to the hallux movements evoked, as

39

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Lennart Grimby

well as to the border areas between the skin regionswhere stimulation produces hallux dorsiflexion, viz.,the ball of the toes, and the regions where plantarflexion of the hallux results, viz., the hollow of thefoot (sect. 3). The investigations have been limitedto a selected material of 25 subjects who wereneurologically healthy, to judge from their medicalhistory and physical examination, and who provedto have typical and brisk flexor plantar responses.The primary aim of the work has been to establishthe reflex variants that can be observed in a typicalnormal material under varying experimental con-ditions, and only secondarily to get an idea of thefrequency of the different reflex variants in generalin healthy humans.

METHODS

The technique employed in the experiments to be des-cribed below is essentially the same as that used byKugelberg et al. (1960), and as a detailed descriptionhas been given in their work only the salient data willbe given here.

Electric stimuli were applied to the skin on the plantarsurface of the foot by means of a pair of needle electrodesinsulated except for the tip and inserted a few millimetresapart into the horny layer of the skin. In most experimentsthe stimulus consisted of a series of shocks delivered overa period of 20 msec. and of a single duration of 1 msec.and a frequency of 500/sec. The maximum strength of thecurrent was approximately 25 mA. The stimuli wereexperienced as a pinprick and, when they produced areflex response, were as a rule slightly painful. Most ofthe experiments could, however, be carried out withstimulus strengths causing only slight discomfort.The reflex responses obtained were recorded electro-

myographically. A pair of needle electrodes insulatedexcept for the tip were inserted 1 or 2 cm. apart into eachmuscle. Most of the muscles studied were easily accessibleto examination, but for some of them that were not soeasily found the following procedure had to be adopted.In the short hallux flexor the electrodes were placedimmediately medial to the tendon of the long halluxflexor, a few centimetres behind the ball, at a depth ofabout 1 cm. In the long hallux flexor the electrodes wereinserted from the lateral side of the leg, about 5 cm.proximal to the malleoli; the muscle could then be foundat a depth of about 2 cm. In the long hallux extensor andthe extensor digitorum longus the electrodes were placedimmediately below the respective tendons, a few centi-metres proximal to the malleoli. Confirmation of theplacing was obtained by the following checks:-Passiveextension of the muscle should cause displacement of theelectrode; voluntary contraction of the muscle shouldgive rise to several action potentials; no activity shouldresult on voluntary contraction of nearby muscles. Wheninserting the electrodes into the different muscles, greatcare was taken to attain the most favourable recordingconditions. In this manner our previous recordingtechnique could be considerably improved. At least as far

as the short hallux extensor and flexor are concerned, ithas become possible to record contractions that are tooweak to result in discernible movements and to keep thestimulus strength low enough not to make the subjectexposed to the procedure averse to further cooperation.

In the following, when referring to the strength of theelectromyographic reflex, this expression is used to denotethe amount of activity evoked.

In addition to the electromyographic recording of thereflexes, the movements evoked have also been observedby ocular inspection. In one series of experiments theywere also recorded photographically by means of astroboscope giving a series of pictures at intervals of50 msec. This recording was, however, abandoned as ithad no advantage over the electromyographic recordingplus inspection beyond giving a picture of the movements.

RESULTS

1 LATENCY VALUES OF REFLEXES STUDIED A strongpainful stimulation of the plantar surface of the footmay give rise to a series of discharges in the muscleinvolved, of latencies varying between 50 and 500msec. Responses of short latencies were found to becomparatively stable, whereas the later responses arevarying and very susceptible to habituation. If, in asubject with brisk reflexes, a strong stimulus is appliedto a skin area from which no reflex activity can beevoked in the short hallux flexor, and if the subjectis instructed to bend his toes as soon as he experiencesthe stimulus, the latencies of the voluntary flexoractivity thus obtained have been found to be highlyvariable from one stimulation to the other, the sub-ject being never capable of reacting within 150 msec.and only exceptionally within 200 msec. The presentwork is mainly concerned with reflex responsesobtained within 100 msec. and of constant latencyfrom one stimulation to another. There is thus abroad margin between the latencies of the reflexresponses studied here and the purely voluntaryresponses.A threshold stimulus often results in reflex re-

sponses of comparatively long latency, and whenusing weak stimuli it is difficult to judge whetherthe recorded activity is to be considered as a reflex.A progressive increase of the stimulus strengthgradually shortens the latency of the reflex up to acertain limit which differs somewhat in differentindividuals. The shortest latency observed for thereflexes in the short hallux flexor and extensor is55 msec. With the present technique it is, however,as far as most subjects are concerned, impossible toobtain shorter latencies than 70 to 80 msec. Theseminimum latencies require strong stimuli and canoften not be obtained until a strong stimulus hasbeen repeated once or twice. The shortest latencieswere observed in subjects with brisk reflexes, and in

