Activity of single motor units from human forearm muscles during voluntary isometric contractions.

Activity

of Single

Il/luscles

During

Motor

Units

Voluntary

From

Isometric

H.-J. FREUND, H. J. BijDINGEN, AND V. DIETZ Department of Neurophysiology, University of Freiburg, RESPECT TO spike train analysis, motor unit recordings from human subjects provide the same information as recording from the motoneuron itself. The advantage of this approach is the facility of studying motor units during voluntary innervation. The disadvantage is the lack of additional information on the units recorded. This situation was recently improved by establishing a method for the measurement of the contractile properties of single motor units during voluntary contractions (44). Another relevant parameter became accessible by the measurement of the conduction velocities of the nerve fibers innervating the motor units under study (21, 23). Using the latter approach, this study is concerned with two questions: I) Is the activity of single motor units studied during isometric contractions correlated with the conduction velocity of their nerve fibers? 2) Is it possible to classify units as tonic and phasic and link this classification to motoneuron size estimated by conduction velocity? The two questions are closely related. They refer to the problem of whether the a-motoneuron pool for a hand muscle is divided into two classesof cells specialized for different functional requirements, or rather represents one homogeneous population. From animal experiments, evidence has been obtained supporting both views. On one hand, a correlation between the excitability and the size of motoneurons, no matter what kind of synaptic drive was used for their activation, has been reported (31, 34, 56). In the case of a unimodal distribution of cell sizes in a motoneuron WITH

Received

for

publication

September

24, 1974.

Human

Forearm

Contractions

Freiburg,

West Germany

pool innervating one muscle, this would indicate an approximately linear excitability gradient among the cells. On the other hand, a classification of a-motoneurons into tonic and phasic types (24) has been made on the basis of differences between cells observed during electrical and reflex excitation. The analogy to such a specialization in the y-motor system and the existence of red and pale muscle fibers supported this view. A convincing demonstration of functional differences between a-motoneurons of one muscle, justifying a distinction into these two groups during natural use of the neurons has-to our knowledge-not yet been made. This study refers mainly to recordings made from the first dorsal interosseus muscle (FDI). This muscle was thought to fit particularly well for this investigation since the nerve fibers innervating its extrafusal muscle fibers are known to be unimodally distributed (20). But in contrast, histochemically the muscle is composed of type I (57.4%) and type II (42.6%) fibers (35), so that both possibilities considered above appear equally possible. Our results suggest that the features of the motor units studied in the course of these exPeriments can be exPlained on the basis of excitability differences between cells of different size. METHODS

A total of 210 single muscle recorded from the first dorsal cle (FDI) in the hands of 45 aged 20-40 yr and 65 muscle m. extensor indicis (EI) of 20 Informed consent was obtained ject and each was aware that draw from the experiment at

fibers have been interosseus musnormal subjects fibers from the normal subjects. from each subhe could withany time. Pain 933

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (139.184.014.150) on August 6, 2018. Copyright © 1975 American Physiological Society. All rights reserved.

FREUND,

934

BUDINGEN,

was minimal, equivalent to that experienced during routine clinical electromyography. Electrical

recordings

Recordings were made with tungsten microelectrodes electrolytically sharpened in a NaNO, solution and, thereafter, the approximately I-#pm tip was coarsened by rubbing against fine sandpaper to about 5-15 pm. Then the electrode (shaft diameter 1 mm) was insulated 8 times. Usually, the electrode resistance was then infinite, since the roughened tip was totally coated. Therefore, the tip was again very carefully rubbed over the sandpaper until the AC resistance measured at 182 Hz was l-5 MQ. The reason for this procedure was to obtain high-impedance, but mechanically resistant electrodes. In most cases, the electrodes were directly inserted through the intact skin. In some “hard” cases the skin was initially penetrated with a small injection needle. A skin electrode at the proximal phalanx of the third finger served as reference electrode and a band electrode round the wrist as earth connection. The surface EMG was recorded with skin electrodes in belly-tendon position, so that one electrode was placed over the endplate zone and the tendon electrode placed on the knuckle at the base of the index finger. The belly electrode was adjusted to the endplate zone by searching the optimal potential by means of a movable electrode. For stimulation of the ulnar nerve, two sets of two skin electrodes (3 mm diameter, 1 cm distance) were fixed over the nerve at the wrist and elbow. Care was taken that the cathode was always distally located. Stimulus duration was 0.1 ms. Electrical activity was fed into a cathode follower input stage, amplified by a Tijnnies differential amplifier, and displayed on an oscilloscope. Both visual and auditory feedback of the signals was provided. The signals were directly fed to an interface (WDV, Munich) with digital trigger inputs. Triggering was carefully controlled in the usual way by observing that each spike produced one Schmitt trigger impulse. Tension

recording

In the case of FDI, the hand and forearm were fixed on a comfortable support. The hand was packed in modeling clay and firmly held by rigid bands. The brackets were ‘mounted on the force transducer, which was rigidly fixed to the support. They were held between the proximal phalanx of the thumb and the proximal interphalangeal joint of the index. In the case of the EI muscle, the forearm

AND

DIET2

and the hand were rigidly fixed to a flat support which left a 3-cm-wide slit for the index finger. A rigid band was attached to the proximal phalanx of the index, and this band was fixed to a strain gauge mounted rigidly 11 cm below the hand. Mechanical recordings from both muscles allowed very little movement so that the muscle actions were nearly isometric. Two resistance strain-gauge transducers supplied by DC bridge amplifiers were used. The range was 10 kg for the interosseus muscle and 2 kg for the extensor indicis. Their resolution was 3 log units. The output of the strain gauge was fed into a Tijnnies DC amplifier and displayed on a four-channel oscilloscope. Tension and spike recordings, as well as the trigger signal for the computer, were continuously monitored on the oscilloscope. For the measurement of the rate of rise of isometric tension and the firing onset of single units in relation to muscle force, the amplifier outputs were also connected to a fast ink recorder (Mingograph). Paper speed was either 100 or 500 mm/s (for faster contractions). Data processing Electrical activity and force were fed into an interface with digital trigger inputs and an analog-digital converter. The time between the trigger impulses was measured and stored on the disk of an IBM 1130 computer. The following statistical displays could be chosen for calculation and for readout either onto a storage oscilloscope or on an x-y plotter: frequency histogram, interval histogram of the first to third order, and joint-interval histogram. For further details see Freund et al. (23). RESULTS

