J. Phypiol. (1977), 265, pp. 163-174 With 7 text-figure8 Printed in Great Britain
163
THE ORDERLY RECRUITMENT OF MOTOR UNITS OF THE MASSETER AND TEMPORAL MUSCLES DURING VOLUNTARY ISOMETRIC CONTRACTION IN MAN
By R. YEMM* From the Departments of Oral Biology and Dental Prosthetics, University of Bristol, Medical School, University Walk, Bristol BS8 1TD
(Received 24 May 1976) SUMMARY
1. The contractile properties of the motor units of the masseter and temporal muscles of human subjects were studied during voluntary isometric contractions, using a method previously employed to examine units of a small hand muscle. 2. Over the range of forces studied (0-6 kg), the units of both muscles were recruited in an orderly fashion, with a nearly linear relationship between the voluntary force at which units were recruited and their measured twitch tensions. 3. The range of contraction times (25-90 msec) was similar to that observed for the hand muscle. In some subjects it seemed that small units, recruited at low forces, exhibited shorter contraction times. INTRODUCTION
In earlier papers, a method has been described which enables the determination of the contractile properties of human motor units during voluntary isometric contraction (Stein, French, Mannard & Yemm, 1972; MilnerBrown, Stein & Yemm, 1973a, b). The muscle used for this investigation was the first dorsal interosseous muscle of the hand. Information concerning magnitude and time course of the twitch tension of units was obtained and in addition, it was reported that recruitment of motor units during increasing voluntary contractions took place in an orderly fashion with units of increasing power being recruited after the lightest forces were achieved by means of activation of the smallest units (Milner-Brown et al. 1973b). Present address: Department of Dental Prosthetics, Dental School and Hospital, The University, Dundee DD1 4HN. *
R. YEMM The present experiments on the masseter and temporal muscles were undertaken to look for further evidence of the existence of orderly recruitment in a different region of the body. In addition, the ability to record forces of contraction between the teeth provides a superior connexion to the force transducer than anywhere else in the body. A preliminary report of some of the results from these experiments has been made elsewhere (Yemm, 1976). The technique requires that electrical activity associated with the discharge of single motor units be recorded during a voluntary contraction, together with the output of a transducer positioned to record the total contractile force. The impulses of the motor unit are used to trigger a signal averager, which sums forces recorded from the whole muscle. Provided that other units, contributing to the contraction, discharge at random intervals relative to the unit being studied, the averaging process will extract those tension fluctuations correlated to the single unit from those due to other units and other factors. 164
METHODS
A total of 149 motor units have been studied; of these, ninety-four were from the masseter muscle, recorded from four subjects. Fifty-five units from the temporal muscle were examined from three subjects. For two subjects, data have been obtained from both muscles. In each experimental session, data were recorded on an FM tape recorder for subsequent analysis. Four channels of the latter were employed to record (1) motor unit activity from a needle electrode, (2) output of the force transducer at low gain, (3) the force signal at higher gain after filtering out low frequency fluctuations, and (4) the electromyogram (e.m.g.) of the muscle obtained with surface electrodes. Motor unit activity. A bipolar needle electrode was employed. Two fine Teflon insulated silver wires 75 ,um in diameter were contained inside the needle (25 gauge), for differential amplification. The needle provided the earth and was filled with epoxy resin. The signal was amplified by means of a conventional differential pre-amplifier as in previous experiments. Tension recording. The force transducer, employing strain gauges on a stiff beam, was placed between the upper and lower first premolar teeth on the same side as the needle electrode. The mouth was opened by about 8 mm. An impression of the upper teeth in a thermoplastic compound was formed on the upper beam to enable accurate repositioning of the device from time to time. The natural frequency of the transducer was approximately 400 Hz. Voluntary contraction by the subject caused flexion of the transducer beams of the order of 1 mm/kg so that forces in the experimental range of 0-6 kg could be regarded as taking place essentially isometrically. The filter used to permit high gain recording of the force record by elimination of low frequency fluctuations in the force had a time constant of 2-5 sec, chosen to avoid distortion of the subsequent records. Surface e.m.g. Two 9 mm diameter silver disk electrodes were applied to the skin over the muscle about 2 cm apart. The signal from these electrodes was subjected to similar amplification to that used for the needle electrode. Recording procedure. For each motor unit, data were recorded for a period of about 3 min. During this, the subject was asked to maintain a steady discharge of the unit
