Motor unit spike counts in the biceps brachii muscle were evaluated using a monopolar needle electrode during maintenance of antigravity posture with the elbow flexed to 45" before and afier maximal voluntary contraction (MVC). After MVC, the number of motor unit spikesisecond needed to maintain this posture was about 50% less than prior to MVC. Surface rectified integrated EMG activity declined in parallel to the decline in spike counts. To determine whether different muscles were compensating for the reduction in spike counts, synergistic muscles were examined simultaneously. Similar reductions were noted. In one deafferented subject, EMG activity increased rather than decreased after MVC. Fatigue decreases the contraction- relaxation rate of muscle fibers, which lowers fusion frequency. Thus, lower rates of motor unit activation can result in the maintenance of constant force. A feedback system from muscle to CNS likely senses this slowing and leads to the spike count reduction. Key words: motor unit maximal contraction fatigue spike counts
.
MUSCLE & NERVE
13~1146-1151 1990
MOTOR UNIT SPIKE COUNTS BEFORE AND AFTER MAXIMAL VOLUNTARY CONTRACTION JUDITH L. GOOCH, MD, BRUCE V. NEWTON, MD, and JACK H. PETAJAN, MD, PhD
During a maxinial voluntary contraction (MVC:), a s maximum force declines, the firing rate ot' nio-
toneurons l ' a l l s . ~ , 4 . ~ . ~ . i " 1 8 . Specifically, 1~ in the adductor pollicis muscle using Tungsten niicroelectrodes, mean rates wcrc noted to fall from about 27 €12 to 15 Hz in the first 60 secoiicls of c ~ n t r a c t i o n Fatipic .~ has also been shown to prolong the active state ol' the rnuscle, decreasing muscle contractile speed and relaxation rate.5, 10,15,1 G , 2 2 , 2 3 It has been found that the percentage change in EM(; activity approximately equals the percentage change in muscle c(m1ractile speed (50%)to Slowing of muscle fiber- c o i l tractile speed allows fewer motor unit spikes per second to maintain a specific level of force. It has
From the University of Utah Health Sciences Center, Physical Medicine and Rehabilitation, Salt Lake City, Utah Acknowledgments. The authors would like to thank Sharon Schuur for preparation of the manuscript. Presented in part at annual American Association of Electrornyography and Electrodiagnosis Meeting, San Diego, California, October, 1988 Address reprint requests to Judith L Gooch, MD, University of Utah Health Sciences Center, Physical Medicine and Rehabilitation, 50 North Medical Drive. Salt Lake City, UT 84132 Accepted for publication January 9, 1990 CCC 0148-639X190101201146-06 $04.00 0 1990 John Wiley & Sons, Inc.
1146
Motor Unit Spike Counts
been suggested that during M V C the CNS ariapis the moti)i. unit firing I'requency t o the contractile state of the muscle t o maintain optimal tension development and possil~ly avoid fitiliire of iieuroin iiscu Iar trans r n issioi 1. .', "*") ?I.tie pi.esence of ;i fketlback mec1i;inisrn requirinq scrisory iripiii from the rnusc1cs t o t tlc CNS is & i y . : ~ ~ l + ~ ~ -. I he studies mentioned were performed during M V C wliile force was declining. Measurement of 1iiotc)t' u n i t spike coiints is clil'licult at high lcvcls of foixe due t o the I;ii.ge number of motor units fi r-i rig. It has been shown that antigravity posture can t x used a s ;i specific lcvcl of iiiiicrvation, representing minimal efi'ort, with motor unit firing rate set A about t h l of pliysiological tremor ( I 0 €37 to 1 3 Hz)." T h e rriaiiileiinrice of this postui-e recluires ~ h caclivatiori of' ;t ~ ~ I f ) ~ ) ( ' p l ~ I ;of' ~ t i low oi~ ~ l m d i o l dsrnall , ( t y p e I ) ~iio~oiieuroiis. 111 a s n i w h a s ttic s i x of tlie priniar) agonist muscle is "atljusled" to the size of' the limb, it is cxpccted that 1 . 1 1 ~relative size of the motoneuron population required t o iiiiiiiitiiiii a given posture will he nearly
'
Mairiteiiaiice of antigravity posture during 45" joint flexion, whilc i-et:ortling from the pi-imary agonist riiiisclc, has txeri used to provide ;I srahlc, miiiimal level of innewition during ncctilc elec-
MUSCLE & NERVE
December 1990
trotle recording in iiorriial sit tijects and thosc with orciers of' the motor. iiriit.'.' Initial otxct-wtions using this technique havc I-evei~letlthat after ;I maxiinal voliiiitary contraction, i i i o t o r uiiit spikes per- srt:oncl decline. To further evaluate this phenonietioii, tiiotor unit spikc (:outits were ~iie:is~iIccl 1xFore arid after fatiguing contractioiis in rii;iny srrhjects. To tletermiiie whethcx- other riiusclcs increased spike counts conipcnsntirig for the noted reduction, synergistic miiscles w e r e also stuclicrl. Because n rtretlle electrode tiieasriies elrctricd a(.tivity in a v e r y small area of' thc mirscle, experiments WCI-C repealed with siir-l'~ice-i.cc.tifie~l integrated EMC; (RlEMG) rnonitoi-ing. Also, t o further explore tlic prol~osed fcctlhack ittechanisrn, studies were perfor-rned o ~ one i siihject with clcaf fc 1 - e 11t nt ic )I 1. MATERIALS AND METHODS
Sutiects w c ~ ehealthy voluritciers. Iriformed coiiscnt was ol.)tainecl from e;ic:h sul,jcc:t atid studies wci-e appr0vt:d hy h e 1nstitution;il Kcview Board at tlic finiversity 01' Utah Healtli S c erices Center.
Subjects.
