Journal of Neurology, Neurosurgery, and Psychiatry, 1980, 43, 917-924

Axonal refractory period of single short toe extensormotor units in manJORGEN BORG

From the Neurological Clinic, Karolinska Hospital, Stockholm, Sweden

SUMMARY The aim of this study was to test the reliability of the commonly used blockingtechnique for measuring the conduction velocity spectrum of the peripheral motor nerve in man,as first described by Hopf in 19621. Electromyographic recordings were carried out with a selectivitypermitting identification of single motor unit potentials. The conduction velocity of the singlealpha motor axon was determined by a direct technique and by the blocking technique. Therefractory period was determined by comparison of the results of these two techniques. Therefractory period was studied in relation to the strength of the testing stimulus, axonal conductionvelocity, skin temperature and age of the subject. When the strength of the test stimulus increasedfrom 10% to 50% above the axonal threshold, the refractory period decreased from 1-7 to 0-6 ms.An inverse relationship between the axonal conduction velocity and the refractory period wasobserved. When skin temperature was 4°C subnormal the refractory period was significantlyprolonged; at 10°C subnormal it was more than doubled. The same relationship between axonalconduction velocity and the refractory period was observed in both young and elderly subjects.The significance of these findings for the clinical application of the blocking technique is discussed.

By electrical stimulation of a human motor nerveat two different points and electromyographicrecordings, the conduction velocity of the fastestconducting alpha motor axons of the stimulatednerve can be calculated by dividing the lengthbetween the stimulus points by the latencydifference of the electromyographic responses.2 3This "direct" technique gives no detailed in-formation of the conduction velocity of the small,slowly conducting motor nerve fibres.Methods using blocking nerve impulses have

been used for calculating conduction velocity ofthe slow nerve fibres.' 4The "blocking" techniquedescribed by Hopf' has been used in severalclinical studies.1 5-12 Two supramaximal stimuliare applied at a short time interval to a nerve.The first stimulus is applied close to the muscleevoking a maximum muscle response and also anantidrome nerve impulse volley. The secondstimulus is applied at a greater distance from themuscle evoking a muscle response which, due toblocking by the antidromic impulse volley, altersin amplitude when the interstimulus time is

Address for reprint requests: Dr Jorgen Borg, Neurological Clinic,Karolinska Hospital, Fack, 10401 Stockholm, Sweden

Accepted 11 April 1980

changed. By determining the shortest inter-stimulus time for the maximum and the minimumsecond muscle response and correcting for therefractory period of the nerve fibres, the con-duction velocity of the slowest and the fastestconducting fibres can be calculated.

Conflicts are found in the reported findingsinvolving the use of this blocking technique.Hopfl first reported a difference of 4 to 7 m/sbetween the conduction velocity of the fast andthe slow motor fibres in the ulnar nerve, but aftermodifying the technique he described a differenceof 6 to 10 m/s.5 Blackstock et all" found thedifference in the same nerve to be 15 to 20 m/s.In the peroneal nerve Miglietta6 found a dif-ference of 4 to 8 m/s. Betts et al'2 (1976) stressedthe importance of the correction made for therefractory period and regarded discrepancies inthis respect as one important reason for the con-flicting results.The refractory periods of whole human

peripheral nerve trunks have been studied ex-tensively.12-18 The absolute and relative re-fractory periods of different kinds of single nervefibres have been studied in animal experiments(see review by Paintal'9). Due to methodologicaldifficulties this has not been possible in man.

917

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from

918

Bergman20 studied the recovery cycle of singlemotor axons in man by percutaneous recordings,but could study only part of the cycle.

In a previous study21 the axonal conductionvelocity and discharge properties of single motorunits in the short toe extensor in man were

investigated using special muscle preparations(see Methods below). Such muscle preparationsprovided an opportunity of comparing the con-

duction velocity of the single alpha motor axon

obtained by separate proximal and distal nervestimulation, that is, by the direct technique, withthe conduction velocity of the same motor axonobtained by the blocking technique. By com-

paring the results of these two techniques an

index of the refractory period of the motor nerve

fibre could be obtained. For the clinical appli-cation of the blocking technique, with regard tothe variation in terms of stimulus strength,temperature and age of the subject, 71 motorneurons falling within the whole conductionvelocity spectrum were studied.

