pins and needles observations on some of sensations

"' PINS AND NEEDLES ": OBSERVATIONS ON SOME OF THESENSATIONS AROUSED IN A LIMB BY THE APPLICATION OF

PRESSURE

BY

G. WEDDELL and D. C. SINCLAIRFrom the Departments ofAnatomy and Orthopaedic Surgery, University of Oxford

(RECEIVED 4TH OCTOBER, 1946)

IntroductionWHEN the limb of a normal subject is compressed bymeans of a sphygnomanometer cuff inflated to a

pressure beyond the systolic blood pressure, a seriesof spontaneous sensations can be produced in it,both during and after the release of pressure,provided the compression lasts long enough.Investigators in the past appear to have paid littleattention to such sensations, although some of these,under the generic term of " pins and needles ", are a

common feature of everyday life. In 1931, Lewis,Pickering, and Rothschild published an account ofthe sensory changes occurring during and aftercompressibn of the arms of normal subjects bysphygmomanometer cuffs. At the end of thiscomprehensive paper is given a description of a" form of tingling", occurring after release of com-pression, which was investigated by the authors,together with some general conclusions as to itsnature and the mechanism of its production.

It appeared possible that an investigation of thesensations experienced as a result of compressingthe arm might throw some light on the mechanismby which spontaneous pain originates in the humanbody, and accordingly it was determined to repeatand amplify the observations of Lewis and hiscollaborators with this end in view. The presentpaper gives an account of some investigations on

the subjective sensory phenomena associated withcompression of the upper limb by means of a

sphygmomanometer cuff.

TechniqueOrdinary sphygmomanometer cuffs, 12 cm. in width,

were used, and inflated to a pressure of 150 mm. Hg forvarying times in varying situations on the upper limb.In order to determine the effect of the degree of com-

pression on the results, a few experiments were done with

pressures ranging from 60 mm. Hg to 300 mm. Hg, but150 mm. Hg was selected as standard since it was wellabove the systolic pressure in all our subjects, and didnot cause undue discomfort. The pressure was throwninto the cuff by pumping up the instrument in the usualway, but as rapidly as possible. The time taken toreach 150 mm. Hg pressure by this means was usuallyless than 3 seconds. At the end of compression therubber tube was disconnected, and the pressure fell tozero in approximately 1 second. The armlet was thenrapidly removed. The experiments were, in general,done at room temperature, which throughout the periodof the investigation varied between 170 C. and 220 C.A few observations were also made on one subject withthe limb immersed in a thermostatically controlled water-bath at 370 C. In the experiments at room temperature,care was taken, so far as possible, to exclude draughts.

It was thought that, if any of the sensations producedwere wholly or partly due to events taking place, in theperiphery of the limb, localized pressule on any area ofskin in the periphery might modify the results. Accord-ingly, throughout each compression the arm was allowedto hang down vertically so that no part of the limb com-pressed made contact with other objects. No specialattention was paid to the condition of the limb beforecompression, except that the pressure was always appliedwith the arm at the side and the elbow bent to a rightangle, to avoid any undesirable congestion of the limb.Immediately after release of the pressure the elbow wasagain bent to a right angle to facilitate observations.Timing was done with a stop-watch, which could-not beread by the subject of the experiment. The main bulkof the work was done by the authors, using each otheras experimental subjects, but every experiment reportedin this paper has been confirmed on at least one othersubject. The total number of subjects examined is 16;of these 7 were normal in all respects, and 9 had variouspathological changes involving one upper limb.

In the case of all subjects other than the authors, noindication was given by the experimenter as to whatsensations might be expected during or after compression;they were merely told to observe the compressed limb

26

Protected by copyright.

on June 7, 2022 by guest.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.10.1.26 on 1 F

ebruary 1947. Dow

nloaded from

http://jnnp.bmj.com/

PINS AND NEEDLESclosely and to report any sensations they might feel assoon as these arose. As soon as any sensation wasreported, the time was noted, and the subjects were thenquestioned as to the exact nature, intensity, and distribu-tion of the sensation, records being made of the subjectsown words. They were then warned that the sensationmight or might not disappear, and were asked to give theend-point as accuratelv as possible if it did disappear.This procedure, which was unavoidable if the effects ofsuggestion were to be eliminated, caused one or two ofthe untrained observers, who had no idea what to expect,to fail to report a given sensation until it was fullyestablished, instead of reporting it as soon as it becamejust perceptible. In such cases the experiment wasrepeated after a suitable interval. This technique differsin some respects from that adopted by Lewis et al. (1931).Though these authors do not explicitly state the positionof the limb during experimentation, it can be inferredthat it lay horizontally in a water-bath. This bath waskept at a constant temperature of 35" C. throughout. Itwill be seen later that in spite of these differencescomparable results were obtained.

Sensations ArousedBefore discussing the results of individual experi-

ments, it is desirable for clarity to give an accountof the subjective sensations produced during andafter compression of the arm just above the elbowby a sphygmomanometer cuff at 150 mm. Hg. Thissummary represents the general experience of theauthors, and is based on consideration of the notesof all the subjects so far tested. In any compressionat this site in a normal limb, providing (a) thecompression is of sufficient duration, and (b) nocompression has been done on the arm for at least2 hours previously, three types of sensation can beelicited almost invariably.

1.-A characteristic tingling (" compression tingling ")in the hand, coming on between I and 2 minutesafter compression has started, and lasting anaverage of 3 to 4 minutes. This tingling has beenvariously described as " a faint comfortable soda-water sensation ", " a buzzing ", " a fine lighttingle", " ants running up and down inside theskin", etc. In some cases a few sharp prickingelements have been reported, but in the main thesensation appears subjectively to be composed of" bright " touch impulses, and the general effectproduced is one of touch rather than pain. Thesensation begins gradually, builds up toamaximum,and gradually fades away. External stimuli, suchas touching or rubbing the fingers, do not modify"compression tingling " in any way. There isusually very little difficulty in giving an accuratetime for the onset of the tingling, but its disap-pearance is more difficult to fix precisely, and thistime has in all cases been taken to the nearestquarter of a minute.

2.-A " velvety" sensation (" velvety numbness ")arising much later, rarely before 13 minutes andusually after 15 minutes from the initiation ofcompression. This sensation is usually notspontaneous at first, but can be elicited by lighttouch, stroking, or pressure. It appears first in thetips of the fingers and spreads proximally up thehand fairly rapidly. It gives the impression thatthe finger is touching the pile of a very fine velvet,and, though it has no apparent pain component,it gives rise to a peculiarly unpleasant subjectiveinterpretation of a sickening character. Afterprolonged compressions (over 16 minutes), and incases where compressions are repeated at shortintervals, this sensation may become spontaneous.No compressions in this series have been of longerduration than 20 minutes, and in all cases where" velvety numbness " has been aroused it haspersisted until the release of compression or sometime later-on occasions for as long as 45 secondsafter release. It is possible, however, that ifcompression were maintained for longer periodsthe sensation would subside before release.

3.-A very strong coarse pricking (" release pricking ")which arises in the hand after the compression isreleased and lasts for a variable time depending on.the length of compression. This is completelycharacteristic, and differs from the " compressiontingling" in being much coarser, having a com-ponent of large numbers of sharp bright pricks,as of a needle entering the skin, and in beinginterpreted subjectively as pain rather than touch.The invariable description of the sensation givenby untrained subjects was " pins and needles ". Onbeing pressed for a closer description, they gave oneof two replies: "like a lot of sharp needles goingvertically through the skin " or " like a lot of smallelectric shocks ". No subject had the slightestdifficulty in distinguishing subjectively between theinitial phase of the " release pricking " and the" compression tingling ". The sensation wasdescribed as " painful" in all cases, but only 3 ofthe 16 subjects gave an affirmative reply to thequestion " Would you like to get away from it ifyoucould ? " The others, including the authors, con-sidered that while the sensation was undoubtedlypainful, there was no unpleasant affective com-ponent, and the total emotional attitude to thesensation was one of interest rather than pain.There is, however, a considerable underlyingtouch component in the release pricking, whichis evident from the first as a continuous diffusetingling sensation in between the pricks. Ini-tially this is not readily perceived, owing to theinsistence with which the pricks engage the attention,but after some time, depending on the duration ofthe compression, the pricks " die down", leavinga general tingling sensation difficult to distinguishfrom compression tingling. This in its turn fadesvery gradually indeed, so that the exact end-pointwhen sensation in the hand again becomes normalis difficult to determine. In spite of this difficulty,

27



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


G. WEDDELL AND D. C. SINCLAIRhowever, repeated experiments indicate that,though the interpretation ofthe end-point may varyfrom subject to subject, the time given is remarkablyconstant in the same individual. This time hasin all cases been taken to the nearest quarterof a minute. The onset of the pricking is in allcases definite and sharp, and can be given with anaccuracy of a few seconds after compressions of10 minutes or longer. It is found that, previousto the onset of spontaneous pricking, -the samesensation can be induced by lightly tapping thefingers against a solid object. Similarly, aftercessation of the spontaneous tingling which followsthe disappearance of the pricking, tapping willusually revive the tingling for a little longer. Thetimes given for the duration of release pricking inthis paper include the short periods, both beforeand after the time of spontaneous sensation,during which the sensation can be produced byexternal stimuli such as tapping or rubbing thefingers. During the period of spontaneouspricking, such stimulation temporarily greatlyincreases the intensity of the sensation, a showerof pricking impulses being produced locally and,though to a lesser extent, throughout the dis-tribution in which the release pricking is felt.Once established as a spontaneous sensation, therelease pricking mounts rapidly to a plateau ofintensity, then gradually fades, as previottslydescribed.

In addition to these three clearly defined sensations,several others may be aroused in the limb during com-pression. These do not invariably occur, and, whenthey do, there is considerable variation in their time ofappearance and duration. All subjects reported anaching discomfort under the pressure cuff, which gradu-ally, became less marked as the experiment proceeded.In some cases this ache spread to include the elbow anda small area of the forearm just distal to the cuff. Inseveral subjects the hand felt " tight " and " swollen "during compression, and two reported a well markedaching pain " deep in the wrist". Another two subjectsconsistently felt a diffuse vague ache in the area in whichcompression tingling was felt. This ache came ongradually during the tingling, and eventually replaced it:in both cases the ache persisted until release of compres-sion. Vague transient feelings of cold and warmth oftenoccur during compression; they are usually diffuse andfelt chiefly on the back of the hand. One of us, however(G. W.), has experienced definite punctate flashes ofcold on the back of the hand. These waves of tempera-ture sensation are to be distinguished from the generalizedcold sensation which comes on in a prolonged compres-sion, and is accompanied by a considerable fall in skintemperature.On release of pressure after moderately long compres-

sions a fairly definite sequence of temperature sensationsis felt almost immediately. The first of these is a waveof warmth which appears to " hit" the hand alrnost assoon as compression is released. This is succeededabruptly by a strong sensation of cold, which then

gradually "swells" into a warm glowing comfortablefeeling. The whole sequence takes place in about 15seconds, and precedes the onset of release pricking. Itis interesting that flushing of the hand does not occuruntil about 15 seconds after release, which is aboutthe time of onset of the warm glowing feeling. Of theserelease 'temperature sensations, only the warm glow isinvariable, but the above sequence has been reported byseveral subjects. The sensations of warmth and coldlast a few seconds only, and may readily be overlookedby an untrained subject.The investigations reported in this paper have been

carried out with the object of elucidating the natureand mechanism of production of compressiontingling and release pricking, these sensations beingsubjectively characteristic, reproducible in distribu-tion, quality, duration, and intensity in the samesubject, and reproducible to a considerable degrep indistribution, quality, and duration from subject tosubject.

