refractory period studies in neuromuscular preparation' · absolutely refractory period was...

$: Refractory period studies in neuromuscular preparation' · absolutely refractory period was succeeded by a 'relatively refractory period' (RRP) in which the membrane displayed a rise$
Journzal ofNeurology, Neurosurgery, and Psychiatry, 1978, 41, 54-64

Refractory period studies in a human neuromuscularpreparation'J. KOPEC, J. DELBEKE, AND A. J. McCOMAS2

From the MRC Developmental Neurobiology Research Group and Departmnent of Medicine(Neurology). McMaster University, Hamilton, Ontario, Canada

S U M M A R Y A subtraction technique has been employed to study refractoriness in the mediannerve and thenar muscles in man. The absolutely refractory periods of the distal motor nerve

branches and of the muscle fibres were found to be similar, the majority being within the 2.5-3.0 ms range; the values were distributed in a unimodal manner. The relatively refractory periodsof the motor nerve terminals were about 5 ms. In some experiments it was possible to recordnon-propagated potentials at the endplate zones of the muscle fibres.

The concept of an inexcitable period after thedischarge of an impulse was first described byLucas (1909) in relation to frog sartorius musclefibres and by Gotch (1910) during an investigationof frog sciatic nerve. A more extensive analysis ofthis phenomenon was subsequently made withthe cathode ray oscilloscope by Gasser andErlanger (1925); these authors described an 'abso-lutely refractory period' (ARP) after excitationduring which a nerve or muscle fibre could notdisclharge an impulse to a second stimulus. Laterit was shown that small, non-propagated, mem-brane depolarisations ('local responses') could takeplace during this time (Hodgkin, 1938). Theabsolutely refractory period was succeeded by a'relatively refractory period' (RRP) in which themembrane displayed a rise in threshold but couldbe excited nevertheless, provided relatively strongstimulation was used. During the RRP there wasalso a slowing of impulse conduction which wasmost evident close to the stimulating electrode.This last phenomenon introduces a complicationinto any study of nerve or muscle excitability inwhich recordings are made at a distance from thestimulating electrodes. Because of the slowing ofimpulse conduction in the RRP the time separat-ing the arrivals of two impulses at a distant point

Presented in part at the Ninth Interniational Congress of' Electro-encephalography and Clinical Neurophysiology, Amsterdam.September 1977.2 Address for corresponidence and reprint requests: Dr AlanMcComas, Room 4U7, McMaster University Medical Centre, 1200Main St. West, Hamilton, Ontario, Canada L8S 4J9.Accepted 16 July 1977

will be greater than the interval between the twoinitiating stimuli. Thus, while the ARP may bedetermined accurately in the region of membraneunderneath the stimulating electrode, it is onlypossible to measure an 'irresponsive period' whenrecordings are made at a distance, this periodbeing the shortest time observed between thearrivals of two impulses.Although there have been numerous investiga-

tions of impulse conduction velocity in normal andpathological nerves, there have been comparativelyfew studies of refractoriness in human nerve andmuscle. Among the latter studies. the investigationof peripheral nerve by Gilliatt and Willison (1963)was noteworthy in that the significance of impulseslowing during the RRP was emphasised; morerecent studies of nerve have been undertaken byLowitzsch and Hopf (1972) and by Tackmann andLehmann (1974). In all of the studies cited abovethe main trunks of peripheral nerves were exam-ined and consequently there is no informationpresently available as to the refractoriness of thedistal ramifications of sensory and motor nervefibres. This omission is an important one for inneuropathies of the axonal ('dying-back') type thepathological abnormalities are most marked inthe nerve fibre extremities (Cavanagh, 1954).

In relation to the refractoriness of muscle fibres,Farmer et al. (1960) stimulated bundles of musclefibres with intramuscular needle electrodes andrecorded from single fibres with other needleelectrodes at a distance. This technique has theadvantage over indirect stimulation of muscle inthat the contribution of slowed impulse conduc-