ORDERLY RECRUITMENT OF MOTOR UNITS
165
being investigated, at as slow a rate as possible. The subject was provided with auditory feed-back. Commonly the slowest rate of firing that could be maintained was 10-15 impulses/sec. From time to time, the needle was moved within the muscle, although at many sites more than one unit could be discriminated, in which case the voluntary force was varied to recruit and record data from all units discriminated. Single units were identified on the basis of constant amplitude and regularity of discharge. For each unit, the voluntary force at which the unit was recruited was noted at the start of the recording period. Several experiments were performed on most subjects, during each of which between 5-10 units were studied. An attempt was made in each session to sample units with a wide range of force thresholds. Y V V
V
"'5 W
(U
E
TP 0
0)
00
Co
I
00
0
2000
4000
6000
Force (g) Fig. 1. The relationship between force exerted upon the transducer and the integral of the surface electromyogram (e.m.g.) of the masseter muscle. Each symbol represents the integrals obtained from the first minute of a record made of the e.m.g. and force whilst studying six units during a single experimental session. The surface electrodes remained undisturbed throughout the recording period.
The threshold force measured at the transducer was developed by a number of muscles, rather than as is the case in the earlier experiments on the hand muscle (Milner-Brown et at. 1973a, b), where most if not all of the recorded total force was due to a single muscle. Thus voluntary forces exerted upon the apparatus in the present experiments will be derived from the masseter, temporal and medial pterygoid muscles of the side upon which the apparatus was placed, with a contribution from contralateral muscles. Further, it is possible that the proportion of the total force exerted by each muscle varies with degrees of voluntary contraction. In order to investigate this, the recorded surface e.m.g. was integrated electronically. Five-second integrals of the e.m.g. were obtained together with an instant value of the force level at the mid point of the period of integration. Fig. 1 shows the relationship between force and the integrated e.m.g. during the recording of six
166
R. YEMM
units from the masseter muscle of one subject during one experimental session and provides evidence that there is an approximately linear relationship between the two. Similar results were obtained for other muscles of other subjects. Since it has been shown (Lippold, 1952) that, under isometric conditions, the integral of a surface e.m.g. is linearly related to the force developed, it is reasonable to conclude that under the experimental conditions applied in this study, the masseter and temporal muscles each exerted an approximately constant proportion of the total recorded voluntary force throughout the range of forces studied. Analy8is of data. Data were analysed using a signal averager. The motor unit impulses were used to trigger the device and the averages of the motor impulse, the high gain filtered force recorded and the surface e.m.g. were obtained. The motor unit impulse could be replayed in addition through a pre-detection facility on the tape recorder, which enabled the averages to include the period immediately preceding the motor unit discharge (see Fig. 2). A pulse height analyzer was used to derive the trigger pulse from the motor unit impulse where the acitivity of more than one unit was sampled by the needle electrode. The surface e.m.g. was averaged in two forms, rectified and unrectified. The rectified form provided a test for the possibility that the unit under study was one of a group of units tending to discharge together (MilnerBrown et al. 1973a). The average of the unrectified surface e.m.g. provided evidence of the representation of the unit on the surface e.m.g., commonly obscured by other units in a single event. RESULTS