I n 2 I subjects, using a monopolmrieedle elcctrocle, niotor unit spike c:oi.ints were rileastired in the liceps brachii muscla with the subject supine anel tlic: ;itmi sitpinatctl ( : o t n i i ~ t a l ) l y and elhow flexctl t o 15".l'he needle electrotle was ;icl,jiisteci to record the iiiaxitiiiirii itumbci- ol' mot o r iiiiit ; d o n poteiitials. Fhch subject suhseqiirrilly perfornied a 1 i i i i n MVC of' the Iiiceps br-acliii with the cll)ow reniaiiiing flexed at 45". M;inual resistance was pi-ovitlcd h y the examiner arid frequent verbal ciicouragcrrieiit was given. Immecliately f'ollowing contraction, with the subject reniaiiiing iii the origin;il position, spike c o u n t s w e r e again measui-crl. 'Ttic. arm was nlluwctl t o rest i r i the 45" position atid iiio101-units wcr-e agiiii recordctl at 30 scc intervals Lip to 90 sc'c (the duration of each recording 1x1-iod ranged from 5- 10 scc). To st;iridaidizt: the degree o f fatigue, 14 siihjects were tested using a i'orce loss t o 50%)MV(; rallirr than using 1 i i t i t t ol' M V C . To evaliiatc iiiiiscles syncrgislic to biceps hrachii, niolor u n i t spike counts wcr-e mcasiirecl simullarieously in the liccl)s l)i.iicIiii, 1~i~ac:hi;rlis. and 1,r;ic-hioradialis iiiuscles of' 11 siil+x:ts. In 3 subject s, spike counts we re riicasi i red si m u Itaneou sly in tlif'f'ereiit areas (midcllc, proxiriial, distal) of t l i c biceps brachii in iiscle. In the biceps l.)txhii rnitscle of 7 siiljccts, using a myoclectric s i p i l rnonitor (Motioii Control,
Protocols.
Motor Unit Spike Coiints
Salt 1,:ike City, LJT) hefoi-c arid al'ter ;i 1 min MV(:, siirl'acc KIEMG was measurctl in atltlition t o i i i o t o r unit spikcs/seconrl. The output scale of' t.his devicc: records millivolts (0- 100 mV). 'I'he output reading dcpcncls Lipon the gain settiiig. I n t h i s experiment, piior- t o M V C , the gaiii was adj u s t c d suc-11 that the oi11p i i t scalc 1-end 50 mV with the arm cutrifortnbly supinated and the elbow flcxcd to 45". 'l'liis periiiitted changes to hc rriost r(,a rl i 1y I i ()I e d . K e coI-d i t i gs (11' 11ot 11 su 1- face a t 1 d needle EMG activity wcre tiiatle a t 1 miit, 2 i i i i r i , 3 min, 5 m i n , a r i d continuing at 5 inin iiitcrvals u p to 30 min "dowing M V C . Recordings were niatle ;it 5 inin iiitci-vals Lip to (i0 r i i i r i al'ter M V C in 1 suhject. K I E M G xctivity befoi-e and af'tei. a 1 min M V ( ; was 1iie;tsurctl in onc su1,ject with cleaffci-entation m t l tiornial strcngth. EMG Recordings. ' f h e nuniher of motor unit spikes per second was measured iisirig ;I 26-gauge nioiiopolar clectimle itiscrtecl into the Iiiiti-Iielly &f tlic inusclc. M o t o r i i n i t poteiihls wertr evalualecl ;it a sensitivity ol' 200-500 p V per divisiort, with amplifier filter settings of 20 I l z and 10 l i f I z using ;I 'l.l-:LA 'l'E42 clectrotiiyogr-~i~~li (I'leasarilville, N Y ) . Recordings wei-e riiade usiiig a T E C A fiheroptic recorder a t a specd of 50 cm/sec. Siir-l';icerectified integrated EMG (KIEMG) ;ictivity in the liccps brxliii miiscle w;is rneasurc.d using ;I Myolal) I I ittyogtaphic signal rnonitor (Moliori Chnlrol, Salt Lake City, LJT).'l'he prcamplilier of the Myolab was t;ipul securely to tlic mid-t x l I y c) I' the rii ti sclc. Force Measurements. 1)iiririg experiments using ;I Lbrcc loss t o 50% MVC:, rnaximal isometric: elhow
flcxion force was rrieasured iising a11 isometric fijrce traiistliicer (Interface, Motlcl SM-250, Scottsdalc, A%) a i i t l r.ecortlei- (Linseis, Model L M M , Princeton J i l t d o n , N.1). An inelastic: strap ;irouritl the wrist w a s attached to the strain 1?;"uge.
T h e riurriber of spikcs/second was deterrninctl I.i.orn the recordings. 0 1 i l y motor units greater ttiari 200 p V peak-to-peak ;implitiicle were counted. Motor uriits of' less t h a i i 200 pV a t 500 pV/division arc n o t well focused (I-isc time is olirn greater than 0.3 riisec) ;irid can be c:nsily coiifiised with baseline Iluctiratioris arising f'i-on1nrlifact. M o l o r unit spikes/sccontl and K I E M G activity ai-e reported ;IS rrieatis a n d slmtlard deviations. 1 he signific:anc:e of differences between m c i n values b e h - c a r i d after MVC; was evaluated using Studcnt's L test. Thc correlation coel'ficierit be-
Data Analysis.
r 7
MUSCLE & NERVE
December 1990
1147
tween K1 EM(; activity m d spikcs/sccond was also calculated. T h e P < 0.05 level was choscn for significance.
RESULTS
Spike Counts in Biceps Brachii Before and After MVC.
Measurement of' M U A P spikes/second with the clbow flexed to 45" before and atter a 1 niiii MVC revealed thal prior to MVC the rrieaii spikcs pcr second was 48.3 k 18.7. Irnmcdiately following MVC, the means spikes per second was 18.0 ? 12.8 ( P < 0.001) (Fig. 1). Motor unit [-iring r a m ww-e ;rppr-oxiinately I0 Hz. The reduction in spikes per second persisted for h e 90 sec during which it was monitored. Measureiiieiit of spikes/ second before and aftcr an MVC; resulting in a 50% loss of foce, showed that prior to contraction, mean spikes per second was 46.3 2 21.0 arid immediately after contraction it was 23.8 ? 23.1 ( P < 0.01). Figure 2 illustrates a typical record of the rcciuction in motor unit spike counts. Prior to MVC the mean spikes per sccoritl was 48.9 10.3 in the biceps and (57.8 29.0 in thc brachialis. Following a 1 niin MVC, the mean spikes pcr sccond was 24.1 17.9 ( P < 0.001) in the k c p s and 50.6 i 28.9 ( P < 0.05) in the 1)rat:hi;tlis (Fig. 3 ) .