54

group.bmj.com on September 9, 2017 - Published by http://jnnp.bmj.com/Downloaded from

http://jnnp.bmj.com/

http://group.bmj.com

Refractory period studies in a human neuromuscular preparation

tion in motor nerve fibres to the observed resultsis eliminated. Nevertheless, the detection of singlemuscle fibre responses may be a time-consumingprocedure, and it is only practicable to examine afew of the tens of thousands of muscle fibres inany one muscle. There is also the influence ofthe stimulating electrodes to consider, for thethreshold of a muscle fibre will be altered if thefibre has been damaged in any way by the needle.For the various reasons presented above it was

thought opportune to examine refractoriness in ahuman nerve-muscle preparation. The presentpaper includes the report of a method which en-ables the entire ARPs and RRPs of nerve andmuscle fibres to be examined (see also Betts et al.,1976); it also describes the configuration and inter-pretation of the muscle potentials evoked by nervestimulation during recovery from the RRPs ofthe nerve and muscle fibres. In a second paper(Delbeke et a!., 1978) the effects of two physiologi-cal variables, aging and temperature, are describedtogether with the results obtained in a smallseries of patients with neuropathies or myopathies.

Subjects and methods

Seven men and six women, aged 19 to 29 yearsand free of neurological disorders, served as con-trol subjects; some were investigated on severaloccasions. The experiments were performed withthe approval of the Medical Ethics Committee ofMcMaster University, and informed consent wasobtained from the subjects.

RECORDING AND STIMULATION SYSTEMSRecording Evoked action potentials were re-corded from the median-innervated thenar muscleswith a silver disc, 6 mm in diameter, attached tothe skin overlying the endplate zones of themuscles. In each subject the best position for this'endplate electrode' was found as the region yield-ing the largest and earliest response with an in-itially negative deflection. The reference electrodewas a silver strip, 20 mmX6 mm, fastened aroundthe little finger; a longer strip attached around thewrist served as an earth (ground). All the elec-trodes were coated with a conducting cream. Thearm under study was warmed with an infraredlamp and then covered with a blanket to maintainthe skin temperature at 33-340C.Recordings were also made from the median

nerve at the elbow using two chlorided silver discs,10 mm in diameter, their centres being separatedby 30 mm; these electrodes were embedded in aPlexiglass holder. The muscle and nerve actionpotentials were fed into pre-amplifiers with differ-

ential input impedances of 10 MQ, using pass-bands of 2 Hz-3 kHz; the peak-to-peak noiselevels, with the electrodes connected, were ap-proximately 2 ,uV. The potentials were displayedon a storage oscilloscope with variable persistence(Hewlett-Packard (HP) type 141B), and weresimultaneously entered into a signal averager(HP Signal Analyser type 5480B). The outputfrom the averager was taken to a programmablecalculator (HP Type 9810A) which, in turn, wasconnected to an X-Y plotter (HP Type 9862A).Stimulation Stimuli were delivered to the mediannerve at the wrist or at the elbow through silverdisc electrodes similar to those used for recordingfrom the nerve (see previous section). A Velcrostrap was fastened around the limb and attachedto the electrode holder; in order to prevent ischae-mia of the extremity the undersurface of thestrap was padded with foam rubber. Electricshocks, 50 ,us in duration, were delivered from astimulator (Devices Ltd type 3072), triggered by adigital timing instrument (Devices Ltd digitimer,type 32090). Each stimulus trial was separatedfrom the next by at least five seconds. All stimuliwere 10 to 20% above the maximal value; higherintensities caused some ulnar nerve axons to beexcited and prevented analysis of the thenarresponses.

SUBTRACTION TECHNIQUEThe response (R,) of the muscle (or nerve) to asingle supramaximal sLimulus was stored in thefirst quarter of the memory in the signal analyserand was used as a reference. The responses to apair of shocks comprising conditioning and teststimuli (responses R., and R., respectively) were

IS2,3

'2 2,3-

STIMULATE

STORE

SUBTRACT

PLOT

Fig. 1 Method for extracting R3 responses to testingstimuli from paired R2 R3 (conditioning-test)potentials; see text.

55

SIsv

R R2

.I

I

R

3

---




56

then stored in the second quarter of the analysermemory. The programmable calculator read outthe contents of the two memories and subtractedthat of the first quarter from that of the second.Since the response to a conditioning shock (R2)was the same as that to a solitary stimulus (R1),the subtraction gave the response (R3) to the teststimulus alone (Fig. 1). A small error was intro-duced because the successive moments at whichthe amplified responses were sampled differedbetween the two averaging channels, in theseexperiments by 100 us. A correction was madeusing the calculator such that the two successivepoints in one channel were averaged and thencompared with the intervening point in the secondchannel. A residual error remained, its size beingless than 1% for a 200 Hz signal and less than 5%for one of 500 Hz.The results of the subtraction were plotted in

analogue form on the X-Y plotter. In additionthe maximum peak-to-peak amplitude of the R3response was computed, together with the re-sponse area (voltageXtime).

