Journal of Neurology, Neurosurgery, and Psychiatry, 1973, 36, 174-182

Physiological changes in ageing muscles

M. J. CAMPBELL, A. J. McCOMAS, AND F. PETITO

From the Regional Neurological Centre, Newcastle General Hospital, Newcastle upon Tyne,and McMaster University Medical Centre (Neurology), Hamilton 16, Ontario, Canada

SUMMARY Physiological studies have been made of extensor digitorum brevis muscles in 28healthy subjects aged between 60 and 96. Within this elderly population there was evidence ofmuscle wasting and weakness. These changes were shown to result from a loss of functioning motorunits. The surviving motor units were often enlarged and tended to have relatively slow twitches. Insome subjects the maximum impulse conduction velocities were reduced in motor nerves; there was

evidence that slowing of impulse conduction could be especially marked in distal regions of axons.

The findings are considered to indicate the presence of motoneurone dysfunction in old age.

It is well established that there is a decline inmuscular performance with advancing age. Forexample, Burke, Tuttle, Thompson, Janney, andWeber (1953) found that maximum grip strengthfell to almost half between the ages of 25 and 79years. Associated with such findings is the com-mon observation of muscle wasting in theelderly, particularly of proximal limb musclesbut also clearly evident in the small muscles ofthe hand. Undoubtedly many extraneous factorsmay contribute to neuromuscular disease in theelderly, of which the most important are prob-ably malnutrition, disuse, circulatory impair-ment, and occult carcinoma. In the present studywe have been largely able to exclude thesefactors by selecting only healthy and activesubjects. A further feature of the investigation isthat recently described techniques have beenemployed to estimate the isometric twitch tensionand also the number and sizes of motor units in amuscle (the extensor digitorum brevis; seeMcComas, Fawcett, Campbell, and Sica, 1971a;and Sica and McComas, 1971).

Perhaps the most interesting and significantfinding in the present study has been a progres-sive fall in the number of functioning motorunits beyond the age of 60. The nature of thisreduction, and its consequence for the survivingmotoneurones, will be discussed. A preliminaryaccount of this work has been presented else-where (Campbell and McComas, 1970).

174

METHODS

SUBJECTS Seventeen men and 11 women, agedbetween 60 and 96 years were studied (mean age79 4, SD ± 12 2 years). All the subjects were judgedto be in good physical condition for their age and hadled active lives at home before admission to hospital.In some cases, admission had been made on socialgrounds and was for a short period only, while inothers it had been required for the treatment ofminor, non-neurological, ailments. Seven of the menwere about to undergo prostatectomy but in all ofthem the blood urea levels were within normallimits. Each subject was given a careful medicalexamination and particular attention was paid to thevasculature of the leg under study. No patient wasaccepted if there was cutaneous evidence of circula-tory impairment or if neither the dorsalis pedis ormedial plantar arterial pulses could be felt in thesame leg. A total of seventy-two subjects of bothsexes between the ages of 3 and 58 years served ascontrols. Unfortunately it proved especially difficultto obtain an adequately sized population of controlsubjects aged 40 to 60 years. For this reason we haveincluded observations on the estimated numbers ofmotor units in 10 patients who had previously hadhemiplegias but in whom the investigated leg showedno neurological abnormality on careful clinicalexamination. Although the results from thesepatients did not differ significantly from those incompletely normal controls, the values have beendistinguished in Fig. 2. No other observations fromthe hemiplegic patients have been included in thisstudy.

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

Physiological changes in ageing muscles

ELECTROPHYSIOLOGICAL INVESTIGATIONS The num-ber of motor units in the extensor digitorum brevis(EDB) muscle was estimated by the method ofMcComas et al. (1971a) and compared with theisometric twitch tension of the extensor hallucisbrevis muscle (EHB) (most medial subdivision ofEDB: see Sica and McComas, 1971). A collisiontechnique (Thomas, Sears, and Gilliatt, 1959) wasused to compare impulse propagation in the fastestand slowest conducting fibres of the deep peronealnerve. Electromyography was also performed inEDB muscles of 13 elderly and 13 adult controlsubjects using a concentric needle electrode (Disatype 9013K051 1) and an amplifier with a passbandextending from 2 Hz to 10 kHz. In each muscle threesites, separated from each other by at least 1 cm inthe transverse axis of the muscle, were exploredduring slight effort until twenty or more distinctivemuscle action potentials had been analysed foramplitude, duration and configuration. In patientswith severe denervation it was only possible to studyrelatively small numbers of units.