Methods

The study included 12 clinically healthy subjects withnormal motor conduction velocity (MCV), (theauthor, two students aged 25 and 30 years, and nineelderly subjects from a pensioners' gymnastic group

aged 67 to 78 years).The common peroneal nerve was stimulated close

to the head of the fibula. The deep peroneal and theaccessory deep peroneal nerves were stimulated atthe ankle, anterior and posterior to the lateralmalleoli. The stimuli were delivered through surfaceelectrodes 0-6 cm in diameter. The cathode was placedover the nerve and the position was adjusted so thatminimum stimulus strength was required for muscleresponse. The anode was placed 2-3 cm medial or

lateral to the cathode. Rectangular pulse waves of0-2 ms duration were used. Stimulus strength couldbe gradually changed between 0 and 300 volts.

Electromyographic recordings were made by bipolarneedle electrodes (DISA 9013K0802 Electronic,Skovlunde, Denmark). Only motor units withpotentials which could be safely identified by theircharacteristic shape in both voluntary contractionand at nerve stimulation were studied. The potentialswvere amplified and displayed on a Medelec oscillo-scope no 4329, and recorded on Kodak Linographdirect print paper.

Selective electromyographic recordings from singlemotor units were obtained in three different ways.

23 motor units were studied in JB, who had takenpart in earlier electromyographic studies of the samemuscle resulting in an increased muscle fibre densitywithin the motor unit and a decreased number ofmotor units permitting recordings from individualmotor units.21 In the elderly subjects 45 motor units

J6rgen Borg

were studied. In this age group the normal type-grouping of the short toe extensor muscle22 isenough to permit recordings from individual motorunits in some subjects. Three motor units werestudied in two medical students who had an abberantinnervation of the short toe extensor muscle by anaccessory deep peroneal nerve23 passing posterior tothe lateral malleol. When the main deep peronealnerve was blocked by lidocain, only a few motorunits were innervated by the accessory branch andselective recordings from these were possible.

After identification of a motor unit in a maximumsustained voluntary contraction, the effort wasreduced until the motor unit no longer discharged.The direct technique for determining the conductionvelocity was used first. The nerve was stimulatedseparately at the ankle and the fibular head,permitting the latency difference to be determinedand the conduction velocity to be calculated.Stimulus strength was 10% above threshold for thetest unit. The blocking technique was then appliedto the same motor unit. The delay of the stimulusat the fibular head, usually 10-15 ms, was initiallylong enough to permit the antidromic impulse fromthe ankle to pass the fibular head before stimulationthere. The interstimulus time was then reduced by0-2 ms stages until blocking took place at the fibularhead due to refractory delay. The shortest inter-stimulus time without blocking occurring could thenbe determined. By comparing this time with thelatency difference at separate proximal and distalstimulation, the refractory period of the test motorunit could be calculated. The stimulus strength atdouble stimulation was at first 10% above threshold.If there was no intolerable pain, the stimulus strengthat the fibular head was made 25%, 50% and 100%ZOabove threshold.Room temperature was held at 23-25°C. Skin

temperature at the fibular head was 30-32°C. Duringthe examination the temperature was regulatedcontinuously by a Disaheater (DISA Electronic,Skovlunde, Denmark). The effect of low temperatureon the refractory period was studied by cooling anarea about 10 x 10 cm around the fibular head usingcold packs (ColPac).

Results

Figs 1 and 2 illustrate the axonal conductionvelocity measurement in a motor unit by thedirect technique and by the blocking technique.