Results

CHARACTERISTICS OF RESPONSESFigure 2 demonstrates the ability of the sub-traction technique to detect very small muscle

J. Kopec, J. Delbeke, and A. J. McComas

responses to testing stimuli after brief conditioning-test stimulus intervals. In the left hand column areoscilloscope traces showing the combined R2 andR3 responses at varying conditioning-test (S2 S3)stimulus intervals; in the right hand column theR2 responses have been extracted leaving the R3potentials. It can be seen that the R3 response wasabsent for an S2 S3 interval of 1.15 ms but wasclearly present when the interval was 1.3 ms. Atsuccessively greater intervals the R3 response be-came larger, growing rapidly for S2 S3 valuesbetween 1.3 and 3 ms and more slowly afterwards.The pooled results for 20 experiments on 18

hands are displayed in Fig. 3; for each S2 S3 inter-val the R3 response has been plotted as an ampli-tude (peak-to-peak) and as a trace area(voltage-time integral), both measurements beingmade by computer. The forms of both types ofresponse curves are seen to be S-shaped, with theinitial segments corresponding to non-propagatedmuscle responses (see below). The 'amplitude'curve differs from the 'area' curve in being sig-nificantly larger at S2 S3 intervals 3-100 ms, and itexceeds the control value at S2 S3 values of 10 to200 ms. This overshoot of the amplitude responseoccurred during the phase of supernormalitywhich follows the end of the RRP in a musclefibre; during the supernormal phase the initiationand propagation of a further impulse (in R3) was

INTERVALS2-..S3

ms

2

- 3

- 5

Fig. 2 R3 responses to testing stimuli (S3) delivered at different times afterconditioning (S2) shocks; stimulus intervals (1.15-8.0 ms) given at right. Atleft are corresponding oscilloscope traces showing the paired R2 R3(conditioning-test) potentials.





speeded up. Although the potentials of individualfibres were not increased in the supernormalperiod, the increased synchronisation of the fibredischarges resulted in larger compound actionpotentials; possibly the shortening of the musclefibres during the contraction resulting from theR2 potential also contributed to the enlargementof the R3 potential. Because of the complicationintroduced by the supernormal period the 'area'curve provides a better indication of the numberof fibres participating in a response. It was ofinterest that, even at the relatively long SO S3interval of 200 ms, the R. response had not fullyrecovered to the control size, being approximately5% diminished.

Figure 3 also shows the mean latencies of theR3 responses after S3 stimuli; these correspond tothe 'terminal motor latencies' as measured inclinical electromyography. At short S S3 intervalsthe R3 responses were delayed, being 4.36 ms fora stimulus interval of 1.1 Ims. As the S2 S, intervalwas increased the latency of R3 shortened, reach-ing the control value of 2.80 ms at a stimulusinterval of approximately 7 ms.The final curve in Fig. 3 is the amplitude of the

median nerve compound action potential whenrecorded at the elbow. It can be seen that, in con-trast to the muscle fibres, some nerve fibres were

120

(n)

j00~60

1o40-wU

capable of responding to a testing stimulus given0.5 to 0.6 ms after a conditioning shock; the re-sponse had regained virtually the full controlamplitude at a stimulus interval of 1.4 ms.

EFFECT OF STIMULUS POSITIONIn six experiments paired stimuli were applied tothe median nerve at the elbow, and the muscleresponses were compared with those obtainedafter stimulation at the wrist with similar shockintervals (see above). It was found that the mini-mum S2 S3 interval for which an R3 responsecould just be detected was always less if the stimuliwere applied to the elbow rather than at thewrist. An example of this difference is given inFig. 4, in which it can be seen that the minimuminterval was 0.8 ms for paired stimulation at theelbow and 1.1 ms for that at the wrist. The resultsdiffer in another respect in that the curve for'elbow' stimulation arises abruptly from the ab-scissa whereas the initial segment of the curve for'wrist' stimulation develops slowly, giving thegraph an S-shape. When the individual records inFig. 5 are examined it can be seen that the earliestR3 responses after elbow stimulation were diphasic(negative-positive), while those after wrist stimula-tion were monophasically negative (compare theresults for 0.8 and 1.10 ms stimulus intervals). A

LATENCY (ms)_6

4

2

2 0200

STIMULUS INTERVAL (ms)Fig. 3 R3 responses at different conditioning-test stimulus intervals(abscissa). Values shown are of peak-to-peak amplitude (0), potential area(e), terminal motor latency ( o ), and median nerve action potentials atelbow (E, n). Points represent means of 20 experiments ±+ SD.