Reproducibility of SensationsINDIVIDUAL VARIATION.-NO figures can, of

course, be quoted to show the reproducibility of thecompression tingling and release pricking in qualityand intensity. It may be said, however, that in anygiven, subject these are practically constant for anygiven length of compression. Some figures may beadduced to indicate the degree of reproducibility ofthe sensations in time. The table gives the meantimes of appearance and the mean durations ofcompression tingling and release pricking in 4subjects as a result of 10-minute compressions of thearm. In subject 1 the figures are the mean of 8 suchcompressions; in the remainder, the mean of 4experiments. In all cases no compressions had beendone on the arm for more than 2 hours previously.As stated above, the times of disappearance of

both compression tingling and release pricking havebeen taken only to the nearest quarter of a minute.Accordingly, these times and standard deviations inthe table are given only to the nearest 5 seconds.The figures show that of the two sensations, releasepricking is the more exactly reproducible in time,both in the same subject and from subject to subject.There are, nevertheless, considerable differences inthe duration of both sensatiQns from subject tosubject, and the relatively high standard deviationsfor the duration of compression tingling demon-strate the variability of the duration of this sensationin the same subject. For these reasons it wasnecessary, in any experiment dealing with the timingof these sensations, to control the observations bydirect reference to the other arm ofthe same subject.In most subjects it was possible to take several suchcontrol readings to establish normal mean values, but

28P

rotected by copyright. on June 7, 2022 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.10.1.26 on 1 February 1947. D

ownloaded from


TABLE.-MEAN TimEs OF APPEARANCE AND MEAN DURATIONS OF COMPRESSION TINGLING AND RELEASE PRICKINGIN 4 SUBJECTS AS A RESULT OF 10-MINUTE COMPRESSIONS OF THE ARM

Compression tingling Release pricking

Mean time Mean Mean time MeanSubject of onset duration of onset duration

(after compression) (after release)

1. D. S... .. I.1' 42"±20t 3' 30"+55" 50"I4" 5' 45"4±15"2. G. W. .. .. .. 1' 04"± 18" 4' 55"+ 15" 50"± 1" 4' 55"±35"3. R. M. W. .. .. .. 1' 18"±9" 5' 45"±45" 47"15" 4' 15"±20"4. L. D. B. .. . . 1 13"'±3" 3' 40,"±20" 60" 4 3" 4' 10"'±15"

in some of the subjects with pathological conditionsonly one control observation could be made.EFFECT OF ENVIRONMENTAL TEMPERATURE.-In

one subject (D. S.) inspection of the records of 85compressions experiments in which the room tem-perature was carefully recorded failed to show thatvariation in the air temperature between 17° C. and220 C. had any effect on either sensation. Several16-minute compressions ofthe arm were done on thissubject, the limb being placed in a thermostaticallycontrolled water-bath at 370 C. for 2 minutes beforecompression, and maintained vertically in the bathduring compression. Comparison of these experi-ments with other 16-niinute compressions done atthe same level on the same subject at room tem-perature failed to show any difference in the intensity,Iistribution, quality, and duration of the sensations.

As a check on this observation, the left arm was

placed in a water bath at 370 C., the right arm beingat room temperature (180 C.). Both arms were thencompressed simultaneously. No difference in thesensations experienced in the two limbs could bedetected. Repetition of this experiment on twoother normal subjects gave precisely similar results.No other environmental factors were investigated.For the remainder of the paper compression

tingling and release pricking will be consideredseparately, as it will be seen later that they are

evidently due to different mechanisms.

Observations on Compression Tingling

EFFECT OF SITE OF COMPRESSION.-It has beenfound that, for typical compression tingling to beproduced, the cuff must be above the elbow. Infour subjects so far tested to establish this point,while compression above the elbow invariablyproduced the sensation, compression at any leveldistal to the elbow almost invariably failed to produceit. One subject, however (G. W.), has occasionallyfelt a very feeble tingling as a result of a compressionof the forearm just below the elbow.

EFFECT OF DURATION OF COMPRESSION.-As pre-viously stated, compression tingling begins between1 and 2 minutes after compression is instituted, andlasts for 3 to 4 minutes before spontaneously fading.If the pressure is released while tingling is beingexperienced, the sensation completely disappearsin a matter of a few seconds, irrespective of the stageof tingling at which release occurs.

DISTRIBUTION OF SENSATION.-Compression ting-ling is felt mainly in the hand: in none of the 16subjects tested did it extend more than 2 inches.abovethe wrist joint. It is invariably felt more strongly onthe palmar surface than on the dorsal surface, butthere is no evidence of any gradation of the sensa-tion in a proximo-distal direction. In this respectcompression tingling differs considerably fromrelease pricking, as will be seen later. In 11 of the16 subjects the tingling was felt diffusely throughoutthe hand, and involved the whole of the palmar sur-face of the hand and fingers, sometimes extending alittle above the ulnar side of the wrist. In 3 of these11 subjects tingling invariably began in the distribu-tion of the ulnar nerve, and gradually spread toinvolve the rest of the hand. It should be notedthat in no subject examined did tingling occur in thearea supplied ty the radial nerve. Two of theremaining 5 subjects experienced tingling confinedto the ulnar distribution and the terminal phalanxof the thumb, while in the other 3 tingling wasrigidly restricted to the area supplied by the ulnarnerve, which could, indeed, be readily mapped outin consequence.

EFFECT OF MODIFICATION OF LOCAL CONDITIONS.As already stated, compression of the arm above theelbow by a cuff pressure of 150 mm. Hg invariablyproduces compression tingling. It is found, how-ever, that compression tingling, differing from the" normal " sensation only in being slightly reducedin intensity, can be elicited by a cuff pressure of 100mm. Hg. On the other hand, a pressure of 60 mm.Hg fails to arouse it. Pressures greater than 150mm. Hg, up to a maximum of 300 mrnm. Hg, are

29PINS AND NEEDLESP


http://jnnp.bmj.com

/J N

eurol Neurosurg P


ownloaded from


G. WEDDELL AND\D. C. SINCLAIR

found to give rise to compression tingling in allrespects identical with that produced by a pressureof 150 mm. Hg. It may, therefore, be said that thepressure " threshold " necessary to produce com-pression tingling lies somewhere between 60 and100 mm. Hg, and that the full response occurs at apressure of 150 mm. Hg, no further increase inpressure beyond this level having any effect on thesensation.The " normal " sensation described above is

obtained by compressing a cylinder of the arm12 cm. long. If two overlapping cuffs are placed onthe arm above the elbow and inflated together so asto compress a cylinder of the arm approximately20 cm. long, the resultant compression tingling isindisiinguishable in all respects from the normalresponse. Similarly, if the length of the limb com-pressed is reduced to 1 cm. by employing an ordinaryroller bandage wound round the arm at a tensionjust sufficient to obliterate the radial pulse, a " nor-mal " result is again obtained.EFFECT OF MODIFICATION OF PERIPHERAL CON-

DITIONS.-If the circulation to the hand and fore-arm is impeded by a cuff just below the elbowmaintained at 150 mm. Hg for 16 minutes, notingling is produced during compression. If nowanother cuff be placed just above the elbow andinflated to 150 mm. Hg 10 minutes after the startof the distal compression, without releasing thedistal cuff, a normal compression tingling resultsat the usual time after applying the upper cuff, andlasts for the normal period.EFFECT OF REPEATED COMPRESSIONS.-If a

regular series of compressions is made on the samearm at intervals of less than 2 hours, the onset ofcompression tingling tends to become more andmore delayed in each successive compression. Atthe same time the intensity of the sensation becomesprogressively less and less, till, after several com-pressions, it can no longer be elicited. The shorterthe interval between compressions, the more rapidlydoes this diminution and eventual disappearanceof compression tingling occur. For example, in aseries of 10-minute compressions, with an intervalof 10 minutes between each compression (cycleinterval, 20 minutes), tingling can usually be elicitedin the first two or three compressions, but by thefourth or fifth compression the sensation producedis no longer a tingling, but merely an abnormal"s awareness" of the area in which the tingling isusually felt. Later compressions evoke no sensationwhatever. If, on the other hand, 80 minutes areallowed to elapse between individual 10-minutecompressions (cycle interval, 90 inutes), an ade-quate reproduction of the original tingling may be

obtained for four or five compressions, but thereafterthe sensation " fades" in the same manner, thoughmuch more gradually. Because of this phenomenon,in the investigation of compression tingin com-pressions were not repeated on the same armunless an interval of 2 hours had elapsed since thelast compression.

SITE OF ORIGIN OF THE IMPULSES.-It is clear thatthe impulses of compression tingling must arise inthe nervous tissues of the limb peripheral to theupper margin of the stretch of limb directly com-pressed. Theoretically, therefore, they could arisein several sites: (a) at the peripheral nerve endings;(b) in the main nerves peripheral to the compressioncuff; (c) in the nerves directly pressed on by thecuff; (d) in a combination of the above sites.The first of these possibilities need only be con-

sidered briefly. It has been noted that, in three ofthe subjects tested, compression tingling is restrictedto the area supplied by the ulnar nerve. It isdifficult to conceive of any factor involving theperipheral nerve endings which would selectivelyaffect all those in the ulnar area without also affectingthose in the median and radial areas. Directevidence that the peripheral nerve endings are not thesite of origin of compression tingling was obtainedby compressing the arm of an individual (subject 8)whose forearm had been amputated 6 inches belowthe level ofthe olecranon eighteen months previously.Though this man had not felt his " hand " for 9months, within 30 seconds of compression a com-plete phantom hand became evident, and the fingersbegan to " move ". Two minutes after compressionbegan, tingling, indistinguishable from that felt inhis control arm, appeared in the whole of the palmaraspect of the missing hand, and lasted for. 74minutes. In the control arm tingling began at 24minutes and lasted for 54 minutes. Similar resultswere obtained in another case of amputationthrough the forearm (subject 9). The findings inthese two subjects, while they exclude the peripheralnerve endings as the site of origin of the impulses,still leave the possibility that compression tinglingarises either in the nerves under the cuff, or in thenerves distal to the cuff, or in both sites.

If the tingling arises in the nerves peripheral to thecuff, then alteration of the condition of those nervesby compression would be expected to influence theresults. It has been seen, however, that, even afterten minutes' compression of the limb below theelbow by a cuff at 150 mm. Hg, a normal compressiontingling may be obtained by applying a second cuffimmediately above the first. Further evidenceagainst the site of origin of compression tinglingbeing located distal to the compression cuff was

30



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


PINS AND NEEDLFS

obtained by carrying out a series of 10-minutecompressions on the same limb, the arm just abovethe elbow and the forearm just below the elbow beingcompressed alternately. The cycle interval was25 minutes, and thus the cycle interval for successivecompressions of the arm was 50 minutes. Now thetissues distal to the proximal cuff were in this casesubjected to the effects of compression every 25minutes, and, if the compression tingling producedby the arm compressions is due to impulses takingorigin in these tissues, it would be expected to " fade "in a manner characteristic of a 25-minute cycle. Itwas found, however, that the tingling produced bythe successive compressions above the elbow" faded"and disappeared in a manner characteristic of a cycleinterval of 50 minutes rather than 25 minutes. Theinference can be drawn that the tingling is due toimpulses produced in the nerves directly pressedon by the cuff, rather than in the nerves peripheralto the cuff.The findings, therefore, are strongly in favour of

the suggestion that compression tingling is due inits entirety to impulses produced in the nervesdirectly pressed on by the compression cuff. Onthis hypothesis the differences in the distributionsof the sensation in different subjects may be readilyexplained. The radial nerve, lying deeply undercover of powerful muscles, would not be expectedto be so readily affected as the median and ulnarnerves, and of the latter two, the ulnar, being liableto compression against the medial intermuscularseptum, is probably the more vulnerable.The differences observed according to the site of

compression still remain to be explained. Fromthe anatomical configuration of the limb just belowthe elbow it is clear that the nerves are less vulnerableto the effects of compression at this site than theyare in the arm. In spite of this, however, a minimalcompression tingling has occasionally been elicitedby a cuff below the elbow in one subject (G. W.).At the wrist the ulnar and median nerves are againwell protected by their position in relation to thebones, but it is more difficult to understand why thesuperficial branch of the radial nerve does not showthe effects of compression. It will be seen laterthat, in order to produce compression tingling, it isnecessary to compress an adequate length of nerve,and the escape of the radial nerve in this situation isprobably due to the fact that an insufficient lengthof nerve is pressed upon.MECHANISM OF PRODUCTION.-If the impulses

originate in the nerves under the cuff they could,theoretically, be produced in several ways: (a)by stimulation of the nerve trunk caused by (i)mechanical deformation, (ii) oxygen lack, (iii)accumulation of local metabolites ; (b) by a lower-

ing of the threshold to stimulation of the nerve due toone or more of the above causes ; (c) by a com-bination of these factors.Any cuff experiment necessarily involves both

mechanical deformation of the nerves and inter-ruption of their blood supply, and it is difficult toseparate the effects of these two factors. Accord-ingly, further information was sought by compressingdirectly varying lengths of individual nerves.