40

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Normal plantar response: integration offlexor and extensor reflex components

these cases the minimum latency of the subject canoften be reached without having to resort to maxi-mum stimulus strength. In subjects with higherreflex thresholds it is generally impossible to obtainsuch short latencies, but it would seem as though thelatency might be further reduced in these cases ifstronger stimuli could be set in. As a rule, the mini-mum latency is slightly shorter when the hollow ofthe foot is stimulated than on stimulation of otherareas. The reason is probably that the stimulusstrength available is not sufficient to bring forth theactual minimum latency when less sensitive parts ofthe foot are stimulated. There is no difference inlatency between a short hallux extensor reflexelicited by hallux stimulation and a short halluxflexor reflex evoked by stimulation of a part of thesole with the same reflex threshold.The short latency values mentioned require very

high conduction velocities in the reflex arc and avery short central reflex time. A determination of theafferent conduction velocity is of special interest, asthis may give an idea of the types of impulses givingrise to the reflex type studied. It has, however, notbeen possible to make an exact determination, as novalues were available as to the efferent conductionvelocity and the central reflex time of the subjectsexamined. Hodes, Larrabee, and German (1948)have shown, however, that the efferent conductionvelocity of the fibres to the short hallux flexor are notlikely to exceed 60 m./sec. Kugelberg and Hagbarth(1958) have shown that the central reflex time for theabdominal reflex, which is also polysynaptic, is notlikely to be below 3 5 msec. In one subject examined,with an afferent and efferent conduction distance ofabout 135 cm. each way, the minimum latency forthe short hallux flexor reflex was 55 msec., calcu-lating from the stimulus onset. This short latencycould, however, not be obtained by a single shockbut required two to five shocks at intervals of 2msec.; the actual latency must thus be at least 2msec. shorter than 55 and the maximum afferentconduction velocity can then be calculated to be notbelow 50 m./sec.

2 VARIATIONS OF REFLEXES WITH SITE AND STRENGTHOF STIMULUS The conclusions drawn in this sectionhave mainly been based on the electromyographicrecording, as the latencies of the movements evokedhave not been measured. The intensity of a reflexin a certain muscle has been compared with that ofan antagonist muscle while shifting the site or thestrength of the stimulus. All conclusions are based onexperiments in which the recording electrodes werenot displaced in the course of the experiment. Underthese experimental conditions, a change in therelation between the reflex intensities in the two

muscles must be identical with a change or a ten-dency to a change in the direction of the movementevoked.

Plantar and dorsal flexion of the great toe As hasbeen shown previously (Kugelberg et al., 1960),stimulation of the ball of the great toe always resultsin dorsiflexion of the toe and electrical activity in theshort and the long hallux extensor but not, especiallywhen using weak stimuli, in the short and the longhallux flexor (Fig. 1). A certain activity may, how-ever, sometimes be observed in the short halluxflexor (sect. 3). As the stimulus is successively shiftedbackwards towards the hollow of the foot, vaguehallux movements are evoked within a rather broadarea corresponding to the ball of the foot. Anelectromyographic investigation reveals (see Fig. 3)that the short hallux extensor reflex is graduallyreduced in this area and that there appears insteada progressively intensified reflex in the short halluxflexor. In the same way, by shifting the stimulustoward the lateral side of the sole, via the ballsof the lesser toes, the short hallux extensor reflexis gradually substituted by a flexor reflex. Stimulationof the hollow of the foot or the lateral side of thesole evokes (Fig. 1) plantar flexion of the great toeand electrical activity in the short hallux flexor butnot, especially when using weak stimuli, in the shorthallux extensor (Kugelberg et al., 1960). Sometimes,however, a certain activity can be seen in the shorthallux extensor (sect. 3).As appears from Fig. 1, there is no marked change

in the activity of the long hallux extensor when thestimulus is shifted from the hallux to the hollow ofthe foot, viz., the muscle is involved in the reflexmovement both on plantar and dorsal flexion ofthe hallux. This seems to be due to the doublefunction of the muscle as a dorsal flexor both at thehallux joint and the ankle joint (cf. Kugelberg et al.,1960). To what extent a contraction of the musclewill result in dorsiflexion of the great toe and towhat extent it will result in dorsiflexion of the footprobably depends on what other muscles are simul-taneously engaged in the reflex movement.

Electromyographic recording in the long halluxflexor is a rather difficult undertaking, and previousattempts to record reflex activity from this musclehave not been successful. In the material studied inthe present investigation it has, however, been pos-sible to record a reflex activity in the long halluxflexor on stimulation of the medial side of the solein a few cases under especially favourable recordingconditions. In these cases the latency values were,however, about 30 to 40 msec. longer than for otherreflex responses (Fig. 1). No satisfactory explanationfor this has been found, but it can be establishedthat the latency of the activity in the long hallux

41

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Lennart Grimby

A

H Pr

A

EDB :

DBP H Lt

FHSS~ J-'_ v

FIG. 1. Reflex obtained in one muscle is substituted by reflex in antagonist after shifting position of stimulus. Stimuliapplied to points marked A, B, C, D; recordings from FHB=flexor hallucis brevis, EHB =extensor hallucis brevis,FHL=flexor hallucis longus, EHL= extensor hallucis longus, FDB=flexor digitorum brevis, EDB= extensor digitorumbrevis, TA = tibialis anterior, PL=peroneus longus, EDL= extensor digitorum longus. Time 10 msec.

flexor is so short (down to 90 msec.) that it islikely to be of spinal origin. No activity in the longhallux flexor could be recorded by stimulation of theball of the great toe, nor by stimulation of the lateralside of the foot. This is of a certain interest, as itmay perhaps give a hint why the Babinski signs are

practically always easier to evoke by stimulation ofthe lateral than of the medial side of the sole.As appears from the foregoing, the hallux move-

ment evoked is closely reflected in the relationbetween the strength of the reflexes in the shorthallux flexor and extensor, respectively, as observedby simultaneous recordings in the two antagonists.