Fig. 2 shows the average of the impulse of a unit and the correlated changes in the force. The pre-detection facility of the tape recorder enabled the examination of a period of about 10 msec before the occurrence of the motor unit impulses. Before the unit's discharge, the force is seen to be declining and to rise sharply after a short latency. The peak increase in force is 4-4 g occurring about 55 msec after the motor unit impulses. No evidence of synchronization of the units was found in the present experiments. Fig. 3 shows a typical record of the average of the needle and surface electromyograms, the latter in both rectified and unrectified form (same unit as Fig. 2). A tendency for synchronization of other units to that from which the recording was derived would be seen as an elevation of the average rectified e.m.g. of longer duration than the wave form seen on the average unrectified trace (see Milner-Brown et at. 1973a). No such feature is seen and it is reasonable, therefore, to assume that the force change is caused by the single motor unit. The rate of discharge was such that, for almost all units, a further discharge occurred during the time course of the twitch seen on the average record. This means that, even at the slowest voluntary rate, the units were contracting to a partly fused state. The possibility that this greatly affects the magnitude and time course of the twitch was examined as in previous experiments by obtaining records from a small number of units, first, at the slowest rate of discharge and second, at a higher rate. This was achieved for a total of seven units. The results are shown in Table 1 where it can be
167 ORDERLY RECRUITMENT OF MOTOR UNITS seen that whilst there is some variation in the values obtained, the effect of increased firing rate is usually small, at least over the range achieved voluntarily by the subject in these experiments. The example in Fig. 2 shows a decline in the force record immediately prior to the unit discharge and a short latency afterwards before the force record rises. Occasionally, and in one subject in particular, the records
1 00
UVI
20,uV
G
4g
50 msec
Fig. 2. The average data for 256 discharges of a single unit. The upper trace shows the signal from the needle electrode, with the trace initiated about 1O msec before the discharge of the unit. Below are shown the corresponding averages of the signal from the surface electrodes and the filtered force record. The unit had a force threshold of about 1000 g. A
I-m
100 U~V
20 pV
. D
I 50 msec
Fig. 3. The average of the surface e.m.g. The traces are: A, the average impulse from the needle electrode; B, the average of the surface e.m.g.; C, the average of the rectified surface e.m.g.; D, the zero reference for the rectified e.m.g. The record, which is from the recording of the same unit as shown in Fig. 2 and is the average of 256 sweeps, is typical of those obtained from other units and reveals no tendency for synchronization of other units (see Methods).
168 R. YEMM revealed a rising force level in the early part of the record (Fig. 4). In these cases, it was found that the firing rate of the unit had been very slow with intervals of up to 1 sec. On inspection of the average of the rectified surface e.m.g. it was deduced that the subject was voluntarily allowing the TABLE 1. Contractile properties of seven units determined from two separate recordings. The peak of an interval histogram obtained from each recording is also given, as an indication of the change in firing rate
Slow firing record A
Muscle Masseter
Temporal
1*8 0-6 0*7 40 pV
-
Contrac- Interspike tion time interval (msec) (msec) 31 85 72 120 52 90 64 105 30 80 37 75 55 105
Twitch tension (g) 09 2-4 2-6 4-6
Twitch tension (g) 1.1 2-6
3*2 3-5 1-4 2-1 1-3
Contrac- Interspike tion time interval (msec) (msec) 45 27 65 61 54 60 64 75 30 55 39 55 60 61
f
15g/
2FvI I~4i\|Rwglf
2 pV
100 msec
Fig. 4. Records from a unit which discharged very slowly. The upper record is the average of the needle electrode signal, derived from 64 single events. Below are the average force record and the average of the rectified e.m.g., the latter at high gain. The force level rises prior to the unit impulse and falls rapidly to its lowest level after reaching its peak.The rectified surface e.m.g. shows a fluctuation during the time course of the record, rising in the period immediately before the unit discharge and falling as the force record exhibits its peak.
169 ORDERLY RECRUITMENT OF MOTOR UNITS unit to stop firing, gradually raising the force of contraction to induce one or a few discharges and then relaxing (Fig. 4). This produced a gross increase in the force variation apparently associated with the unit discharge. Fig. 5 shows an example of the average force record for a unit firing in this slow, irregular fashion, compared with the record obtained for the same unit discharging more rapidly. Note that the apparent twitch measured
1 00 /IV
1 0 g|
1 g
.
