J. Neurol. Neurosurg. Psychiat., 1963, 26, 314

Pathological plantar response: disturbances of thenormal integration of flexor and extensor

reflex componentsLENNART GRIMBY

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

Previous reports (Kugelberg, Eklund, and Grimby,1960; Grimby, 1963) have presented the results ofelectromyographic studies ofspinal reflexes elicited inhealthy subjects by painful brief electric stimuli ap-plied to the plantar surface of the foot. Special atten-tion was paid to the 'reflex pattern' obtained onsimultaneous recording in the short hallux flexor andextensor muscles, since the strength of the responsein the flexor has been shown to be a sensitive indexof the tendency to plantar flexion, and the responsein the extensor of the tendency to dorsiflexion of thegreat toe. The investigations of the reflexes obtainedin healthy individuals revealed that the reflex patternhas a basic composition typical of the stimulus site.Thus, stimulation of the hollow of the foot gives riseto activity predominantly in the short hallux flexor,whereas hallux stimulation results in dominantextensor activity; as the stimulus is successivelyshifted from planta to hallux, the flexor activity isgradually replaced by extensor activity. In an areawhich was called the 'transition zone' and whichroughly corresponds to the ball of the foot, the reflexpattern is composed of alternating flexor andextensor activity of equal strength; the site andwidth of the transition zone vary from individual toindividual. Finally, changes in the subject's attentionand expectancy may result in marked, but onlyoccasional, deviations from the basic reflex patternat a given stimulus site.The electromyographic responses to stimulation

of the planta have previously been studied in certainpatients with severe spinal cord lesions (Kugelberget al., 1960); in these cases the reflex patternsconsisted of pure short hallux extensor activityindependently of the stimulus site. The present studyhas been undertaken to investigate the types ofdeviation from the normal reflex patterns that mayappear in subjects with less severe lesions of thereflex mechanism, and a direct analysis of theBabinski sign in the common clinical sense is notwithin the scope of this work. The material has been

chosen so as to give the widest possible range ofvariations, and no conclusions as to the frequencyof the different types of deviation in unselectedpathological material can be drawn from the results.

MATERIAL AND METHODS

The investigation is based on material consisting of twogroups of cases with spastic para- or hemipareses, viz.,one group of 28 patients with clinically confirmedBabinski responses, referred to as the 'Babinski group',and another group of 12 cases in which the clinicalBabinski response was unexpectedly absent; this groupwill be referred to as the 'non-Babinski group'. Since thereflex patterns studied should be as widely varying aspossible, the patients of the first group were selected onthe basis of the wide range of variations found in the type,site, and extent of the nervous lesions and in their clinicalreflex responses. In some of the patients of the secondgroup the Babinski sign had been present in previousphases of their disorders, but at the clinical examinationin connexion with this investigation flexor plantarresponses were obtained, although slight traces of halluxdorsiflexion might sometimes be observed.The technique employed was the same as that used in

previous investigations (Kugelberg et al., 1960; Grimby,1963). Thus, various spots on the plantar surface of thefoot were exposed to painful repetitive electric shocksover a period of 20 msec., and the reflex responsesobtained were recorded simultaneously in the shorthallux flexor and extensor. All remarkable results werecontrolled by repeated experiments. The study has beenlimited to discharges of latencies short enough to ruleout voluntary responses, attention being particularlyfocused on discharges possible to obtain at such brieflatencies, i.e., below 100 msec. (cf. Grimby, 1963) thatthey must be presumed to be mediated by spinal reflexarcs.When selecting suitable subjects for the investigation,

atrophic short toe extensors were found to be commonamong persons suffering from pronounced centralpareses of long duration, and in several cases the electro-myogram revealed significant peripheral nerve lesions.The same observation was made by Landau and Clare(1959) in their studies of traumatic paraplegia. No

314

Pathological plantar response

attempt was made to find out the cause of the peripheralnerve lesions, which may, however, be secondary to theforced inactivity of these patients. Subjects with atrophicshort hallux extensors have not been included in thematerial since the extensor activity obtained in such casesis no satisfactory index of the tendency to dorsiflexion ofthe great toe.