In their studies of normal and pathological plantarreflexes Landau and Clare (1959), however, regularlyobtained simultaneous responses of fairly equalintensity in the short hallux flexor and extensor bothon plantar and dorsal flexion of the great toe. This

may be due to the fact that they used surfaceelectrodes; these have not nearly the same selectivityas needle electrodes but may pick up activity alsofrom the interossei which are engaged in the reflexresponse independently of the direction of the halluxmovement (Kugelberg et al, 1960). Besides, whenusing surface electrodes a strong activity in the deepshort hallux flexor muscle may very well appear to beweaker than a feeble response from the short halluxextensor which is more superficial and thus easier torecord from. They have, also, mainly been concernedwith reflexes of long latency and, as will be furthermentioned below, the tendency to simultaneousactivity in the short hallux flexor and extensor ismore pronounced for reflexes of long than for thoseof short latencies.The hallux movement evoked is of course also

influenced by the activity prevailing in the strong

L}

rHL

A1

K

P L

l

.: ...:..... .o--'FDL .-,

%. :,. i. ...

4211-111-

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Normalplantar response: integration offlexor and extensor reflex components

long hallux muscles but these are not so closelycorrelated to the hallux movement as are the shortmuscles; most likely this is due to the fact that thelong muscles are also involved in the movements ofthe ankle.

In contrast to the results described here, Landauand Clare, on stimulation of the planta, succeededin recording long hallux extensor activity only inconnexion with the vigorous dorsiflexion of the footand hallux obtained in pathological cases, and theydrew the conclusion that 'the extensor reflex is not adifferent reaction from the flexor but rather a hyper-active flexor response in which the extensor hallucislongus is included by irradiation'. As, however,concentric needle electrodes were used in theirexperiments, their recordings emanate from a verysmall part of the muscle; besides, they have chieflystimulated the lateral side of the sole, which has ahigher threshold for the long hallux extensor responsethan has the medial side in normal cases.

In view of the great clinical importance of thereflex movement of the great toe, a special study hasbeen performed concerning the variability of therelation between the strength of the short halluxflexor and extensor reflex (sect. 3).

Supination and pronation of the foot When astimulus is applied to the medial side of the planta,the result is generally a tendency to supination of thefoot (cf. Babinski, 1898), whereas stimulation of thelateral side often results in a tendency to pronation.As the stimulus is shifted from the medial to thelateral side of the planta a gradual shift can beobserved in the intensity of the reflex responses frommuscles partly involved in supination toward musclespartly involved in pronation, the former subsidingin proportion as more vigorous responses appear inthe latter (Fig. 1). Reflex activity is also set up in thetibialis posterior on stimulation of the medial (butnot of the lateral) side of the planta. The differencesthus observed vary greatly from one individual toanother. They may be large enough to be observedby mere inspection; in other cases, although notapparent to the eye, they are reflected in the electro-myographic recording as a distinct change in reflexpattern as the stimulus is shifted; in still other casesno differences are recorded even in the electro-myogram.

In many individuals also the toes are involved inthe supination-pronation process when a stimulusis applied to the ball of the foot and especially whenapplied to the balls of the toes. On stimulation of thegreat toe, dorsiflexion of this toe and plantar flexionof the lesser toes can be observed in connexion withthe general supination. On stimulation of the fifthtoe, plantar flexion of the great toe and dorsiflexionof the lesser toes can be seen connected with the

general pronation. The electromyographic recordingshows (Fig. 1) that when the stimulus is shifted fromthe first to the fifth toe the short hallux extensorreflex is substituted by a flexor reflex and the reflexin the flexor digitorum brevis is replaced by one inthe extensor digitorum brevis. A distinct plantarflexion of the great toe following stimulation of thefifth toe, and a similar flexion of the lesser toes afterhallux stimulation were, however, observed only inpart of the material studied and mainly on applica-tion of relatively weak stimuli. The stronger thestimulus, the greater is the tendency to simultaneousdorsiflexion of all toes. That in the previous in-vestigations dorsiflexion was a regular finding,irrespectively of which toes were stimulated(Kugelberg et al., 1960), is probably due to the strongstimuli employed.The differences in reflex responses obtained on

medial and lateral stimulation are often very vaguewhen the stimulus is set in without previous warningbut distinct when the subject knows in advance whenand where the stimulus is going to be applied.

Plantar and dorsalflexion of the foot As has beendemonstrated previously (Eklund et al., 1959;Kugelberg et al., 1960; Hagbarth, 1960), stimuliapplied to the ball and hollow of the foot producedorsiflexion at the ankle joint and activity in thetibialis anterior, whereas stimuli to the posteriorparts of the planta, and particularly to the plantarsurface of the heel, result in plantar flexion at theankle joint and activity in the gastrocnemiusmuscle.A closer electromyographic study of these reflexes

shows that the tibialis anterior reflex is most readilyevoked by stimulation of the anterior and medialparts of the planta and that there is a gradual rise inreflex threshold as the stimulus is shifted toward thelateral or posterior parts of the foot. In someindividuals the tibialis anterior reflex is graduallysubstituted by a gastrocnemius reflex as the stimulusapproaches the plantar surface of the heel. In otherindividuals stimulation of the heel may produceeither a weak gastrocnemius reflex or a weak tibialisanterior reflex, and it would seem as though thetibialis reflex is more common if the heel stimulationis preceded by a strong stimulus to the anterior partof the foot.