5-

0 2 4 6

FIG. 1, Upper. Comparison of twitlegs of elderly subjects and in 81(means 210± 131 g and 310± 88P <0001).Lower. Comparison ofM wave ampof elderly subjects and in 43 legs of2-7± 17mVandS-7±2 1 mVrespectObservations in control subjects shoumns and in elderly subjects by hatch

Twitch Tension (kg)

n]

STATISTICAL TREATMENT Unless stated otherwise,all means have been expressed with their standarddeviations; the significance of a difference betweenmean values was calculated using Student's t test.

RESULTS

GROSS EVIDENCE OF MUSCLE DYSFUNCTION IN THEELDERLY Two simple methods may be used toassess the functional status of a muscle and bothdepend on the total cross-sectional area of theactivated muscle fibres; they are the measure-ments of M wave amplitude and of maximaltwitch tension. It can be seen from Fig. 1 thatthese parameters were usually, but not invariably,reduced in elderly subjects in comparison withcontrols. This diminution was most evident forthe M wave, the mean results for the elderly andcontrol subjects being 2-7 + 1-7 mV and 5-7+ 2-1 mV respectively (P < 0 001). It has alreadybeen shown that an ageing EHB muscle is ableto develop maximal twitch tension when theinitial length of the muscle is only slightlyincreased (Sica and McComas, 1971). This find-ing is in contrast with the situation in a youngperson, in whom the EHB twitch tension maynot be maximal even when the great toe is fullyplantarflexed. This difference in the length-tension relationship between elderly and youngsubjects would tend to obscure the full loss offorce which occurs with age. Nevertheless, in thepresent study the mean twitch tensions in thecontrols and elderly still differed significantly(310 + 88 g and 210 +131 g respectively;P < 0-001).

control subjects the mean number of motorunits was 197 + 58 and there was no evidence of

any decline between the ages of 3 and 58 years8 10 12 (r= -0 04; see also McComas et al., 1971).M wave (mv) Beyond the age of 60 years, however, many

subjects exhibited a loss of functioning units andch tensions in 26 this reduction became more apparent with

legs of controls advancing age. It is clear from Fig. 2 that withing respectively; the elderly population there was considerable

litudes in 34 legsvariation in residual innervation; for example

P controls (means one 83 year old lady had no functioning EDBlively;P< 0 001). units in one leg and only one in the other. At thePwn by open col- other extreme a sprightly 93 year old lady wasked columns. estimated to have about 112 units in one EDB

Ca)

en

2

Ez

175

r

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

M. J. Campbell, A. J. McCoinias, anid F. Petito

S*

00

* S

* * 00 0 ~ 0

0 U

* 0.0 0

0000-

20

0

**

0

0

40

me

0 0

*

60 80

0*

01

I%

FIG. 2. iAVumbers offiunction-ilg EDB motor units in 94subjects aged betweeni 3 atid96 y'ears (101 legs altogether).Ver-tical lin(es lilik r-eslults inboth legs of samiie subject.Squares demiotc vaalues ob-tailied( in nlornial' legs of 10hemliplegic patieuits (seemethiods). Upper- initerrlutcptedhorizoulital lihie shows 71neauinumi,iiber of unlits in controlsubjects aged 3-58 years (197units), lower inlterru-lWpted lilieindicates smi?allest valite illsaumae subjects (121 units).

100

AGE (yrs)

50o

40

30

20

10

D n

I.L

C ° 20

L,L 30

2O~

120

10

0

mLuscle. In addition to the subject already men-

tioned. a further five had EDB muscles investi-gated in both legs; the paired estimates were 100and 70. 90 and 73, 61 and 45. 51 and 31. and 21and 6 units respectively.