Measurement of conduction velocity (CV) of single nerve fibers The correlation between the size of a motoneuron and the CV of its axon has been generally assumed and could recently be demonstrated directly by comparing the size of dye-injected motoneurons with the CV of their axons (3). In our experiments on human subjects, the CV of the nerve fiber innervating the muscle fiber recorded was used as an estimator of the size of the motoneuron. This was done by standardized measurements because there is considerable variation of CV between different normal subjects. For calculation of the relative values, the mass EMG potential of the FDI was recorded simultaneously with the single-fiber potential. Following supra-


ACTIVITY

OF

HUMAN

maximal stimulation of the ulnar nerve at the wrist, the duration of the gross potential was normalized to 1 and the time of occurrence of the peak of the single-fiber potential was expressed as a fraction of 1. The identity of the voluntarily activated spike and the stimulus-evoked spike was checked carefully. One further confirmation that the action potential was recorded from a single unit was its constant amplitude during varying stimulus strength. This is shown in Fig. 1, where single-fiber potential and mass potential are displayed at two different stimulus intensities. In order to minimize errors due to different locations of the microelectrode along the muscle length, the tip of the microelectrode was closely adjusted to the end-plate zone (see METHODS). Variation in

3mV

MOTOR

UNITS

935

Recruitment

not

properties

Recruitment was determined bv measuring the total force during slowly &creasing isometric contractions as the motor unit began to fire. The speed of contraction was 50 g/s. Repeated measurements of recruitment threshold showed only small random variations which were in the same range as described by Tanji and Kato (59). From five units measured twice a minute for 6 min, a continuous trend was seen only in one low-threshold unit, which showed a steady decline of threshold from 100 to 70 g The relation between threshold force and CV is plotted in Fig. 2 for 65 units recorded from the FDI. Relative values of CV are given at the abscissa.Zero on the standardized scale refers to the onset of the compound muscle potential (fastest fibers) and 1 to the end (slowest fibers). The close correlation between CV and threshold force is evident. This orderly recruitment of motor units according to size has also been observed if the contractile properties of motor units were used as an estimator of unit size (45). In multiunit recordings by means of lowimpedance electrodes, it was constantly found that during decreasing force the units ceased firing in the reversed order of

.

.

.

.j*.

.

0 .

.

l m

.

. .

.

.

l *oo

5 msec

.

0

0

0 0 .‘=.

. 2.

.

0

0 0

0%

FPG, I. Single-unit action potential and surface EMG recorded from the first dorsal interosseus muscle following stimulation of the ulnar nerve at the elbow. Upper two traces: single-unit spike and surface EMG at weak stimulus intensity. Lower two traces: same recordings during stronger stimulation. Negativity upward. The negative peak of the single-unit potential was used for the calculation of relative CV.

. did

the action-potential amplitude affect the location of the peak.

t

I

0.2 relative

1

8

0!4 nerve

I

016 conduction

* .

I

0.8

1

l!O

velocity

FIG. 2. Threshold force of recruitment of 65 units recorded from the first dorsal interosseus muscle during slowly increasing isometric contractions (50 g/s) as plotted against relative conduction velocity of their nerve fibers.


FREUND,

936

BUDINGEN,

recruitment. If the speed of contraction and relaxation was the same, the unit started and stopped firing frequently at different force levels. As already observed by others (14, 46, 49), most units ceased firing during descreasing force at a higher force level than they start firing while increasing force. This is probably mainly the consequence of the time delay between the discharge and peak contraction. As shown on Fig. 3B, all units studied were recruited at force levels up to 700 g. The number of units recruited at a particular force level decreases rapidly with increasing force. The recruitment thresholds of the units of the extensor indicis have a similar distribution (Fig. 3A). For this muscle the maximum force can be determined without the contribution of other synergistic muscles and varied between 1,200 and 1,500 g for male subjects. Since no unit was recruited at force levels above 400 g, the recruitment range covers about one-third of the total force range. Milner-Brown et al. (45), measuring the threshold force of recruitment in the FDI, observed units which were recruited at forces near 2 kg. The distribution of threshold forces displayed in their Fig. 2 is, however, similar to our Fig. 3 and shows that about 930/, of 145 units were recruited below 1 kg and 80% below 600 g. In the abductor digiti minimi, Tanji and Kato (59) found 45.8% of their units recruited below 207, MVC, but the recruitment 40 “x l-l

AND

DIETZ

domain in this muscle extended up to 80% MVC. Since no unit of our sample was recruited at force levels higher than 700 g, the possibility of missing high-threshold units during slowly increasing force had to be considered. Possibly, these experiments were inadequate to activate the large so-called phasic units. Therefore, recruitment was tested during various phasic and tonic motor tasks. No units could be isolated which were only recruited by brisk contractions, but not by slow ones. The results of the systematic variation of the rate of rise of isometric tension are unpublished data. In these experiments, all units were recruited at successively lower force levels if the contraction became faster. The order of recruitment remained, however, unchanged also during brisk phasic contractions. Tonic

and phasic

discharge

patterns

As a consequence of the decrease of the threshold force of recruitment at increasingly faster contractions, the discharge pattern during stepwise increasing isometric contractions shows a characteristic pattern. This is illustrated in Fig. 4. In A, a high-threshold unit from the EI always becomes activated at the small force increments. During the subsequent plateaus the unit stops firing. But if a certain force level is attained, the unit fires continu-

r @

@ kextensor

indicis

m.interosseus

dorsalis

]i

I, 250

A./)

g i

1 set

threshold

force

of ‘recruitment

(g)

FIG. 3. Distribution of the threshold force of recruitment of 65 units recorded from the extensor indicis (A) and 210 units recorded from the first dorsal interosseus muscle (B>. Ordinate: percentage of units recruited at forces indicated on the abscissa. For this figure, units were pooled in classes of 50 g for the extensor and 100 g for the interos-

FIG. 4. A: activation of a high-threshold unit of the extensor indicis during stepwise increasing isometric contractions. B: similar condition, but simultaneous record from a low- and a higher threshold unit of the first dorsal interosseus muscle.