Spike Counts in Synergistic Muscles.
*
*
*
70
I
T
1
I
I
4 I i
100 msec
toopv
FIGURE 2. Typical record of motor units recording from the biceps brachii muscle during 45" of elbow flexion before and after MVC.
This was ;i reduction of 5. 1%) in thc hiccps ;ind 25% in brachialis. T h e hrachioradialis muscle s h o w e d no motor unit activity in any of the subjccts before or after MVC. All locatioiis in the hiceps bracliii muscle revealed a rctiuction in spike coiin~safter MVC. Needle EMG vs. Surface EMG: Duration of Reduction.
Prior LO MVC, the Mycllab gain wits adjusted such tli;it the output scale read 50 m V with the elbow flexed to 45". Mean spikes per secoiid prior LO MVC was 62.7 i 42.2. Imnietlia~ely following MVC, the RIEMG activity I ~ I to I 28 2 ~3 r n and ~ the spikes per scconcl tell to 27.0 & 25.9, a dccrease of 44%) in KIEM(; a i d 5774 in spikes per scconti. After 30 rnin, thc mc;iii RlEMG activity was 41 2 23 iriV (lS%, less than pre-fatigue) (Fig. 4). Tfic rate 01 recovery varied greatly. KIEMG
50 0
2
40 fn
-.
'O0TT-----
;30 a
VI
20
Y,
,
w
v)
Y
10
40
n UI
20
0
pre-
0
30
60
90
contract ion
T I M E AFTER CONTRACTION ( s e c )
FIGURE 1. Mean motor unit spikes/second of 21 normal subjects recording from the biceps brachii muscles during 45" of elbow flexion. Recordings were made prior to a 60 sec MVC, immediately after MVC, and at 30 sec intervals after MVC.
1148
Motor Unit Spike Counts
0 pre contraction
0-45
60-90
TIME AFTER CONTRACTION (set)
FIGURE 3. Mean motor unit spikes/second recording from the biceps brachii and brachialis muscles during 45" of elbow flexion. Recordings were made prior to MVC and from 0-45 sec and 60-90 sec after MVC.
MUSCLE & NERVE
December 1990
- 60
I\ I
T
9
n
-50
0
-40
s
I
0
-30
5[r
- 20 -10
COntrOCtlOn
TIME AFTER CONTRACTION (min)
FIGURE 4. Comparison of mean motor unit spikes/second and surface RIEMG activity recording from the biceps brachii muscle during 45" of elbow flexion before and for 30 rnin following MVC.
and needle EMG activity ret ui-nctl t o prc-l'atigue levcls b y 10 min in 1 subject.. In ;inother suliject who was followed for 60 rnin, EMG activity did not return to prc-fatigue levels. Comparison of each subject's surface EMG activity to motor unit spike counts rcvealetl a high correlation ( r 0.74-0.99) (I' < 0.01).
=
111 the one subject with deafferentation and not-ma1 strength, the Myolab output increased frorri 15 iriV t o 30 iriV following a 1 min M V C . Ohcrvatioti revealed t h a ~after the contraction the sulijjcct h a d a p-onouncetl tericleiicy to flex the elbow, and had rnuch dil'liciilty holding the a r m in the 45" p s i tion.
Surface EMG in Deafferented Subject.
DISCUSSION
After a fatiguing muscle contraction, fewer motor unit spikes 1x1second are rieeded to niairitain ii certain low level of force than before the contraction. 'I'he surfxx RI EMG declines in par;tllel to the decrease i n MUAP spike count and this el'fect has been rioted to last u p to an lioui.. Orie rriiglit hypothesize that nio tor i i n its in sy ne rgist ic rii uscles or in diffkrcnt. ar-cas of' the biccps hracliii muscle show an incrcasc in spike counts, conipensating for the fatigued tested area. However, a significarit rcdiictiori iri spikes per second was found to occur iri rrtultiple regions of the biceps muscle and in other arm flexors. In addition, surface RIEMG decliried in p:irallel to the decrease iii spike counts. 'This measures electrical activily i l l ;I
Motor Unit SDlke Counts
gi-eatcr arcs of the muscle tliari the needle clcctrocle. ni us, other areas of muscle do not appear t o lie cornpensating for more fatigued portion. This implies that after fatigue, force can be maiiit;iiricd with a lower level of' excitation than before fatigue. I n exploring t liis plienoriienon, it is irnportant to i i o t c t h a t during MVC motor rieuron firing rates sccni t o match the concractile s p e d o f the muscle.1-5 * 11.12 Bellemare has found that during an MVC, meail motor unit firing rates i n biceps 1)racIiii ;md adductor pollicis muscles were 3 1 .I Hz and 29.9 Hz, respecLively while for the soleus niuscle they were only 10.7 H7. The twitch coritractinii arid half rclaxation times were a l ~ o u 50%> t longer for the s01c:us muscle t h a n for tlic hiceps arid adductor pollicis muscles.' Slowing of' muscle coritractilc spccd and relaxuliOrl rate does occur during f ~ t i g u e . ~ . ~ ( ) . l ~ ' . ~ ( j . ~ ~ , ~
Specifically, Bigland-Ritchie has s h o w n that after a 60 sec MVC, muscle twitch relaxation time was prolonged by about 50'r." Slowing of contractile speed allows fewcii- motor unit spikes per set:ond to maintain a spccific level 01 l'otw. Iri our experiments, motor i i n i t spikr counts and surlace LMG activity nccticd t o maintairi aiitigr;tvity posture (45" elbow flexion) decrcasctl by ahout 50%, with fatigue. 'l'his is in agrccmcnt with the studies of Bigland-Kitchie, which also showed 5OC;i: I-eduction in E M G activity during a (50 sc(' MVC.4*15 'Thus, it is likely that. our notccl reduction in n101or unit spike couriLs at a spccilic level of forcc is rclaced to the slowing of muscle relaxation rate associated with fatigue. Tlie reduction in motoi- unit spike counts associated with fatigue likely henelits the niotoisystel,l~2.53 13.20 High frequency stirnulation of a motor ncrw results in failure o f electrical p r o p gation at the rnuscle fiber membraric. Although still a su1,jec:t of controversy, volulitary contixctioris do not seem ID cause this failure of ~iropngatioii."'