75 msec
Fig. 5. The average unit impulse (upper record), and the average force records obtained (centre) from a recording where the unit was discharging at long intervals and irregularly, and (lower record) from a period where the unit was discharging at shorter intervals and more regularly. Each record is the average of 128 events.
from the moment of the unit discharge is some eight times larger in the record from the slow firing period. Results from fourteen units (nine from the masseter muscle) where the average force record showed a rise before the motor unit discharge, often with a sharp fall in force record after the apparent peak (Fig. 4), were eliminated from the study. From all other units, values for the twitch tension were obtained. Also, from the majority, the contraction time (time from onset of motor unit impulses to the peak of the force record) could be measured. Halfrelaxation-time, determined for about a third of the units detected in the first dorsal interosseous muscle in the earlier experiments (Milner-Brown et al. 1973b), was not measured. In too few units was the rate of discharge sufficiently slow and regular for the falling phase of the average force change to be reliable. Table 2A summarizes the results from four subjects for the masseter muscle and Table 2B for three subjects from the temporal muscle.
R. YEMM
170
Twitch tension Fig. 6A and B shows the relationship between twitch tension and threshold force for recruitment for one subject for units of the two muscles. Similar evidence of a relationship was found for all other subjects. Slopes were fitted to each set of data. The values, with corresponding coefficients of linear correlation, are given in Table 3 A, B. The finding is consistent, for both muscles, with that for the first dorsal interosseous muscle. There is a well organized orderly recruitment of units of increasing size during the rising levels of voluntary isometric contraction. TABLE 2. A, summary of data obtained for masseter motor units
Y)
Recruitment threshold (I ,
Subject
Range
R.L R.Y. H.F. H.G.
440-5900 50-5700 300-3700 20-3840
~~~~A
Mean + S.D. of observation 1770 + 1550 1630 ± 1310 1435 + 795 1060 + 1050
Twitch tension (g)
Contraction time (msec)
No. of unitIs 25 22 18 20
Range 0-2-15-9 0*4-33-6 1-8-24-6
0*1-10-7
A
A,%
,
-_1
Mean + No. Mean + No. S.D. of of S.D. of of observation units Range observation units 25 28-84 3-5 + 4 0 55 + 15 23 9*0 + 10-4 22 24-91 61 + 18 21 18 42-80 7.5 ± 5.7 63 + 13 16 4-2 + 3-6 20 29-67 48 + 10 19
B, summary of data obtained for temporal motor units Twitch tension (g) Contraction time (msec) Recruitment threshold (g) 11
Mean + SD. of Subject Range observation R.L. 120-2760 1210+ 1020 40-3800 1610+ 1135 R.Y. 40-2300 565+ 640 S.L.
No. of units 7 25 18
-A
5
Mean + No. of S.D. of Range observation units 7 0*6-8-7 3-3 ± 3-1 1*9-30 3 12-5 + 8*9 25 0*3-9*1 3-0+3±1 18
Range 30-75 35-66 30-76
Mean + No. S.D. of of observation units 48+ 16 6 50 + 9 25 49+ 12 17
Contraction time Fig. 7 shows the relationship between contraction time and force threshold for units of both muscles of one subject. In contrast to the finding in the hand muscle, there was in some cases a tendency for some of the units recruited at low forces to exhibit faster contraction times. Table 4 summarizes the slopes of the fitted lines for the data for all subjects. For two subjects, this relationship reached statistical significance. With twelve exceptions, the units were found to have contraction times of less than 75 msec and the mean values were similar to those for the units of the hand muscle (Milner-Brown et al. 1973b). In none of the subjects was there evidence of a discontinuity in the distribution of contraction times to suggest a division into fast and slow twitch units.
ORDERLY RECRUITMENT OF MOTOR UNITS
171
TABLE 3. A, the relationship between twitch tension and force threshold for units of the temporal muscle
Subject R.L R.Y H.F H.G.
No. of units 25 22 18 20
Slope of fitted line 1-07 0*97 0 75 0-79
s.E. of mean
±0.15 ± 0*18 +0.19 +0*10
Correlation coefficient 0*83 0*77 0 70 0 88
For each subject, the slope, s.E. of mean of the slope, and correlation coefficient are given for the line fitted to a plot of the type shown in Fig. 5. In each case the correlation coefficient was significant at the 1 % level.