Fig IA shows the isolated motor units activeon maximum voluntary effort. The motor unitwith the highest amplitude potential was used asthe test unit. Fig lB shows the test unit atstimulation at the ankle and fig 1C, D at stimu-lation at the fibular head. The stimulus strengthat the fibular head was in C 10% above thresholdand in D 100% above threshold. At the highstimulus level there was a contamination of the

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from

Axonal refractory period of single short toe extensor motor units in man

Fig 1 Axonal conduction velocity measured bydirect technique. A: the motor units active inmaximum voluntary effort. B: test unit at stimulationat the ankle. C: test unit at stimulation at the fibu!ahead 10%, above threshold. D: test unit atstimulation at the fibula head 100% above threshold.Time bar 10 ms.

test unit potential due to synchronisation ofdistant activity. The latency difference at the10% level was 12-3 ms, and it was the same atthe 25% and 50% levels. At the 100% level itwas reduced to 12 1 ms presumably due tostimulus spread (see below). The conductionvelocity calculation was based upon the latencydifference at the low stimulus level. The distancebetween the stimulus points was 39 cm and thusthe axonal conduction velocity was 32 m/s.As can be seen in fig 1, a slight difference

occurred in the shape of the action potentialduring a voluntary contraction compared withthe same action potential evoked by electricalnerve stimulation. This was due to minor changesin the position of the needle electrode whenmuscle contraction was reduced from maximumvoluntary effort to relaxation before nervestimulation. Only a slight change of the shape

was accepted and the motor unit identity wasestablished by the absence of any other similaraction potential on maximum voluntary effort,or at supra-maximum nerve stimulation, and alsowas confirmed by the blocking procedure (seebelow).

Fig 2 shows the results of blocking experimentsusing the same test unit as above. The stimulusstrength at the fibular head was 10% abovethreshold in A-C, 25% above threshold in D-F,50% in G-I and 100% in J-L. At each stimuluslevel the interstimulus time was reduced by0-2 ms stages until blocking took place (C, F, I,L). The shortest time interval without blockingwas 14-0 ms at the 10% level (B), 13-0 ms at the25% level (E), 12 8 ms at the 50% level (H) and

AD

BE

4ff~W .~frL4

H K

Fig 2 Axonal conduction velocity of the same testunit as in fig I measured by the blocking technique.A-C: stimulus strength at the fibula head 10%above threshold, D-F: 25% above threshold,G-I: 50% above threshold, J-L: 100% abovethreshold. At each stimulus level the interstimulustime was reduced on 0-2 ms stages. Time bar 10 ms.

919

A 9-1

III

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from

920

12-6 ms at the 100% level (K). With the directtechniques the latency difference was 12-3-12-1 ms. Thus, the shortest time interval withoutblocking was always longer than the latencydifference of the direct techniques.

Since the axonal conduction velocity isregarded as being the same in the antidromic andorthodromic directions of a nerve fibre and sincethe measurements were made under the samecircumstances, the difference in the results usingthe two methods should be due mainly to therefractory period. Thus, the refractory period ofthe axon in figs 1, 2 was 17 ms (14-0-12-3) at the10% level, 07 ms (13-0-12-3) at the 25% level,0-5 ms (12-8-12-3) at the 50% level and 05 ms(12-6-12-1) at the 100% level.

Refractory period and strength of testingstimulusIt was possible to study the refractory period of25 motor units in five subjects when the strengthof the testing stimulus, at the fibular head, was10%, 25% and 50% above threshold. The con-

duction velocity of these motor units ranged from30 to 41 m/s.

Fig 3 shows the mean values of the axonalrefractory periods of these 25 motor units at eachstimulus level. Different numbers of motor unitswere studied in each subject and the standarddeviations in the figure are based upon the meanvalues from each of the subjects. The length ofthe axonal refractory period, when the stimulusstrength was 10% above threshold, was 1-72 +

20

18

16inE 14

*° 1 2

> 1 0N

a 08

06

04

10 25 50Stimulus strength (suprathreshold ) ('1.)

Fig 3 Refractory period and stimulus strength.Mean values of 25 motor units in five subjects. Onestandard deviation of the five individual meanvalues is marked.