~~)~~' Also, high frequency stiiiiulatiori of ;I iiervc rcsults in a faster rate of force loss tkiaii secn in MVC.",'' rills, the reciuction in iuotor unit spike counts m a y protect the musclc fro111 pcriphcral I'ailutr. Also, if' this ptierioiiieiion did not occur, somc degree o f iriotoi- c : o i i t r o l may he lost. Biglanci-Ritchie has suggcstctl that i l ' iiiotor units fire above the frcqiwticy needed duriii MVC, then a large rcduc:tion in neur~al rat wou~ttIK necessary to reduce force.'" Slowing of contractile speed allows fi-wcr motor- u n i t spikes per second to maintain a specific levcl 01' l'orce. Tlir preserice of ii feedback nicckia-
MUSCLE & NERVE
December 1990
1149
nism that allows the CNS to perceive fatigue and/ or contractile slowing and to dccrease motor unit spike counts accordingly has been proposed. Hannerz has shown that when the all'erent iriflow from the muscle is decreased by compression of a nerve, an increase in firing rate of some motor units occurs." Also, Bigland-Kitchie has shown that in fatigued muscle, no recovery of firing rates occurs after 3 min of rest if the muscle is kept ischemic, but near full recovcry occurs 3 min after the blood supply is restored. This suggests the presence of a periptcral reflex from the fatigued muscle rather than a change in central motor neuron excitability (which would show recovery under both conditions).' Also the lack ol' recovery during ischemia is not likcly related to failure of neuromuscular transmission or impaircd subject cff0rt.l' For well-motivated subjects, twitch occlusion showed no reduction in the dcgree to which the muscles could be activated voluntarily during MVC after 3 min of ischemic rest. Also, no reduction in M wave amplitude was noted. Our deafferented suqjcct did not show the usual reduction in RIEMG aclivity. In fact, after a fatiguing contraction, an increase was noted. Also interesting was the tendency for this subject's arm to flex involuntarily after MVC. If more t.han the optimum number of motor units were firing, force would increase and tend to flex the arm. The findings in this subjcct suggest that the af'ferent system is necessary to achieve the normal de-
cliric in electrical activity. €lowever, as the reduction in electrical activity is variable even in normal sutjects, study of additional deafferented subjects would be necessary to confirrri this. We have found that after a fatiguing contraction fewer motor unit spikes per second are needed to maintain a certain low level of force than before fatigue. These findings are in accord with those of prcviaus studies which documcnted a reduction in motor unit firing rates during MVC.3,4.5+8+12,.".19However, thcse previous studies were performed at high levcls of force while I'orce was declining. Ours werc performed at a constant and minimal pre- and post-fatigue level of innervat.ion. 'To standardize this level of innervation, we have used the antigravity posture during 45" ofjoint flexion. This low level of force allows one to easily count the motor units, avoiding the difficulties encountered with counting units at high levels of force. Also, the use of antigravity posture is a very simple method of standardizing innervation level. Limitations include the fact that with a nionopolar electrode, orily one sinall area of the muscle is evaluated. Also, due to niovenicnt of the muscle during MVC, the exact same motor units are not likcly to be evaluated before and after the contract.ion. However, as the rcduction in spikes per second is noted in all areas of the muscle, it is not necessary to monitor one specific motor unit to document changes in overall spike COutltS.
REFERENCES 1. Bellemarc I;, Wocds JJ, Jolra~isson K, Ri larid-Kitchie R:
Motor-unit discharge rates in maximal vifuntaty contractioris of three human musclcs. J Neuuu&&l 1985; 50:1580- 1392.
2. Bighnd-Kitchic B: EMC and fatiguc of human voluntary arid stimulated contractions, in Porter K. Whelan J (eds): Cibu Ikundatiun Synposium 82,Hunicm Mu.wL FaiKup; Yhyviulugical Meclmtisms. London, Pi~niaoMedical, I98 I , pp 130-148.
3. Bigland-Kitchie RR, Uawson NJ,Johansson KS, I.ippold OCT: Reflex origin for the slowing of rnotoneuron firing rates in fatigue of human voluntary contractions. J Plysiul 1986;379:453 -459. 4. Rigland-Kitchie A, Jolransson
R, Lippold OCJ. Smith S, Wocds JJ: Changes in motoncuron firing ratcs during sus-
tained maximal voluntary contractions. J Phyiul 1983; 340:335-346.
5. Bigland-Ritchic B,Johansson R, Lippold OCJ, Woods JJ: Contractile spccd arid EMC changcs during fatiguc of sustained maximal voluntary contractions. J Noirot&siur( I983;50:313- 321. 6. Bigland-Ritchie B, Joncs CIA, Wcmds JJ: Excitation lrequency and muscle fatigue: electrical responses duririg humaxi voluntary and stimulated coniractions. h$ Neuuul I979;64:414-427. 7. Bigland-Ritchie B, Kukulka CG, Lippolrl OCJ,
1150
Motor Unit Spike Counts
Woods JJ:
Thc absence of ncurorriuscular transmission failure in sustained maximal voltiulary contractions. .) Physiui 1982; 530:265-278.