CONTROLS (16-58)

1800V 1 2 3 4mV

0 20n 40 60 80yuV 1 .2 3 *4mV

MOTOR UN T nOTENTIAL AMPLITUDE

FIG. 3. Amplitudes of 268 m?1otor- unit potentials inelder/Iv slubjects (meam 476+ 432 MV) amid in 415potenitials of contrtols aged 17-58 years (mizeami299±255 MV, P<O0001). All measurements madefia11z recor dinigs with surface electrodes.

PROPERTIES OF SURVIVINGNMOTOR UNITS Themotor units remaining in the elderly were investi-gated in several ways. In Fig. 3 the amplitudes of'

the individual motor unit action potentials havebeen compared with those of the control popula-tion. Although there is considerable overlap be-tween the two populations. many of the poten-tials in the elderly were greater than 100 /V andthe mean value, 47 6 + 43 2 MV. differed signifi-cailtly from that of the controls (29 9 + 25 5 ,uV;P<0001). This result indicated that the cross-

sectional area of the suLIiving illotor uLnits hadinlcreased either tIlrouglh adoption of denervatedfibres or by fibre hypertrophy or by both mechan-isms. In this study, only six single unit twitcheswere recorded in the elderly but one of these hada tensioln of 62 g which was considerably larger-than the upper limit of the control range (14 g;Sica and McComas, 1971).

In one patient, a 92 year old man, it was pos-sible to compare the mechanical and electrical

176

300 -

zD200-

00

Li0

m 100-2z

0

i

w

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

Phvsiological changes in ageing muscles

17. 7 69

_ I]10uin......],

FIG. 4. Studies on the sole survivingmotor unit in the EDB muscle ofa 92 yearold man. Left hand column shows observa-tions made at initial examination; all-or-nothing responses on indirect stimulation(top), isometric twitch (middle) andelectromyogram during maximal efjort

:15g (bottom). Right hand column displaysresults obtained five weeks later, showingmuch smaller electrical and mechanicalresponses. Electrical recordings made withsurface electrodes. See text.

properties of the sole surviving unit. In Fig. 4 itcan be seen that an all-or-nothing response wasobtained on indirect stimulation and the volun-tarily-induced electromyogram confirmed thatonly a single unit was present. This unit had apotential amplitude of 340 ,uV, slightly greaterthan any found in the control population of 415units; the large twitch tension of this unit hasalready been referred to (62 g; see above). Whenthe same muscle was examined five weeks later asingle unit response was obtained once more butwas now much smaller, the potential amplitudeand twitch tension being 32 ,uV and 5 g respec-tively (Fig. 4). Further examination at seven and11 weeks after the initial one disclosed no furtherchange. It was conceivable that the unit studiedat the first examination had ceased to functionand that a second nerve fibre had regenerated.On the other hand, certain observations sug-gested that all the responses had been obtained

from the same motor unit and that the territoryof this unit had been reduced by impulse block-age in a major branch of the motor axon con-cerned. In the first place, it was noted that thenumber of units remained constant, whereas ifnerve fibre regeneration was occurring to anyextent, the number of functioning units mighthave been expected to vary. Secondly, themotor unit potential amplitude did not increasein size after the second examination, suggestingthat the numbers and sizes of the innervatedfibres remained the same. A regenerated motoraxon might have continued to adopt denervatedfibres, while newly innervated muscle fibreswould have increased their diameters over aperiod of time. Finally, on all four occasions theend-plate zone of the remaining unit wassituated in the most medial part (EHB) of EDB.If, in fact, all the observations were made on thesame unit, then it is possible that they signalled

11.6.69

10msec

]50g

I

l00msec

--- - r uI

Lm50 msec

177

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

M. J. Campbell, A. J. McComas, and F. Petito

570 440

* *** I

* *I0

* 0

S

B

15 r

10 [

.

* 0

*

g3 5

*O 3

w 15m

_z

50 70

10 [

51

r 190 110 130 150

CONTRACTIOI

r fP7777 7718

O . - r30 50

I I0 50 100 150

NUMBER OF EDB MOTOR UNITS

200

70 90 110

HALF-RELAXATION

N TIME (msec)

13015170 190

TIME (msec)

FIG. 6. Upper. Contraction times of isometrictwitches in 24 legs of elderly subjects (hatchedcolumns) and in 79 legs of controls (open columns);respective mean values were 93±22 msec and64± 7-0 msec (P= <0-001). Lower. Comparison ofhalf-relaxation times in elderly and control popula-tions; means were 109±45 and 53± 9 6 msec re-spectively (P < 0 001).

an early stage of a 'dying back' process(Cavanagh, 1964). In this connection it is ofinterest that the latency of the response to in-direct stimulation increased by rather more than1 msec before the second examination (Fig. 4).