In the first series of experiments the ulnar nerve at theelbow was compressed against the medial epicondyle ofthe humerus by leaning the elbow on a thin, round rulerplaced between the medial epicondyle and the olecranon,the elbow being bent to a right angle. By this means avery short length of nerve, of the order of 0-5 cm.,was directly compressed, while the circulation to theperiphery was completely unimpeded. Similarly, ashort stretch of the lateral popliteal nerve was compressedagainst the neck of the fibula by crossing the legs and thusbringing to bear the pressure of the lateral condyle of theopposite femur. Repeated experiments with both nervesin three subjects failed to elicit any compression tingling.In these experiments it is definite that the nerves con-cerned were directly pressed upon, since continuance ofthe same pressure led eventually to complete paralysisof the muscles which they supply, and loss of light touchsensibility over their cutaneous distribution. Attemptswere then made to increase the length of ulnar nervecompressed by leaning the elbow on a thin strip of wood,sufficiently narrow to fit into the groove between theolecranon and the medial epicondyle, and curved to fitthe contour of the elbow in a semiflexed position. Theseexperiments also failed to produce compression tingling.Finally, the ulnar nerve was directly compressed in thearm by means of a rectangular block of wood 10 cm. long.This block was placed on the medial side of the arm,behind the medial intermuscular septum, the distal end ofthe block being just proximal to the medial epicondyle,and pressure was exerted forwards and inwards on thearm. Compression of the ulnar nerve in this manner.regularly gave rise to compression tingling normal inonset, intensity, and duration, but restricted to the ulnararea of the hand.

In all these experiments involving compression of theulnar nerve there was little or no interference with theperipheral circulation. This implies that the circulationto the nerve above and below the length directly pressedupon was probably normal. In all cases, however, thenerve was subjected to severe deformation. In the rulerexperiments the length of nerve pressed upon was of theorder of 0 5 cm. Providing that the circulation aboveand below this region is normal, it can readily be supposedthat diffusion might take place to an appreciable extentacross so small a " gap ", and thus the nutrition of thissmall length of nerve might not be seriously interferedwith. In the experiments in which the ulnar nerve was

compressed above the elbow, a considerable length ofnerve was pressed upon, and while diffusion mightreadily occur at the upper and lower ends of this region,in the centre there is probably a region in which the

31



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


G. WEDDELL AND D. C. SINCLAIRprocesses dependent on normal circulatory nutrition arecompletely inhibited.

These findings, therefore, strongly suggest thatthe essential stimulus to the production of com-pression tingling is not solely mechanical deforma-tion of the nerve, but is rather some factor connectedwith the inhibition of normal circulation in anadequate length of nerve, possibly acting in com-bination with mechanical deformation.

Returning to the cuff experiments, it has been seen thatcompression ofa cylinder of-the arm 1 cm. long by meansof a tight roller bandage is capable of producing com-pression tingling. Such a result, at first sight somewhatat variance with the above interpretation, can be explainedby the fact that in this experiment the peripheral cir-culation is interfered with. While diffusion into thestretch of nerve under the cuff can thus readily take placefrom above, little blood can enter the region from below,and in consequence a stretch of nerve " adequate " toproduce compression tingling is asphyxiated. Providingthat an adequate length of nerve is involved, no furtherincrease in the length of nerve asphyxiated has any effecton the result; it has been seen already that the results ofcompressing 20 cm. of nerves by overlapping cuffs areidentical with these produced by a 1 cm. bandage, and bylocal compression ofa stretch ofulnar nerve not exceeding10 cm. in length. It has already been pointed out thatcompression tingling can be produced by a cuff pressureof 100 mm. Hg, but not by one of 60 mm. Hg, the lengthof nerve compressed being the same in both cases.When a cuff at a given pressure is applied to the arm,there is no information available to show what fractionof that pressure is transmitted directly to the nerves underthe cuff. It is reasonable to suppose, however, that witha cuff pressure of 60 mm. Hg the direct pressure on thenerves is insufficient to obstruct the normal circulatoryrequirements of the nerve to such an extent that com-pression tingling is aroused, whereas a pressure of 100mm. Hg is just sufficient to do so.

It still remains to consider whether the impulsesare aroused by direct stimulation of the nerve fibresunder the cuff, or whether they are due to a loweringof the threshold to stimulation in the stretch ofnerveunder the cuff. Lehmann (1937), using a gassingchamber technique, found that, when the phrenicand peroneal nerves of the cat are asphyxiated withnitrogen, the threshold rapidly falls and remainslow until about the sixth minute; during the periodof lowered threshold, spontaneous firing occurs,which is marked about 5 minutes after asphyxia.The duration of the lowered threshold and the timeof maximal spontaneous firing accord reasonablywell with the time factors in compression tinglingin the human arm. Heinbecker (1929), by asphyxi-ating frog nerves, found that the threshold tostimulation is at first lowered, and if asphyxia isprolonged, subsequently rises. Thompson andKimball (1936), using the intact human arm,

obtained results closely paralleling those of Hein-becker, though the period of asphyxia was lessprolonged. In their experiments ischaemia wasproduced by sphygmomanometer cuffs above theelbow, a pressure of 160 mm. Hg being maintainedfor 16 minutes. Electrical stimulation of the mainnerves at the wrist showed a rapid and progressivelowering of the threshold for about the first 12minutes. In 5 subjects the drop in threshold pro-gressed throughout, but in 8 cases there was a termi-nal rise to values equalling or exceeding the initial.figure.

It thus appears that compression of the arm bymeans of a sphygmomanometer cuff can produce alowering of threshold in the main nerves of thehuman arm, and that this result is brought about byasphyxia of the nerve trunks. At first sight, there-fore, it would appear reasonable to attribute com-pression tingling to the presence of a loweredthreshold in the nerves under the cuff. But thereare two facts which render this interpretationunlikely. If the tingling is due to a lowered thres-hold, it would be expected that external stimuli,such as tapping or rubbing the fingers, by feeding astream of normal impulses into the affected segmentof nerve, would intensify the sensation by originatingadditional impulses in the affected segment. Wehave found that nothing of this sort occurs.Secondly, in the experiments of Thompson andKimball, the threshold of the radial nerve wasapparently as much depressed as those of the ulnarand median nerves, and, in fact, these authorsillustrate the marked depression which they reportby a protocol of an experiment on the radial nerve.If tingling is due merely to a drop in the threshold,therefore, there appears to be no reason why itshould not occur as much in the radial nervedistribution as in the distribution of the ulnar andmedian nerves. This, we have already stated, isnot the case.NATURE OF THE IMPULsES.-It has been seen

already that compression tingling is interpretedsubjectively as touch rather than pain., In order toinvestigate the inference that compression tinglingis the result of stimulation of touch rather than painfibres, a subject in whom there was complete loss ofpain sensibility in one upper limb was examined.The history and clinical examination of this subjectmay be summarized as follows.

Subject 5.-A spinal cord injury on December 7, 1945,resulted in paresis of both legs and the right armwith extensor plantar response. There was hemianaes-thesia above the umbilicus on the right side. Gradualrecovery of motor function took place, and left upperthoracic sympathetic ramisectomy was performed onJanuary 3, 1946. On examination on May 30, 1946,

32



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


PINS AND NEEDLES

there was found to be loss of pain and temperaturesensibility over the whole of the right upper limb and rightside of the chest. Touch and vibration were completelyunimpaired. Light and deep pinpricks were felt as

touch; deep pressure pain was not appreciated apartfrom sensation of pressure. Deep reflexes were absentin the right arm, and were increased in the right leg.Plantar reflexes were indefinite. On the left side goodfunctional result was obtained from sympathectomy.The patient was an excellent witness.Numerous investigations on this man at intervals

over a period of 2 months showed that compressiontingling could be produced in both arms by a pressure

of 150 mm. Hg. On the left (sympathectomized)side, tingling began on the average 55 seconds aftercompression, and lasted 8j minutes. It was felt allover the hand, and occasionally sharp prickingelements were appreciated. In the right (pain-deficient) hand, compression tingling began on theaverage 65 seconds after compression and lasted 8minutes. The distribution was similar on both sides,and the only difference lay in the quality of thesensations. No sharp elements were felt on theright side, and the sensation was described as being" dead " rather than " bright " and " as though a

feather were rubbing his fingers ". The findings inthis subject strongly suggest that no more than a

small fraction of the impulses concerned in thesubjective sensation of compression tingling can bemediated by pain fibres, and that the major part ofthe sensation is due to impulses originating in thetouch fibres. Similar results have been obtainedin 3 more subjects in whom there was partial or

complete loss of pain sensibility in one upper limbwithout any considerable interference with touchsensibility.

DISCUSSION

It is curious that Lewis et al. (1931) make no

mention of the occurrence of tingling arising duringcompression of the arm. Bourguignon and Laugier(1923) note the occurrence of tingling in suchcircumstances, and Bazett and McGlone (1931)refer to " intense " tingling sensations producedduring compression when the periphery of the limbis maintained at a temperature of 39°C. Theobservations made by these authors were incidentalto an investigation on temperature sensibility, but,on the basis of the difference in the tingling sensationwhich they state occurs with variation in the tem-perature of the periphery of the limb, they concludethat a peripheral chemical factor is responsible.We have been unable to detect any difference in thesensation of compression tingling with variationsin peripheral temperature, and on other grounds, as

has been seen, are unable to accept the hypothesisof Bazett and McGlone. Reid (1931) fails to men-

E

tion the occurrence of tingling during-compression,and attributes the greater part of the various sensa-tions noted by other authors during compressionto " the excess of armlet pressure on nerves". Wehave, however, found compression tingling to be aconstant feature of any compression carried outabove the elbow in normal subjects, if the pressureemployed is over 100 mm. Hg.On the evidence detailed in this paper, certain

suggestions as to the nature of this tingling may beput forward:

(1) it is due to local stimulation of certain fibresof the main nerves under the compression cuff;

(2) this stimulation is probably brought about byasphyxia rather than deformation;

(3) the sensation produced probably originates inthe main nerve trunks under the cuff, and is chieflydue to impulses arising in the fibres normallyconveying the sense of touch.Two points merit further discussion. The gradual

disappearance of compression tingling in successivecompressions is difficult to explain except on thebasis of some form of adaptation taking place in thetissues under the cuff. Such adaptation might occurin the nerve fibres themselves, the number offibres responding becoming less and less. If,however, the essential stimulus is asphyxia, thenan adaptation of the circulatory conditions under thecuff would have precisely the same effect, the degreeof asphyxia to which the nerves are subjectedbecoming progressively less and less. This hypo-thesis will be further discussed in the considerationof the mechanism of production of release pricking.It is interesting that in only one subject has com-pression tingling been elicited, even occasionally,by any compression distal to the elbow. In thissubject the forearm contained considerable amountsof fatty tissue, and it is possible that this circum-stance leads to a more even distribution of pressurewithin the limb, and thus to a more effective com-pression of the nerves. In any case, it appears that,while a cuff round the arm is effective in compressingthe tissues under it, cuffs on the forearm or wrist arerelatively less effective. This point will also beconsidered further in connexion with releasepricking.