8. ISigland-Kitchie I%,Lippold OCJ: Changes in musclc activation during prolonged niaxirnal voluntary contractions J Phy.vid 1979;292: 141'- 15Y. 9. Ihchaicau J, HainauL K: Electrical and mechanical failure during sustained and intermittent contractions in humans. J ~ p p ritpbi i 19a~;.w~2-945. 10. Edwards RH'I', llill DK,Joncs DA: Mclahlic changcs associated with thc slowing of r e h a t i o n in fatigued mouse rnusc1e.J Phyiul 197.5;45 1:287-301. 11.
.
MUSCLE & NERVE
13~1146-1151 1990
MOTOR UNIT SPIKE COUNTS BEFORE AND AFTER MAXIMAL VOLUNTARY CONTRACTION JUDITH L. GOOCH, MD, BRUCE V. NEWTON, MD, and JACK H. PETAJAN, MD, PhD
During a maxinial voluntary contraction (MVC:), a s maximum force declines, the firing rate ot' nio-
toneurons l ' a l l s . ~ , 4 . ~ . ~ . i " 1 8 . Specifically, 1~ in the adductor pollicis muscle using Tungsten niicroelectrodes, mean rates wcrc noted to fall from about 27 €12 to 15 Hz in the first 60 secoiicls of c ~ n t r a c t i o n Fatipic .~ has also been shown to prolong the active state ol' the rnuscle, decreasing muscle contractile speed and relaxation rate.5, 10,15,1 G , 2 2 , 2 3 It has been found that the percentage change in EM(; activity approximately equals the percentage change in muscle c(m1ractile speed (50%)to Slowing of muscle fiber- c o i l tractile speed allows fewer motor unit spikes per second to maintain a specific level of force. It has
From the University of Utah Health Sciences Center, Physical Medicine and Rehabilitation, Salt Lake City, Utah Acknowledgments. The authors would like to thank Sharon Schuur for preparation of the manuscript. Presented in part at annual American Association of Electrornyography and Electrodiagnosis Meeting, San Diego, California, October, 1988 Address reprint requests to Judith L Gooch, MD, University of Utah Health Sciences Center, Physical Medicine and Rehabilitation, 50 North Medical Drive. Salt Lake City, UT 84132 Accepted for publication January 9, 1990 CCC 0148-639X190101201146-06 $04.00 0 1990 John Wiley & Sons, Inc.
1146
Motor Unit Spike Counts
been suggested that during M V C the CNS ariapis the moti)i. unit firing I'requency t o the contractile state of the muscle t o maintain optimal tension development and possil~ly avoid fitiliire of iieuroin iiscu Iar trans r n issioi 1. .', "*") ?I.tie pi.esence of ;i fketlback mec1i;inisrn requirinq scrisory iripiii from the rnusc1cs t o t tlc CNS is & i y . : ~ ~ l + ~ ~ -. I he studies mentioned were performed during M V C wliile force was declining. Measurement of 1iiotc)t' u n i t spike coiints is clil'licult at high lcvcls of foixe due t o the I;ii.ge number of motor units fi r-i rig. It has been shown that antigravity posture can t x used a s ;i specific lcvcl of iiiiicrvation, representing minimal efi'ort, with motor unit firing rate set A about t h l of pliysiological tremor ( I 0 €37 to 1 3 Hz)." T h e rriaiiileiinrice of this postui-e recluires ~ h caclivatiori of' ;t ~ ~ I f ) ~ ) ( ' p l ~ I ;of' ~ t i low oi~ ~ l m d i o l dsrnall , ( t y p e I ) ~iio~oiieuroiis. 111 a s n i w h a s ttic s i x of tlie priniar) agonist muscle is "atljusled" to the size of' the limb, it is cxpccted that 1 . 1 1 ~relative size of the motoneuron population required t o iiiiiiiitiiiii a given posture will he nearly
'
Mairiteiiaiice of antigravity posture during 45" joint flexion, whilc i-et:ortling from the pi-imary agonist riiiisclc, has txeri used to provide ;I srahlc, miiiimal level of innewition during ncctilc elec-
MUSCLE & NERVE
December 1990
trotle recording in iiorriial sit tijects and thosc with orciers of' the motor. iiriit.'.' Initial otxct-wtions using this technique havc I-evei~letlthat after ;I maxiinal voliiiitary contraction, i i i o t o r uiiit spikes per- srt:oncl decline. To further evaluate this phenonietioii, tiiotor unit spikc (:outits were ~iie:is~iIccl 1xFore arid after fatiguing contractioiis in rii;iny srrhjects. To tletermiiie whethcx- other riiusclcs increased spike counts conipcnsntirig for the noted reduction, synergistic miiscles w e r e also stuclicrl. Because n rtretlle electrode tiieasriies elrctricd a(.tivity in a v e r y small area of' thc mirscle, experiments WCI-C repealed with siir-l'~ice-i.cc.tifie~l integrated EMC; (RlEMG) rnonitoi-ing. Also, t o further explore tlic prol~osed fcctlhack ittechanisrn, studies were perfor-rned o ~ one i siihject with clcaf fc 1 - e 11t nt ic )I 1. MATERIALS AND METHODS
Sutiects w c ~ ehealthy voluritciers. Iriformed coiiscnt was ol.)tainecl from e;ic:h sul,jcc:t atid studies wci-e appr0vt:d hy h e 1nstitution;il Kcview Board at tlic finiversity 01' Utah Healtli S c erices Center.
Subjects.