Influence of stimulus strength on reflex movementAs shown in the previous section, the site of thestimulus determines the direction of the various com-ponents constituting the reflex movement, but thestimulus strength also plays a certain role in thisconnexion. It is a clinically well-known phenomenon(cf. Riddoch, 1917) that very strong and suddenstimuli applied to the planta may, even in apparentlyhealthy individuals, provoke brisk flexor reflexes

43

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Lennart Grimby

FIG. 2. Very strong stimulifavour the EHB reflexmore than the FHB reflex.Upper record result of weak,lower record of strongstimuli applied to anterior

part ofplanta; recordingsfrom FHB (top) andEHB (lower tracings).Time 10 msec.

with hallux dorsiflexion, but so far evidence islacking whether this phenomenon is of a spinalnature or not.The reflex changes obtained by varying the strength

of the electric stimulus are as a rule too small to beapparent to mere inspection but as recorded in theelectromyogram they may be significant. When astimulus is applied to an area where it gives rise toreflexes both in the short hallux flexor and extensor,it is often possible to see quite distinctly how,following a strong increase in stimulus strength, theextensor reflex increases in intensity at a faster ratethan the flexor reflex (Fig. 2). Correspondingly,a very strong stimulus favours the extensor digi-torum brevis reflex more than the flexor digitorumbrevis reflex; it favours the tibialis anterior reflexmore than the gastrocnemius reflex. As the stimulusstrength is increased, the activities in musclesengaged in the pathological reflex will thus tend to bepredominant.

3 VARIABILITY OF REFLEX PATTERN IN SHORT HALLUXFLEXOR AND EXTENSOR In view of their clinicalimportance, the reflex movements of the halluxelicited by painful stimulation of the plantar surfaceof the foot deserve a closer investigation. Of specialinterest is the variability of the reflex movements indifferent individuals and, also, in different recordingsfrom one and the same individual. This variabilityought to be most pronounced in the border areabetween the toes and the hollow of the foot. Thevague movements elicited in these areas cannot bediscerned by ocular inspection, but by means ofelectromyographic recording it is possible to follow,step by step, how the short hallux extensor reflexis gradually being substituted by a short halluxflexor reflex as the stimulus is shifted from the

hallux to the hollow of the foot and as the reflexmovement changes from dorsal to plantar flexion(Fig. 3). When a stimulus gives rise to reflex activityboth in the short hallux flexor and extensor, thedischarges elicited in the two antagonists are hardlyever of exactly identical latency but tend to bealternating. Purely synchronous contractions occuronly exceptionally or following excessively strongstimuli. However, simultaneous contractions (oflatencies of 200 to 500 msec.), can often be observedafter the initial reflex responses, and they may alsooccur as a result of strong stimuli of the usualclinical type, such as heavily dragging a pin along theplanta.

In the following, the expression 'reflex pattern'will be used to denote the combination of alternatingdischarges obtained in the short hallux flexor andextensor by a given stimulus and by simultaneousrecording in the two antagonist muscles; this reflexpattern will be presumed to give a true reflection ofthe hallux movement evoked.On application of very strong stimuli there are

practically always signs of activity both in the shorthallux flexor and extensor; in a border area betweenthe hallux and the hollow of the foot even weakstimulation results in a combination of flexor andextensor activity. This border area will in thefollowing be called the 'transition zone'.The stimulus strength necessary to obtain an

electromyographic response varies with the positionof the electrodes. As these experiments are based ona comparison between the thresholds of the shorthallux extensor and flexor responses, the recordingconditions in the two muscles have to be absolutelyidentical otherwise no comparison is possible betweenindividual reflex patterns. The technique used in thisstudy has made it possible to attain satisfactoryrecording conditions in this respect, and repeatedexperiments on a few subjects have shown that thedifferences caused by the recording have been veryslight and of no consequence for the results.As a rule, the initial reflex pattern has the same

composition as the final pattern, but in severalcases a marked difference has been observed in so faras there has been an early part within 100 msec. anda later part after 150 msec., in which later part flexoractivity has been more dominant than in the earlypart of the reflex pattern (see Fig. 3). The followingdescription will deal exclusively with the early partof the reflex pattern, and when this part consists ofalternating discharges in the short hallux extensorand flexor it is often accidental or a matter of therecording technique whether the discharge in theextensor or the flexor has the shortest latency.

Variations in reflex pattern from one individual toanother The site as well as the width of the tran-

44

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Nornmalplantar response: integration offlexor and extensor reflex components

/ 1~, /1

/ f

vidual to another (cf. Fig. 4).

In one group of cases, the transition zone has a

relatively far distal site, close by or even on, the hallux;in these cases there is a considerably greater ten-*dency toward short hallux flexor activity in thehallux pattern than toward short hallux extensoractivity in the pattern of the hollow of the foot.In another group, the transition zone has a relativelyproximal site, close by or even behind, the ball of thefoot; in these cases the tendency toward short halluxflexor activity in the hallux pattern is much lesspronounced than is the tendency to short halluxextensor activity in the pattern of the hollow of thefoot. The course of the zone in relation to the trans-verse axis of the foot may also vary from one indi-vidual to another. When the zone is located proxi-mally, short hallux extensor activity is generallymore pronounced in the pattern on stimulation ofthe anterior medial part of the sole than of its anteriorlateral part.