B, the relationship between twitch tension and force threshold for units of the temporal muscle Correlation No. of Slope of units fitted line coefficient s.E. of mean Subject 0 93 1-04 R.L. +0-18 7 + 0 08 R.Y. 25 0*39 0*70 +0*14 18 0*71 S.L. 0*59 The data are as for A. The correlation coefficients were all significant at the 1 % level. TABLE 4. A, the relationship between force threshold and contraction time for units of the masseter muscle No. of units 23 21 16
Correlation Slope of fitted line coefficient 0*57 25-6 R.L. - 0*4 0*01 R.Y. - 10-7 0-22 H.F 19 9-2 0*58 H.G. Data are for lines fitted to semilogarithmic plots as shown in Fig. 7. The correlation coefficient is significantly different from zero for both cases where the slope of the line is positive (R.L. and H.G.) at the 1 % level.
Subject
B, the relationship between force threshold and contraction time for units of the temporal muscle No. of Correlation Slope of units coefficient fitted line Subject R.L. 0-18 6 5*8 - 4.7 R.Y. 25 0 34 S.L. 0-43 17 9-7 Data are as shown in A. The slopes were not significantly different from zero at the 5 % level.
172
R. YEMM A
o
0
10OB
10 0
@10 _
oW
0~~~~~
1.0 0
00
6
0.1 100
/
0
0 1010, Threshold force (g)
00
100
1000 10,000 Threshold force (g)
0 U twitch tensions and force thresholds for single units of the Fig. 6. A, masseter and B. temporal muscles of one subject (R.L.). The scales are logarithmic and the fitted straight line has a slope close to unity (see
Table 3). 00
100
Thrshod frce(g) 0 0 EJ 100 Maue
80
Threshold force (g)
0 Fig. 6. oe U 00 ~ muscles ~ massleter and Btemporal (0ildcrls oesbcth(R.L.) of0 0 ofthescalesr e logaithmct and tho ofFitte line s.Tragh hhwas afsloed lostemsto munt(scee daaol n Ca 20lp infcnl ifrn rmzr seTable 3). . 0000 1000 1 thresold 010,000 tichotnsrionsnoc for units singlemsee of the
0
U 0
100
1000 Threshold force (g)
10,000
Fig. 7. Measured contraction times of motor units of the masseter (open circles) and temporal muscles (filled circles) of the units of the same subjects as those of Fig. 6. The line shown was fitted to the masseter muscle data only and has a slope significantly different from zero (see Table 4). DISCUSSION
Twitch tension The force fluctuation determined for each unit may be an underestimate of the true twitch tension, because of the partially fused nature of the unit contraction under the conditions of voluntary activation. The observation that an increase in firing rate, and hence a greater degree of fusion, does not result in a consistent diminution in the force fluctuation, suggests that the underestimate is not large. The correlation observed between the twitch tension of the motor units of both masseter and temporal muscles and the force at which the units
173 ORDERLY RECRUITMENT OF MOTOR UNITS were recruited during voluntary isometric contraction seems to confirm the mechanism of orderly recruitment reported for the first dorsal interosseous muscle of the hand. As suggested previously, the arrangement may be considered appropriate to provide optimum precision of control of light forces but may not apply under other conditions, such as rapid movement. Some variation was found in the slopes of the straight lines fitted to the logarithmic plots of twitch tension against force threshold (Table 3) with the two lowest values obtained for the temporal muscle data of two subjects. It is possible that such variations indicate a real difference between the muscles or between subjects. However, they may also arise from variations in experimental situation, where the geometry of the muscles and of the site of force recording cannot be controlled. For these reasons also, no valid conclusions can be drawn on the absolute values obtained for the twitch tensions measured for units of different muscles. No evidence was obtained to indicate regional differences in the distribution of units within the relatively large muscles studied. For all but one muscle, the data were accumulated in several recording sessions; the site of entry of the needle electrode was varied, to provide a wide sample of unit locations. The only region from which no units were sampled was the extreme