ACTIVITY

OF

HUMAN

ously also during steady force. In B, the same result is shown for a simultaneous record from two units from the FDI. Both units discharge at steady force, but the small-spike low-threshold unit fires continuously at a lower steady force level than does the large-spike, higher threshold unit. The transient activation during the high contraction velocities reflects the low firing threshold at this condition. Maintained firing is achieved if considerably higher force levels are reached. Because firing threshold depends on the rate of rise of tension but the mean force level required for continuous firing is constant (9), we found that a special term for the stable condition was useful. The steady tension, where a particular unit remains tonically activated no matter what speed of contraction was used to recruit the unit, was called tonic threshold (TT). It was consistently found that TT is positively correlated to CV, as is the threshold force of recruitment. Its actual value is approximately the same as the threshold force of recruitment if tested during slowly increasing contractions (50 g/s; see Zig.

2).

With the exception of the smallest cells, which remain tonically activated after the smallest force increments, all cells that could be activated at all could be phasitally activated during rapid contractions as long as the total force was below TT. On the other hand, all units could be tonically activated if the steady force level was above TT. Thus, the type of discharge, tonic or phasic, simply reflects whether the unit operates below or above TT. According to the correlation between TT and CV, fast conduction units have a larger force range below TT than slow conduction units. Therefore, larger units discharge more frequently phasically than small units. These apparent differences in the discharge pattern of small and large cells can thus be explained by quantitative differences in the excitation of cells.

Firing properties during contractions of different strength Firing rates were calculated from 30-s time segments during stationary voluntary contraction of different strength. For this

MOTOR

UNITS

937

duration, the firing rates were uniform throughout the record and also the highest threshold units showed no adaptation.-The results of the examination of 65 units recorded from the FDI are summarized in Fig. 5. In order to analyze the amount of firing rate modulation for units of different size and threshold, groups of units with similar thresholds around the values indicated on the ordinate were pooled together. The mean values are fairly representative since variance was reasonablv small. The firing-rate modulation is much more pronounced for the low- than for the highthreshold units. This is in agreement with (7), the results of Bigland and Lippold Petajan and Philip (50), Person and Kudina (49), and Tanji and Kato (60). It is different from the results of Milner-Brown et al. (46) who found similar firing rate

-E

500

g $ z E b z 2 E -

400 3co 200 100 6

1000

FIG. 5. Dependence of firing rates (abscissa) of classes of units of different recruitment threshold (ordinate) on the force of isometric contractions (2 axis). Curves represent mean values of units pooled together in classes of similar threshold force of recruitment so that each curve shows the mean firing rates for units being recruited at values within a force range of 50 g to either side of the ordinate values (except the lowest curve which contains only cells recruited between 0 and 50 g). The firing rates -of the individual units were calculated for a measurement period of 30’ s during steady force levels indicated on the z axis. For this figure a total of 65 units was used.


FREUND,

938

BUDINGEN,

modulation for all units. This difference is probably due to different experimental conditions, since the latter authors calculated the firing rates during continuously increasing contractions. The largest change of firing rate per unit force was seen in the just-suprathreshold range, the smallest at the higher force levels tested. This nonlinear increase in firing rate was also observed in other (7, 14, 15, 36). A steady-state experiments final increase of firing rates, if force rose has been defrom 75 to 100% of MVC, scribed by Bigland and Lippold (7) and Kaiser and Petersen (36). Constant firing rates at constant forces without rotation of units were described by Smith (55), Lindsley (40), Weddell et (7), Petajan al. (62), Bigland and Lippold and Philip (50), Bracchi et al. (8), Clamann (14), and Milner-Brown et al. (46). Decreasing rates during steady contractions

30

AND

DIET2

were observed by Person and Kudina (49) and Kranz and Baumgartner (39). During our usual recording time of 30 s and at force levels up to 1,000 g, we did not observe systematic trends indicating nonstationary firing rates. This held true also for five units recorded at stationary contractions for 5 min. In those cases where two units were recorded simultaneously either by one or by two electrodes, no rotation between units has been observed during steady contractions. The features of the normal discharge pattern during constant isometric contractions (Fig. GB) are summarized in Fig. 6. The unsmoothed frequency profile (A> is regular. The nonsequential-interval histogram shows an approximately normal distribution for the first-order intervals (C), and the joint-interval histogram (D) shows a concentric point distribution. Thus, the interval length varies randomly

0A

imp. set IO

I I

I

I I

2oow II msec II

1 .

I

FIG.

6.

1 1

1

30

D

0

1

fi

I

I

I

1

’

1

1

1

I

I

1

1

I

E

I

1

1

1

1

’

’

80 interval traction.