I n 2 I subjects, using a monopolmrieedle elcctrocle, niotor unit spike c:oi.ints were rileastired in the liceps brachii muscla with the subject supine anel tlic: ;itmi sitpinatctl ( : o t n i i ~ t a l ) l y and elhow flexctl t o 15".l'he needle electrotle was ;icl,jiisteci to record the iiiaxitiiiirii itumbci- ol' mot o r iiiiit ; d o n poteiitials. Fhch subject suhseqiirrilly perfornied a 1 i i i i n MVC of' the Iiiceps br-acliii with the cll)ow reniaiiiing flexed at 45". M;inual resistance was pi-ovitlcd h y the examiner arid frequent verbal ciicouragcrrieiit was given. Immecliately f'ollowing contraction, with the subject reniaiiiing iii the origin;il position, spike c o u n t s w e r e again measui-crl. 'Ttic. arm was nlluwctl t o rest i r i the 45" position atid iiio101-units wcr-e agiiii recordctl at 30 scc intervals Lip to 90 sc'c (the duration of each recording 1x1-iod ranged from 5- 10 scc). To st;iridaidizt: the degree o f fatigue, 14 siihjects were tested using a i'orce loss t o 50%)MV(; rallirr than using 1 i i t i t t ol' M V C . To evaliiatc iiiiiscles syncrgislic to biceps hrachii, niolor u n i t spike counts wcr-e mcasiirecl simullarieously in the liccl)s l)i.iicIiii, 1~i~ac:hi;rlis. and 1,r;ic-hioradialis iiiuscles of' 11 siil+x:ts. In 3 subject s, spike counts we re riicasi i red si m u Itaneou sly in tlif'f'ereiit areas (midcllc, proxiriial, distal) of t l i c biceps brachii in iiscle. In the biceps l.)txhii rnitscle of 7 siiljccts, using a myoclectric s i p i l rnonitor (Motioii Control,
Protocols.
Motor Unit Spike Coiints
Salt 1,:ike City, LJT) hefoi-c arid al'ter ;i 1 min MV(:, siirl'acc KIEMG was measurctl in atltlition t o i i i o t o r unit spikcs/seconrl. The output scale of' t.his devicc: records millivolts (0- 100 mV). 'I'he output reading dcpcncls Lipon the gain settiiig. I n t h i s experiment, piior- t o M V C , the gaiii was adj u s t c d suc-11 that the oi11p i i t scalc 1-end 50 mV with the arm cutrifortnbly supinated and the elbow flcxcd to 45". 'l'liis periiiitted changes to hc rriost r(,a rl i 1y I i ()I e d . K e coI-d i t i gs (11' 11ot 11 su 1- face a t 1 d needle EMG activity wcre tiiatle a t 1 miit, 2 i i i i r i , 3 min, 5 m i n , a r i d continuing at 5 inin iiitcrvals u p to 30 min "dowing M V C . Recordings were niatle ;it 5 inin iiitci-vals Lip to (i0 r i i i r i al'ter M V C in 1 suhject. K I E M G xctivity befoi-e and af'tei. a 1 min M V ( ; was 1iie;tsurctl in onc su1,ject with cleaffci-entation m t l tiornial strcngth. EMG Recordings. ' f h e nuniher of motor unit spikes per second was measured iisirig ;I 26-gauge nioiiopolar clectimle itiscrtecl into the Iiiiti-Iielly &f tlic inusclc. M o t o r i i n i t poteiihls wertr evalualecl ;it a sensitivity ol' 200-500 p V per divisiort, with amplifier filter settings of 20 I l z and 10 l i f I z using ;I 'l.l-:LA 'l'E42 clectrotiiyogr-~i~~li (I'leasarilville, N Y ) . Recordings wei-e riiade usiiig a T E C A fiheroptic recorder a t a specd of 50 cm/sec. Siir-l';icerectified integrated EMG (KIEMG) ;ictivity in the liccps brxliii miiscle w;is rneasurc.d using ;I Myolal) I I ittyogtaphic signal rnonitor (Moliori Chnlrol, Salt Lake City, LJT).'l'he prcamplilier of the Myolab was t;ipul securely to tlic mid-t x l I y c) I' the rii ti sclc. Force Measurements. 1)iiririg experiments using ;I Lbrcc loss t o 50% MVC:, rnaximal isometric: elhow
flcxion force was rrieasured iising a11 isometric fijrce traiistliicer (Interface, Motlcl SM-250, Scottsdalc, A%) a i i t l r.ecortlei- (Linseis, Model L M M , Princeton J i l t d o n , N.1). An inelastic: strap ;irouritl the wrist w a s attached to the strain 1?;"uge.
T h e riurriber of spikcs/second was deterrninctl I.i.orn the recordings. 0 1 i l y motor units greater ttiari 200 p V peak-to-peak ;implitiicle were counted. Motor uriits of' less t h a i i 200 pV a t 500 pV/division arc n o t well focused (I-isc time is olirn greater than 0.3 riisec) ;irid can be c:nsily coiifiised with baseline Iluctiratioris arising f'i-on1nrlifact. M o l o r unit spikes/sccontl and K I E M G activity ai-e reported ;IS rrieatis a n d slmtlard deviations. 1 he signific:anc:e of differences between m c i n values b e h - c a r i d after MVC; was evaluated using Studcnt's L test. Thc correlation coel'ficierit be-
Data Analysis.
r 7
MUSCLE & NERVE
December 1990
1147
tween K1 EM(; activity m d spikcs/sccond was also calculated. T h e P < 0.05 level was choscn for significance.
RESULTS
Spike Counts in Biceps Brachii Before and After MVC.
Measurement of' M U A P spikes/second with the clbow flexed to 45" before and atter a 1 niiii MVC revealed thal prior to MVC the rrieaii spikcs pcr second was 48.3 k 18.7. Irnmcdiately following MVC, the means spikes per second was 18.0 ? 12.8 ( P < 0.001) (Fig. 1). Motor unit [-iring r a m ww-e ;rppr-oxiinately I0 Hz. The reduction in spikes per second persisted for h e 90 sec during which it was monitored. Measureiiieiit of spikes/ second before and aftcr an MVC; resulting in a 50% loss of foce, showed that prior to contraction, mean spikes per second was 46.3 2 21.0 arid immediately after contraction it was 23.8 ? 23.1 ( P < 0.01). Figure 2 illustrates a typical record of the rcciuction in motor unit spike counts. Prior to MVC the mean spikes per sccoritl was 48.9 10.3 in the biceps and (57.8 29.0 in thc brachialis. Following a 1 niin MVC, the mean spikes pcr sccond was 24.1 17.9 ( P < 0.001) in the k c p s and 50.6 i 28.9 ( P < 0.05) in the 1)rat:hi;tlis (Fig. 3 ) .