Observations on Release PrickingEFFECT OF SITE OF COMIPRESSION.-Release prick-

ing can be elicited after compression ofany part of thelimb, provided the pressure has been maintainedfor long enough. There are, however, markeddifferences in the results according to the site of thecompression. A 16-minute compression of the armabove the elbow gives rise to severe release prickingcoming on approximately 45 seconds after release,

33P


http://jnnp.bmj.com

/J N

eurol Neurosurg P


ownloaded from


G. WEDDELL AND D. C. SINCLAIR

and lasting for approximately 9 minutes, the prick-ing at its height extending proximally to aboutan inch above the level of the wrist joint. A similarcompression just above the wrist produces prickinglasting about 7 minutes, coming on about 1 minuteafter release, greatly reduced in intensity, and usuallynot extending proximally further than the middle ofthe palm. Compression just below the elbow givesan intermediate result in duration, intensity, anddistribution. Experiments in which a rubber band1 cm. wide is tied tightly round the base of themiddle finger show that a 16-minute compression atthis level is completely ineffective, but that com-pression for 25 minutes gives a faint release prickingin the finger lasting for about 8 minutes. Theseresults are in general agreement with those of Lewiset al., except that in their paper they state that inorder to produce " tingling " on release, compressionsat the wrist must tisually last 20 minutes or longer,and compressions of the fingers must usually last 40minutes or longer.

EFFECT OF DURATION OF COMPRESSION.-Com-pression of the arm above the elbow by a cuff at150 mm. Hg for 61 minutes is not followed byrelease pricking. Compressions of 7 minutes orlonger at this site give rise to definite pricking, theintensity, duration, and distribution of the sensationexperienced varying directly with the length of thecompression. After a 7-minute compression thepricking is faint, comes on about 1 minutes afterrelease, is limited to the palmar surfaces of thefingers, and lasts less than 2 minutes. After a 16-minute compression, as already stated, pricking isvery strongly experienced, comes on about 45 secondsafter release, extends up to the wrist, and lastsapproximately 9 minutes. Aftercompressions ofinter-mediate duration intermediate results are obtained.

It has already been stated that even from thebeginning of release pricking a faint tingling sensa-tion can be detected " in between " the pricks. As,the duration of compression is increased from 10to 16 minutes, however, other sensations beginto obtrude on consciousness. The first of theseis a condition described by Lewis et al. as " pseudocramp", in which the muscles of the hand feelas though they were contracted and paralysed.There is no real paralysis, but a considerable mentaleffort is required to move the fingers. As long asthe hand is kept perfectly still during this phasethere is no sensation additional to the pricking andtingling previously described. Movements of thehand or fingers while " pseudo-cramp " is presentnot only greatly intensify the pricking, but producea third type of sensation. This consists of apronounced fine vibratory " buzzing " felt locallyin the fingers either on movement or on stimulation

by tapping. It is noteworthy that movement ortapping intensifies the release pricking both locallyand also over the whole area in which pricking isfelt, the intensification lasting for some seconds afterstimulation has ceased. In contrast, the " buzzing "sensation described above is never spontaneous,lasts only for the time stimulation by movement ortapping is applied, and is felt locally, i.e. only inthe parts actually moved or tapped. These twoadditional sensations, " pseudo-cramp " and " buzz-ing " occur only in the first stages of release pricking,and rapidly disappear in the course of one or twominutes. Following a 16-minute compression of thearm by means of a pressure cuff it is often verydifficult to separate the elements of the sensationcomplex presented to consciousness, but it will beseen later that, by employing other means, " pseudo-cramp " and " buzzing " may be produced together,without the complicating factors of release prickingand its underlying tingling sensation. Onceanalysis of the " pseudo-cramp " and " buzzing"has been made under these circumstances, it is easyto detect them following a 16-minute compressionin the majority of subjects. It should be noted,however, that in some subjects a 16-minute compres-sion is insufficient to produce these sensations, which,however, occur if compression is continued for longenough.

All these observations closely confirm the findingsof Lewis et al., except that these authors do notappear to have detected the additional " buzzing "element occurring on stimulation after prolongedcompression.

DISTRIBUTION OF SENSATION.-We have alsoconfirmed the findings of Lewis et al. that releasepricking invariably begins in the finger-tips and ismost intense in the palmar aspect of the fingers andin the palm. Usually, pricking in the dorsum isslight, and after the shorter compressions it may beabsent. As has been pointed out, the proximalboundary of the area in which pricking is felt isdependent on the site and duration of compression,and this boundary advances up the limb as a hori-zontal level, rather than in a manner correspondingto the distribution of the cutaneous nerves supplyingthe hand. We have not seen a case in which thedistribution of release pricking is restricted to theterritory of any one nerve, as occasionally happenswith compression tingling. It is quite common,however, for pricking to be more intensely felt onthe radial side of the hand than on the ulnar; thereverse only infrequently occurs. It is also commonfor pricking to begin on one side of the hand anappreciable time before it starts on the other,and the disappearance of the sensation may obey asimilar or the reverse sequence.

34P


http://jnnp.bmj.com

/J N

eurol Neurosurg P


ownloaded from


PINS AND NEEDLESEFFECT OF MODIFICATION OF LOCAL CONDITIONS.-

By varying the pressure in the sphygmomanometercuff we have found that a pressure of 60 mm. Hgwill not give rise to release pricking, even whencompression is applied for 16 minutes. On theother hand, a pressure of 100 mm. Hg applied for10 minutes is regularly followed by release prickingof normal quality, distribution, and duration,though of slightly reduced intensity. The full" normal " response is obtained with a cuff pressureof 150 mm. Hg, and no further increase of pressure,up to a limit of 300 mm. Hg, affects the resultantprickcing. Release pricking, slightly reduced inintensity and duration, follows a 10-minute com-pression of the arm by a bandage 2 cm. wide appliedat a tension just sufficient to obliterate the radialpulse. If the release pricking following compressionby a cuff 12 cm. wide is taken as the " normal "result, then it is found that an increase beyond thisfigure in the length of the limb directly compressedappears to have little influence on the pricking.Thus, if two cuffs, one above the elbow and theother below, are so arranged as to overlap, and bothare then inflated for 16 minutes, the result producedis indistinguishable from that produced by a10-minute compression of the arm above the elbowby one cuff alone.

Lewis, Pickering, and Rothschild state that tingling "isof similar intensity and duration when the cuff has beenapplied at 150 or 300 mm. Hg, and over 12 or 24 cm. ofnerve ", but later they state that ";the actual length ofnerve released matters. . . . If a cuff 12 cm. wide is dis-tended for 10 minutes and a second cuff of the same widthis distended directly below it, and the upper cuff is nowreleased, blood supply is restored to the nerves for astretch of 12 cm. This gives tingling. Similarly, byoverlapping two cuffs, 6 cm. of upper arm with containednerves can be released; tingling in this instance, thoughquite definite, is very slight and transient."

It is not quite clear what is meant by these state-ments, as at first sight they appear mutually exclusive.We have taken them to mean that the authorsconsider that "the actual length of nerve releasedmatters " only up to a length of 12 cm. and that nofurther increase in the length of the nerve compressedhas any effect in increasing release pricking.EFFECT OF MODIFICATION OF PERIPHERAL CONDITIONS.-I.-If two cuffs, one above the elbow and the other

below, are so arranged as to overlap, both may be inflatedtogether to 150 mm. Hg. Ifnow the upper cuff be releasedafter 10 minutes, the pressure in the lower cuff beingmaintained for a further 6 minutes after this time, it isfound that two periods of characteristic release prickingare obtained. The first appears approximately 1minute after release of the upper cuff, lasts for about 4minutes, and is of very low intensity. It can hardly bedoubted that this first period ofpricking is to be attributedin some way to the release of the upper cuff, and at first

sight it might be assumed that it is due to a return to anormal condition of the tissues under the upper cuff, thelower cuff preventing such a return to normality of thetissues peripheral to the upper cuff. Comparison of thepricking thus produced with that occurring after therelease of a 10-minute compression by one cuff at 150 mm.Hg on the arm above the elbow shows that the prickingis slightly reduced in duration and distribution, and verymarkedly reduced in intensity. The second period ofrelease pricking follows the release of the cuff on theforearm, 16 minutes after the start of compression, and isalmost identical in intensity, distribution, and durationwith the release pricking which normally follows releaseof a 16-minute compression of the forearm alone.

II.-If this experiment be repeated, the lower cuff thistime being arranged just above the wrist, the results areslightly different. Two periods of pricking are againobtained but this time the first period is almost the same,in intensity, distribution, and duration, as the prickingproduced by the release of a 10-minute compression ofthe arm above the elbow in the absence of a peripheralcuff. The second period of pricking in this experiment,occurring after release of the lower cuff, is less than thefirst, and in intensity, distribution, and duration resemblesthe effect normally following the release of a 16-minutecompression of the wrist. Lewis et al. state that tingling " isfelt just as strongly in fingers to which the return of bloodis prevented by rubber straps tied around their bases, andalmost, if not quite as strongly when the return of bloodto the whole hand is prevented." With this observationwe are in agreement, provided only that the cuff preventingreturn of blood to the hand is situated at the wrist andnot just below the elbow.III.-A variant of experiment I provides further

information. A cuff was applied to the arm above theelbow for 10 minutes and then released. Half a minutebefore release of this cuff, another one was inflated on theforearm just below the elbow, and this peripheral cuffremained inflated for a further 6 minutes after release ofthe proximal cuff. This experiment produced results forpractical purposes indistinguishable from the results ofexperiment I. A feeble period of pricking followed therelease of the proximal cuff, and a severe pricking followedthe release of the distal cuff. Now in this case the distalcuff was inflated for a total time of 61 minutes, and acompression of this duration below (or indeed above) theelbow normally produces no effect whatever by itself.Yet the pricking following its release was indistinguishablefrom that produced by a 16-minute compression of theforearm.

IV.-Repetition of experiment III with the lower cuffat the wrist, as in experiment 11, provides results indis-tiniguishable from those of experiment II. If it is assumedthat the pricking is due to the return of normal conditionsto the tissues in which the impulses giving rise to prickingare initiated, then the observations on the first period ofrelease pricking in these four experiments are capable ofthe following interpretation.

In an experiment with one cuff on the arm, bloodreturns to the whole limb peripheral to the uppermargin of the cuff, giving what may be termed the

35



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


G. WEDDELL AND D. -C. SINCLAIR

"normal " response. In experiments I and III, on

release of the upper cuff, blood returns to a cylinderof tissue 12 cm. in length, giving a greatly reducedand feeble response. It is at once apparent thatreturn of circulation to the tissues directly underthe cuff is not the only factor in the production ofrelease pricking. In experiments II and IV releaseof the upper cuff leads to the return of blood to a

cylinder of the limb approximately 25 cm. in length(the distance from the upper margin of the upper

cuff to the upper margin of the lower cuff). Thisproduces a result almost indistinguishable from thenormal response. It might, therefore, be arguedthat the normal response is produced by the returnof blood to the tissues under the cuff plus thetissues between the cuff and the wrist. This argu-

ment leads to the conclusion that it is the length oflimb to which blood returns which is the operativefactor in producing release pricking. It is supportedby the observed facts that, as the site of compressionis moved distally on the limb, the resultant releasepricking is progressively diminished. This appears

to be the conclusion reached by Lewis et al., thoughwe cannot agree with them when they state that" the portion of the nerve trunk . . . lying withinthe cuff is chiefly responsible for subsequenttingling."But the second period of release pricking in these

experiments also requires an explanation. If, inaccordance with the above hypothesis, we assume

that the first period of pricking is due to the returnof blood to the tissues as far as the proximal marginof the distal cuff, then this effect is exhausted beforethe distal cuff is released, since the pricking hasdisappeared by this time. It therefore follows thatthe second period of pricking must be due to thereturn ofblood to the tissues under, and peripheral to,the distal cuff. In other words, the second periodof pricking must be due to factors connected onlywith the distal cuff, such as duration and site ofcompression. It has already been seen that nor-

mally a 61-minute compression of the forearm justbelow the elbow produces no release pricking what-ever. But in experiment III release of a 64-minutecompression at this site gives rise to release prickingsimilar in all respects to that produced by a 16-minute compression. We are therefore forcefullydriven to conclude that the presence of a cuff abovethe elbow for 10 minutes of the total compressiontime has completely altered the response followingthe release of the lower cuff.