posterior part of the temporal muscle. Contraction times The measured contraction times for the units of the masseter and temporal muscles ranged from 24 to 91 msec (masseter), and 30 to 75 msec (temporal muscle). For the hand muscle, the corresponding range was 30-100 msec (Milner-Brown et al. 1973b). In the earlier report it was suggested that the units might therefore be classified as being composed of fast twitch fibres. Such a tentative conclusion may be equally justified for the units of the muscles examined in the present study. No evidence was obtained to suggest that there was a difference in the contraction times of the units of the masseter and temporal muscles. For the hand muscle, a weak relationship was found between contraction time and force threshold of the units. Units recruited at high forces tended to reach peak tension more rapidly. The units from the jaw muscles did not show this. Indeed, for four of the seven muscles tested, the data suggested the reverse relationship, with the smaller, low threshold units showing the faster contraction (Table 4 fitted lines with positive slopes). The relationship was statistically significant for two of these muscles (masseter muscle units, R.L., H.G.). This finding was not expected; it conflicts with that for the hand muscle and with the results obtained in animal experiments. McPhedran, Wuerker & Henneman (1965), Wuerker, McPhedran & Henneman (1965) and Burke (1967) have shown a tendency for small units
174
R. YEMM
to exhibit longer contraction times than large units. However, Stephens & Stuart (1975) have made similar observations but show that the relationships between unit size and speed of contraction depend largely on the presence of a mixed population of slow and fast twitch units in the muscle studied (cat medical gastrocnemius). Examination of the groups separately was shown to eliminate or weaken the correlations. In a few experiments, they observed a just significant increase in contraction time of units with higher conduction velocity of the axon of the motor neurone (it is generally accepted that axon conduction velocity increases with motor unit size). In any event, it seems that the relationship between force threshold or unit size and contraction time is weak, at least in those human muscles studied so far (Sica & McComas, 1971; Milner-Brown et al. 1973b). Thanks are due to Professor D. J. Anderson and Professor A. I. Darling for provision of research facilities. Dr R. J. Linden rendered great assistance, both as the first subject and during recordings of my own units. REFERENCES
BuRKE, R. E. (1967). Motor unit types of cat triceps surae muscle. J. Phy8iol. 193, 141-160. LIPPOLD, 0. C. J. (1952). The relation between integrated action potentials in a human muscle and its isometric tension. J. Phy8iol. 117, 492-499. MCPHEDRAN, A. M., WUERKER, R. B. & HENNEMAN, E. (1965). Properties of motor units in a homogeneous red muscle (soleus) of the cat. J. Neurophygiol. 28, 71-84. MILNER-BROWN, H. S., STEIN, R. B. & YEMM, R. (1973a). The contractile properties of human motor units during voluntary isometric contractions. J. Phygiol. 228, 285-306. MILNER-BROWN, H. S., STEIN, R. B. & YEMM, R. (1973 b). The orderly recruitment of human motor units during voluntary isometric contractions. J. Phy-giol. 230, 359-370. SICA, R. E. P. & MCCOMAS, A. J. (1971). Fast and slow twitch units in a human muscle. J. Neurol. P.ychiat. Lond. 34, 113-120. STEIN, R. B., FRENCH, A. S., MANNARD, A. & YEMM, R. (1972). New methods for analyzing motor function in man and animals. Brain Re8. 40, 187-192. STEPHENS, J. A. & STUART, D. G. (1975). The motor units of cat medial gastrocnemius: speed-size relations and their significance for the recruitment order of motor units. Brain Res. 91, 177-195. WUERKER, R. B., MCPHEDRAN, A. M. & HENNEMAN, E. (1965). Properties of motor units in a heterogeneous pale muscle (m. gastrocnemius) of the cat. J. Neurophysiol. 28, 85-99. YEMM, R. (1976). The muscles of mastication: the properties of their motor units, and length-tension relationships of the muscles. In Mastication, ed. ANDERSON, D. J. & MATTHEWS, B. pp. 25-32. Bristol: Wright.