57





-o

0i-z0

olU0

0ILC,)LuI

(ms)

STIMULUS INTERVAL (ms)

Fig. 4 R3 responses after stimulation of median nerve at wrist (opensymbols) or at elbow (filled symbols) in one subject. Circles and squaresdenote measurements of potential amplitude and of terminal latencyrespectivelh.

further difference between the two sets of resultswas that, in relation to the respective controllatencies, the R:, responses for small S., S:, intervalswere delayed to a greater extent when the stimuliwere applied to the elbow than to the wrist.

At this point it can be stated that the two dif-ferences noted above depend on the slowing ofimpulse conduction in the RRP and on the addi-tional length of nerve available for conductionbetween the wrist and elbow. Thus, until the RRPis terminated, the testing nerve volley (evoked byS) will lag further and further behind the con-ditioning volley (evoked by S2), provided nerve isstill available for impulse propagation. The conse-quences of slowed impulse conduction are dealtwith in more detail in the Discussion.

In a further subject several attempts were madeto obtain a set of recovery curves of ulnar-innervated thenar muscles after paired stimuliapplied to the motor point of the adductor pollicismuscle. These attempts were mostly unsuccessfulbecauIse the S, stimulus artefact could never beeliminatecl completely from the R, response. Inthe best of these experiments a minimum S. S:interval of 1.3 ms was obtained; the terminal

latency of a control response was 1.5 ms and thedistance between the stimulating catlhode and thecndplate electrode was 10 mm.

INITIAL MUSCLE RESPONSES

As already observed, when the paired shockswere delivered to the median nerve at the elbowthe earliest R3 responses had a diphasic (negative-positive) appearance as in Fig. 5. The interpreta-tion of such potentials was that the initialnegativity represented the development of an im-pulse from an endplate potential while the suc-ceeding positivity indicated propagation of theimpulse, the area membrane under the stigmaticelectrode now acting as a 'source' of current fora receding 'sink'.The results after nerve stimulation at the wrist

were usually different in that the earliest R:,responses were monophasically negative, as inFig. 5 for an S., S:, interval of 1.10 ins. Thesepotentials could be shown not to propagate fromthe endplate since no evoked activity could bedetected in simultaneous recordings made from adistant region of the muscle (Fig. 6, upper). Theseobservations served to identify the negative poten-

58





l ~~~INTERVAL (ms)0.70

A ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~M' 1.10

I a"!C \C_..

tials as endplate potentials; they had been revealedbecause the refractoriness of the muscle fibresprevented them from responding to the endplatepotentials with impulses. On some occasions asmall diphasic component in the endplate response,in the absence of a remote potential, suggestedthat an abortive spike had been formed (as inFig. 6. upper; 1.00 ms interval).

LATENCIES OF RESPONSESAs described earlier the impulse initiated in amotor nerve axon by stimulus S2 is followed byan ARP and then by an RRP. If stimulus S:, isdelivered to the nerve during the RRP, the evokedaction potential initiated will travel along an axonwith a reduced conduction velocity. Because ofthis slowing, the interval separating the arrivals atthe motor nerve terminals of the impulses evoked

1.20

:E Fig. 5 R3 responses at differentconditioning-test stimulus intervals(numbers at right). Thle upper and lowertraces in each pair of records correspondto stimulation at the wrist and elbowrespectively. Same subject as in Fig. 4.

2.5ms/div

by S2 and S3 respectively will be greater than theS., S3 inter-stimulus period. There are no tech-niques available at present for reliably detectingaction potentials in human motor nerve terminalsin vivo and the best that can be achieved is torecord the onset of the postsynaptic responses fromthe muscle fibres instead. With reference to Fig. 9,it can be seen that this time (T') is equal to thesum of the S, S, stimulus interval and the terminallatency of the R. response, less the terminallatency of the R2 response-that is, T'=T(S2 S3)+T(R:,) - T(R2). As long as the testing nerve impulsespends some part of its journey in the RRP, T(R3)will exceed T(R,) and T' will consequently belonger than S2 S:,. Once S:, falls outside theRRP T(R2) will be the same as T(R3) and henceT' will equal T(S2 S3).