RESULTS

GENERAL PROPERTIES OF THE REFLEX PATrERN Onstimulation of the planta in normal individuals, thelatency of the first discharges in the short halluxmuscles may be as brief as 55 msec., the reflex attainsits maximum strength almost immediately, and aftera few hundred milliseconds only little activity is left,at least after habituation. In the pathological casesthe latency values were fundamentally the same butthere was often a progressive build-up of the reflexstrength and a much longer duration of the activityelicited.When a stimulus gives rise to reflexes both in the

short hallux flexor and extensor, the normaldischarges have not the same latency but arealternating. Also in the pathological cases suchreciprocal discharges were generally observed in thetwo antagonists. In two cases, however. simultaneousbursts of activity regularly occurred in the shorthallux flexor and extensor; the absence of a distinctreciprocity in these cases may be explained by theinvestigation by Holmqvist and Lundberg (1961) ondecerebrate cats, showing that the reciprocitybetween flexor and extensor motor neurons maybe disturbed at a suprasegmental level.

In normal individuals, cerebral factors mayinfluence the basic composition of the reflex patternobtained at a given stimulus site and strength. Suchvariations, due to changes in the subject's attentionand expectancy, were observed also in the patho-logical material, except in those extreme cases wherepure extensor activity was induced independently ofthe stimulus type. In the present paper, however,only the average pattern typical of a given stimulussite and strength in a given individual will beconsidered.

STIMULATION OF THE MIDDLE OF THE PLANTA Innormal individuals short hallux flexor activity is byfar the most dominant in the reflex pattern onstimulation of the middle of the planta, andsignificant extensor activity is regularly obtainedonly in subjects with a very proximal transitionzone.

Babinskigroup In the Babinski group practicallyall transitional forms, from the normal pure shorthallux flexor to the extreme pathological pure

FIG. 1Reflexpatternsobtained onstimulation of

_ _ _ m id dlW middle ofplanta,illustratingvarioustransitionalforms frompure flexor topure extensor

-A, | responses.

RecordsTITwIlpt from Patients

_______ _____________ in theBabinskigroup, all

L~~~J~~with pure_ ,+11. V eextensoractivity inhalluxpatterns.Recordingsfrom flexorhallucisbrevis (top)and extensor

\1

hallucis

l#I|i1 tracings).Time 10 msec.

G .~~~~~~ ~~~~~ ~~~~~~-.> -

extensor reflex, were observed (Fig. 1). Almost pureflexor responses were obtained in a remarkablylarge group (nine cases), even including subjects whoat the clinical examination exhibited brisk Babinskisigns. Pure extensor responses were obtained in asmall group of patients (four cases) with severelesions of the spinal cord but were not a characteristicfeature in this type of lesion. In the remaining 15

315

Lennart Grimby

cases in the Babinski group the reflex patternsconsisted of various combinations of flexor andextensor discharges, as a rule alternating, and closelyresembling those obtained on stimulation near thetransition zone in normal subjects. The initial andthe later parts of the pattern often differed somewhat,and in some cases two different components couldbe distinguished, viz., either an initial part withdominant flexor activity and shorter latency than100 msec., and a distinctly separate later part withdominant extensor activity and about twice as longlatency (four cases, cf. Fig. 3A), or an initial partwith dominant extensor and a later part withdominant flexor activity (two cases, cf. Fig. 3B).

Non-Babinski group In the non-Babinski groupthe extensor activity in the planta pattern was oftenstronger than in a typical normal case. In eight ofthe cases it was, however, not stronger than innormal individuals with a proximal transition zone,but in the remaining four cases the plantar patterndiffered significantly from normal and consistedof an initial part of dominant extensor and a laterpart of dominant flexor activity. As was to beexpected, no plantar patterns with long-lastingdominant extensor activity, viz., patterns resulting

in protracted dorsiflexion of the great toe, wereobserved in this group.No opportunity was provided to observe the

transformation of normal plantar patterns intopathological patterns, but four patients could befollowed up while their plantar responses changedfrom extensor to flexor; in the course of theirrecovery the receptive field of the extensor reflexeswas successively reduced in the direction of thehallux ball and the extensor activity in the plantarpattern gradually subsided and in two cases finallydisappeared. In three of these cases the developmentof the plantar pattern during recovery can best beillustrated by Fig. 1 as read inversely, i.e., frombottom to top. In the fourth case (Fig. 3A), thepathological plantar pattern consisted of an initialcomponent with dominant flexor activity and alater part with dominant extensor activity; in thecourse of the recovery of the patient this later partdisappeared completely while the initial partremained unchanged.