It is extremely difficult to interpret this finding onany variant of the above hypothesis. It might beargued that the compression of the arm by the uppercuff induces a hyperexcitability of the nerve underit, which -lasts for some time after pricking has

subsided and allows impulses reaching that segmentfrom below to be multiplied. But if hyperexcit-ability of the nerve trunk is called in to explain theresults ofexperiments HI and IV we must then assumethat the same factor does not operate in experimentsI and II, where there is no " enhancement" of thesecond period of pricking. Since in all four experi-ments the duration of the arm compression and theinterval between the release of the upper and lowercuffs are the same, such an assumption is manifestlyuntenable. In the absence of some such phenom-enon there appears to be no reason why the prickingfollowing the release of the lower cuff in experimentsIll and IV should differ in any way from the prickingnormally produced by a compression of similarduration at the same sites. It is, therefore, alreadyclear that the hypothesis of Lewis et al. is unsatis-factory, and that an alternative one is necessary.The only common factor between experiment I,

experiment 111, and a normal 16-minute compressionofthe forearm just below the elbow, is that in all threethe circulation in the tissues distal to the elbow wasimpeded for 16 minutes. Yet the results ensuingafter final release of compression of the forearm justbelow the elbow are in all three experimentsessentially similar. It may, therefore, be inferredthat events taking place in the periphery of the limbplay a leading part in the production of releasepricking.V.-A further experiment of the same nature was also

carried out. Two cuffs were applied, one above and theother below the elbow, as in experiment I, and inflatedtogether. After 10 minutes' compression the lower cuffwas released, pressure in the upper cuff being maintained-for a further 6 minutes. The procedure was thusexactly the reverse of that in experiment I.The results showed that no release pricking followed

the release of the distal cuff, and that the release prickingfollowing release of the proximal cuff was equivalent tothat produced by a normal 16-minute compression abovethe elbow. In other words, the presence of the distalcuff produced no effect, either on its release, or on therelease of the proximal cuff. Now in this experimentrelease of the distal cuff removed the direct pressure on a12 cm. cylinder of forearm, but did not alter the conditionof the circulation in the tissues peripheral to the proximalcuff. If the stimulus to release pricking is the removal ofdirect mechanical pressure on the limb, then a period ofpricking would be expected following the release of thedistal cuff. The fact that no such pricking occurredindicates that the stimulus is probably not of this nature.It is possible that the presence of the proximal cuff, bydepressing conduction in the nerves, prevented the im-pulses arising after release of the distal cuff from reachingconsciousness. The results are, however, consonant withthe supposition that release pricking is largely occasionedby a mechanism having its origin in the tissues peripheralto the site of compression, and initiated by the return ofnormal circulatory conditions to these tissues.

36



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


PINS AND NEEDLESVI.-If experiment V be repeated in the manner of

experiment HI-i.e. by applying the lower cuff for 10minutes, and half a minute before its release inflating theupper cuff for a total time of 6* minutes-again no prick-ing follows the release of the lower cuff. On release ofthe upper cuff, however, severe pricking is produced.This response is more marked than that following a16-minute compression below the elbow, but rather lessthan that following a 16-minute compression of the armabove the elbow. Here again, a 61-minute compressionis normally totally ineffective in producing releaseprickingand it appears that events taking place in the forearm andhand due to the inflation of the lower cuff for 10 nminuteshave completely altered the normal 61-minute responseabove the elbow.

In this case it is not so acceptable to say that the presenceofthe upper cuffhas obstructed the passage ofperipherallyarising impulses, since by the time any pricking occasionedby the release of the lower cuff would be expected-i.e.about 1 minute after release-the proximal cuff has onlybeen inflated for 11 minutes, and it is difficult to believeh at depression of conduction in the nerves, to the extentof suppressing a vigorous sensation such as release

,pricking, could occur in this short time. It appearsunlikely, therefore, that the essential stimulus to releasepricking is the removal of direct pressure, and the evidenceagain favours the importance of events taking place inthe periphery during compression.

EFFECT OF REPEATED COMPRESSIONS.-The effecton compression tingling of carrying out repeatedcompressions at the same site has already beendescribed. These experiments also yielded infor-mation on the behaviour of release pricking. Thetype of result obtained may be indicated by means,of an example. Ten-minute compressions at150 mm. Hg on the arm above the elbow weremade in series on the same subject, the cycle interval(interval between releases) being 20 minutes. lTheresult of the first compression was a " normal "period of pricking, felt all over the hand to the levelof the wrist and lasting 6j minutes. The secondcompression gave a similar result, but, after thethird compression, pricking, though of normalduration, was considered to be reduced in intensityand was not felt proximal to the level of the out-stretched thumb. After the fourth and fifth com-pressions pricking was restricted to the area distalto the metacarpo-phalangeal joints, and the sharppricking elements in the sensation were greatlyreduced. Release pricking following the fifthcompression lasted 4j minutes. After the sixth andseventh compressions the sensation was one of puretingling, rather like a faint compression tingling,and was not felt above the proximal interphalangealjoints. After the eighth and succeeding -compres-sions tingling did not appear spontaneously, butcould be elicited, in the tips of the fingers only, bytapping. The intensity and duration of this tingling

produced on stimulation gradu1lly " faded," untilafter the twelfth compression no abnormal sensationof any kind could be elicited. In all these compres-sions the time of onset of the sensation, whetherpricking or tingling, and whether spontaneous oronly elicited by tapping, remained remarkably con-stant. Very similar results were obtained with othersubjects at the same cycle interval. It may, therefore,be said that, if the cycle interval is short, releasepricking becomes progressively reduced in intensity,distribution, and duration, until eventually itdisappears altogether. As the cycle interval isprogressively lengthened, this process takes placemore and more slowly with successive compressions,until with a cycle interval of 60 minutes there is littlechange perceptible in the release pricking even after13 successive compressions. Because of this phe-nomenon, compressions were not repeated on thesame arm in the investigation of release pricking,unless an interval of 1 hour had elapsed since thelast compression.

SrrE OF ORIGrN OF THE IMPULmEm.-Just as inthe case of compression tingling, it must be assumedthat the impulses of release pricking take origin inthe nervous tissues of the limb peripheral to theupper margin of the stretch of limb compressed.There are, therefore, the same four possibilities to beconsidered: (a) that the impulses arise at the peri-pheral nerve endings; (b) that they arise in themain nerves peripheral to the compression cuff;(c) that they arise in the nerves directly pressed on bythe cuff; or (d) in a combination of the above sites.

It has already been seen that modification ofconditions in the periphery of the limb during com-pression can exercise a profound effect upon theresulting release pricking. There is thus considerablesupport for the view that the impulses of releasepricking take origin in the periphery, and not, asLewis et al. suggested, under the cuff. Tworelevant facts may be re-emphasized here. Thefirst is that release pricking is never confined to theterritory of any one nerve. Secondly, in a fullyestablished response, pricking is always mostpronounced in the pads of the digits, while it is lessso in the more proximal parts of the palmar aspectof the digits and in the palm, and is faint in thedorsum -of the hand and digits. In other words, theintensity of the sensation experienced is distributedin a manner roughly corresponding to the densityof the peripheral innervation of the hand. Boththese facts can readily be explained if it is assumedthat release pricking is due to stimulation of aproportion of nerve endings in the periphery. Afurther point in connexion with the distribution ofthe sensation is also important. It has been seenthat if conditions are such that pricking of minimal

37



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from



distribution is produced it will occur in the finger-tips.As conditions are altered so that pricking is feltover an increased area, the upper margin of thatarea advances up the limb as a more or less hori-zontal line. This observation makes it exceedinglydifficult to attribute the sensation to factors affectingthe fibres of a main nerve trunk in any situation inthe limb, since such a hypothesis would imply aselective action on individual nerve fibres on adistributional basis. Lewis et al. appear to haveaccepted this unlikely view, for they state: " Justas sensory loss occurs through involvement in theupper arm of the nerve fibres to the finger-tips, sotingling of the corresponding parts is similarlybrought about." There is, however, no soundphysiological basis for such a hypothesis, and thefindings can be much more plausibly explained if it isaccepted that the impulses originate at or near theperipheral nerve endings.The effects ensuing after direct compression of various

peripheral nerves were investigated by the same means asalready described under compression tingling. In thefirst series of experiments pressure was maintained onthe ulnar nerve by leaning the elbow on a round rulerplaced between the medial epicondyle and the olecranon.Pressure was maintained in several subjects for varyingperiods up to the point where complete paralysisof all muscles supplied by the ulnar nerve in the handhad occurred. By this time there was considerableimpairment of light touch sensibility over the wholeof the ulnar cutaneous distribution. Release of pres-sure at any time up to this point was not, however,followed by any suggestion of release pricking. In allcases complete recovery of the motor paralysis andsensory loss occurred within two minutes of release,motor recovery occurring first and being apparentlycomplete in less than one minute. During or after thisrecovery phase- no spontaneous sensations occurred inthe hand except that on two occasions a mild sensationof warmth was noted. Rather larger stretches of theulnar nerve were then subjected to pressure at thesame site by using a thin block of wood shaped tothecurvature ofthe elbow, as describedundercompressiontingling. Owing to the difficulty of knowing what lengthof nerve was pressed upon in different subjects and whatpressures were actually exerted upon the nerve, therewas naturally considerable variation in the results withdifferent subjects. One such experiment may, however,be quoted as typical.

Twenty-three minutes after the institution of compres-sion " velvety numbness " first became evident in theulnar area. Eight minutes later this sensation had spreadto involve the whole ulnar area, and was very marked onstimulation, though not spontaneous. At this time, also,hypomsthesia had appeared in the fifth digit. Thirty-five minutes after the start of compression the interosseiwere partially paralysed, and the tip ofthe fifth finger wasanesthetic to stimulation by a number 4 nylon suture."Velvety numbness" became spontaneous in the ulnar

area one minute later. On release, 40 minutes after thestart of compression, the whole of the fifth finger wasanaesthetic to touch, and position sense in this digit wasextremely poor. The interossei and hypothenar muscleswere completely paralysed, and the ulnar half of flexordigitorum profundus was very weak. Thirty secondsafter release the "velvety numbness" had completelydisappeared, and thirty-five seconds after this theparalysis had gone, to be succeeded almost immediatelyby marked pseudo-cramp, which was accompanied bythe " buzzing" sensation previously referred to asoccurring after prolonged compressions of the arm bymeans of a sphygmomanometer cuff. Both sensationswere strictly localized, the cramp to the muscles suppliedby the ulnar nerve, and the " buzzing" to the ulnarcutaneous area. It must be emphasized that no spon-taneous sensation occurred, the " buzzing " being feltonly during the time of stimulation and only in the areasstimulated, just as the " cramp " only became evident onattempting to exercise the muscles concerned. Owing tothe evanescent character of the " buzzing " it is difficult todescribe its quality accurately, but it might be said toresemble an accumulation of very minute " prickingtouch " impulses, much harsher and more violent thanthe tingling normally accompanying release pricking.Five minutes after release of compression, sensation inthe hand had returned to normal. At no time was thereany suggestion of release pricking or spontaneous tinglingin the hand.