1 I

1

’ set

20

Discharge histogram

I

I

1

160

profile (A), isometric (D> of a unit recorded

force from

msec

100

(13), the

distribution (C), and jointmuscle duringsteady con-

nonsequential-interval first dorsal interosseus

msec 200


ACTIVITY

OF

HUMAN

around a mean value and shows no relevant serial dependencies between adjacent intervals. Serial correlation was not calculated, but the results of Kranz and Baumthat the interval gar tner (39) indicate sequence does not represent a renewal process. Interval histograms showing a right-hand tail at low firing rates (39, 49) were also observed in our material. But this was only the case if the unit discharged at a force level just suprathreshold for that unit. The variance of low- and high-threshold units showed no consis tent differences. A classification into tonic and kinetic types of units on the basis of differer t variances (61) could not be confirmed. This is in accordance with the results of Person and Kudina (49) and Kranz and Baumgartner (39). The majority of the recordings from the FDI were made at a force range between zero and 1,000 g. The limitation of the systematic examination of firing rates to 1,000 g was done for the following reasons: 1) During or near MVC, the discharge rates are no longer stationary but decline rapidly. For this muscle, this is accompanied by the rapid decrease of force to about 50% of MVC within 1 min (58). 2) Many subjects experienced intense pain if stationary contraction was above 1,000 g. 3) Nonlinearities in the relation between smoothed rectified EMG and force during stronger contraction indi cated the contribu tion of other synergistic muscles to total force. The reader can demonstrate this to himself, if the contraction of the muscles synergistic to the first dorsal interosseus are palpated during different strengths of contraction. Near MVC the whole arm and hand are stiff and minimal changes of position and axis of the fixed hand and fingers cannot be avoided. In the EI, the situation is less complicated by synergistic muscles and some units could be recorded up to 100% MVC. Their firing rates showed the same nonlinear increase as the units from the FDI. All units started firing at approximately the same rate (6.8/ s t- 1.4) without differences between units of different threshold. Thus, the firing rates were well within the range which was found by most authors

MOTOR

UNITS

939

recording single-unit activity from hand muscles (8, 14, 29, 46, 62). The same conformity of data was obtained for firing rates at the point of recruitment (5-10/s). The maximum firing rates (60-100/s) observed in the abductor digiti minimi by Marsden et al. (42) following the isolation of one motor unit after nerve blockade show that under certain conditions an1 in some muscles, the motoneurons can fire at higher rates. Tanji and Kato (60), recording from the same muscle, found maximum firing rates between 25 and 35/s if steady contractions were examined, but between 80 and 100/s if the contractions were performed as fast as possible. In the FDI and in the extensor indicis we did not observe firing rates higher than 35/s even if the rate of rise of isometric tension was as fast as 2,000 g/s (unpublished observa tions). A close correlation between conduction velocity, recruitment threshold, and firing rate modulation was also observed if continuous contractions were used instead of stepwise force increments. The change in firing rate per unit force was largest for the small, slow-conducting units ;tnd was successively smaller for larger units. This traces show is illustrated in Fig. 7. Top the instantaneous firing rates; lower traces, isometric force. In A, a low-threshold unit recruited at 80 g and in R, a higher threshold unit recruited at 370 g are displayed. In the case of both units the subject performed rhythmic contractions of approximately equal amplitude (100 g) at justsuprathreshold force levels. Although the contractions in A vary, on the average, by 15 g more than in B, there are certainly some contractions in both records in which the amplitudes of the contractions are equal. In addition to the differences between the units, the firing rate modulation at this condition is more pronounced than in the case of steady force levels of comparable strength (Fig. 5). This is also due to the fact that the change in firing rates per unit force becomes greater with increasing rate of rise of isometric tension (unpublished observations). The differences in firing rate modulation between units of different size do not


940

FREUND,

;

BUDINGEN,

So

AND

60~

? si &a0 )r .ul 5 p 20 ti 3

06

40*

2oqb

.-E”

& 10

700

DIET25

set

20

10

30

T

-

2-o

‘set

c 3’0

20

set

30

700 T

+ 10

FIG. 7. Instantaneous and spike intervals, (23) during rhythmic

firing isometric isometric

+

set

20

4

10

30

rates (top traces) calculated as reciprocals from force (bottom traces) of a low-threshold (A) and contractions.

change basically if the unit is engaged in more tonic or phasic motor tasks. This is illustrated by Fig. 8 where two units of different threshold were recorded simultaneouslv with two electrodes. The upper profiles of a traces show the frequency force = low-threshold un i t (recruitment unit (re100 g) and of a high-threshold

the individual-interhigher threshold

a 10

b

b

5-

I

10

I

r,

1 s=c

unit

cruitment force = 380 g), the bottom traces show isometric force. In Fig. EM, the variation of force was slow so that one contraction period lasted about 8 s; in B, the same units were recorded during faster contractions of about the same amplitude. The difference in firing rate between the two units tends even to become greater

a

I

+

‘,’

20

,

I

I=

10

set

20

10

set

20

8001

8007

z 5 *-oL 600 zi E 2 *- 400 * *

FIG. 8. Instantaneous firing ric contractions of comparable IO0 g, that of unit b 380 g.

600-

. 10

rates (top amplitude

4oo-L set

20

traces) of two units (a, b) during (lower traces). Threshold force

slow (A) and of recruitment

faster (B) isometof unit a was


ACTIVITY

during there is old unit ing the

OF

HUMAN

the faster contractions. Certainly no evidence that the high-threshis more adequately activated durquicker contraction.

DISCUSSION

Our results show that the units in a motoneuron pool for a human hand or forearm muscle are functionallv used in a very simple and consistent manner. The two mechanisms determining the form of any contraction of one muscle, recruitment and firing-rate modulation, are closely correlated -to the conduction velocity of the units. This is a close similaritv to the results of animal experiments, where the excitability of motor units as reflected by their recruitment order showed a similar correlation to unit size, as did the susceptibility to discharge (34). In animals, the size principle could be shown to be functionally used during stereotyped natural movements in the swimmeret system of the lobster (16) and in the rabbit diaphragm (10). In man, the close correlation between firing threshold and size of a motoneuron during voluntary isometric contractions has been independently demonstrated by two different methods. In one set of experiments, the conduction velocity of the nerve fiber (2123) and in another, the contractile properties of the muscle unit (45), were correlated with recruitment threshold. Both conduction velocitv and contractile properties are correlated with each other (1, 6, 32, 43, 63) as well as one of them, or both, with a variety of membrane properties of the motoneuron (12, 13, 19, 38) supposed to be related to the size of the motoneuron. The final proof of the correlation between the surface area of mo toneurons and their input resistance has been provided by Lux et al. (41) and by Barrett and Crill (3). The latter authors also established the correlation between motoneuronal surface area and the conduction velocity of its axon. The correlation between the amplitude of the motor unit action potential and recruitment is less precise, since electrode properties and distance between electrode and the muscle fibers influence this parameter. Nevertheless, Olson et al. (47) found