In the two groups described above variations mayalso be observed in the width of the transition zone.Thus, in one type of case the transition zone isrelatively narrow, a stimulus shift of only one or twocentimetres being sufficient to change the reflexpattern; in these cases the contrast between thepattern elicited by stimulation of the hollow of the-foot and that evoked by hallux stimulation ispractically maximal. As appears from Fig. 4,stimulation of the hollow of the foot results in astrong, short hallux flexor reflex but none in the

FIG. 3. Short hallux reflexis gradually substituted byshort hallux entensor reflex asstimulus is shiftedfrommiddle ofplanta to hallux ball.Stimuli applied as shown on theschematic drawing; recordingsfrom FHB (top) and EHB(lower tracings).Time 10 msec.

..........

extensor; hallux stimulation results in a strong,short hallux extensor reflex but none in the flexor.In another type of case the transition zone isrelatively broad; it may even be necessary to movethe stimulus from the hallux to the hollow of thefoot before any significant change can be observedin the reflex pattern. In these cases there is also aless pronounced contrast between the reflex patternresulting from stimulation of the hollow of the footand that resulting from hallux stimulation. As shownin Fig. 4, the reflex pattern of the hollow of the footincludes, besides the dominant short hallux flexoractivity, also signs of extensor activity; the halluxpattern includes, besides the dominant short halluxextensor activity, signs of flexor activity. There is nodirect correlation between the strength of the clinicalplantar reflex and the width of the transition zone,but it would seem as though remarkably broadtransition zones were more common in individualswith high reflex thresholds than in those with briskreflexes.On an average, the width of the transition zone

in the material studied has been some centimetres,and it has been located between the base of the toesand the ball of the foot. The reflex patterns depictedin Fig. 4 should be regarded as extreme; generallythe plantar pattern and the hallux pattern are in therange between those shown in Fig. 4.

Variations in reflex pattern with subject's attentionand expectancy The reflex pattern elicited by a givenstimulus does not only differ from one individual toanother but may also vary in one and the same indi-

45

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

46

* .eA

t

r .X .S

FIG. 4. Individual variationsapplied to hallux ball (left-ha)planta (right-hand column); i

and EHB (lower tracings). Tiiwith distally located transitionshort hallux flexor reflex; cas

located transition zone and relkextensor reflex; case C, subzone andgood contrast betweecase D, subject with broad tracontrast between hallux andp

vidual, as has been shownusing stimuli of a givenelectrode positions througlThe reflex response obta.

ted stimulation is generall:On repeated stimulation;occurs, resulting in a weakbrief-lasting response; eve]lations the responses howeboth in regard to intensit2series of about ten stimulation effects can, as a rule,The initial stimulation i

give rise to a reflex pattei

Lennart Grimby

stimulus site; as a rule, these atypical features haveAs =:disappeared on the second stimulation. The tendency

to atypical responses to the initial stimulation is,besides, much more pronounced when using weakthan when using strong stimuli. In some subjects

________________ with very brisk reflexes a weak but pure short halluxextensor reflex may be observed as the result of aweak, scarcely painful stimulation of the hollow ofthe foot; or a weak but pure flexor reflex may appearon weak, scarcely painful hallux stimulation.

Except for the first stimulations in a series, nosignificant successive changes can be observed in thereflex patterns elicited. Some fluctuations in thecomposition of the pattern may, however, occur inthe series; these variations are more pronouncedwhen using comparatively weak stimuli and onstimulation in the transition zone. As appears fromFig. 5, a series of stimuli applied to this zone maygive rise to reflex patterns in which either the shorthallux flexor or the short hallux extensor is dominant.In a series of stimulations of the hollow of the footor of the great toe, however, such marked differencesare exceptional, very likely due to the fact thatactivity in the short hallux flexor or extensor, res-pectively, is so strongly dominant at these stimu-lation points that minor fluctuations in the reflexpattern do not become apparent.

If a strong hallux stimulation is interpolated in aseries of stimulations in the transition zone, thesubsequent reflex pattern is often characterized by amore dominant short hallux extensor activity than

in reflex patterns. Stimuli was the reflex pattern preceding the interpolated

recolding)from FHB (top) stimulation. A strong interpolated stimulation of the

me 10 msec. Case A, subject hollow of the foot results in a corresponding shiftzone and relatively dominant toward predominance of short hallux flexor activity.e B, subject with proximally In a few easily suggestible subjects, such shifts toatively dominant short hallux dominant short hallux extensor or dominant flexorwject with narrow transition activity could be evoked merely by a verbal threat ton hallux andplanta patterns; set in strong stimuli on the planta or the hallux.mnsition zone and less marked It is evident that cerebral factors, such as the sub-lanta patterns. ject's attention or expectancy, have a certain

influence on the reflex pattern. By adopting the fol-in a series of experiments lowing experimental procedure, these cerebral factorsstrength and unchanged could be varied according to the intentions of thehout the experiment. investigator and studied more systematically. Theined on the first, unexpec- subject was instructed to concentrate his thoughts ony strong and long-lasting. making a certain movement with his toes withouta very rapid habituation actually contracting the muscles, which could beer and, particularly, more checked by the subject himself by listening to a loud-n after one or two stimu- speaker connected to the amplifier. Suddenly aaver seem to be stabilized relatively weak, unexpected stimulation was elicited;y and duration, and in a by exposing the subject to strong stimuli he mayations no further habitu- become averse to further cooperation in the course, be observed. of the experiment. Experiments of this kind aren a series may sometimes rather difficult and have, in fact, only been performedrn that is atypical of the on the author and two other subjects who were