STIMULATION OF LATERAL SIDE OF THE PLANTA Innormal individuals the same reflex pattern is generallyobtained on stimulation of the lateral as of the

JV

FIG. 2. Changes in reflex patterns on shift of stimulus from middle to lateral side of planta. Stimuli applied tohallux ball (left-hand column), middle of planta (middle column), and lateral side of planta (right-hand column).Recordings from flexor hallucis brevis (top) and extensor hallucis brevis (lower tracings). Time 10 msec.Case A On lateral planta stimulation, pattern predominantly extensor, thus resembling hallux pattern; on medialplanta stimulation, pattern predominantly flexor.Case B On lateralplanta stimulation, patternpredominantlyflexor; on medialplanta stimulation, patternpredominantlyextensor, thus resembling hallux pattern.

316

...!

.. .19;;4- I.. ..:!:.A .4:

"P

'.20-wA. 4 i-O!

Pathological plantar response

middle or medial side of the planta. If the stimulusis applied to the anterior parts of the sole there is,however, often a greater tendency to extensoractivity on medial than on lateral stimulation. Inthe pathological material. stimulation of the lateralside resulted in the same range of variations aswere observed on stimulation of the middle of theplanta, viz., from pure flexor to pure extensorresponses. In most cases no definite changes wereinduced by shifting the stimulus from the medialto the lateral side of the planta. In the Babinskigroup, however, five cases exhibited a markedincrease of extensor activity (Fig. 2A) which agreeswith the well-known clinical observation that theBabinski sign is more readily evoked from thelateral than from the medial side of the planta. Twocases were anyhow observed where extensor activitywas dominant on medial and flexor activity onlateral stimulation (Fig. 2B); although contrary toclinical experience, these findings are less surprisingin view of the fact that extensor activity is normallymore readily elicited from the medial than from thelateral side of the planta.

STIMULATION OF THE HALLUX BALL Normally,short hallux extensor activity is strongly dominantin the hallux pattern; pronounced flexor activityoccurs only in individuals with a very distal transitionzone. Also in most pathological cases extensoractivity was by far dominant in the reflex patternsinduced by hallux stimulation, but signs of flexoractivity often appeared which, although not strongerthan is normal in subjects with distal transitionzones, were anyhow remarkable insofar as theyoccurred also in subjects with a considerably reducedflexor activity in the plantar pattern. In some casesthe flexor activity was slightly more prominent inthe early part of the hallux pattern and in other casesin its later part. In three cases, all in the Babinskigroup, weak hallux stimulation regularly resultedin patterns throughout dominated by flexor activityand even in a distinct plantar flexion of the greattoe, a reaction never observed in any of the normalsubjects examined.

CONTRAST BETWEEN PLANTA AND HALLUX PATTERNSIn normal individuals, flexor activity is stronglydominant in the planta pattern and extensor activitystrongly dominant in the hallux pattern, and thecontrast between the two patterns is almost maximal,except in subjects with extremely broad transitionzones. In most pathological cases, both in theBabinski and the non-Babinski group, this contrastwas reduced or even abolished.

In cases where the reflex pattern consisted of twodistinctly separate components, the contrast between

-~~~~~Ir ;)I

1*

1 '''''''''' ' > *~~~~

IILJL-.I11

a R

I F ..4 Ipa- .li.-kw 4;-I "' ''.

m

AB.

4mf

W%VWAWV0AVV

FIG. 3.Initial and latercomponents ofreflex patterns

j! Wt Iexhibitingsignificantdifferences.

ji; Upper recordsin A, B, and C,hallux stimu-lation; lowerrecords

j stimulationa applied at

middle ofplanta.Recordingsfrom flexorhallucisbrevis (top)and extensor

t hallucis brevis(lowertracings).Time 10 msec.