MOTOR

UNITS

941

78.6% of the units recorded were recruited in an orderly fashion according to actionpotential amplitude. Tanji and Kato (59) found a correlation coefficient of 0.62 between action-potential amplitude and recruitment. A fixed order in the recruitment of several simul taneouslv recorded human motor units had alreadv been reported in the older literature (8, 17, 40, 55, 62). Excepti ons from this orderlv recruitment were also reported. Ashworth et al. (2) and Grimby and Hannerz (27, 28) described a rotation in the recruitment of units which was observed during voluntary and reflex activation. Basmajian (4, 5), Simard and Basmajian (54), and Kato and Tanji (37) reported that subjects can learn to single out a particular unit by audiovisual feedback. This, however, was only possible during gentle contractions. In the experiments of Kato and Tanji (37), no subject was able to single out large-spike units which were activated at stronger efforts. Person (48) compared the rank order for the recruitment of pairs of units recorded from the rectus femoris muscle and found a fixed order if the isometric contractions were performed while the position of the leg was fixed. She observed a more variable order if similar contractions were made during different “free” positions of the leg. ts, Wym .an et al. (64) In animal experimen could elici t different orderlv recru itments of pairs of cat hindlimb motoneurons by using a variety of noxious inputs-scratching the cat’s paw or bending the toes. This variability in the order of recruitment is not in contrast to the close correforce of lation between the threshold recruitment and the conduction velocity or twitch contraction of motor units. It could well be explained by the results of Tanji and Kato (59). They observed the recrui tm .ent order of up to five si.multaneouslv recorded motor units and found no reversal of the order between units which were recruited at considerably different tension ranges. But the recruitment order was variable among units whose tension range of recruitment overlapped. The implications of the size principle with respect to the force output of the muscle have been discussed as well as the


942

FREUND,

BiiDINGEN,

analogy to the Weber-Fechner relation in sensory systems (33, 45). In this context, it will only be mentioned that also the fastest conducting units of our sample could be tonically activated. Above tonic threshold, they fired as regularly as slower conduction units. This indicates that the hand muscles use their units providing the largest tension output also during maintained contractions. Although we cannot exclude the possibility that the largest units of the sample were not the largest units of the population, Fig. 2 suggests that some units had their peak very near to the first deflection of the surface EMG so that they belonged to one of the fastest fibers. For this consideration one has to take into account that the single-unit spike recorded from the surface is broader than recorded within the muscle (cf. Fig. 8 of ref 44). Correlating the conduction velocity of single motor units with some of their firing characteristics, we found no evidence for a different mode of activation or activity suggesting the existence of a tonic and a phasic type of a-motoneuron. As described in the RESULTS and shown in Fig. 4, the more tonic or phasic discharge pattern reflects quantitative rather than qualitative differences between units of different size. Since this statement is confined to isometric contractions, the extension of the results by further experiments allowing isotonic movements and initiation of contraction by peripheral inputs, for example, by counterbalancing a motor-driven pivot, is necessary. The same limitation applied to the muscles studied since human hand and forearm muscles are functionally different from axial and leg muscles. It is obvious that the duality in the requirements for hindlimb muscles-to provide a constant innervation for standing and transient activation for movements-has no counterpart in the human hand, which is poorly fitted to accomplish comparable tonic contractions. Simultaneous recordings from pairs of motor units, one from the soleus and the other from the gastrocnemius muscle of normal subjects showed, however, no units which could only be tonically or phasically activated (unpublished observations). The histochemical examination of

AND

DIETZ

human muscles showed that both fiber types I and II, which are commonly attributed to tonic and phasic motor units, are present in the hand and forearm muscles as well as in the leg muscles. In the FDI, type I fibers were encountered in 57.4% and type II fibers in 42.6% (35). In contrast to the existence of two histochemical fiber types, the recruitment and firing properties of the motor units in the human FDI showed a uniform distribution (21, 23, 45), as did the fiber spectrum of the nerve supplying the motor units of this muscle, morphologically (20). As already suggested by Buller et al. (11), Burke (12), Dubowitz (IS), Romanul and Van Der Meulen (53), and Robbins et al. (52), the motoneuronal factors determining the development of type I and II muscle fibers may be metabolic or chemical rather than physiological in nature. On the basis of different usage or differences in firing pattern, muscle fibers should either be uniformly distributed or the transition between the fiber types should be smooth. Some of the differences between motoneurons which supported the subdivision of la-motoneurons into a tonic and phasic type apply only for firing rates considerably higher (13, 57) than the highest rates observed in natural conditions. The maximum firing rate observed in most human muscles during stationary contractions (20-30/s) are surprisingly low if one regards the contraction times of the motor units. It could, however, be shown that asynchronous stimulation of five ventral root filaments imitating normal innervation better than synchronous stimuli achieved fusion already at 7/s and maximal tension development at 20/s (51). Both the different strength of Renshaw inhibition (25) and the different duration of afterhyperpolarizations (19) have been assumed to play a frequency-limiting role, reducing the firing rates of the smaller tonic cells more than those of the larger, phasic cells. The reverse is true during voluntary contractions of human hand, forearm, and leg muscles, so that these inhibitory processes obviously do not play this differential frequency-limiting role. Possibly, during natural movements they are more important for the precise timing of