Spike Counts in Synergistic Muscles.
*
*
*
70
I
T
1
I
I
4 I i
100 msec
toopv
FIGURE 2. Typical record of motor units recording from the biceps brachii muscle during 45" of elbow flexion before and after MVC.
This was ;i reduction of 5. 1%) in thc hiccps ;ind 25% in brachialis. T h e hrachioradialis muscle s h o w e d no motor unit activity in any of the subjccts before or after MVC. All locatioiis in the hiceps bracliii muscle revealed a rctiuction in spike coiin~safter MVC. Needle EMG vs. Surface EMG: Duration of Reduction.
Prior LO MVC, the Mycllab gain wits adjusted such tli;it the output scale read 50 m V with the elbow flexed to 45". Mean spikes per secoiid prior LO MVC was 62.7 i 42.2. Imnietlia~ely following MVC, the RIEMG activity I ~ I to I 28 2 ~3 r n and ~ the spikes per scconcl tell to 27.0 & 25.9, a dccrease of 44%) in KIEM(; a i d 5774 in spikes per scconti. After 30 rnin, thc mc;iii RlEMG activity was 41 2 23 iriV (lS%, less than pre-fatigue) (Fig. 4). Tfic rate 01 recovery varied greatly. KIEMG
50 0
2
40 fn
-.
'O0TT-----
;30 a
VI
20
Y,
,
w
v)
Y
10
40
n UI
20
0
pre-
0
30
60
90
contract ion
T I M E AFTER CONTRACTION ( s e c )
FIGURE 1. Mean motor unit spikes/second of 21 normal subjects recording from the biceps brachii muscles during 45" of elbow flexion. Recordings were made prior to a 60 sec MVC, immediately after MVC, and at 30 sec intervals after MVC.
1148
Motor Unit Spike Counts
0 pre contraction
0-45
60-90
TIME AFTER CONTRACTION (set)
FIGURE 3. Mean motor unit spikes/second recording from the biceps brachii and brachialis muscles during 45" of elbow flexion. Recordings were made prior to MVC and from 0-45 sec and 60-90 sec after MVC.
MUSCLE & NERVE
December 1990
- 60
I\ I
T
9
n
-50
0
-40
s
I
0
-30
5[r
- 20 -10
COntrOCtlOn
TIME AFTER CONTRACTION (min)
FIGURE 4. Comparison of mean motor unit spikes/second and surface RIEMG activity recording from the biceps brachii muscle during 45" of elbow flexion before and for 30 rnin following MVC.
and needle EMG activity ret ui-nctl t o prc-l'atigue levcls b y 10 min in 1 subject.. In ;inother suliject who was followed for 60 rnin, EMG activity did not return to prc-fatigue levels. Comparison of each subject's surface EMG activity to motor unit spike counts rcvealetl a high correlation ( r 0.74-0.99) (I' < 0.01).
=
111 the one subject with deafferentation and not-ma1 strength, the Myolab output increased frorri 15 iriV t o 30 iriV following a 1 min M V C . Ohcrvatioti revealed t h a ~after the contraction the sulijjcct h a d a p-onouncetl tericleiicy to flex the elbow, and had rnuch dil'liciilty holding the a r m in the 45" p s i tion.
Surface EMG in Deafferented Subject.
DISCUSSION
After a fatiguing muscle contraction, fewer motor unit spikes 1x1second are rieeded to niairitain ii certain low level of force than before the contraction. 'I'he surfxx RI EMG declines in par;tllel to the decrease i n MUAP spike count and this el'fect has been rioted to last u p to an lioui.. Orie rriiglit hypothesize that nio tor i i n its in sy ne rgist ic rii uscles or in diffkrcnt. ar-cas of' the biccps hracliii muscle show an incrcasc in spike counts, conipensating for the fatigued tested area. However, a significarit rcdiictiori iri spikes per second was found to occur iri rrtultiple regions of the biceps muscle and in other arm flexors. In addition, surface RIEMG decliried in p:irallel to the decrease iii spike counts. 'This measures electrical activily i l l ;I
Motor Unit SDlke Counts
gi-eatcr arcs of the muscle tliari the needle clcctrocle. ni us, other areas of muscle do not appear t o lie cornpensating for more fatigued portion. This implies that after fatigue, force can be maiiit;iiricd with a lower level of' excitation than before fatigue. I n exploring t liis plienoriienon, it is irnportant to i i o t c t h a t during MVC motor rieuron firing rates sccni t o match the concractile s p e d o f the muscle.1-5 * 11.12 Bellemare has found that during an MVC, meail motor unit firing rates i n biceps 1)racIiii ;md adductor pollicis muscles were 3 1 .I Hz and 29.9 Hz, respecLively while for the soleus niuscle they were only 10.7 H7. The twitch coritractinii arid half rclaxation times were a l ~ o u 50%> t longer for the s01c:us muscle t h a n for tlic hiceps arid adductor pollicis muscles.' Slowing of' muscle coritractilc spccd and relaxuliOrl rate does occur during f ~ t i g u e . ~ . ~ ( ) . l ~ ' . ~ ( j . ~ ~ , ~
Specifically, Bigland-Ritchie has s h o w n that after a 60 sec MVC, muscle twitch relaxation time was prolonged by about 50'r." Slowing of contractile speed allows fewcii- motor unit spikes per set:ond to maintain a spccific level 01 l'otw. Iri our experiments, motor i i n i t spikr counts and surlace LMG activity nccticd t o maintairi aiitigr;tvity posture (45" elbow flexion) decrcasctl by ahout 50%, with fatigue. 'l'his is in agrccmcnt with the studies of Bigland-Kitchie, which also showed 5OC;i: I-eduction in E M G activity during a (50 sc(' MVC.4*15 'Thus, it is likely that. our notccl reduction in n101or unit spike couriLs at a spccilic level of forcc is rclaced to the slowing of muscle relaxation rate associated with fatigue. Tlie reduction in motoi- unit spike counts associated with fatigue likely henelits the niotoisystel,l~2.53 13.20 High frequency stirnulation of a motor ncrw results in failure o f electrical p r o p gation at the rnuscle fiber membraric. Although still a su1,jec:t of controversy, volulitary contixctioris do not seem ID cause this failure of ~iropngatioii."'~~)~~' Also, high frequency stiiiiulatiori of ;I iiervc rcsults in a faster rate of force loss tkiaii secn in MVC.",'' rills, the reciuction in iuotor unit spike counts m a y protect the musclc fro111 pcriphcral I'ailutr. Also, if' this ptierioiiieiion did not occur, somc degree o f iriotoi- c : o i i t r o l may he lost. Biglanci-Ritchie has suggcstctl that i l ' iiiotor units fire above the frcqiwticy needed duriii MVC, then a large rcduc:tion in neur~al rat wou~ttIK necessary to reduce force.'" Slowing of contractile speed allows fi-wcr motor- u n i t spikes per second to maintain a specific levcl 01' l'orce. Tlir preserice of ii feedback nicckia-