In Fig. 5 the mechanical efficiency of the unitssurviving in this and other subjects have beendisplayed by comparing the maximum EHBtwitch tensions with the numbers of EDB unitsremaining. It can be seen that the observed ten-sions are greater than those which would havebeen obtained had no compensatory changetaken place in the surviving motor units. Al-though part of this improvement will have beena consequence of the altered length-tensioncurve in the elderly, it is probable that collateralreinnervation and muscle fibre hypertrophy werethe major factors.The isometric twitch recordings were of

interest in another respect, for they yieldedinformation about the time courses of contrac-tion and relaxation in the activated fibres. It wasfound that-both phases of the twitch were usually

prolonged in the elderly (Fig. 7); the mean con-traction and half-relaxation times (93 ±22 and109 + 45 msec respectively) differed significantlyfrom corresponding values in the controls (64 + 7and 53 + 10 msec; P<0 001 in each case).

OTHER STUDIES In 25 elderly subjects measure-ments were made of impulse velocity in thefastest conducting nerve fibres and of thelatencies of the evoked EDB responses followingindirect stimulation at the ankle (terminal latency).From Fig. 7 it can be seen that all the lattervalues fell within the normal range (up to 5.0msec), and the mean values determined for theelderly and control populations were identical(4-0 msec). In contrast the mean (maximal)motor nerve impulse conduction velocity wassignificantly lower in the elderly (44-2 + 4-1m/sec; control mean 48-1 + 3.9 m/sec; P < 0 001).When the conduction velocities were consideredindividually, five subjects had values which fell

400 r-

300 M

z0zw

I

mI

200 -

100

(i~77 mm7

FIG. 5. Relationship between EHB twitch tensionsand numbers of surviving EDB motor units in 24legs ofelderly subjects. Interrupted lines indicate lowerlimits of respective observations in controls.

178

r

^ L

-0

7 /7

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

Physiological changes in ageing muscles

z

29

flnn,4 ELDERLY

30 40 50 60

CONTROLS

30 40 50 60

CONDUCTION VELOCITY (m/sec)

ELDERLY

3.0 4.0 5.0 6.0

5-_TER_INAL CONTROLS

3.0 4.0 5.0 6.0

TERMINAL LATENCY (ms*c)

FIG. 7. Upper. Maximal impulse conduction veloci-ties in deep peroneal nerves of 25 legs in elderly sub-jects (hatched columns) and in 30 legs of controls(open columns); respective mean values were 44-2± 4 1 m/sec and 48J1 ± 3*9 m/sec (P< 0001). Lower.Comparison of terminal latencies (see text) in elderlyand control populations; means were 4-0 msec in eachcase.

outside the control range of 42-60 m/sec. In thislaboratory velocities lower than 40 m/sec are

regarded as definitely abnormal in subjects belowthe age of 60. In the elderly subject with thelowest velocity (32 m/sec) only a single motoraxon remained. In 11 of the elderly subjectsdeterminations were made of the latencies of themuscle responses mediated by the least excitablenerve fibres after stimulation at the knee, usingthe technique of Thomas et al. (1959). The leastexcitable fibres would be expected to includethose with the lowest conduction velocities(Erlanger and Gasser, 1937). However, it isknown that other factors, such as the relativeaccessibility of the fibres to the stimulatingcurrent, are also important in determining thethresholds of the fibres when surface electrodesare employed (Bergmans, 1970; McComas et al.,1971a). In our experience the difference between

the latencies of the muscle responses evoked bythe least excitable and most excitable fibres doesnot exceed 4 0 msec in normal subjects. In threeof the 11 subjects tested values larger than thiswere obtained (23'0, 12-9, and 12-0 msec). Inthese three instances the responses evoked by theslowest conducting fibres were sufficiently de-layed to permit their identification after maximalstimulation of the nerve at the ankle. In thesethree subjects it appeared that, although impulseconduction was slowed in the affected fibresbelow the ankle, it was normal in the segment ofnerve between the knee and ankle.