ACTIVITY

OF

HUMAN

MOTOR

UNITS

943

impulses, as previously suggested by Person virtually all cells can be phasically actiand Kudina (49). vated if the force increments are below the A reconsideration of different kinds of tonic threshold, but fire tonically if the activation which led to the distinction of force level is above tonic threshold. cat hindlimb la-motoneurons into a tonic On the basis of the functional characand phasic type shows a remarkable simiteristics observed in human motor units larity to our results on human cervical during voluntary contraction, the recruitunits. Granit et al. (24, 26) made this disment and firing properties, including tinction on the basis of different modes of phasic and tonic modes of activity, can be activity of small and large cells. Using postthoroughly explained by a graded excittetanic potentiation as a method for condiability of motoneurons according to size. tioning the motoneurons for optimal perSUMMARY formance in subsequent stretch reflexes, they found only the small cells discharging 1. Microelectrode recordings from single tonically during maintained stretch, motor units of the first dorsal interosseus whereas the large, phasic units showed and the extensor indicis muscles of normal preferentially single discharges during the human subjects were studied during volundynamic phase of stretch. Similar results tary, isometric contractions. The conducwere obtained if repetitive stretches, pinna tion velocity of the nerve fiber innervating twitches, or crossed stretch reflexes were the muscle unit was used as an estimator used for conditioning. If the a-motoneuron of the size of the motoneuron. responses to the dynamic and static 2. During slowly increasing contracstretches in Fig. 1 of the paper of Granit tions, the units were recruited at force et al. (26) are compared with the dynamic levels which were closely correlated to conand static periods of voluntary contractions duction velocity. The units associated with in our Fig. 4, the similarity between the low conduction velocity were recruited two different recordings is obvious. During first, those with high conduction velocity, the phasic condition, the small and the last. large unit are discharged and during a 3. If small, stepwise force increments subsequent period of maintained stretch or were used instead of slowly, continuously contraction, only the small unit remains increasing contractions, the units were first active. activated during the steps and became inFor the case of the human motor units, active during the subsequent plateaus. If the firing thresholds depend on the rate of higher steady-force levels were reached, the rise of tension (9). In contrast, the tonic activity was maintained also during the threshold is constant. If this concept is plateaus. This steady force, where a unit extended to the segmental input, the exremained continuously active independent perimental conditions in the reflex studies of the rate of rise of tension, represents its of Granit and his colleagues would provide tonic threshold. the excitatory drive to reach the tonic 4. The tonic threshold is positively corthreshold of small cells but not of the large related with conduction velocity, as is the cells, possibly because the decerebrate and threshold force of recruitment. As a consedeefferented preparation provides no inquence, high-threshold units have a large nervation. The large cells were only force range below tonic threshold where phasically activated, since for this condition they can only be transiently activated, whereas low-threshold units have a large the firing threshold of all motor units becomes considerably lower than during physiological force range above tonic steady tension (9). The results of Henatsch threshold where they operate tonically. et al. (30) have demonstrated that increasThe phasic or tonic appearance of dising the background activation of motoneucharge pattern reflects quantitative differrons chemically could change the phasic ences in tonic threshold between units of different size. All units examined could be into a tonic response type to the same reactivated phasically (below) and tonically flex activation, and vice versa. Correspondduring voluntary innervation, (above tonic threshold). No evidence was ingly,


944

FREUND,

BUDINGEN,

found indicating the existence of two qualitatively different classes of units corresponding to a tonic and phasic type, although both muscles investigated consist of about equal numbers of type I and type II muscle fibers. 5. The change in firing rate per unit force was inverselv related to conduction velocity: the slower conducting units showed larger changes in firing rate per unit force than faster conducting units. This corresponds to the larger excitability

AND

DIETZ

of the smaller units indicated by their earlier recruitment. 6. The data of this studv are consistent with the hypothesis that -the functional characteristics of human motoneurons are determined by the graded excitability of mo toneurons according to size. ACKNOWLEDGMENTS

We gratefully acknowledge N. Schneiderhan. This work was supported by schungsgemeinschaft (SFB 70).

the the

assistance Deutsche

of For-

REFERENCES 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

APPELBERG, B. AND EMONET-D~NAND, F. Motor units of the first superficial lumbrical muscle of the cat. J. Neurophysiol. 30: 154-160, 1967. ASHWORTH, B., GRIMBY, L., AND KUGELBERG, E. Comparison of voluntary and reflex activation of motor units. J. Neural. Neurosurg. Psychiat. 30: 91-98, 1967. BARRET, I. M. AND CRILL, W. E. Specific membrane resistivity of dye-injected cat motoneurons. Brain Res. 28: 556-561, 1971. BASMAJIAN, J. V. Muscles Aliue: Their Functions Revealed by Electromyography. Baltimore: Williams 8c Wilkins, 1962. BASMAJIAN, J. V. Control and training of individual motor units. Science 141: 440-441, 1963. EMONET-DI?NAND, F., AND LABESSOU, P., PORTE, Y. Relation entre la vitesse de conduction des fibres nerveuses matrices et le temps de contraction de leurs unites matrices. Compt. Rend. 256: 5625-5627, 1963. BIGLAND, B. AND LIPPOLD, 0. C. J. Motor unit activity in the voluntary contraction of human muscle. J. Physiol., London 125: 322-335, 1954. BRACCHI, F., DECANDIA, M., AND GUALTIEROTTI, T. Frequency stabilization in the motor centers of spinal cord and caudal brain stem. Am. J Physiol. 210: 1170-1177, 1966. B~~DINGEN, H. J., HILLESHEIMER, W., AND FREUND, H.-J. Orderly recruitment of human motor units during tonic and phasic activation. Arch. Ges. Physiol. Suppl. 343: R 77, 1973. B~~DINGEN, H. J. AND YASARGIL, G. M. Die funk tionelle Organisation der motorischen Einheiten des Zwerhfells. Arch. Ges. Physiol. 332: 218-231, 1972. BULLER, A. J., ECCLES, J. C., AND ECCLES, R. M. Interactions between motoneurones and muscles in respect of the characteristic speeds of their responses. J. Physiol., London 150: 417439, 1960. unit types of cat triceps BURKE, R. E. Motor surae muscle. J. Physiol., London 193: 141-160, 1967. BURKE, R. E., LEVINE, D. M., ZAJAK, F. E., W. K. Mammalian TSAIRIS, P., AND ENGEL,