Jorgen Borg

0-29 ms. When the stimulus strength was 25%above threshold the refractory period was 0 99+j0-26 ms and when stimulus strength was 50%above threshold the refractory period was 0-64+0-22 ms. The difference of the refractory periodbetween the 10% and the 25% above thresholdstimulus levels was a mean of 073 ms (p<001).The difference of the refractory period betweenthe 10% and 50% levels was a mean of 1-08 ms(p<0r001).

It was not possible to check if there was anysignificant stimulus spread when the stimulusstrength at the test point was increased from 10%to 50% above threshold; often it was necessaryto reduce the number of stimulations as muchas possible because of pain and displacement ofthe needle electrode. However, in 10 motor unitsthe significance of the stimulus spread wasinvestigated. No change of the latency differencebetween the distal and the proximal stimuluspoints was detected when stimulus strength atthe fibular was increased from 10% to 25% abovethreshold. With an increase to 50% above thresh-old a shortening of 0-2 ms was seen in two ofthe ten motor units, which means that the meanvalue of the refractory period at the 50% levelnoted above might be a little too low but never-theless significantly lower than the mean valueat the 10% level. For 19 of the 25 motor unitsthe stimulus strength at the testing point couldbe further increased to 100% above the axonalthreshold. The refractory period was then 0-52+0-18 ms. However, when this stimulus strengthwas used, the latency difference between thestimulus points exhibited a shortening of 02 oreven 04 ms in six of the ten motor units.

Refractory period in relation to conductionvelocityFigs 1 and 2 show the findings from a motor unitwith low axonal conduction velocity of 32 m/s.This motor unit had an axonal refractory periodof 1 7 ms at 10% above threshold blockingstimulus.

Fig 4 shows the corresponding findings in astudy of a motor unit with higher axonal con-duction velocity. Fig 4A shows the only highamplitude potential on maximum voluntaryeffort. Fig 4B shows the results of electricalstimulation at the ankle and C at the fibularhead. The latency difference was 9-2 ms corres-ponding to an axonal conduction velocity of 44m/s. In D-F are shown the recordings during theblocking experiment with stimulus strength 10%above threshold. Interstimulus time was decreasedfrom 8-2 ms (D) to 8-0 ms (E) and 7-8 ms (F).

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from

Axonal refractory period of single short toe extensor motor units in man

A B

:i I..I

2-4 A2-322212-01.91-8 01-71-61 51-41-31-21.11-00 9

C ,: F

i ~~~~~~~~~~~2-2

2-0

Fig 4 Axonal conduction velocity and refractory E 19period. A: test unit on maximum voluntary effort. 18B: test unit at stimulation at the ankle. C: test unit , 17

at stimulation at the fibula head. D-F: test unit at the X 16

blocking experiment with decreasing interstimulus ° 5

time. Time bar 10 ms. 1-11-3

The shortest time interval without blocking was8-0 ms and thus the refractory period at thisstimulus level was 12 ms (9, 2-8, 0).

Fig 5A shows the relation between the axonalconduction velocity and refractory period studiedwith the stimulus strength at the fibular head at10% above threshold. The total number of motorunits were 71 in 12 subjects. The conductionvelocity of these motor units ranged from 25 to48 m/s. Of 15 motor units with an axonal con-duction velocity below 35 m/s, nine had refractoryperiods longer than 1-8 ms and only one hada refractory period below 1-3 ms. Of nine motorunits with an axonal conduction velocity of 45m/s or more, five had refractory periods at orbelow 1-3 ms and none had a refractory periodof 18 ms or longer.

Fig 5B shows the mean values of the axonalconduction velocities and refractory periods ineach one of the 12 subjects. These data in thefigure were correlated with a statistical signifi-cance at the 1% level (p<001), the correlationcoefficient being -0-71. Fig 5C shows the axonalconduction velocity and refractory period of23 motor units studied in one subject, JB. Theseindividual data in the figure were correlated witha statistical significance at the 1% level (p<0-01).The correlation coefficient was -0 55.