TABLEMEAN MUSCLE ACTION POTENTIAL PARAMETERS IN

CONTROLS AND ELDERLY SUBJECTS

Amplitude Duration Phases(g V) (msec) (no.)

Controls (193) 878 ± 907 9.1 ± 3-1 2-7 ± 1-3Elderly (146) 739 ± 874 9 9± 3 0 2-7± 1-4P >0-1 0-01 >0-9

Numbers of potentials measured are given in parentheses inleft hand column.

In the early part of the study quantitativeelectromyography was performed in 13 elderlysubjects and in the same number of controls.This investigation was discontinued when it be-came clear that information about motor unitfunction could be obtained more easily and pre-cisely from the unit estimating technique ofMcComas et al. (1971a). Of the 146 muscleaction potentials investigated altogether in theelderly population, the mean durations, ampli-tudes and numbers of phases have been given inthe Table. Although no significant change wasnoted in the last two parameters, there was asmall but significant increase in the mean actionpotential duration in the elderly, confirming pre-vious observations (Petersen and Kugelberg,1949; Sacco, Buchthal, and Rosenfalck, 1962).If, however, the results were considered forindividual subjects rather than for the wholepopulation, then in only one of the elderly sub-jects with more than 50 units remaining was theelectromyogram abnormal. In this subject a highincidence of 'myopathic' potentials was detected

179

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

M. J. Campbell, A. J. McComas, and F. Perito

in one region of the EDB muscle. In spite of theloss of functioning motor units in many elderlysubjects, fibrillation potentials were not a featureof the present study.

DISCUSSION

Of all the causes of muscle wasting, ageing is byfar the commonest and it is therefore surprisingthat so little attention has been paid to itsmechanism. Moreover, in the few studies avail-able, opinion appears to be equally divided be-tween a myopathic process (Betourne, 1953;Verzar, 1959; Serratrice, Roux, and Aquaron,1968) and a neuropathic one (Lhermitte, 1901;Rubinstein, 1960; Tomlinson, Walton, andRebeiz, 1969). It is probable that this conflictreflects the difficulties inherent in a study of thiskind. Old people are prone to debilitating illnesswhich, as in the case of diabetes mellitus, maysometimes remain unsuspected in life. The illnessitself may necessitate bed rest or other forms ofimmobilization and these in turn may provokedisuse atrophy. Malnutrition is also common inthe elderly and finally there remains the contribu-tion of vascular insufficiency to myoneural de-generation. For these reasons it is uncertainwhat significance may be attached to studies ofpost-mortem material from patients who hadsuffered from various disorders and in whom thevascular and neurological status of the limbswas unknown before the terminal illness oraccident. Furthermore, in those studies employ-ing histological methods, the interpretation ofmuscle fibre architecture remains an indirectmeans of assessing motor nerve function andcan give no indication of the numbers and sizesof surviving motor units. Attempts to recognizemyopathic lesions in senile muscles are alsoquestionable, since it is now accepted that 'myo-pathic' features may occur in any longstandingneuropathic process (for example, see Mumen-thaler, 1970). Although histological examina-tions of motoneurones, ventral roots, peripheralnerves, or neuromuscular junctions provide moreacceptable approaches to the problem of ageing,it is possible that the results of such studies mayunderestimate functional deficits (see McComas,Sica, and Campbell, 1971b).

In the present study an attempt has been madeto exclude these factors by restricting the study

to elderly people who were in good health andleading active lives. In addition, a relativelydirect method has been employed to estimate thenumbers and sizes of motor units. The choice ofEDB as a suitable muscle for this kind ofinvestigation might be questioned, since changessuggestive of denervation were found in arecent post-mortem study of this muscle in'controls' (Jennekens, Tomlinson, and Walton,1971). While acknowledging that some degree ofdenervation might occur in apparently healthysubjects, we had previously pointed out that sucha process was unlikely to involve a significantproportion of the motor unit population (Mc-Comas et al., 1971a). It is therefore felt that thepresent approach is not only justifiable, but thatit offers real advantages over methods formerlyemployed to study the cause of muscle wastingin the elderly. Although this study has beenconfined to EDB, it is reasonable to assume thatthe abnormalities detected in this muscle arequalitatively similar to those occurring in othermuscles in any generalized neuromuscular dis-order. So far as the intensity of the disease pro-cess is concerned, it is probable that the EDBmuscle will sometimes show particularly strikingchanges because of the increased susceptibility oflong axons to degeneration of the 'dying-back'type (Cavanagh, 1964).