motor relation Science

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

units: physiological-histochemical corin three types in cat gastrocnemius. 174: 709-712, 1971. CLAMANN, H. P. Activity of single motor units during isometric tension. Neurology 20: 254-260, 1970. DASGUPTA, A. AND SIMPSON, J. A. Relation between firing frequency of motor units and muscle tension in the human. Electromyography 2: 117-128, 1962. DAVIS, W. J. Functional significance of motoneuron size and soma position in swimmeret system of the lobster. J. NeurophysioZ. 34: 274-288, 1971. DENNY-BROWN, D. AND PENNYBACKER, J. B. Fibrillation and fasciculation in voluntary muscle. Brain 61: 311-334, 1938. DUBOWITZ, V. Cross-innervated mammalian skeletal muscle; his tochemical, physiological and biochemical observations. J. PhysioZ., London 193: 481-496, 1967. ECCLES, J. C., ECCLES, R. M., AND LUNDBERG, A. Duration of after-hyperpolarization of motoneurones supplying fast and slow muscles. Nature 179: 866-868, 1957. FEINSTEIN, B., LINDEGARD, B., NYMAN, E., AND WOHLFAHRT, G. Morphologic studies of motor units in normal human muscles. Acta Anat. 23: 127-142, 1955. FREUND, H.-J., DIETZ, V., SCHNEIDERHAN, N., AND B~~DINGEN, H. J. Die Beziehung zwischen GriiBe und Rekrutierungsverhalten menschlicher Mo toneurone. Arch. Ges. Physiol. Suppl. 335: R 67, 1972. FREUND, H.-J., DIETZ, V., WITA, C. W., AND KAPP, H. Discharge characteristics of single motor units in normal subjects and patients with supraspinal motor disturbances. In: New Developments in Electromyography and Clinical Neurophysiology, edited by J. E. Desmedt. Basel: Karger, 1973, vol. 3, p. 242-250. FREUND, H.-J., WITA, C. W., AND SPRUNG, C. Discharge properties and functional differentiation of single motor units in man. In: Neurophysiology Studied in Man, edited by G. G. Somjen. Amsterdam: Excerpta Medica, 1972, p. 305-313. GRANIT, R., HENATSCH, H.-D., AND STEG, G.


ACTIVITY

25.

26.

2’7.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

OF

HUMAN

Tonic and phasic ventral horn cells differentiated by post-tetanic potentiation in cat extensors. Acta Physiol. Stand. 37: 114-126, 1956. GRANIT, R., PHILLIPS, C. G., SKOGLUND, S., AND STEG, G. Differentiation of tonic from phasic alpha ventral horn cells by stretch, pinna and crossed extensor reflexes. J. Neurophysiol. 20: 470481, 1957. GRANIT, R., PASCOE, J. E., AND STEG, G. The behaviour of tonic alpha and gamma motoneurones during stimulation of recurrent collaterals. J. Physiol., London 138: 381-400, 1957. GRIMBY, L. AND HANNERZ, J. Recruitment order of motor units in voluntary contraction: changes induced by proprioceptive afferent activity. J. Neural. Neurosurg. Psychiat. 31: 565573, 1968. GRIMBY, L. AND HANNERZ, J. Differences in recruitment order of motor units in phasic and tonic flexion reflex in “spinal man.” J. Neural. Neurosurg. Psychiat. 33: 562-570, 1970. GYDIKOV, A. AND KOSAROV, D. Some features of different motor units in human biceps brachii. Arch. Ges. Physiol. 347: 75-88, 1974. HENATSCH, H. D., SCHULTE, F. J., AND BUSCH, G. Wandelbarkeit des tonisch-phasischen Reaktionstyps einzelner Extensor-Motoneurone bei Variation ihrer Antriebe. Arch. Ges. Physiol. 270: 161-173, 1959. HENNEMAN, E. Relation between size of neurons and their susceptibility to discharge. Science 126: 1345-1347, 1957. HENNEMAN, E. AND OLSON, C. B. Relations between structure and functior: in the design of skeletal muscle. J. Neurop+ 01. 28: 581-598, 1965. HENNEMAN, E., SOMJEN, G., AND CARPENTER, D. 0. Functional significance of cell size in spinal motoneurons. J. Neurophysiol. 28: 560-580, 1965. HENNEMAN, E., SOMJEN, G., AND CARPENTER, D. 0. Excitability and inhibitibility of r-notoneurons of different sizes. J. NeurophysioZ. 28: 599-620, 1965. JOHNSON, M. A., POLGAR, J., WEIGHTMAN, D., AND APPLETON, D. Data on the distribution of fiber types in thirty-six human muscles. An autopsy study. J. Neurol. Sci. 18: 111-129, 1973. KAISER, E. I. AND PETERSEN, I. Muscle action potentials studied by frequency analysis and duration measurement. Acta Neural. Stand. 41, Suppl. 13: 213-236, 1965. KATO, M. AND TANJ-I, J. Conscious control of motor units of human finger muscles. In: Neurophysiology Studied in Man, edited by G. G. Somjen. Amsterdam: Excerpta Medica, 1972, p. 330-342. KERNELL, D. Input resistance, electrical excitability and size of ventral horn cells in cat spinal cord. Science 152: 1637-1640, 1966. H. AND BAUMGARTNER, G. Human KFUNZ, alpha mo toneurone discharge, a statistical analysis. Brain Res. 67: 324-329, 1974. LINDSLEY, D. B. Electrical activity of human motor units during voluntary contraction. Am. J. Physiol. 114: 90-99, 1935.