Refractory period and ageThe results presented in fig 5A and B are basedon studies of 45 motor units in elderly subjectsand 26 motor units in young subjects (23 in the

201-91-81-7

1-61-5

1-411-31-21.-1

a

0 0

* 0

0

O3

001.0

o a

3

* 0

a.

*:* .

0

25 30 35 410 45

B 0

0

25

J c

30 35 40 45

* 0

S0 0 0

0 0 00

0 0

* *0

0 * 0

30 35 40 45Conduction velocity ( mis)

Fig 5 Axonal conduction velocities and refractoryperiods for 71 motor units. Forty-five motor unitsfrom nine elderly subjects (squares), 26 motor unitsfrom three young subjects (dots). B: Mean values ofthe conduction velocities and refractory periods ofeach of the 12 subjects. Correlation coefficient-0 71. C: Axonal conduction velocities andrefractory periods of 23 motor units in one subject.Correlation coefficient -0-55. In B and C theregression lines are marked as illustrations.

author and three in two medical students). Re-gression analysis showed no significant differencebetween the age groups, suggesting that the samerelationship between the conduction velocity andrefractory period exists in both age groups.

The refractory period in relation to temperatureIf the room temperature was held below 20°Cand no arrangements were made to keep thelower leg warm it was noted that during theexamination the temperature at the fibular head

921

.-

0

0

0 0

0

IIjI

II

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from

922

often fell below 30'C, and that simultaneouslythe refractory period became longer even thoughthe conduction velocity remained the same.The effect of cooling the region around the

fibular head to a skin temperature of 20'C wasstudied in two motor units in JB. The refractoryperiod with the stimulus strength 10% abovethreshold at the fibular head was prolonged from1-8 to 4-6 ms in one unit and from 1-8 to 3 0 msin the other. After spontaneous rewarming to28'C the refractory period was 2-8 ms and 2-2ms, that is, was prolonged by 1-0 and 0-4 ms res-pectively. The refractory period became normalwhen skin temperature increased to 32'C. Afterwarming to a skin temperature of 380C a slightshortening of the refractory period took place.The conduction velocity was not significantlychanged perhaps because only a short length ofthe actual nerve segment was superficial enoughto be affected by,the skin temperature variation.

Discussion

A systematic study of the refractory period ofsingle human nerve fibres has not been madebefore. The actual method for measuring therefractory period does not, of course, achievethe exactness of animal experiments. The re-fractory period measurements were made aftera propagated antidromic nerve impulse, in con-trast to those measurements made after a condi-tion,ing electrical nerve stimulation. Moreover,it must be pointed out that muscle preparationsresulting in a reduced number of motor unitsand an increased fibre density were necessaryprerequisites and secondary changes in the proxi-mal nerve segment, where measurements weremade, cannot be excluded. Finally, stimulusspread might be a source of error in determiningrefractory period, even though it was shownto play a minor role in determining conductionvelocity.On increasing the test stimulus strength above

the axonal threshold there was a decrease of therefractory period and an increase of the stimulusspread. Both these effects must be taken intoaccount when the blocking technique is usedwith percutaneous recordings from wholemuscles. If ,the stimulus strength at the testingpoint is 10% above the maximum percutaneousmuscle response, those fibres having the highestthreshold are stimulated 10% above threshold.According to the present findings the correctionfor the refractory period of these fibres shouldbe about 1-7 ms. Since there is often a thresholddifference of more than 100% between different

Jdrgen Borg

motor units this means that the motor units withthe lowest thresholds are stimulated more than100% above threshold when the nerve trunk isstimulated 10% above the maximum muscle res-ponse. The correction for the refractory periodfor these fibres should be about 0-6 ms.