Case ADifferentiation

n..rjaus_ of initial reflexcomponenttypical of

*9.. stimulus site,latercomponentstereotypedextensoractivity.

Case BInitial reflexcomponentstereotypedextensoractivity, latercomponentnormaldifferentiation.Case CBoth reflexcomponentspure extensor,independentlyofstimulus site.

C:.

the planta and hallux patterns might be normal inthe initial part of the pattern but abolished in itslater part (Fig. 3A); conversely, the first part mightconsist of stereotyped extensor activity whereas thelater component had a normal differentiation(Fig. 3B). In some cases, finally, a shift of thestimulus site did not result in any change, either in

L~~~~~~~~~~~~~~~~~~~--

b2.iL,_

317

a

. ......-

I .

I

..T

I

Lennart Grimby

FIG. 4. Varying amounts of contrastbetween hallux and planta patternsdespite similar planta patterns.Stimuli applied to hallux ball (left-handcolumn) and middle ofplanta(right-hand column).Recordings from flexor hallucis brevis(top) and extensor hallucis brevis(lower tracings). Time 10 msec.

:

-"N>

the early or in the later part of the reflex pattern(Fig. 3C).The contrast between planta and hallux patterns

may vary considerably, i.e., be more or less pro-

nounced, in cases with rather similar planta patterns(Fig. 4). Besides, there may be a marked contrastbetween the two patterns although extensor activityis strongly dominant in both; or the two patternsmay be similar although there is no general extensorpredominance. There is thus no direct relationshipbetween the amount of contrast between the tworeflex patterns and the increased level of extensoractivity in the planta pattern.

SIGNIFICANCE OF STIMULUS STRENGTH With increas-ing stimulus intensity the strength of the short hallux

extensor reflex normally increases at a somewhatfaster rate than the flexor reflex. On an average, thisdifferentiation was more pronounced in the patho-logical material, especially in the Babinski group,

and in one case in this group an increase of thestimulus strength caused not only a relative but an

absolute weakening of the flexor reflex (Fig. 5A).In this case the reflex patterns obtained were thesame independently of the stimulus site, but flexoractivity was dominant on weak. and extensoractivity on strong, stimulation; while the normalpattern is almost completely determined by thestimulus site, the stimulus strength was altogetherdecisive in this case.Two cases in the Babinski group formed an

exception to the general tendency to increased

3 1%/ 7 V.

318

......

..:.%:..., -

1. -4'---

Pathological plantar response

.\

B

}3A

k _ *

U~~~~~~~~~~~l-AY:

fK .vo

1

r

!3t$ :.

,{ .: s o ;*:: A(W$.

r

f.h / < .s' hp}O

/!i,?J C. A

', b

i * #

extensor predominance with increasing stimulusstrength; in these cases very weak planta stimulationresulted in pure extensor activity, although strongerstimulation provoked dominant flexor activity(Fig. 5B); in one of the cases, however, extensoractivity was again dominant on application of verystrong stimuli.

DISCUSSION

This investigation of the integration of the shorthallux flexor and extensor reflexes in pathologicalcases is a direct sequel to the previous correspondingstudy performed on healthy individuals (Grimby,1963), and the results should thus be correlated tothose obtained in normal cases.Even in a normal material, ruling out all cases of

clinically atypical reflexes where an injured reflexmechanism may be suspected, significant deviationsfrom the average normal reflex pattern can beobserved, and it was presumed that an unselectedpopulation of apparently healthy individuals wouldinclude single cases exhibiting still more pronounceddeviations than those actually found in the recentinvestigation. Since there is, however, no safe methodof judging whether the reflex mechanism is actuallyintact in such cases, no exact limits can be drawnfor the normal variations. It would anyhow appearto be justified to assume that very marked deviationsfrom normal observed in a high proportion of thepathological cases are, in fact, caused by the nervousdisorders present, and the following discussion willbe based on this assumption.