1,

)rKLO'

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Normal plantar response: integration offlexor and extensor reflex components

willing and capable of a high degree of cooperation.On stimulation in the transition zone, the short

hallux extensor reflex becomes stronger and theflexor reflex weaker if the subject is intent on dorsi-flexion of the great toe; in the same way, the shorthallux flexor reflex becomes stronger and the extensorreflex weaker if he is intent on plantar flexion of thetoe. A relatively strong short hallux extensor reflex,without any perceptible activity in the flexor, oftendevelops when the subject is intent on dorsiflexion ofthe hallux, and a relatively strong short halluxflexor reflex, without any perceptible activity in theextensor, often results when the subject is intent onplantar flexion of the hallux.

If the hollow of the foot is stimulated while thesubject is intent on dorsiflexion of the hallux, reflexactivity develops in the short hallux extensor, eitherimmediately preceding or immediately followingupon the flexor reflex which at the same timebecomes weaker (Fig. 5). Occasionally, short halluxextensor activity is predominant in the reflex patternevoked, and in these cases a brief-lasting but distinctdorsiflexion of the hallux may also be observed.

*8$~ ~ *N* Ni=iSssPis->.A:

FIG. 5. Reflex pattern variations in a given individual, atgiven stimulus site and strength. Records from FHB (top)and EHB (lower tracings). Left-hand column shows 'spon-taneous' variations in series of10 stimuli to transition zone.Right-hand column shows normal reflex pattern on stimu-lation of middle of planta (upper record) and deviationsresulting when the subject is intent on hallux dorsiflexion atmoment of stimulation (lower records). Time 10 msec.

If the hallux is stimulated while the subject isintent on plantar flexion of this toe, reflex activitydevelops in the short hallux flexor side by side withthe simultaneously reduced extensor reflex. In a fewrecords, short hallux flexor activity has been slightlydominant in the reflex pattern evoked, but no distinctplantar flexion of the hallux has been discernible.

Changes in reflex pattern under voluntary contrac-tion of the short hallux flexor or extensor In theirstudies of the abdominal reflex, Kugelberg andHagbarth (1958) demonstrated that reflex responsesin a muscle are facilitated by voluntary contractionof the muscle and inhibited by voluntary contractionof its antagonist. They also showed that the facilit-atory effect does not increase in parallel with thestrength of the voluntary contraction. As a rule,voluntary contraction of the short hallux flexorresults in facilitation of the reflexes in that muscleand inhibition of reflexes in the short hallux extensor,and vice versa. Occasionally, however, strong con-traction of the short hallux extensor does not resultin any discernible facilitation of the extensor reflex,nor in inhibition of the flexor reflex. Besides, asshown above, the facilitatory and inhibitory effectsmay develop if the subject is intent on making amovement without actually contracting the muscles.In some experiments performed on one subject, thefollowing observations were made. If during amaintained weak voluntary plantar flexion of thetoes an unexpected hallux stimulation was set in, theshort hallux extensor reflex became substantiallyreduced and the voluntary background activity inthe short hallux flexor considerably increased. If, onthe other hand, the hallux stimulation was set induring plantar flexion of the toes and while thesubject expected the stimulus, the short halluxextensor reflex was practically normal and thevoluntary flexor activity inhibited. It is evident thatthe facilitatory and inhibitory effects do not onlydepend on the contraction as such, and it seems asthough the subject's expectancy might be of decisiveimportance.

DISCUSSION

When strong stimuli are applied to the sole, theearliest reflex discharges evoked are of such shortlatencies (down to 55 msec. for the reflexes in theshort hallux flexor and extensor) that only purelyspinal reflex arcs can be involved. When weakerstimuli are used the latency of the response becomeslonger, but the change is gradual and even when thelatency is in the vicinity of or slightly above 100 msec.the initial part of the reflex must be presumed to beof spinal origin. This does not of course exclude thatthe later part of the reflex may be conducted through

47

-'.. ,.F...-

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Lennart Grimby

a cerebral reflex arc although this is hardly probablefor the discharges to be discussed below, viz., thoserecorded within 100 msec. from the onset of thestimulus. When reflex discharges of very shortlatencies are elicited both in the agonist and theantagonist, the briefest latency may be recordedalternatively from both muscles, and both reflexesshould be considered as being fundamentally equi-valent. Part of the volleys may perhaps be some kindof rebound phenomenon, but no attempts have beenmade to study this problem more closely, as themain subject of the present investigation has been tostudy the direction of the toe movement evoked, andthis direction is influenced by all the dischargesrecorded. As mentioned above, those parts of thereflex that are oflonger latency than about 150 msec.are of a more uncertain origin; of special interest arethe reflexes composed of two distinctly separateparts, viz., an early part of a latency between 50 and100 msec., and a later part of a latency between 150and 200 msec. The two parts of the reflex mayrepresent two different responses, one of spinal andone of cerebral origin, but it may also be possiblethat both are spinal, the latter part being transmittedby slower afferent impulses or via more polysynapticreflex arcs. In a reflex pattern obtained from a'transition zone', as defined above, and composedof these two parts, short hallux flexor activity isoften more dominant in the later than in the earlypart of the reflex. On stimulation of the anteriorpart of the planta by the brief electric stimulusused in the present experiments, there is often atendency toward hallux dorsiflexion, which is onlyrarely seen in connexion with the long-lastingmechanical stimulation used in clinical routine work.It would seem as though the later parts of the reflexare more conspicuous in the clinical type of stimu-lation than when using the electric stimuli.The latencies measured in these experiments are