Stimulus strength at 10% above the percutan-eously recorded maximum muscle response isthus accompanied by an unsafety factor of morethan 1 ms when the correction for the refrac-tory period is made. This means a possible errorof more than 10% in calculating the conductionvelocity of the peroneal nerve and even higherwhere shorter nerve segments are studied, forexample the ulnar nerve. Much of this errorcan be avoided if the stimulus strength is high.In previous reports the stimulus strength oftenwas not defined, but usually was 10% abovemaximal probably due to discomfort at higherstimulus strength. According to the present studythe stimulus strength should be 50% above thepercutaneously recorded maximal muscle res-ponse and the correction for the refractoryperiod then - 0-6 ms.An inverse relationship between the axonal

conduction velocity and the relative refractoryperiod was observed when the strength of thetest stimulus at the fibular head was 10% abovethreshold. Proper temperature control excludesthe possibility that this was secondary to in-fluence of temperature. This kind of relationshipis compatible with findings in animal experi-ments.'9 24 The difference between the refrac-tory period of an axon with a conduction velocityof 45 m/s and of an axon with a conductionvelocity of 30 m/s is about 05 ms based uponthe regression analysis in fig 5B. Thus, if themean value of the refractory period at the 10%level is used when calculating the conductionvelocity of both the slow and fast fibres an errorof about 2 m/s is introduced. Theoretically thefastest conducting fibres have the lowest thresh-old, which means that the error related to dif-ferent stimulus strength is accentuated by theerror related to different conduction velocities.

If there is the same relationship between theabsolute refractory period of slow and fast con-ducting fibres and their relative refractory period,the absolute difference between the refractoryperiod for the fast and slow fibres should be lessat h-igh stimulus level. When test stimulus strengthwas 50% above threshold, no statistically signi-ficant difference between the refractory periodof fibres with different conduction velocities wasobserved. However, only a small number ofmotor units within only one segment of the con-

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from

Axonal refractory period of single short toe extensor motor units in man

duction velocity spectrum were studied at thisstimulus level. Thus, no recommendations canbe made concerning different corrections for therefractory period of the different fibres. How-ever, the error introduced by using 06 ms forboth slow and fast fibres is unlikely to beimportant.The influence of temperature on the refrac-

tory period, well known from animal experi-ments25 26 is of great significance when using theblocking technique. With only a slight decreaseof the skin temperature the refractory period isprolonged even when the conduction velocity isnot significantly affected. This is because thenerve segment in which the refractory period ismeasured is more superficial, and thus moreaffected by external temperature variations thanthe rest of the nerve.The relationship between the refractory period

and the conduction velocity was the same in theyoung and the old age groups. This means thatno change of the refractory period, other thanthat correlating to changes in conduction velo-city, was observed in elderly persons when thismethod was used. It must be emphasised, how-ever, that the elderly subjects were selected froma pensioners' gymnastic group and probablyrepresent the most healthy persons in their agegroup. Moreover, the actual refractory periodmeasurements were made at the fibular head andtell us nothing about the situation more distally.Furthermore, the young age group is mainlyrepresented by one person.To summarise, it seems possible to improve

the reliability of the blocking technique in clini-cal practice by using a high stimulus level, whichunfortunately often is painful, and by perfecttemperature control. Of course factors otherthan refractory period are of great importancewhen optimising the blocking technique de-scribed by Hopf, but these aspects were beyondthe scope of the study. Finally, since a propercorrection for the refractory period is of suchimportance for the reliability of the method,knoweldge of the refractory period in patho-logical states is necessary if this method is to beused in the clinic.

References

1 Hopf HC. Untersuchungen uber die Unterschiedein der Leitgeschwindigkeit motorischer Nerven-fasern beim Menschen. Dtsch Z Nervenheilk1962; 183:579-88.

2 von Helmholtz H. Messungen uber den zeitlichenVerlauf der Zuckung animalischer Muskeln unddie Fortpflantungsgeschwindigkeit der Reizung in

den Nerven. Jona Mallers, Arach Anat Physiol1850; 276-364.

3 Hodes R, Larrabee MC, German W. The humanelectromyogram in response to nerve stimulationand the conduction velocity of motor axons.Arach Neurol Psychiat 1948; 60:340-95.