Since the extensor activity in the reflex pattern

FIG. 5. Influence of stimulusstrength on reflex pattern.Left-hand column weak, right-hand column strong stimulation.Recordingsfrom flexor hallucisbrevis (top) and extensorhallucis brevis (lower tracings).Time 10 msec.

Case A Pattern predominantlyflexor on weak, extensor onstrong, stimulation (records fromanterior part ofplanta but resultsindependent of stimulus site).

Case B Pattern predominantlyextensor on weak, flexor onstrong, stimulation (stimulusapplied to lateral side ofplanta).

obtained on stimulation of the planta in subjectswith pathological reflexes gradually subsides andsometimes disappears as the patients recover, adirect relationship must be presumed to exist betweenthe varying degrees of extensor activity and thedegree of pathological change in the reflexmechanism. The extensor activity sometimes presentin the planta pattern in all normal cases is onlyoccasional and disappears on repeated stimulation;the intact reflex mechanism may thus, in contrast tothat injured, be able to correct deviations from thenormal reflex pattern. As, however, a relatively weakbut constant extensor activity may sometimes occurin the plantar pattern also in single normal cases,this type of reflex pattern must be considered to beeither a normal variant or the effect of an injuredreflex mechanism. No exact margin can thus bedrawn between normal and pathological plantarpatterns.When a pathological plantar pattern is successively

converted into a normal reflex type, the extent ofthe receptive field of the extensor reflex graduallyshrinks in the direction of the hallux ball, and itmust be presumed that, inversely, the receptive fieldsuccessively spreads to a larger area as a normalpattern is converted into a pathological; this conceptwould lend further support to the view advanced byKugelberg et al. (1960) that the extensor reflexobtained on hallux stimulation in normal cases isfundamentally the same as that elicited on plantastimulation in pathological cases. Since in somecases pronounced extensor activity can be elicitedsolely by lateral, and in other cases solely by medial,planta stimulation, there must be individual

319

Lennart Grimby

differences in the extension of the receptive fieldof' the extensor reflex, and these may be accountedfor either by constitutional individual differences orby differences in the nature of the nervous lesionsinvolved.Two distinctly separate components may be

distinguished in the reflex pattern; the two responsesmay differ in habituation tendency (cf. Grimby,1963) and in composition (cf. Fig. 3) and areprobably transmitted in somewhat different ways.The latency of the first response in the reflex patternis so short that it must be of spinal origin andmediated by A-fibres. Kugelberg (1948) has shownthat the pathological plantar response may consistof an early A- and a later C-fibre response, but thesecond response in the reflex pattern discussed inthis work has such a short latency that it cannot bemediated by C-fibres. The longer latency of thesecond response is more likely to be due to a longercentral reflex time than to slower afferent impulses,but whether this response is of spinal or of cerebralorigin cannot be decided judging by the latencyvalues; nor can it be ruled out that it is of a differentorigin in different cases. That the deviation fromnormal observed in pathological cases may belimited either to the first or to the second responseindicates that a lesion may affect one response,leaving the other unaffected, and this differentiationis probably due to the difference in transmission,whatever this difference may be.

In cases where a Babinski response is obtained onclinical stimulation, the plantar pattern may oftenbe normal; conversely, markedly deviating patternsmay be observed in pathological cases with clinicallynormal reflexes. It would seem as though the briefstimulus used in the experiments favours the initialpurely spinal response, whereas the later phenomenaare more apparent in the long-lasting stimulationused in the clinical examination. Besides, thestimulus strength may be of fundamental significancefor the composition of the plantar pattern (Fig. 5),and the tendency to dominant extensor activityseems to be least pronounced on stimuli of mediumstrength and to increase both with stronger (Fig. 5A)and weaker (Fig. 5B) stimulus intensity. This is ingood agreement with the finding that in normal casesmarked extensor activity in the plantar pattern ismore common on very strong or very weakstimulation than when stimuli of medium strengthare set in.When the reflex mechanism is intact, the reflex

pattern is closely correlated to the site of thestimulus so that the resulting response alwaysrepresents the most appropriate defence reaction;when it is injured this ability may be reduced orabolished, and the reflex pattern is then less adaptive