MUSCLE & NERVE
December 1990
1149
nism that allows the CNS to perceive fatigue and/ or contractile slowing and to dccrease motor unit spike counts accordingly has been proposed. Hannerz has shown that when the all'erent iriflow from the muscle is decreased by compression of a nerve, an increase in firing rate of some motor units occurs." Also, Bigland-Kitchie has shown that in fatigued muscle, no recovery of firing rates occurs after 3 min of rest if the muscle is kept ischemic, but near full recovcry occurs 3 min after the blood supply is restored. This suggests the presence of a periptcral reflex from the fatigued muscle rather than a change in central motor neuron excitability (which would show recovery under both conditions).' Also the lack ol' recovery during ischemia is not likcly related to failure of neuromuscular transmission or impaircd subject cff0rt.l' For well-motivated subjects, twitch occlusion showed no reduction in the dcgree to which the muscles could be activated voluntarily during MVC after 3 min of ischemic rest. Also, no reduction in M wave amplitude was noted. Our deafferented suqjcct did not show the usual reduction in RIEMG aclivity. In fact, after a fatiguing contraction, an increase was noted. Also interesting was the tendency for this subject's arm to flex involuntarily after MVC. If more t.han the optimum number of motor units were firing, force would increase and tend to flex the arm. The findings in this subjcct suggest that the af'ferent system is necessary to achieve the normal de-
cliric in electrical activity. €lowever, as the reduction in electrical activity is variable even in normal sutjects, study of additional deafferented subjects would be necessary to confirrri this. We have found that after a fatiguing contraction fewer motor unit spikes per second are needed to maintain a certain low level of force than before fatigue. These findings are in accord with those of prcviaus studies which documcnted a reduction in motor unit firing rates during MVC.3,4.5+8+12,.".19However, thcse previous studies were performed at high levcls of force while I'orce was declining. Ours werc performed at a constant and minimal pre- and post-fatigue level of innervat.ion. 'To standardize this level of innervation, we have used the antigravity posture during 45" ofjoint flexion. This low level of force allows one to easily count the motor units, avoiding the difficulties encountered with counting units at high levels of force. Also, the use of antigravity posture is a very simple method of standardizing innervation level. Limitations include the fact that with a nionopolar electrode, orily one sinall area of the muscle is evaluated. Also, due to niovenicnt of the muscle during MVC, the exact same motor units are not likcly to be evaluated before and after the contract.ion. However, as the rcduction in spikes per second is noted in all areas of the muscle, it is not necessary to monitor one specific motor unit to document changes in overall spike COutltS.
REFERENCES 1. Bellemarc I;, Wocds JJ, Jolra~isson K, Ri larid-Kitchie R:
Motor-unit discharge rates in maximal vifuntaty contractioris of three human musclcs. J Neuuu&&l 1985; 50:1580- 1392.
2. Bighnd-Kitchic B: EMC and fatiguc of human voluntary arid stimulated contractions, in Porter K. Whelan J (eds): Cibu Ikundatiun Synposium 82,Hunicm Mu.wL FaiKup; Yhyviulugical Meclmtisms. London, Pi~niaoMedical, I98 I , pp 130-148.
3. Bigland-Kitchie RR, Uawson NJ,Johansson KS, I.ippold OCT: Reflex origin for the slowing of rnotoneuron firing rates in fatigue of human voluntary contractions. J Plysiul 1986;379:453 -459. 4. Rigland-Kitchie A, Jolransson
R, Lippold OCJ. Smith S, Wocds JJ: Changes in motoncuron firing ratcs during sus-
tained maximal voluntary contractions. J Phyiul 1983; 340:335-346.
5. Bigland-Ritchic B,Johansson R, Lippold OCJ, Woods JJ: Contractile spccd arid EMC changcs during fatiguc of sustained maximal voluntary contractions. J Noirot&siur( I983;50:313- 321. 6. Bigland-Ritchie B, Joncs CIA, Wcmds JJ: Excitation lrequency and muscle fatigue: electrical responses duririg humaxi voluntary and stimulated coniractions. h$ Neuuul I979;64:414-427. 7. Bigland-Ritchie B, Kukulka CG, Lippolrl OCJ,
1150
Motor Unit Spike Counts
Woods JJ:
Thc absence of ncurorriuscular transmission failure in sustained maximal voltiulary contractions. .) Physiui 1982; 530:265-278.
8. ISigland-Kitchie I%,Lippold OCJ: Changes in musclc activation during prolonged niaxirnal voluntary contractions J Phy.vid 1979;292: 141'- 15Y. 9. Ihchaicau J, HainauL K: Electrical and mechanical failure during sustained and intermittent contractions in humans. J ~ p p ritpbi i 19a~;.w~2-945. 10. Edwards RH'I', llill DK,Joncs DA: Mclahlic changcs associated with thc slowing of r e h a t i o n in fatigued mouse rnusc1e.J Phyiul 197.5;45 1:287-301. 11.