In the present study the results have been un-equivocal; the most important factor contribu-ting to wasting and weakness of ageing musclesis a reduction in the number of functional motorunits. The study has also shown that the severityof the denervating process varies considerablyamong individuals but that it does not usuallycommence before the age of sixty years. Incertain other studies of peripheral nerve eitherthe density or the total number of nerve fibreshave been counted (Swallow, 1966; O'Sullivanand Swallow, 1968). Although these last observa-tions were made on cutaneous nerves, they havealso revealed a loss of nerve fibres which wasparticularly marked in and beyond the seventhdecade. Other authors have dissected single nervefibres and have shown that abnormalities ofmyelination, suggestive of degeneration and re-generation, are more common in elderly subjects(Vizoso, 1950; Lascelles and Thomas, 1966;Ochoa and Mair, 1969; Arnold and Harriman,1970). However, the physiological nature of the

180

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

Phlysiological changes in ageing muscles

present investigation has demonstrated that thetrue extent of the loss of neural function is fargreater than that suggested by morphologicalstudies of peripheral nerve (see above), ventralroot fibres (Corbin and Gardner, 1937) andmotoneurones (Gardner, 1940). The presentfindings are in contrast with the situation in the30 months old rat, in which the careful studies ofGutmann and Hanzlikovfa (1966) have not showna loss of motor axons. In these animals miniatureend-plate potentials persist although their fre-quency of discharge is reduced (Gutmann,Hanzlikova, and Vyskocil, 1971). These lastauthors regard senile muscle atrophy as a specificentity with both pre- and post-synaptic elements.In man, however, we have shown that the atro-phic process does not select neuromuscularfunctions randomly, since entire motor unitsmay cease to function. This finding indicates thatnot only is the underlying mechanism one ofdenervation but that the primary lesion issituated centrally, within the main axon ormotoneurone (see below). In the present study ithas also been possible to estimate the rate of thisprocess. From inspection of Fig. 2 (see Results)the rate of motor unit loss is such that denerva-tion would have usually been complete in EDBhad the investigated subjects lived to the age of90 years. The true rate in the general populationwill have been greater than this, for, at agesbeyond the mean life expectancy, the populationunder study becomes increasingly artificial.Thus, it is reasonable to suppose that those sub-jects in whom the denervating process is mostsevere would, because of their weakness, be morevulnerable to intercurrent illness and would notsurvive for study.

The present investigation also throws light ontwo properties of the surviving motoneurones.Firstly, from the muscle contraction experiments,it would appear that most motoneurones inner-vate muscle fibres of the slow-twitch type. Areduced speed of contraction was also found inaged rats by Gutman et al. (1971). Whether theageing process involves fast-twitch units prefer-entially or whether the slowing results fromexcessive activity in the surviving innervatedmuscle fibres cannot be determined (see Salmonsand Vrbova, 1967; Olson and Swett, 1969). A

second property of the surviving motoneuronesemerges from the measurements of motor unitpotential amplitude and of twitch tension. Theseindicate that, even in the elderly, the residualmotoneurones can participate in compensatorychanges, either by sustaining hypertrophiedmuscle fibres or by adopting denervated fibres.Although these responses are not as marked asthose of motoneurones in younger patients withlongstanding denervation (McComas, Sica,Campbell, and Upton, 1971c) they would stillserve to reduce loss of strength in an elderlysubject. A third type of compensatory neuralresponse, recently investigated in detail byTuffery (1971), would also be expected to occurthough there is at present no electrophysiologicalmethod of assessing its prevalence in man. Inthis process ageing motoneurones send supple-mentary axonal sprouts to reinforce the existinginnervations of muscle fibres.