In the 12 experiments in which the first R3

X

I'

0.80 k

a .n, - =- 3

0.906kid

k0s'

E

59

2:q

014El




INTERVAL (ms)

0.95

1.00

1.05

1.10

0.5mV/div

1.60

I 2.5ms/div

Fig. 6 Examples of non-propagated R, responses totesting stimuli. The upper and lower traces in eachpair correspond to recordings made at the endplatezone and at the distal end of the abductor pollicisbrevis respectively. A t a stimulus interval of 1.00 msonly an endplate response is present; at the longerintervals (1.05, 1.10 ms) the evoked activity reachesthe distal electrode after a conduction delay. Thelowest record is taken from a separate experiment ata muscle temperature of 28°C; the local response hasthe characteristic form of an endplate potential.

response was an endplate potential, the meanvalue for T' was 2.24+0.24 ms. After an addi-tional delay of 0.1-0.2 ms the monophasic end-plate potentials had successfully initiated actionpotentials in the muscle fibres. The subsequentrecovery of the full muscle action potential as afunction of T' is shown in Fig 7 (curve a); in this


figure the mean response areas of the 20 experi-ments have been pooled. For comparison the fig-ure also shows the recovery curve in terms of theS2 S3 interval (curve b), as previously plotted inFig. 3. As would be expected from the precedingdiscussion the initial segment of curve a is shiftedto the right of that of curve b; subsequently thetwo graphs fuse. Curve c in Fig. 7 displays thegrowth of the muscle response as a function ofthe time to the negative peak of the potential. Inpractice this measurement was easier to makethan that of terminal motor latency, as used forcurve a. The 'time-to-peak' measurement has afurther value in that it provides information con-cerning the majority of muscle fibres; in contrast,the estimation of terminal latency is determinedfrom only a small proportion of fibres, namelythose generating the earliest detectable responses.In Fig. 7 it can be seen that the initial sectionsof curves a and c differ from b not only in beingdisplaced to the right but in being very muchsteeper. Thus, as the S2 S3 interval is lengthenedthere is a great increase in muscle response withalmost no accompanying change in T'. This situa-tion arises because the decrease in terminal latencyalmost matches the increase in S2 S3 interval; thepossible significance of this behaviour is con-sidered later (see Discussion).The steepness of the recovery curve is reflected

in Fig. 8 in which successive increments in re-sponse area have been shown as functions of T'.Using measurements of terminal latency, thelargest increments occur within the 2.40-2.80 msepochs; when measurements of latency-to-peakare used ins'ead, the unimodality of the results ismore striking still, 75% of recovery taking placebetween 3.0 and 3.2 ms.

Discussion

The events in time and space which follow theinitiation of an impulse in a motor axon can bedisplayed diagrammatically, as in Fig. 9. In thisfigure the abscissa represents the distance betweenthe site of stimulation of a motor axon and theendplate zone of a muscle fibre within the motorunit. The two ordinates represent time after a con-ditioning stimulus (S2 being delivered at zerotime). The lowest curve represents the passage ofthe impulse evoked by S2 from the stimulatingelectrode to the muscle fibre. For most of itsjourney from the wrist the velocity of the impulseis constant, and it is not until the boundary of themuscle has been reached that appreciable slowingoccurs. Thus, when the stimulating electrodeswere applied in the palm over the motor point of

60





100

0tseo140

~20

0~~~~~~0.5 1 2 3 5 10 20 50 100 200

INTERVAL (ms)

Fig. 7 R3 response areas (voltageX time) plotted as functions of intervalsbetween stimuli (b), onsets of R2 and R3 responses (a), and negative peaks ofR2 and R3 responses (c); see text.

_ _

20 3.05

4.0 5.0

RESPONSE INTERVAL (ms)

Fig. 8 Increases in R3 potential as functions of intervals between onsets(filled columns) and peaks (open columns) of R2 and R3 responses. Same dataas in Fig 7.

(.