MOTOR 41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

UNITS

945

P., AND KREUTZBERG, Lux, H. D., SCHUBERT, G. W. Direct matching of morphological and electrophysiological data in cat spinal motoneurones. In: Excitatory Synaptic Mechan isms, edited by P. Andersen and I. K. S. Jansen. Oslo: Universitetsforlaget, 1970, p. 189-198. MARSDEN, C. D., MEADOWS, J. C., AND MERTON, P. A. Isolated single motor units in human muscle and their rate of discharge during maximal voluntary effort. J. Physiol., London 217: 12 P, 1971. MCPHEDRAN, A. M., WUERKER, R. B., AND HENNEMAN, E. Properties of motor units in a homogeneous red muscle (soleus) of the cat. J. Neurophysiol. 28: 71-84, 1965. MILNER-BROWN, H. S., STEIN, R. B., AND YEMM, R. The contractile properties of human motor units during voluntary isometric contractions. J. Physiol., London 228: 285-306, 1973. MILNER-BROWN, H. S., STEIN, R. B., AND YEMM, R. The orderly recruitment of human motor units during voluntary isometric contractions. J. Physiol., London 230: 350-370, 1973. MILNER-BROWN, H. S., STEIN, R. B., AND YEMM, R. Changes in firing rate of human motor units during linearly changing voluntary contractions. J. Physiol., London 230: 371-390, 1973. OLSON, C. B., CARPENTER, D. O., AND HENNEMAN, E. Orderly recruitment of muscle action potentials. Arch. NeuroZ. 19: 591-597, 1968. PERSON, R. S. Rhythmic activity of a group of human motoneurones during voluntary contraction of a muscle. Ekctroencephalog. Clin. Neurophysiol. 36: 585-595, 1974. PERSON, R. S. AND KUDINA, L. P. Discharge frequency and discharge pattern of human motor units during voluntary contraction of muscle. Electroencephalog. Clin. Neurophysiol. 32: 471483, 1972. PETAJAN, J. H. AND PHILIP, B. A. Frequency control of motor unit action potentials. Electroencephalog. Clin. Neurophysiol. 27: 66-72, 1969. RACK, P. M. H. AND WESTBURY, D. R. The effects of length and stimulus rate on tension in the isometric cat soleus muscle. J. Physiol., London 204: 443460, 1969. ROBBINS, N., KARPATI, G., AND ENGEL, W. K. Histochemical and contractile properties in the cross-innervated guinea pig soleus muscle. Arch. Neural. 20: 318-329, 1969. ROMANUL, F. C. A. AND VAN DER MEULEN, J. P. Slow and fast muscles after cross innervation. Arch. Neural. 17: 387-402, 1967. SIMARD, T. G. AND BASMAJIAN, J. V. Methods in training the conscious control of motor units. Arch. Phys. Med. Rehabil. 48: 12-19, 1967. potentials from single SMITH, 0. C. Action motor units in voluntary contraction. Am. Jo Physiol. 108: 629-638, 1934. SQMJEN, G., CARPENTER, D. O., AND HENNEMAN, E. Responses of motoneurons of different sizes to graded stimulation of supraspinal


946

FREUND, centers 958-965,

57.

STEG,

gidity. 1964. 58.

59.

60.

BUDINGEN,

of the brain. J. Neurophysiol. 28: 1965. G. Efferent muscle innervation and riActa Physiol. Stand. 61, Suppl. 225,

STEPHENS, J. A. AND TAYLOR, A. Fatigue of maintained voluntary muscle contraction in man. J. Physiol., London 220: l-18, 1972. TANJI, J. AND KATO, M. Recruitment of motor units in voluntary contraction of a finger muscle in man. Exptl. Neural. 40: 759-770, 1973. TANJI, J. AND KATO, M. Firing rate of individual motor units in voluntary contraction of abductor digiti minimi muscle in man. Exptl. Neurol. 40: 771-783, 1973.

AND 61.

DIETZ

T. AND SHIMAZU, H. Functional differentiation of human skeletal muscle. In: Corticalization of Morlement. Tokyo: Univ. of Tokyo Press, 1964. 62. WEDDELL, G., FEINSTEIN, B., AND PATTLE, R. E. The electrical activity of voluntary muscle in man under normal and pathological conditions. Brain 67: 178-257, 1944. 63. WUERKER, R. B., MCPHEDRAN, A. M., AND HENNEMAN, E. Properties of motor units in a heterogeneous pale muscle (m. gastrocnemius) of the cat. J. Neurophysiol. 28: 85-99, 1965. 64. WYMAN, R. J., WALDRON, I., AND WACHTEL, G. M. Lack of fixed order of recruitment in cat motoneuron pools. Exptl. Brain Res. 20: 101-114, 1974. TOKIZANE,


Prediction of maximal surface electromyographically based voluntary contractions of erector spinae muscles from sonographic measurements during isometric contractions.

Ballistic contractions in fast or slow human muscles: discharge patterns of single motor units.

Firing rate behavior of human motor units during isometric voluntary contraction: relation to unit size.

The orderly recruitment of motor units of the masseter and temporal muscles during voluntary isometric contraction in man.

The control of blood flow through human forearm muscles following brief isometric contractions.

Position sense at the human forearm after conditioning elbow muscles with isometric contractions.

Improved identification of dystonic cervical muscles via abnormal muscle activity during isometric contractions.

Instability of motor unit firing rates during prolonged isometric contractions in human masseter.

Recruitment of motor units in human forearm extensors.

Central nervous pathways underlying synchronization of human motor unit firing studied during voluntary contractions.

Fast motor units are not preferentially activated in rapid voluntary contractions in man.

The distribution of activity among the muscles of a single group during isometric contraction.

Absence of gender differences in the fatigability of the forearm muscles during intermittent isometric handgrip exercise.

Plyometric training improves voluntary activation and strength during isometric, concentric and eccentric contractions.

Neuromuscular transmission in human single motor units.

Motor units of the fourth deep lumbrical muscle of the adult rat: isometric contractions and fibre type compositions.

Planckian Power Spectral Densities from Human Calves during Posture Maintenance and Controlled Isometric Contractions.

Ballistic contractions in man: characteristic recruitment pattern of single motor units of the tibialis anterior muscle.

Pacing strategies during repeated maximal voluntary contractions.

Cross-talk in mechanomyographic signals from the forearm muscles during sub-maximal to maximal isometric grip force.

A model for a motor unit train recorded during constant force isometric contractions.

Synchronization of motor unit activity during voluntary contraction in man.

Reflexes evoked in human thenar muscles during voluntary activity and their conduction pathways.

Maximal discharge rate of motor units determines the maximal rate of force development during ballistic contractions in human.