Finally, the nature of the pathological processresponsible for the loss of functioning units re-mains to be considered. The only observations inthe present study relevant to this problem are themeasurements of impulse propagation velocity.The finding of a moderate reduction in the maxi-mum velocities in the elderly confirmed earlierobservations made on motor axons by Norris,Shock, and Wagman (1953) and on sensoryaxons by Downie and Newell (1961) andBuchthal and Rosenfalck (1966). The verymarked slowing of impulse conduction in distalregions of some motor fibres would suggest thatthese fibres had undergone segmental demyelina-tion and could indicate a primary dysfunction ofSchwann cells. Recently, however, it has becomerecognized that Schwann cell dysfunction, likeaxonal degeneration, may result from a lesion ofthe perikaryon (for example, Dyck, Johnson,Lambert, and O'Brien, 1971). In the absence ofevidence to the contrary, our present hypothesisis that the denervating process which we havedemonstrated in the elderly arises from moto-neurone dysfunction, with or without recogniz-able morphological abnormality (cf. McComaset al., 1971b).

Financial support was received from the MedicalResearch Councils of Great Britain and Canada. Weare grateful to Mrs. B. MacLean for secretarialservices and to Mr. T. Blogg for technical assistance.

181

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from

M. J. Campbell, A. J. McComas, and F. Petito

REFERENCES

Arnold, N., and Harriman, D. G. F. (1970). The incidence ofabnormality in control human peripheral nerves studied bysingle axon dissection. Journal of Neurology, Neuro-surgery, and Psychiatry, 33, 55-61.

Bergmans, J. (1970). The Physiology of Single Human NerveFibres. Vander: Louvain.

Betourne, C. (1953). S6nescence et arthrose. Revue duPraticien, 3, 2143-2146.

Buchthal, F., and Rosenfalck, A. (1966). Evoked actionpotentials and conduction velocity in human sensorynerves. Brain Research, 3, 1-122. Special Issue.

Burke, W. E., Tuttle, W. W., Thompson, C. W., Janney,C. D., and Weber, R. J. (1953). The relation of gripstrength and grip-strength endurance to age. Journal ofApplied Physiology, 5, 628-630.

Campbell, M. J., and McComas, A. J. (1970). The effects ofageing on muscle function. In Abstracts of 5th Symposiumon Current Research on Muscular Dystrophy and RelatedDiseases. London, 1970. Communications. Abstract No. 6.Muscular Dystrophy Group of Great Britain: London.

Cavanagh, J. B. (1964). Peripheral nerve changes in ortho-cresyl phosphate poisoning in the cat. Journal ofPathologyand Bacteriology, 87, 365-383.

Corbin, K. B., and Gardner, E. D. (1937). Decrease innumber of myelinated fibres in human spinal roots withage. Anatomical Record, 68, 63-74.

Downie, A. W., and Newell, D. J. (1961). Sensory nerveconduction in patients with diabetes mellitus and controls.Neurology (Minneapolis), 11, 876-882.

Dyck, P. J., Johnson, W. J., Lambert, E. H., and O'Brien,P. C. (1971). Segmental demyelination secondary to axonaldegeneration in uremic neuropathy. Mayo Clinic Proceed-ings, 46, 400-431.

Erlanger, J., and Gasser, H. S. (1937). Electrical Signs ofNervous Activity. University of Pennsylvania Press:Philadelphia.

Gardner, E. D. (1940). Decrease in human neurones with age.Anatomical Record, 77, 529-536.

Gutmann, E., and Hanzlikova, V. (1966). Motor unit in oldage. Nature, 209, 921-922.

Gutmann, E., Hanzlikova, V., and Vyskocil, F. (1971). Agechanges in cross striated muscle of the rat. Journal ofPhysiology, 216, 331-343.

Jennekens, F. G. I., Tomlinson, B. E., and Walton, J. N.(1971). Data on the distribution of fibre types in five humanlimb muscles. An autopsy study. Journal ofthe NeurologicalSciences, 14, 245-257.

Lascelles, R. G., and Thomas, P. K. (1966). Changes due toage in internodal length in the sural nerve in man. Journalof Neurology, Neurosurgery, and Psychiatry, 29, 40-44.