163
THE ORDERLY RECRUITMENT OF MOTOR UNITS OF THE MASSETER AND TEMPORAL MUSCLES DURING VOLUNTARY ISOMETRIC CONTRACTION IN MAN
By R. YEMM* From the Departments of Oral Biology and Dental Prosthetics, University of Bristol, Medical School, University Walk, Bristol BS8 1TD
(Received 24 May 1976) SUMMARY
1. The contractile properties of the motor units of the masseter and temporal muscles of human subjects were studied during voluntary isometric contractions, using a method previously employed to examine units of a small hand muscle. 2. Over the range of forces studied (0-6 kg), the units of both muscles were recruited in an orderly fashion, with a nearly linear relationship between the voluntary force at which units were recruited and their measured twitch tensions. 3. The range of contraction times (25-90 msec) was similar to that observed for the hand muscle. In some subjects it seemed that small units, recruited at low forces, exhibited shorter contraction times. INTRODUCTION
In earlier papers, a method has been described which enables the determination of the contractile properties of human motor units during voluntary isometric contraction (Stein, French, Mannard & Yemm, 1972; MilnerBrown, Stein & Yemm, 1973a, b). The muscle used for this investigation was the first dorsal interosseous muscle of the hand. Information concerning magnitude and time course of the twitch tension of units was obtained and in addition, it was reported that recruitment of motor units during increasing voluntary contractions took place in an orderly fashion with units of increasing power being recruited after the lightest forces were achieved by means of activation of the smallest units (Milner-Brown et al. 1973b). Present address: Department of Dental Prosthetics, Dental School and Hospital, The University, Dundee DD1 4HN. *
R. YEMM The present experiments on the masseter and temporal muscles were undertaken to look for further evidence of the existence of orderly recruitment in a different region of the body. In addition, the ability to record forces of contraction between the teeth provides a superior connexion to the force transducer than anywhere else in the body. A preliminary report of some of the results from these experiments has been made elsewhere (Yemm, 1976). The technique requires that electrical activity associated with the discharge of single motor units be recorded during a voluntary contraction, together with the output of a transducer positioned to record the total contractile force. The impulses of the motor unit are used to trigger a signal averager, which sums forces recorded from the whole muscle. Provided that other units, contributing to the contraction, discharge at random intervals relative to the unit being studied, the averaging process will extract those tension fluctuations correlated to the single unit from those due to other units and other factors. 164
METHODS
A total of 149 motor units have been studied; of these, ninety-four were from the masseter muscle, recorded from four subjects. Fifty-five units from the temporal muscle were examined from three subjects. For two subjects, data have been obtained from both muscles. In each experimental session, data were recorded on an FM tape recorder for subsequent analysis. Four channels of the latter were employed to record (1) motor unit activity from a needle electrode, (2) output of the force transducer at low gain, (3) the force signal at higher gain after filtering out low frequency fluctuations, and (4) the electromyogram (e.m.g.) of the muscle obtained with surface electrodes. Motor unit activity. A bipolar needle electrode was employed. Two fine Teflon insulated silver wires 75 ,um in diameter were contained inside the needle (25 gauge), for differential amplification. The needle provided the earth and was filled with epoxy resin. The signal was amplified by means of a conventional differential pre-amplifier as in previous experiments. Tension recording. The force transducer, employing strain gauges on a stiff beam, was placed between the upper and lower first premolar teeth on the same side as the needle electrode. The mouth was opened by about 8 mm. An impression of the upper teeth in a thermoplastic compound was formed on the upper beam to enable accurate repositioning of the device from time to time. The natural frequency of the transducer was approximately 400 Hz. Voluntary contraction by the subject caused flexion of the transducer beams of the order of 1 mm/kg so that forces in the experimental range of 0-6 kg could be regarded as taking place essentially isometrically. The filter used to permit high gain recording of the force record by elimination of low frequency fluctuations in the force had a time constant of 2-5 sec, chosen to avoid distortion of the subsequent records. Surface e.m.g. Two 9 mm diameter silver disk electrodes were applied to the skin over the muscle about 2 cm apart. The signal from these electrodes was subjected to similar amplification to that used for the needle electrode. Recording procedure. For each motor unit, data were recorded for a period of about 3 min. During this, the subject was asked to maintain a steady discharge of the unit
ORDERLY RECRUITMENT OF MOTOR UNITS
165
being investigated, at as slow a rate as possible. The subject was provided with auditory feed-back. Commonly the slowest rate of firing that could be maintained was 10-15 impulses/sec. From time to time, the needle was moved within the muscle, although at many sites more than one unit could be discriminated, in which case the voluntary force was varied to recruit and record data from all units discriminated. Single units were identified on the basis of constant amplitude and regularity of discharge. For each unit, the voluntary force at which the unit was recruited was noted at the start of the recording period. Several experiments were performed on most subjects, during each of which between 5-10 units were studied. An attempt was made in each session to sample units with a wide range of force thresholds. Y V V
V
"'5 W
(U
E
TP 0
0)
00
Co
I
00
0
2000
4000
6000
Force (g) Fig. 1. The relationship between force exerted upon the transducer and the integral of the surface electromyogram (e.m.g.) of the masseter muscle. Each symbol represents the integrals obtained from the first minute of a record made of the e.m.g. and force whilst studying six units during a single experimental session. The surface electrodes remained undisturbed throughout the recording period.