4 Thomas PK, Sears TA, Gilliatt RW. The rangeof conduction velocity in normal motor nervefibres to the small muscles of the hand and foot.J Neurol Neurosurg Psychiatry 1959; 22:175-81.

5 Hopf HC. Electromyographic study on so-calledmononeuritis. Archives Neurol 1968; 9:307-12.

6 Miglietta 0. Nerve motor fibre characteristics iiJchronic ischemia. Arch Neurol 1966; 14:443-53.

7 Miglietta 0. Motor nerve fibres in amyotrophiclateral sclerosis. Am J Phys Med 1968; 47: No 3,118-24.

8 Chaco J. Conduction velocity of motor nervefibres in progressive spinal atrophy. Acta Neurol(Scand) 1970; 46:119-22.

9 Seppalainen A-M, Hernberg S. Sensitive tech-nique for detecting subclinical lead neuropathy.Br J Ind Med 1972; 29:443-9.

10 Seppalainen A-M. Peripheral neuropathy in forestworkers. A field study. Work Environm Health1972; 9:106-11.

1 I Blackstock E, Rushworth G, Gath D. Electro-physiological studies in alcoholism. J NeurolNeurosurg Psychiatry 1972; 35:326-34.

12 Betts RP, Johnston DM, Brown BH. Nerve fibrevelocity and refractory period disturbances innerve trunks. J Neurol Neurosurg Psychiatry1976; 39:694-700.

13 Gilliatt RW, Willison RG. The refractory andsupernormal periods of the human median nerve.J Neurol Neurosurg Psychiatry 1963; 26:136-47.

14 Lowitzsch K, Hopf HC. Refraktarperioden undfrequente Impulsfortleitung im gemischten n.ulnaris des Menschen bei Polyneuropathien. ZNeurol 1973; 205:123-44.

15 Tackman W, Lehmann HJ. Refractory period inhuman sensory nerve fibres. Europ Neurol 1974;12:277-92.

16 Kimura JA. Method for estimating the refractoryperiod of motor fibres in human peripheral nerve.J Neurol Sci 1976; 28:485-90.

17 Kimura J, Yamada T, Rodnitzky L. Refractoryperiod of human motor nerve fibres. J NeurolNeurosurg Psychiatry 1978; 41:784-90.

18 Kopec J, Delbeke, McComas AJ. Refractoryperiod studies in a human neuromuscular pre-paration. J Neurol Neurosurg Psychiatry 1978;41:54-64.

19 Paintal AS. Conduction in mammalian nervefibres. In: Desmedt, ed. New Developments inElectromyography and Clinical Neurophysiology.vol 2. Basel: Karger, 1973; 19-41.

20 Bergmans J. The physiology of single humannerve fibres. Louvain: Vander, 1970.

21 Borg J, Grimby L, Hannerz J. Axonal conductionvelocity and voluntary discharge properties ofindividual short toe extensor motor units in man.

923

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from

924

J Physiol (Lond) 1978; 277:143-52.22 Jennekens FGI, Tomlinson BE, Walton JN. Data

on the distribution of fibre types in five humanlimb muscles. J Neurol Sci 1971; 14:245-57.

23 Lambert EH. The accessory deep peroneal nerve.A common variation of extensor digitorumbrevis. Neurology (Minneap) 1969; 19:1169-76.

24 Paintal S. Effects of temperature on conduction

Jorgen Borgin single vagal and saphenous myelinated nervefibres of the cat. J Physiol (Lond) 1965; 180:20-49.

25 Adrian ED. The recovery process of excitabletissues, Part 2. J Physiol (Lond) 1921; 55:193-225.

26 Tasaki I. The excitatory and recovery process inthe nerve fibre as modified by temperaturechanges. Biochim Biophys A cta 1949; 3:498-508.

by copyright. on A

ugust 28, 2019 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.43.10.917 on 1 October 1980. D

ownloaded from