I

80-

g 70-0-93

awU)

z

0 X

LUwO 20-

10-

61





the thenar muscles the evoked response had alatency of approximately 1.5 ms, the distance tothe endplate recording electrode being about 10mm; the mean terminal latency after stimulationat the wrist was 2.8 ms, and the correspondingconduction distance was about 80 mm. Theseresults correspond to impulse conduction velocitiesof 54 m/s from the wrist to the motor point andof 8 m/s from motor point to endplate, allowing0.2 ms for the release and diffusion of transmitterfrom the motor nerve terminal (Hubbard andSchmidt, 1963). The value for impulse conductionvelocity beyond the motor point is a gross estimatewhich does not take into account the progressiveslowing which would occur each time an axondivided into thinner branches; nor would it allowfor the extra distances incurred by an axon pur-suing a tortuous rather than a direct path towardsa muscle fibre. It is of interest that similar ter-minal motor latencies (1.2-1.5 ms) were obtainedafter the stimulation of a variety of other musclesat their motor points with the recording electrodesbeing placed over the endplate zones at distancesof approximately 10 mm; these muscles includedthe abductor digiti minimi, extensor digitorumbrevis, and biceps brachii. Returning to Fig. 9, thetime a denotes the ARP of the motor axon underthe stimulating cathode after the initial impulse.In these experiments this value was determinedby recording the compound action potential ofthe nerve at the elbow; the value for the firstaxons to recover was 0.5-0.6 ms. Inspection ofFig. 3 shows that the ARPs were over in about95% of fibres by 1.1 ms. The RRPs of the motoraxons at the point of stimulation could have beendetermined by measuring the minimal S2 S3 inter-val at which the compound action potentialhad regained its control latency. These measure-ments were omitted in the present study but therelated experiments of Lowitzsch and Hopf (1972)yielded a mean value of 3.08 ms for the RRP onthe basis of latency; the corresponding ARP of0.51 ms was similar to the present one.Suppose now that the testing stimulus, S3, is

given at the moment the ARP finishes (at time ain Fig. 9). The ensuing impulse is slowed as ittravels in the RRP and Fig. 9 shows that, to beginwith, it becomes increasingly separated from theARP as it recedes from the stimulation site. Inthe present experiment, however, such an impulsestill failed to evoke a muscle response if it wasinitiated in the early part of the RRP (up to 1 ms).The simplest explanation would be that the neuro-muscular junctions were activated while themuscle fibres were still in their own ARPs afterresponse R,. It is known from the early experi-

Fig. 9 Events after propagation of action potentialin motor axon from wrist (at left, zero distance) tomotor endplate (at right, 80 mm distance). Time isshown simultaneously on the two vertical axes. ARPsof axon and muscle fibre shown by heavy stippling,respective RRPs by light stippling. The fourinterrupted slanted lines represent action potentialsevoked by a conditioning stimulus (S2, at zero time) orby three testing shocks (S3) given at different intervals(b, d, e); see text.

ments of Kuffler (1942) that synaptic transmissionwill not add any extra depolarisation during thespike component of a muscle fibre action poten-tial, and hence the effect of a second nerve impulsearriving at that moment will be undetected. How-ever, if the second nerve impulse is slightly delayedso as to arrive just after the spike component ofa muscle fibre, the latter can respond to synaptictransmission with a local potential. Using theextracellular recording technique of the presentexperiments, these local potentials presented asmonophasic negative deflections and, upon oc-casions, had the characteristic sharply rising ap-pearance of endplate potentials (see Fig. 6, lower).The fraction of the ARP during which only localpotentials could be detected was approximately0.1-0.2 ms (no in Fig. 9). Nerve impulses arrivinglater than this evoked diphasic potentials; theseresponses signified that impulses had been initiatedin the muscle fibres and that the RRPs of thefibres had begun. This type of behaviour, namelya local response followed by a propagated poten-tial at a slightly greater conditioning-test interval,was encountered in 12 of the 20 experiments. Inthe remaining eight experiments it appeared thatthe element with the longest refractoriness maynot have been the muscle fibre but the fine nerve

62




Refractorv period studies in a hulman neuromuscular preparation

twigs. In these experiments the initial responserecorded from the muscle to the testing (S3)stimulus was a diphasic potential, signifying thatthe muscle fibres had already emerged from theirARPs by the time that the testing nerve volley hadreached the axon terminals. In these preparationsone might envisage that a slightly earlier impulsetravelling in the RRP of the motor nerve fibremight be suppressed by an abrupt increase inARP. presumably at a site where fine nerve twigsare formed from the parent axon. Even in the 12preparations in which the initial R, responsesat some of the neuromuscular junctions werelocal potentials. it appeared that at the remainingjunctions the axons had the longest refractoryperiods. Thus, the extreme steepness of curves aand c in Fig. 7 indicates that, as the S2 S3 intervalis lengthened, increasingly large numbers ofmuscle fibres can respond, even though hardly anyextra time has been gained between the arrivalsof the conditioning and testing nerve volleys atexcitable neuromuscular junctions. Such a situa-tion could have arisen if. at most junctions, themuscle fibres were ready to fire again before theiraxon twigs had recovered from their ARPs. Theenlargement of the response would then reflectincreasing numbers of conducting axons, eachexciting previously primed muscle fibres.From Fig. 7 it is possible to derive values for