Lhermitte, J.-J. (1907). Etude sur lesParapligies des Vieillards.These de M6dicine, 1906-1907. No. 160. Paris.

McComas, A. J., Fawcett, P. R. W., Campbell, M. J., andSica, R. E. P. (1971a). Electrophysiological estimation ofthe number of motor units within a human muscle.Journal of Neurology, Neurosurgery, and Psychiatry, 34,121-131.

McComas, A. J., Sica, R. E. P., and Campbell, M. J. (1971b).'Sick' motoneurones. A unifying concept of muscledisease. Lancet, 1, 321-325.

McComas, A. J., Sica, R. E. P., Campbell, M. J., and Upton,A. R. M. (1971c). Functional compensation in partiallydenervated muscles. Journal of Neurology, Neurosurgery,and Psychiatry, 34, 453-460.

Mumenthaler, M. (1970). Myopathy in neuropathy. InMuscle Diseases. Proceedings ofan International Congress,Milan, 1969, pp. 585-598. Edited by J. N. Walton, N.Canal, and G. Scarlato. International Congress Series No.199. Excerpta Medica: Amsterdam.

Norris, A. H., Shock, N. W., and Wagman, I. H. (1953). Agechanges in the maximum conduction velocity of motorfibers ofhuman ulnar nerves. Journal ofAppliedPhysiology,5, 589-593.

Ochoa, J., and Mair, W. G. P. (1969). The normal suralnerve in man. 2. Changes in the axons and Schwann cellsdue to ageing. Acta Neuropathologica (Herlin), 13, 217-239.

Olson, C. B., and Swett, C. P., Jr. (1969). Speed of contrac-tion of skeletal muscle. The effect of hypoactivity andhyperactivity. Archives of Neurology (Chicago), 20, 263-270.

O'Sullivan, D. J., and Swallow, M. (1968). The fibre size andcontent of the radial and sural nerves. Journal ofNeurology,Neurosurgery, and Psychiatry. 31, 464-470.

Petersen, I., and Kugelberg, E. (1949). Duration and form ofaction potential in the normal human muscle. Journal ofNeurology, Neurosurgery, and Psychiatry, 12, 124-128.

Rubinstein, L. J. (1960). Aging changes in muscle. In theStructure and Function of Muscle. Vol. 3, pp. 209-226.Edited bv G. H. Bourne. Academic Press: New York.

Sacco, G., Buchthal, F., and Rosenfalck, P. (1962). Motorunit potentials at different ages. Archives of Neurology(Chicago), 6, 366-373.

Serratrice, G., Roux, H., and Aquaron, R. (1968). Proximalmuscular weakness in elderly subjects. Report of 12 cases.Journal of the Neurological Sciences, 7, 275-299.

Sica, R. E. P., and McComas, A. J. (1971). Fast and slowtwitch units in a human muscle. Journal of Neurology,Neurosurgery, and Psychiatry, 34, 113-120.

Swallow, M. (1966). Fibre size and content of the anteriortibial nerve of the foot. Journal ofNeurology, Neurosurgery,and Psychiatry, 29, 205-213.

Thomas, P. K., Sears, T. A., and Gilliatt, R. W. (1959). Therange of conduction velocity in normal motor nerve fibresto the small muscles of the hand and foot. Journal ofNeurology, Neurosurgery, and Psychiatry, 22, 175-181.

Tomlinson, B. E., Walton, J. N., and Rebeiz, J. J. (1969).The effects of ageing and of cachexia upon skeletal muscle.A histopathological study. Journal of the NeurologicalSciences, 9, 321-346.

Tuffery, A. R. (1971). Growth and degeneration of motorend-plates in normal cat hind limb muscles. Journal ofAnatomy, 110, 221-247.

Verzar, F. (1959). Muscular dystrophy in old age. Geronto-logia Clinica (Basel), 1, 41-51.

Vizoso, A. D. (1950). The relationship between internodallength and growth in human nerves. Journal of Anatomy,84, 342-353.

182

by copyright. on M

ay 21, 2021 by guest. Protected

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.2.174 on 1 April 1973. D

ownloaded from