This buzzing sensation, elicited only by stimulationwhich causes neither irradiation nor after-dischargeof the sensation, which does not spread centripetallyup the limb, and which differs markedly in qualityfrom the sharp bright pricks of release pricking,must be accepted as a totally different sensation.Nevertheless, it appears to have misled Lewis et al.,for these authors, as a result of experiments in whichthe ulnar nerve was compressed at the elbow byleaning it on a narrow inflatable bag, state thRt" tingling of precisely the same quality " (as incompression cuff experiments) " occurs in the ulnarregion of the hand after release of pressure confinedto the ulnar nerve at the elbow. . . . " Thismisinterpretation of the sensations produced insuch circumstances appears to have been chieflyresponsible for their conclusion that releasepricking arises mainly in the nerves locally pressedupon. We are, however, satisfied that the " buzzing"sensation is not the same as release pricking, and infact bears only an incidental temporal relationshipto it. It has been shown, in these experiments, that" buzzing " may be produced without any accom-panying release pricking, and it has also been seenthat, following compression of less than 16 minutes'duration by means of a sphygmomanometer cuff,release pricking is regularly produced without anyaccompanying " buzzing " or " pseudo-cramp ".On the other hand, following compressions of 16

38



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


PINS AND NEEDLESminutes or longer, all three sensations often appeartogether. We consider that the available evidenceindicates that both " pseudo-cramp " and " buzzing"are due to a disturbance of the threshold in the nervesat the site of compression, occasioned by the returnof normal circulatory conditions to a stretch ofnerve previously deprived of blood supply, providedthat an adequate length of nerve has been so affected.This explanation, while irrelevant to a discussion onthe site of origin of release pricking, is inserted hereto emphasize the fact that not only are the " buzzing "and release pricking completply different sensations,but the impulses subserving them arise in differentsites, and, as will be seen later, are probably due todifferent mechanisms.These observations on compression of the ulnar

nerve were supplemented by experiments in whichthe lateral popliteal nerve was compressed againstthe neck of the fibula by crossing the knees aspreviously described. The findings were in com-plete agreement with those of the second group ofulnar experiments. After some time, " velvety numb-ness" appeared in the lateral popliteal distribution,to be succeeded by paralysis of the extensors andloss of light touch sensibility in the area suppliedby the nerve. On release of compression at thisstage, motor and sensory recovery rapidly followed,to be succeeded by " pseudo-cramp " and " buzzing",which in turn gradually disappeared without anysuggestion of pricking or, indeed, any spontaneoussensation occurring. It may be noted in paren-thesis that this sequence of events, so common ineveryday life, is invariably described as " pins andneedles " though, in fact, no sharp pricking elementsare -present. An interesting and important fact,which will be referred to later, is that in theseexperiments, though no pricking developed in thelimb in which the lateral popliteal nerve was com-pressed, frequently faint but typical release prickingoccurred in the opposite foot on uncrossing the legs.Further experiments in which the sciatic nerve wascompressed in the thigh by sitting on a round rulergave results similar in general character to thoseof the lateral popliteal experiments. Finally, themedian nerve was compressed in two subjects byallowing the forearm and hand to hang downvertically over the back of a wooden chair, the armbeing at right angles to the trunk and to the forearm.Pressure was thus brought to bear on the nerveimmediately above the bend of the elbow, and itshould be noted that this procedure also obliteratedthe radial pulse. As the experiment proceeded, theforearm and hand became slightly cyanosed, butnot congested or painful. In one subject marked" buzzing " and " pseudo-cramp ", confined to themedian distribution, followed release of pressure, but

in the other these sensations were also felt in theradial distritution, presumably owing to involve-ment of the radial nerve at the elbow. In neithersubject did repeated experiments give rise to anyrelease pricking, but in both a very faint spontaneoustingling sensation was present for the first two orthree minutes after release. It will be noted that inall the other experiments in which nerve trunks weredirectly compressed there was no interference withthe peripheral circulation, and no spontaneoussensations occurred. In the median nerve experi-ments, however, a spontaneous tingling did occur,and this is to be associated with some obstruction tothe peripheral circulation. This point will be referredto later in the consideration of the mechanism ofproduction of release pricking.The above observations may be summarized by

saying that direct compression of several peripheralnerves in different subjects completely failed to pro-duce any suggestion ofrelease pricking, provided thatthe circulation to the periphery remained unimpaired.It must, therefore, be concluded that the impulses ofrelease pricking do not arise in the nerves directlypressed on by the cuff. There still remains thepossibility that the impulses may arise in part at theperipheral nerve endings and in part in the mainnerves peripheral to the cuff, though from theindirect evidence of the distribution of the sensationit has been seen, that it is unlikely that any largeproportion of the impulses originate in the lattersite. The conclusion that release pricking has itsorigin at the peripheral nerve endings was tested bycompressing the arms of the same amputation casesas took part in the investigation of compressiontingling (subjects 8 and 9). Both these men had-amputations between the elbow and the wrist, andtherefore, on release of pressure maintained for 16minutes above the elbow, a cylinder of arm 12 cm.long, and stretches of forearm six inches and seveninches long respectively, were released from theeffects of compression. In spite of this, however, nosuggestion of release pricking was felt by eithersubject, although in both the missing hand was fullyexteriorized at the time. It seems definite, there-fore, that the vast majority of the impulses conveyingthe sensation of release pricking are due to stimula-tion of the peripheral nerve endings.MECHANISM OF PRODUCTION.-The views of Lewis

et al. on the mechanism of production of releasepricking may be summarized by two quotations fromtheir paper:1.-" The portion of the nerve trunk chiefly responsible

for sensory loss, that lying within the cuff, is chieflyresponsible for subsequent tingling. The furtheraway the section of nerve affected lies from itsperipheral ending, the more readily is conduction

39



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


G. WEDDELL AND D. C. SINCLAIRdepressed in it, and the more does it contributeto subsequent tingling."

2.-". .. [Tingling] is caused by a condition developedin the main nerves during their recovery from lossof blood supply."

It is plain that this explanation does not accountfor all the observed facts, and must, therefore, bemodified or abandoned. From the evidence alreadypresented we have been forced to conclude that theimpulses of release pricking arise in or near theperipheral nerve endings. In elaborating a new

theory to account for the observed facts, the startingpoint must, therefore, be occurrences in the periphery.We have seen that compressions of nerve trunksinvolving no interference with the circulation inthe periphery are not followed by release pricking,whereas those in which the peripheral circulationis impeded invariably give rise to release pricking.As an intermediate case may be cited the experimentsin which the median nerve was compressed- at theelbow, these compressions involving a partialobstruction of the peripheral circulation by oblite-rating both the brachial artery and the basilic andcephalic veins. As has been pointed out, thesecompressions were followed by a faint spontaneoustingling sensation, which could conceivably beinterpreted as a minimal response of the releasepricking type.We may therefore assume, as a working hypothesis,

that the production of release pricking bears somerelation to the state of the circulation in the peri-phery of the limb, and proceed to examine in moredetail the circulatory conditions in the peripheryboth during and after compression of the limb bya sphygmomanometer cuff. When a compressioncuff at 150 mm. Hg is placed round the limb thenormal circulation to the tissues peripheral to thecuff is interrupted. That movement in the capil-laries continues during compression has been shownby a few direct observations of the capillary cir-culation in the nail bed. We have found that, forapproximately 45 seconds after applying the cuff inany site on the upper limb, circulation in the capil-laries completely stops. -After this time, however,a sluggish and intermittent flow starts, in a propor-tion varying from one sixth to one third of thecapillaries in the field, and is maintained untilrelease of compression. The capillaries open andshut down at intervals, and the direction of bloodflow in the patent ones is invariably the same as thatobserved during normal circulation. No differencesin these findings were occasioned by alteration ofthe site of compression, or the degree of pressureexerted. The gross effects observable in the limbduring compression are, however, those of cir-

culatory stasis. The skin assumes a leaden cyanoticcolour, and measurement by means ofa thermocoupleshow that the skin temperature falls. On releaseof compression, the limb flushes in reactive hyper-aemia, and a considerably increased flow of oxy-genated blood drives out the accumulated venousblood from the peripheral tissues. It is during thisphase of reactive hyperrmia that stimulation of theperipheral nerve endings occurs, with the productionof the sensation of release pricking. It is tempting,therefore, to attribute release pricking to a phenom-enon consequent upon the return ofoxygenated bloodto the nerve endings in the periphery asphyxiatedby the previous interference with their normalcirculation. This implies that during the periodof circulatory arrest the condition of these nerveendings has been in some way modified so that thereturn of normal circulatory conditions stimulatesthem to spontaneous firing. This must be so sincehyperiemia, produced for example by heat, does notin itself give rise to any sensation resembling releasepricking.The suggestion which most readily presents

itself is that this modifying factor is anoxvmia ofthe nerve endings. But it has already been seenthat compression of the arm by a cuff at 100 mm. Hgis followed by release pricking only slightly reducedin intensity, and " normal " in all other respects.During such a compression the radial pulse con-tinues to beat at the wrist, and the peripheraltissues must be considerably better oxygenated thanwhen the occluding pressure is above the systolicblood pressure. If a 16-minute compression at150 mm. Hg is carried out on the arm, and every4 minutes the pressure is rapidly lowered to zero andimmediately pumped up again, a period of threeor four seconds results in which there is no obstruc-tion to circulation in the limb. During this timefour or five normal pulse beats occur, and a con-siderable volume of oxygenated blood is conveyedto the periphery. On final release at 16 minutes,pricking ensues which is almost indistinguishablefrom that following an uninterrupted 16-minutecompression of the arm.

These results- indicate that the implication ofanox&mia of the nerve endings as a causal factorwould not afford an entirely adequate explanation.There is, however, another possibility. In a com-pression of the limb at 150 mm. Hg, not only is theoxygenation of the tissues interfered with by ob-struction of their arterial blood supply, but thevenous drainage from these tissues is also obstructed.As a result, substances, the results of normal orabnormal cell activity, tend to accumulate in thetissue spaces. The existence of one such meta-bolite has already been postulated by Lewis and

40



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


PINS AND NEEDLES

others in connexion with the phenomena of re-active hyperaemia (Lewis, 1927). It is interesting,therefore, to examine our results on the hypothesisthat the accumulation of some substance in thetissue spaces during compression is responsible forso altering the condition of the nerve endings thatthey respond to removal of this substance, on releaseof compression, by spontaneous firing. Lewis(1927) has shown that, if a cuff pressure of 60 mm.Hg is thrown on the arm, the pressure in the mainveins rapidly rises to that level, and circulationin the limb presumably proceeds almost unimpaired.It has been seen that 10-minute compressions of thearm using a pressure of 60 mm. Hg are not followedby release pricking. On the other hand, with apressure of 100 mm. Hg there is considerableobstruction to the venous return, as evinced byswelling and congestion of the limb and the appear-ance of punctate hemorrhages. At this pressure,as has already been pointed out, release pricking" normal " in all respects except intensity, is obtainedafter a 10-minute compression.