the ARP of the nerve-muscle preparation in twoways. First, by using terminal latency measure-ments. more than half of the nerve-muscle fibresappeared to have ARPs in the 2.2-2.8 ms range.However, these values depend on activity in thefastest conducting nerve fibres only and, for thisreason, measurements were also made of thelatencies to the negative peaks of the evokedmuscle responses. These measurements give abetter indication of the behaviour of the nerve-muscle fibre populations as a whole; their draw-back is that the rate of rise of the test muscleresponse (R:) will be decreased during the RRPsince not all the sodium conductance channels inthe muscle fibre membrane are available foractivation. Thus the ARP of 3.0-3.2 ms, derivedfor 75% of fibres in this way, should be regardedas an upper limit. On the basis of the abovereasoning a better estimate for the ARP would be2.5-3.0 ms in the majority of fibres. These valuesare significantly smaller than the mean ARP of3.6 ms found by Farmer et al. (1960) using directstimulation of muscle fibres. The same authorsobtained a lower limit of 2.2 ms in the populationof muscle fibres studied; in the present investiga-tion the shortest response interval (T') observedfor a propagated response was 2.0 ms. The present

examination differs from that of Farmer et al.(1960) in finding no evidence of a bimodal distri-bution of ARPs among the muscle fibres. Thisanomaly might well depend on differences of tech-nique but there is the added possibility that, interms of excitability, the intrinsic muscles of thehand contain a more homogeneous population offibres than the larger muscles of the limbs, asstudied by Farmer and colleagues.

Unfortunately, the present study does not en-able an upper limit to be determined for themuscle fibre ARP because of the problem intro-duced by diminution of the action potential duringthe RRP. Thus, beyond the steep part of themuscle recovery curve (Fig. 7) it is not possible todistinguish between the excitation of additionalmuscle fibres and the further enlargement of ac-tion potentials in fibres already recruited. On thebasis of the amplitudes of the compound muscleaction potentials, it would appear that the majorityof the muscle fibres pass from the RRP into thesupernormal period when the S9 S3 interval isabout 10 ms; deducting 2.7 ms for the mean ARP,the RRP would be approximately 7.3 ms. Thesupernormal period was greatest at approximately50 ms and was still present at 100 ms. Similarestimates of supernormality have been made onthe basis of impulse conduction velocities in singlefibres by Stalberg (1966). In contrast Farmer et al.(1960) estimated that the supernormal period oc-cupied only 6-10 ms post-stimulus on the basisof conduction velocity, or 5-30 ms on the basis ofthreshold for excitation.

So far as the motor nerve fibres are concerned,the determination of the ARP and RRP at thepoint of stimulation has already been described(see above). For that segment of nerve betweenthe stimulating electrode and the endplate theanalysis is complicated in some of the experimentsby the refractoriness of the muscle fibres. Thiscomplication did not arise in the 12 experimentsin which the initial muscle fibre response was alocal potential, for clearly the ARPs of the mostexcitable nerve fibres cannot have been greaterthan the corresponding value for T' (2.25 ms). Forthe remaining fibres, however, reasons have al-ready been presented for supposing that the ARPsof the terminal motor axons were either equal to,or rather greater than, those of the correspondingmuscle fibres. The majority of nerve fibres would,therefore, be expected to have ARPs of less than3.0 ms (see above).The RRPs of the most excitable terminal motor

axons can be estimated by finding the least S2 S.interval for which the latency of the R:, responseis equal to that of the control (R2) response. With

63





reference to Fig. 9, an impulse starting at (time)d will depolarise the muscle fibre at p; the terminalmotor latency (dp) will be slightly prolonged, how-ever, since the nerve impulse will have travelledthe last part of its journey within the RRP andwill have been slowed accordingly. In contrast anerve impulse initiated at e will just avoid enteringthe RRP of the distal ramifications of the motoraxons and will depolarise the muscle fibre at q;the corresponding terminal motor latency (eq)will be equal to the control value (m). From Fig. 7it was found that the RRP of the most excitablemotor axon terminals, determined in this way,lasted approximately 5 ms from the end of theARP. This result is double the value of 2.5 msfound for the motor axons in the forearm byGilliatt and Willison (1963); the difference wouldbe expected because it is known that small nervefibres, such as those formed by repeated divisionsand subdivisions of the parent motor axon, havelong refractory periods, together with low con-duction velocities and high thresholds for excita-tion (Blair and Erlanger, 1933).