so short that the maximal afferent conduction velo-city in the reflex arc can hardly be below 50 m./sec.(see Results). The conduction velocity of the purepain fibres of the delta group has been calculated tobe about 20 m./sec. (Gasser, 1943), and eventhough no exact figures are available as to thequickest conduction rate in the human pain fibres,the maximal afferent conduction velocity estimatedin the present experiments might be taken as anindication that, besides pain impulses, touchimpulses are also of importance for the initiation ofthe reflexes studied. A previous sensitization isnecessary before extremely brief latencies can berecorded, and it is possible that touch impulses donot become involved to any greater extent until aftersensitization.The toe movements provoked by painful stimu-

lation of the plantar surface of the foot may beeither dorsal or plantar flexion; at the ankle jointthe resulting movement may be either dorsal orplantar flexion and either supination or pronation.The character of the reflex movement evoked on eachsingle occasion is determined by the stimulus site; itrepresents the most appropriate movement for awithdrawal of the stimulated area from the harmfulstimulus, and this functional differentiation of thereflex response must be presumed to be dependenton suprasegmental control over the spinal reflexcentre (cf. Kugelberg et al., 1960). That the reflexobtained on excessively strong stimulation resemblesthe pathological stereotyped flexor reflex may be dueto the suprasegmental control not being sufficientlyprevalent at these very high stimulus intensities. Themore subtle integration of the reflexes into highlypurposeful movements requires that the subjectexpects the stimulus and knows where it will beapplied. The reflex mechanism is thus to some extentcapable of 'learning' and is evidently under a directinfluence from higher cerebral levels capable ofchanging the direction as well as the intensity of thereflex movements at the various joints. Of greatinterest in this connexion are the recent investigationson decerebrate cats by Holmqvist and Lundberg(1961), demonstrating that the flexor motoneuronactivity elicited by skin stimulation of an extremitycan be both facilitated and inhibited at a supraspinallevel.

If, in a given subject, a given point of the foot isstimulated and the activity thus obtained is recordedsimultaneously in the short hallux flexor and exten-sor, the result is a reflex pattern of a certain basiccomposition. With changes in the subject's attentionand expectancy, significant changes may appear inthis basic pattern, either in the form of increasedflexor or increased extensor activity, and it is reason-able to presume that a still wider range of variationscould be obtained by a more elaborate experimentalarrangement than that used in this investigation.The range of variations is larger on weak than onstrong stimulation, and it would seem as thoughcerebral factors played a more important role onweak than on strong stimulation, whereas thestimulus site is, relatively, more important on strongstimulation. Under certain conditions a weakstimulation applied at the middle of the planta maygive rise to a reflex pattern in which short halluxextensor activity is dominant; in this connexion thereis also sometimes a distinct hallux dorsiflexion, andeven if it is very brief it is anyhow typical of the signof Babinski as described by him (1896), viz., halluxdorsiflexion elicited by painful stimulation of the sole.It is, however, striking how rapidly and completelysuch an atypical reflex pattern is corrected on

48

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Normal plantar response: integration offlexor and extensor reflex components

repeated stimulation, and there is nothing to suggestthat a reflex pattern of a healthy individual could bepermanently changed into the types of pattern seenin pathological cases (Grimby, to be published).The results obtained from experiments on different

subjects differ significantly in several respects. Therange of variations of the reflex pattern obtained ata given stimulus site may be rather wide in someindividuals and very narrow in other cases, but thesedifferences may be due to psychological factors. Inthe basic reflex pattern obtained at each stimuluspoint there are, however, such pronounced individualdifferences, for instance in regard to the width andlocation of the transition zone, that it must be in-ferred that the organization of the reflex mechanismis not quite uniform from one individual to another.These individual variations must be due to differ-ences in the organization of the suprasegmentalcontrol (cf. above). In the material studied, vari-ations in the width of the transition zone areindependent of its location, and the suprasegmentalcontrol must thus be presumed to occur via more thanone pathway.Although it cannot be excluded that there may

have been pathological changes in the reflex mechan-ism of one or two of the subjects studied, such casescannot explain the individual variations observed;the normal range of variations should rather bepresumed to be considerably wider than thatactually observed. The average width of the tran-sition zone has been found to be a few centimetres,and its average location has been between the halluxbase and the ball of the foot; the more extreme thedeviations from these average findings, the morerare have they been in the material studied. Thismaterial is limited but the average values mentionedabove might be considered as fairly representativealso of a larger material chosen in the same manner,although no doubt the range of variations will thenbe wider. The material studied has further beenselected in so far as all the subjects have had clinicallytypical flexor plantar responses, and healthy indi-viduals with exceptionally proximal transition zonesmay have been excluded; all the subjects have alsohad clinically brisk reflexes, and healthy subjects withextremely broad transition zones may have beenexcluded, this type of transition zone being, as itseems, more common in individuals with high reflexthresholds than in those with brisk reflexes. Thedeviations from typical normal cases that may beobserved in various pathological conditions mustthus be judged with the utmost caution. It would beof great interest to fix an extreme limit for thenormal variations, but this would imply testing alsoof individuals who are apparently healthy but haveplantar responses widely deviating from the normal,

and there is no safe method to determine whetherthe reflex mechanisms of these individuals are actuallyintact.