In the experiments quoted above, in which a 16-minute compression of the arm is released every 4minutes, with each successive recompression the limbbecomes more swollen, congested, and painful,indicating that the time of release allowed was notsufficient for the afferent side of the circulation inthe limb to become cleared. It appears, therefore,that the essential prerequisite for release prickingis not obstruction of the arterial supply to the tissues,but interference with the venous return, and theobservations are consonant with the implication ofahypothetical metabolite accumulating in .the tissuespaces. Turning to the observed effect of increasingthe duration of compression, this may readily beexplained on the basis of the above hypothesis.The longer compression is maintained, the greaterwill be the accumulation of the metabolite, and thegreater the subsequent response, in terms both ofthe number of nerve endings affected and the periodof spontaneous firing aroused in each individualending. The observations made on the distributionof release pricking may also be adequately inter-preted. If pricking is due to the accumulation of ametabolite in the tissue spaces consequent uponinterference with circulation, it might be expectedthat it would occur first in the parts in which cir-culation is most readily depressed, i.e. the finger-tips.We have seen that this is precisely what occurs.Again, the gradual spread of the sensation in anirregular horizontal line up the limb as the durationof compression is increased can be attributed toexactly the same cause. Lastly, the differences inthe intensity. of the sensation experienced in thedifferent parts of the hand may easily be considered

as due partly to the differences in concentration ofthe metabolite in a proximo-distal direction as aresult of interference with the venous return beingmore marked at the periphery-of the limb, and partlyto the varying density of peripheral innervation inthe hand.So far, therefore, the hypothesis set up is capable

of explaining satisfactorily all the observed facts.But we have still to consider the effects producedby altering the site of compression. It has beenseen that as the compression cuff is moved distallyfrom the arm to the finger it becomes progressivelyless effective in producing release pricking, andlonger and longer compressions are required beforepricking can be elicited. If the accumulationof metabolites in the periphery is responsible forthese effects, then it follows that these metabolitesmust accumulate more readily when the cuff is onthe arm than when it is on the forearm, and so ondown the limb. In order to explain this inferencewe must make a further assuniption, that a cuff onthe arm is more effective in impeding the. circulationin the hand than one on the forearm, that a cuff onthe forearm is more effective than one on the wrist,and that this in turn is more effective than one onthe finger. The anatomical configuration of thelimb in these four sites lends considerable support tothis assumption. In the arm the main vessels, withthe exception of the profunda artery, lie in exposedpositions where they are vulnerable to the effects ofcompression. In the forearm, however, a consider-able volume of blood flows up and down the limbby means of vessels placed more deeply, and partlyprotected from the effects of compression by lyingcentrally between the two bones. Again, in thefinger it is probable that a considerable circulationcan take place through the bone, the cross-sectionalarea of which is here very large in relation to thecross-section of the limb. It may also be mentionedhere that accumulated surgical experience hasindicated that tourniquets applied round the fore-arm are much less effective than those applied roundthe arm. Again, as a result of consideration of themechanism of production of compression tinglingit has already been suggested that compressions onthe forearm are less effective in transmitting pressureto the underlying tissues than compressions on thearm. Experiments with a finger oncometer, in whichthe finger was maintained at a constant temperatureduring compressions in various situations on thelimb, showed that a small increase in volume' of thefinger usually occurred, but there appeared to be nodifference in the results according to the site ofcompression. These experiments are, however,necessarily misleading, since the point which is beinginvestigated is not the mere accumulation of fluid

41



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


G. WEDDELL AND D. C. SINCLAIRin the periphery, but the amount of circulationoccurring in the peripheral tissues.

It may, therefore, be said that no direct evidencecan be adduced here in favour of the assumption thatcompressions on the arm are more effective inimpeding the circulation in the hand than compres-sions on theforearm, and so on. Very strongevidence will, however, be presented later toindicate that the hypothesis is correct. Armedwith this assumption, we may proceed to interpretthe results of the experiments involving modifica-tion of peripheral conditions during compression(experiments I to VI).

In experiment I, in which two cuffs were employed,one above and the other below the elbow, the upperone being released at 10 minutes and the lower oneat 16 minutes, an " adequate " obstruction to thecirculation in the hand was maintained for 10mninutes. At this time the " adequate " cuff wasremoved, leaving a less, effective block to thecirculation in the shape of the lower cuff. As aresult, a certain amount of the hypothetical meta-bolite, accumulated during the previous ten minutes,was washed away from some of the peripheral nerveendings, thus stimulating them to spontaneousfiring which produced the first period of releasepricking. This pricking was necessarily feeble,since only a very partial recovery of circulation hadtaken place. After removal of the upper cuff, thecirculation in the hand rapidly became adapted to itsnew level of depression, and at the time of releaseof the lower cuff the amount of metabolite accumu-lated in the periphery was practically the same aswould have been produced by a 16-minute compres-sion of the forearm. On completely releasing thecirculation, the resulting release pricking was,therefore, indistinguishable from that produced by a16-minute compression of the forearm.In experiment II, where the lower cuff was at the

wrist, the recovery of circulation on releasing theupper cuff was greater than in experiment I, since wehave assumed that a compression at the wrist is lesseffective than one just below the elbow. The firstperiod of release pricking was, therefore, much morepronounced; and, in fact, the findings show that itwas only very slightly less than if no cuff at the wristhad been present. Here again, the second periodofpricking was exactly comparable to that occasionedby a 16-minute compression at the wrist. Experi-ments Ill and IV may readily be explained byprecisely similar reasoning. In experiment V, inwhich two cuffs were again used, the distal cuff thistime being released before the proximal cuff, it wasfound that release of the distal cuff produced noeffect whatever. This is to be expected, sincethroughout the whole 16 minutes of (compression

an adequate obstruction to the circulation in thehand was maintained by the cuff on the arm abovethe elbow.

In experiment VI the reasoning is more com-plicated. Here the cuff below the elbow wasapplied for 10 minutes, and half a minute before itsrelease another cuff was applied on the arm andmaintained there for a further six minutes afterrelease of the distal cuff. We should expect nopricking to follow release of the lower cuff, since thecirculation in the hand was still obstructed by thecuff on the arm. By the time of release of the uppercuff, 16 minutes after the beginning ofthe experiment,the circulation in the hand had been maintained for91 minutes at the level imposed by a forearm com-pression, but for the remainder of the time had beenstill further reduced by the cuff on the arm. Itcould, therefore, be predicted that the prickingfollowing final release would be more marked thanthat following a 16-minute compression of the fore-arm. On the other hand the " adequate " obstructionto circulation occasioned by the cuff on the arm onlylasted for 6j minutes of the total compression time.We should, therefore, expect that the result would benot quite so marked as that following a 16-minutecompression of the arm. It has been seen thatthese predictions are verified by the experimentalfindings.The effects produced by repeated compressions at

the same site remain to be explained. It has beenshown that, when compressions are carried out onthe arm with a short cycle interval, the prickingensuing after each successive compression becomesprogressively less and less, and ultimately dis-appears. Similar results are obtained by repeatedlycompressing the forearm or wrist. If release prick-ing is the result of the accumulation of a tissuemetabolite following obstruction of the bloodsupply to the hand, this phenomenon is susceptibleof two explanations. First, it might be concludedthat the activity of the cells in the periphery becomesdepressed by successive periods of anoxemia, sothat progressively smaller amounts of the activesubstance accumulate. Secondly, and more prob-ably, we may assume that, as a result of repeatedcompressions in the same site, alternative vascularchannels which are not so subject to direct pressureare opened up under the cuff, with the result thatsuccessive compressions lead to less and lessinterference with the circulation in the hand. Itfollows that removal of the hypothetical metabolitefrom the periphery will thus become progressivelyeasier, and the pricking following release will, there-fore, become progressively less. The possibility ofcirculatory adaptation occurring under the cuff hasalready been put forward as a result of discussing

42



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


PINS AND NEEDLESthe mechanism of production of compressiontingling, and further indirect evidence will be adducedlater in support of this assumption.A few experiments were done in which a series of

alternate 10-minute compressions of the arm andforearm of the same limb were performed at acycle interval of 25 minutes. The interval betweenreleases of the arm compressions was thus 50minutes. It was found that the release prickingfollowing releases above the elbow " faded" in amannercharacteristic ofa cycle interval of50minutes,but that the release pricking following the forearmreleases " faded " in a manner characteristic of a25-minute cycle. For example, after the twelfthcompression in one such experiment (the sixth com-pression of the forearm) release pricking lasted only1 minute and was very feeble, but after the thirteenthcompression (the seventh compression of the arm),release pricking was quite strongly felt and lasted44 minutes. If the second of the explanations putforward above to account for the results of experi-ments involving repeated compression is accepted,this finding may easily be explained by assuming thatnot only are compressions below the elbow lesseffective initially than those above in obstructingcirculation in the hand, but circulatory adaptationtakes place more rapidly in the former site. Lastly,we may refer again to the experiments in which thelateral popliteal nerve was compressed by crossingone leg over the other. It has already been statedthat after the adoption of this posture for periodsofthe order of40 minutes, though no release prickingwas observed in the limb in which the nerve wascompressed, frequently faint but unmistakablerelease pricking occurred in the contralateral foot inthin subjects. Now when one thigh is crossed over theother in this manner, a consideration of the anatomyof the parts shows that in the supporting thigh boththe femoral artery and the long saphenous vein maybe pressed upon. There is probably, therefore, avery slight but definite interference with the cir-culation in the foot of this side, and it is quitereasonable to suppose that, after a prolonged periodof pressure, sufficient of the hypothetical metabolitewould accumulate in this foot to cause a faint andshort period of release pricking subsequent to theremoval of the compressing limb.We are now in a position to state that all the

observations we have made on release pricking canbe satisfactorily explained by postulating the accu-mulation of some chemical substance in the tissuespaces of the periphery during the period of com-pression, as a result of interference with the normalcirculation. On release of pressure, this substanceis washed away, leaving the peripheral nerve endingsin a hyperexcitable state so that they give rise to

impulses. To explain some of the findings, twoadditional assumptions are necessary: (a) thatowing to the anatomy of the parts, as the siteof compression is moved distally on the limb,the interference with the circulation in the handbecomes progressively less effective; (b) that,as a result of repeated compressions, circulatoryadaptation takes place under the compressing cuff,this being achieved slightly more readily in the fore-arm than in the arm.

Neither this theory, nor the assumptions entailedin its establishment, do violence to current physio-logical -or anatomical concepts, and direct or in-direct evidence has been adduced in support of eachstepinitselaboration. Nevertheless, ifevents known,or strongly suspected, to be connected with theaccumulation of metabolites in the periphery duringcompression could be shown to behave in a. mannerexactly comparable with the behaviour -of releasepricking, the evidence in support of the theorywould be enormously strengthened. It was accord-ingly determined to repeat the experiments reportedin this paper with a view to ascertaining the behaviourof the reactive hyperemia occurring after release.

EXPERIMENTS ON REACTIVE HYPERARMIAThe view that metabolites accumulating in the

tissue spaces during circulatory arrest are respon-sible for reactive hyperaemia was suggested by Royand Brown (1879), and has received powerfulsupport from Lewis et al. (Lewis, 1927). The state-ment of Lewis (1927) that " reactive hypereemia illus-trates how vasodilation may be produced by normalmetabolic substances" does not appear to havebeen questioned since, and, particularly as the resultof the recent work of Anrep et al. (1944), we mayaccept the fact that reactive hyperamia is due to theaccumulation of a normal metabolite in the tissuespaces during the period of compression andmanifesting its presence on release of compression.It is apparent that this is a mechanism preciselysimilar to the one we have postulated aboveto account for the production of release pricking.Though the metabolites responsible for the twoeffects are not necessarily the same, it is clear thatany manipulations affecting the response producedby one would also be expected to affect the responseproduced by the other.

Experiments on reactive hyperamia were carriedout on two subjects, the results being photographedin colour. In all cases the other arm of the samesubject was "used as a control. The procedure andprecautions taken were exactly similar to thosepreviously detailed for the experiments on release.pricking. In the first experiment a cuff was placedon the left wrist and another on -the right arm above

43



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from



the elbow. Both were then inflated for ten minutesat a pressure of 150 mm. Hg. Just before releasethe condition of the two hands was compared. Amarked difference was at once evident, the left handbeing grossly congested and of a general purplecolour, but with irregular red patches, while theright hand showed no congestion and was of ageneral slate grey colour (Fig. 1). On release ofcompression reactive hyperemia appeared in bothhands but was rather slower in development in theleft than the right, and at its maximum there wasa pronounced difference between the two hands,hyperemia being much more marked on the rightside (Fig. 2). A similar experiment with a cuff onthe forearm controlled by a cuff on the other armabove the elbow produced similar results, thereactive hypernmia again being more marked onthe side of the arm compression, though in this casethe difference between the two hands was not somarked as when the lower cuff was at the wrist. Itmay therefore be said that, just as a cuff on the armis more effective in producing release pricking thanone on the forearm, and this in turn than one on thewrist, so a similar effect holds good for the produc-tion of reactive hyperaemia. Since the latter effectis strongly suspected to be due to the accumulationof a metabolite in the periphery, the evidence isconsonant with the theory that release pricking is alsodue to such a mechanism.Two cuffs were then placed on the right arm, one

above and the other below the elbow, and experimentIII was repeated: that is to say, the upper cuff wasinflated for 10 minutes, and 30 seconds before releaseof this cuff the lower one was inflated, remainingso for a total time of 6- minutes. This series ofmanipulations was controlled on the other limb by acuff which remained on the forearm for 16 minutes.On release of the upper cuff on the right arm therewas flushing of the right hand in spite of the presenceof the second cuff on the forearm. This hyperwmiagradually "settled down" until, 15 minutes afterthe start of compression on both sides, there waslittle difference between the appearance of the twohands, both being congested and purple, with somered patches, and both exhibiting dilated superficialveins. Following the final release at 16 minutes, nodifference could be detected between the hypernmiadeveloping in the two hands. In this experimentalso, the behaviour of the reactive hypernmiaexactly parallels the behaviour of the releasepricking. It is unnecessary to detail the results ofrepeating the other experiments in which theperipheral circulatory conditions were modified bythe use of two cuffs, since in all cases the behaviourof release pricking and reactive hypernmia wasexactly similar.