Apart from the measurements of refractorinessin terminal motor axons and muscle fibres, thepresent study has enabled for the first time theendplate potentials of human muscle to be re-corded in situ. It remains to be seen whether thesenew types of measurement will be of value in theinvestigation of nerve and muscle diseases.

We are indebted to Norma Zimmerman for secre-tarial help, and to Heidi Roth and Glenn Shinefor technical assistance.

References

Betts, R. P., Johnston, D. M., and Brown, B. H.(1976). Nerve fibre velocity and refractory perioddistributions in nerve trunks. Journal of Neurology,Neurosurgery, and Psychiatry, 39, 694-700.

Blair, E. A., and Erlanger, J. (1933). A comparisonof the characteristics of axons through their indi-vidual electrical responses. A merican Journal ofPhysiology, 106, 524-564.

Cavanagh, J. B. (1954). The toxic effects of tri-ortho-cresyl phosphate on the nervous system. An experi-mental study in hens. Journal of Neurology, Neuro-surgery, and Psychiatry, 17, 163-172.

Delbeke, J., Kopec, S. J.. and McComas, A. J. (1978).Effects of age, temperature, and disease on nerveand muscle refractoriness in man. Journal of Neur-ology, Neurosurgery, and Psychziatry. 41, 65-71.

Farmer, T. W., Buchthal. F., and Rosenfalck, P.(1960). Refractory period of human muscle afterthe passage of a propagated action potential.Electroencephalography and Clinical Neurophysi-ology, 12, 455-466.

Gasser, H. S., and Erlanger, J. (1925). The nature ofconduction of an impulse in the relatively refractoryperiod. American Journal of Physiology, 73, 613-635.

Gilliatt, R. W., and Willison, R. G. (1963). The re-fractory and supernormal periods of the humanmedian nerve. Journal of Neurology, Neurosurgery,and Psychiatry, 26, 136-147.

Gotch, F. (1910). The delay of the electrical responseof nerve to a second stimulus. Journal of Physi-ology, 40, 250-274.

Hodgkin, A. L. (1938). The subthreshold potentialsin a crustacean nerve fibre. Proceedings of theRoyal Society, B, 126, 87-121.

Hubbard, J. I.. and Schmidt, R. F. (1963). An electro-physiological investigation of mammalian motornerve terminals. Journal of Physiology, 166, 145-167.

Kuffler, S. W. (1942). Responses during refractoryperiod at myoneural junction in isolated nerve-muscle fibre preparation. Journal of Neuro-physiology, 5, 199-209.

Lowvitzsch. K.. and Hopf. H. C. (1972). Refraktar-periode und UCbermittlung frequenter Reizserienim gemischten peripheren Nerven des Menschen.Journal of the Neurological Sciences, 17, 255-270.

Lucas, K. (1909). On the refractory period of muscleand nerve. Journal of Physiology, 39, 331-340.

Stalberg, E. (1966). Propagation velocity in humanmuscle fibres in situ. Acta Physiologica Scandi-navica, 70, Suppl. 287, 1-112.

Tackmann, W., and Lehmann, A. J. (1974). Refrac-tory period in human sensory nerve fibres. Euro-pean Neurology, 12, 277-292.

64




preparation.human neuromuscular Refractory period studies in a

J Kopec, J Delbeke and A J McComas

doi: 10.1136/jnnp.41.1.541978 41: 54-64 J Neurol Neurosurg Psychiatry

http://jnnp.bmj.com/content/41/1/54Updated information and services can be found at:

These include:

serviceEmail alerting

online article. article. Sign up in the box at the top right corner of the Receive free email alerts when new articles cite this

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:


http://jnnp.bmj.com/content/41/1/54

http://group.bmj.com/group/rights-licensing/permissions

http://journals.bmj.com/cgi/reprintform

http://group.bmj.com/subscribe/



refractory period studies in neuromuscular preparation' · absolutely refractory period was...

Documents