to the stimulus site. In some pathological cases thereduction of the contrast between hallux and plantapatterns is more conspicuous than the generalincrease of the extensor activity in the reflex system;in other cases, however, the latter phenomenon isprevalent. This discrepancy may be too large to beexplained by constitutional individual differences(Grimby, 1963) and must be due to different typesof lesions of the reflex mechanism in the differentcases. On the basis of the results obtained in therecent study on normal subjects it was assumed thatthe control of the spinal reflex centre is exerted byone pathway determining the general relation betweenextensor and flexor activity and another determiningthe amount of contrast between the planta and halluxpatterns; it seems reasonable to assume that thesepathways may, independently of one another, bemore or less affected by the pathological process.That hallux patterns with dominant flexor activitymay be observed in single pathological cases, evenwhere clinical Babinski signs are present, may thenbe explicable on the ground that the contrastbetween hallux and planta patterns has beenabolished without any simultaneous increase of theextensor activity being apparent.The individual deviations from the normal reflex

patterns observed in pathological conditions may beso manifold and varying that they must be presumedto be influenced not only by the degree of the lesionbut also by the type and site of the injury present.However, a positive approach to these problemsrequires studies of a substantial number of caseswith well-defined injuries of the nervous system,and in this connexion the method described abovemay be a useful tool, as permitting a more exactand detailed reflex analysis than that used in theclinical routine. It should be emphasized, however,that a pathological reflex, although obvious to theclinician in the course of the ordinary tests, may notappear as such when studied by the new technique;on the other hand, this method may in other casesdisclose deviations from normal not apparent at theclinical examination, and may thus be of somepractical value as a complement of the clinicaltests.

SUMMARY

The cutaneous reflexes of the foot were investigatedin a group of 40 patients with spastic hemi- orparapareses, 28 of whom exhibited clinical Babinskisigns. Painful stimuli consisting of a series ofrepetitive electric shocks delivered over a period of20 msec. were applied to various points on theplantar surface of the foot. The 'reflex patterns'obtained on simultaneous electromyographic record-

320

Pathological plantar response

ing in the short hallux flexor and extensor werecompared with those observed in a group of normalcases previously studied.

In the pathological material the reflexes obtainedon stimulation of the planta may exhibit a widerange of transitional forms from the typical normalpure short hallux flexor reflex to the extremepathological pure short hallux extensor reflex; adirect relationship is presumed to exist between thedegree of extensor activity in the plantar pattern andthe degree of pathological change in the reflexmechanism. However, no exact margin can be drawnbetween normal and pathological plantar patterns,since a relatively weak but constant extensor activitymay sometimes occur in the plantar pattern also insingle normal cases.The pathological reflex pattern may consist of two

distinctly separate components. The extensor activityin the plantar pattern, viz., the deviation from normal,is in some cases limited to the early and in othercases to the later component. The early response isof such a short latency that it must be a purely spinalreflex. The origin of the later response cannot bedetermined judging by its latency but it is assumedto be transmitted in such a way that it may be

affected by the pathological process independentlyof the early response.

In pathological cases, the normal contrast betweendominant flexor activity in the plantar pattern anddominant extensor activity in the hallux pattern isreduced, although in some cases more, and in othercases less, than was to be expected from the amountof extensor activity in the plantar pattern; it has beenassumed that different pathways may be involved inthe suprasegmental reflex control and that thesepathways may, independently of one another, bemore or less affected by the pathological process.The individual variations found in the pathological

material are too manifold to be ascribed solely tolesions of varying degrees of severity; it has beenassumed that different types of lesion of the reflexmechanism may result in different types of deviationfrom the normal reflex pattern.

REFERENCES

Grimby, L. (1963). J. Neurol. Neurosurg. Psychiat., 26, 39.Holmqvist, B., and Lundberg, A. (1961). Acta physiol. scand 54,

suppi., 186.Kugelberg, E. (1948). Brain, 71, 304.

, Eklund, K., and Grimby, L. (1960). Brain, 83, 394.Landau, W. M., and Clare, M. H. (1959). Ibid., 82, 321.

321