Repeated compressions were carried out on theleft arm above the elbow, the cycle interval being20 minutes. The reactive hypernmia ensuing afterthe ninth such compression was compared with thatproduced by a single compression at the same sitein the control arm, and was found to be verymarkedly diminished in comparison (Fig. 3).Similarly, the reactive hypernmia following onrepeated alternate compressions of the arm andforearm of the same side was investigated. Thirtyseconds after release, photographs were taken of thehypernmia consequent on the sixth release (release3 of the forearm) and the seventh release (release 4of the arm) respectively. It was found that thehyperemia following the seventh release wasconsiderably greater than that following the sixthrelease, as was the release pricking. Lastly, it wasfound that reactive hyperemia was produced aftera 10-minute compression of the arm by a pressureof 100 mm. Hg, but not by a similar compression at60 mm. Hg. Fig. 4 shows the difference in two armsso treated, 1 minute after release.

All these experiments may be summarized bysaying that in every case the behaviour of reactivehypernmia ran parallel to the behaviour of releasepricking. That is, under circumstances whererelease pricking is marked there is also present aconsiderable reactive hypernmia, and, when' releasepricking is onily faintly experienced or absent,reactive hypernmia is also slight or absent. Now,reactive hyperemia has been accepted by manyworkers as due to the accumulation of a normalproduct of tissue metabolism in the tissue spaces,and these findings, therefore, strongly support thetheory that release pricking is due to a similarmechanism. We therefore conclude that releasepricking is the result of the removal from the tissuespaces of a metabolite, which accumulates thereduring a period of complete or partial circulatoryarrest, and by its accumulation leads to the pro-duction of hyperexcitability of the peripheralsensory nerve endings.

NATURE OF THE IMPULSESLewis et al. (1931) suggested, largely on the basis that

release pricking can be "reinforced " by peripheraltactile or pressure stimulation, that the impulses areconveyed by touch fibres. " The reinforcement oftingling by peripheral tactile stimuli, the absence ofreinforcement when anesthesia of the fingers is completebut pain stimuli are still conducted, are amongst the factstending to suggest that tingling is an affair of the tactilenerves." However, they add the following qualification:" That it is wholly produced in this way is contradictedby the fact that tingling in the fingers is felt after proximalrelease of 12 cm. of upper arm at a time when thesefingers are quite anesthetic."

44



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


Fig. 1 (top left).-Shows the difference in the colour ofthe right and left hands immediately before release ofsimultaneous 10-minute compressions. On the rightside the compression cuff is above the elbow, whileon the left side it is at the wrist.

Fig. 2 (top right).-The same hands 30 seconds afterrelease on the 10-minute compressions, showing thedifference in reactive hyperemia on the two sides.

Fig. 3 (lower left).-Reactive hyperemia 30 seconds afterrelease of simultaneous 10-minute compressionsabove the elbow. On the right side no compressionhad been carried out for more than 2 hourspreviously,but on the left side the compression was the ninthin a series of 10-minute compressions, the intervalbetween the initiation of each compression being20 minutes.

Fig. 4 (lower right).-This figure shows the differencein colour of the two hands 1 minute after release ofsimultaneous 10-minute compressions above theelbow. On the right side the pressure employedwas 100 mm. Hg, while on the left side the pressurewas 60 mm. Hg.

LFacing page 44



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


PINS AND NEEDLES

It has already been stated that, in all the subjectsso far tested by us, the sensation of release prickingwas interpreted as pain rather than touch, but thatthere appeared to be an underlying touch component.Direct evidence on this point was obtained from thefour subjects with loss of pain sensibility on onearm who afforded evidence on the same point inconnexion with the nature of compression tingling.The results in all four were similar, and may besummarized by the experience of subject 5, who wasan extremely accurate observer and the most clear-cut case in the group. This man, after a 10- or16-minute compression, consistently reported a" very fine rushing sensation" in the pain-deficienthand, beginning at the appropriate time for releasepricking and lasting for the normal period. This" rushing " sensation was accompanied by a feelingof " blunt pins pushing through the skin." Atno time did he experience any of the coarse, sharppricking which constitutes the major element ofrelease pricking. The sensations which he did feelwere restricted to the thenar eminence and thepalmar surfaces of the index and middle fingers,occasional " tingles " being felt in the palm. Thesefindings strongly suggest that a large proportionof the impulses in release pricking are carried bypain fibres, and tend to support the subjectiveinterpretation of the sensation.

It has already been noted that the sympatheticsupply to the " control " arm of subject 5 had beeninterrupted surgically six months before examination.After a 10-minute compression above the elbow inthis arm he experienced sharp painful releasepricking, which from his description appeared tobe identical with the sensation produced in a normalarm, and was of similar distribution. (It wasexactly similar to, though more marked than, thesensation produced by release of compression onthe left thigh above the knee, both sensibility andvascular bed being normal in the left lower limb.)The duration of the sensation in the sympathectom-ized arm was, however, shorter than usual, lastingonly 2+ minutes. The sympathectomy in this manwas " functioning " very well, as judged by thesuperficial difference in the circulation and skintemperature between thesympathectomized limb andthe other arm. This result, therefore, renders itimprobable that the pain sensation in releasepricking is carried by the usual sympatheticchannels.We therefore conclude that the impulses in release

pricking are conveyed mainly in the afferentsomatic fibres normally subserving the sensation ofpain, but that there is also a subsidiary elementof touch in the sensation complex.

DIscussioNWe have found that " release pricking " is consis-

tently felt in the hand following compression in anysite on the upper limb, provided that the pressureapplied is 100 mm. Hg or more and that the durationof compression is long enough. On the basis ofthe evidence discussed in this paper we may putforward the following conclusions:

1. " Release pricking" is due to the stimulationof a proportion of the peripheral nerve endings inthe area in which it is felt.

2. This stimulation is probably occasioned by theremoval, from the tissue spaces around these nerveendings, of a chemical substance which accumulatesthere during the period of circulatory depression.

3. The greater proportion of the impulses thusaroused are conveyed in the afferent somatic fibresnormally subserving the sensation of pain.

Several authors (Bourguignon and Laugier, 1923;Bazett and McGlone, 1931 ; Reid, 1931 ; Lewis etal.,1931) have mentioned the occurrence of " tingling"or " pins and needles " following the release ofcirculatory arrest in a limb. Bazett and'McGlone(1931) attribute both the tingling occurring duringcompression and the tingling following releaseto a chemical factor. These authors, however,make no distinction between the two types of" tingling ", and the evidence on which they basetheir hypothesis consists only of the observation thatboth types of " tingling " are more marked whenthe periphery of the -limb is maintained at a hightemperature (390 C.). This observation, as has beennoted previously, we are unable to confirm. It isperhaps curious that, if" release pricking " is due tothe accumulation ofa metabolite in the tissue spaces,as we assume, an increase in peripheral temperaturedoes not affect the characteristics of the sensations.It might be argued that such an increase in tem-perature, by increasing the rate of chemical reactionsin the tissues, would lead to the production of anincreased amount of the metabolite responsible for" release pricking," and thus to an increasedresponse. But there are many explanations whichcould account for the absence of any such increase,perhaps the most acceptable of which is that anincrease of temperature in the periphery will alsomodify the circulatory conditions there. Thiswould possibly result in an increased facility in theprocess of removal of the metabolite from the tissuespaces, which would counterbalance the increasedproduction of the substance.

Bazett and McGlone also make the statementthat the chemical factor which they postulate" presumably is formed more rapidly in the muscles,since stasis involving muscular regions.is much moreeffective than stasis of the fingers, which contains

45



j.com/

J Neurol N


ebruary 1947. Dow

nloaded from


G. WEDDELL AND D. C. SINCLAIRonly smooth muscle." The experiments of Lewiset al. and our own results effectively dispose of thishypothesis. Reid (1931) notes the occurrence of"6 a lively sensation of pins and needles " in thehand and fingers following the readmission of bloodinto the forearm after ischaemias of 15 to 35 minutes'duration. In addition, he mentions the occurrenceof fibrillary twitchings in the muscles of the handand forearm in the early stage of the "' pins andneedles." We have frequently observed such fibril-lation, and, like Reid, find it to be most common in.the small muscles of the hand. Reid attributes the"6 pins and needles " and the fibrillation to " phasesin the recovery of nerve endings and of muscle fibresrespectively," but without bringing forward anyevidence to support his contentions. The work ofLewis et al. on " release pricking " has been referredto throughout this paper, and it has already beenstated that for the reasons cited in the text we areunable to accept their conclusions. It is interestingthat these authors advance precisely the same theoryto account both for the occurrence of " releasepricking" and also for the centripetal spread ofsensory loss in prolonged compressions of the limb.It has been seen that "release pricking " can bemuch more readily explained on the basis of achemical mechanism acting at the periphery, and itis possible that a repetition of the work of Lewiset al. on sensory loss might show that these experi-ments also are capable of a similar interpretatlbn.

SummaryWhen a human limb is compressed by means of a

sphygmomanometer cuff at a pressure of 100 mm.Hg or over, a number of sensations are consistentlyaroused in the compressed limb, both during andafter compression. Two of these sensations, whichhave been termed " compression tingling" and" release pricking " respectively, have been describedin detail.

" Compression tingling" is a fine tingling felt inthe periphery of the limb during the period of com-pression, and evidence has been put forward to showthat it is due mainly to stimulation, by asphyxia, ofsome of the touch fibres in the main somatic nervesdirectly pressed on by the cuff.

" Release pricking " is a coarse sharp prickingsensation felt in the periphery ofthe limb after releaseof the compression cuff. It has been concludedthat this sensation is conveyed mainly by somaticpain fibres, and that it is due to the stimulation of aproportion of the peripheral nerve endings in thearea in which it is felt. This stimulation is probablyoccasioned by the removal, from the tissue spacesaround the nerve endings, of a chemical substancewhich accumulates there during the period ofcirculatory depression.The term " pins and needles ", as commonly applied

in everyday life, is a loose generalization which coversa number of clearly defined subjective sensations.

The cost of the investigation was defrayed by a grantto one of us (G. W.) by the Medical Research Council.This is gratefully acknowledged.

REFERENCESAnrep, G.V., Barsoum, G. S., Salama, S., and Souidan, Z.

(1944). J. Physiol., 103, 297.Bazett, H. C., and McGlone, B. (1931). Proc. Soc. exp.

Biol. N. Y., 29, 87.Bourguignon, G., and Laugier, H. (1923). Arch. internat.

Physiol., 21, 265.Heinbecker, P. (1929). Amer. J. Physiol., 89, 58.Lehmann, J. E. (1937). Ibid., 119, 111.Lewis, T. (1927). "The Blood Vessels of the Human

Skin and their Responses." Shaw and Sons.London.

-Pickering, G. W., and Rothschild, P. (1931).Heart, 16, 1.

Reid, C. (1931). Quart. J. exp. Physiol., 21, 243.Roy, C. S., and Brown, J. G. (1879). J. Physiol., 2, 323.Thompson, I. M., and Kimball, H. S. (1936). Proc. Soc.

exp. Biol. N. Y., 34, 601.

46'P


http://jnnp.bmj.com

/J N

eurol Neurosurg P


ownloaded from


pins and needles observations on some of sensations

Documents