SUMMARY

In an investigation of the skin reflexes of the foot,experiments were carried out on a group of 25neurologically healthy subjects with brisk and typicalflexor plantar responses. Painful stimuli consistingof a series of repetitive electric shocks delivered overa period of 20 msec. and of a single duration of1 msec. and a frequency of 500/sec. were applied tovarious points on the skin of the plantar surface ofthe foot. The resulting reflex movements were ob-served and the discharges obtained in variousmuscles of the foot and lower leg were recordedelectromyographically.

1 The study has mainly been limited to reflexesof shorter latency than 100 msec.; these were pre-sumed to be of purely spinal origin. On stimulationof the hollow of the foot, the shortest latencyobserved for the reflexes recorded in the short halluxflexor and extensor was 55 msec. The maximalafferent conduction velocity in the reflex arc wasestimated to be not below 50 m./sec.

2 As the stimulus was shifted from the hollowof the foot to the hallux ball, the hallux movementprovoked gradually changed from plantar to dorsalflexion. As the stimulus was shifted from the hollowof the foot to the plantar surface of the heel, theresulting reflex ankle movement gradually changedfrom dorsal to plantar flexion. As the stimulus wasshifted from the medial to the lateral side of the sole,the ankle movement evoked gradually changed fromsupination to pronation.

3 The coordinated reflex movement resultingin each single case represents a defence mechanismby which each skin area is removed from the noxiousstimulus. On sudden, unexpected stimulation onlythe coarse features of this functional organizationappear; the more subtle integration of the reflexesinto highly purposeful movements requires that thesubject expects the stimulus and knows where it willbe applied.4 On excessively strong stimulation, the differ-

ences in reflex responses caused by changes of thestimulus site become less pronounced, and thecoordinated reflex movement evoked tends to re-semble the pure flexor reflex seen in pathologicalcases.The relation between the strength of the short

hallux flexor and extensor reflexes, as observed in the'reflex pattern' obtained by simultaneous electro-myographic recording from the two antagonists, hasproved to be a sensitive index of the hallux move-

49

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from

Lennart Grimby

ment elicited by the stimulus. In view of the clinicalimportance of the hallux movements a special studyhas been performed on the variability of this reflexpattern.

1 If, in a given individual, a given point of thefoot is stimulated, a reflex pattern of a certain basiccomposition results. Fairly wide deviations from thisbasic pattern may result from changes in the cerebralinfluence on the spinal reflex centre. Thus, for in-stance, a stimulus applied to the hollow of the footin healthy individuals may give rise to a reflexpattern dominated by short hallux extensor activityand even to a distinct hallux dorsiflexion. Onrepeated stimulation the normal reflex pattern is,however, always restored, viz., short hallux flexoractivity is dominant and the hallux plantar flexed.2 The variations displayed in the reflex pattern

as the locus of stimulation is varied differ significantlyin more than one way from one individual toanother. Thus, the boundaries between the skinareas where short hallux flexor and where shorthallux extensor activity is dominant may be moreor less distinct and, further, the receptive field of theshort hallux extensor reflex may be relatively small

or relatively large as compared with that of the shorthallux flexor reflex. It has been concluded that theorganization of the suprasegmental control over thereflex centre in different individuals is not uniformand that this suprasegmental control is exerted viamore than one pathway.

3 The variability of the reflex pattern in typicalnormal subjects has been explored, and the resultsobtained are intended to provide a basis for furtherinvestigations to be performed on pathological cases.

REFERENCES

Babinski, J. (1896). C. R. Soc. Biol. (Paris), ser. 10, 3, 207.- , (1898). Sem. mid. (Paris), 18, 321.Eklund, K., Grimby, L., and Kugelberg, E. (1959). Acta physiol. scand.,

47, 297.Gasser, H. S. (1943). Res. Publ. Ass. nerv. ment. Dis., 23, 44.Hagbarth, K. E. (1960). J. Neurol. Neurosurg. Psychiat., 23, 222.-, and Kugelberg, E. (1958). Brain, 81, 305.Hodes, R., Larrabee, M. G., and German, W. (1948). Arch. Netroal.

Psychiat., 60, 340.Holmqvist, B., and Lundberg, A. (1961). Acta physiol. scand. 54, suppl.

186.Kugelberg, E., Eklund, K., and Grimby, L. (1960). Brain, 83, 394.

, and Hagbarth, K. E. (1958). Ibid., 81, 290.Landau, W. M., and Clare, M. H. (1959). Ibid., 82, 321.Riddoch, G. (1917). Ibid., 40, 264.Walshe, F. (1956). Ibid., 79, 529.

50

Protected by copyright.

on March 21, 2020 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.26.1.39 on 1 February 1963. D

ownloaded from