The threshold force measured at the transducer was developed by a number of muscles, rather than as is the case in the earlier experiments on the hand muscle (Milner-Brown et at. 1973a, b), where most if not all of the recorded total force was due to a single muscle. Thus voluntary forces exerted upon the apparatus in the present experiments will be derived from the masseter, temporal and medial pterygoid muscles of the side upon which the apparatus was placed, with a contribution from contralateral muscles. Further, it is possible that the proportion of the total force exerted by each muscle varies with degrees of voluntary contraction. In order to investigate this, the recorded surface e.m.g. was integrated electronically. Five-second integrals of the e.m.g. were obtained together with an instant value of the force level at the mid point of the period of integration. Fig. 1 shows the relationship between force and the integrated e.m.g. during the recording of six
166
R. YEMM
units from the masseter muscle of one subject during one experimental session and provides evidence that there is an approximately linear relationship between the two. Similar results were obtained for other muscles of other subjects. Since it has been shown (Lippold, 1952) that, under isometric conditions, the integral of a surface e.m.g. is linearly related to the force developed, it is reasonable to conclude that under the experimental conditions applied in this study, the masseter and temporal muscles each exerted an approximately constant proportion of the total recorded voluntary force throughout the range of forces studied. Analy8is of data. Data were analysed using a signal averager. The motor unit impulses were used to trigger the device and the averages of the motor impulse, the high gain filtered force recorded and the surface e.m.g. were obtained. The motor unit impulse could be replayed in addition through a pre-detection facility on the tape recorder, which enabled the averages to include the period immediately preceding the motor unit discharge (see Fig. 2). A pulse height analyzer was used to derive the trigger pulse from the motor unit impulse where the acitivity of more than one unit was sampled by the needle electrode. The surface e.m.g. was averaged in two forms, rectified and unrectified. The rectified form provided a test for the possibility that the unit under study was one of a group of units tending to discharge together (MilnerBrown et al. 1973a). The average of the unrectified surface e.m.g. provided evidence of the representation of the unit on the surface e.m.g., commonly obscured by other units in a single event. RESULTS
Fig. 2 shows the average of the impulse of a unit and the correlated changes in the force. The pre-detection facility of the tape recorder enabled the examination of a period of about 10 msec before the occurrence of the motor unit impulses. Before the unit's discharge, the force is seen to be declining and to rise sharply after a short latency. The peak increase in force is 4-4 g occurring about 55 msec after the motor unit impulses. No evidence of synchronization of the units was found in the present experiments. Fig. 3 shows a typical record of the average of the needle and surface electromyograms, the latter in both rectified and unrectified form (same unit as Fig. 2). A tendency for synchronization of other units to that from which the recording was derived would be seen as an elevation of the average rectified e.m.g. of longer duration than the wave form seen on the average unrectified trace (see Milner-Brown et at. 1973a). No such feature is seen and it is reasonable, therefore, to assume that the force change is caused by the single motor unit. The rate of discharge was such that, for almost all units, a further discharge occurred during the time course of the twitch seen on the average record. This means that, even at the slowest voluntary rate, the units were contracting to a partly fused state. The possibility that this greatly affects the magnitude and time course of the twitch was examined as in previous experiments by obtaining records from a small number of units, first, at the slowest rate of discharge and second, at a higher rate. This was achieved for a total of seven units. The results are shown in Table 1 where it can be
167 ORDERLY RECRUITMENT OF MOTOR UNITS seen that whilst there is some variation in the values obtained, the effect of increased firing rate is usually small, at least over the range achieved voluntarily by the subject in these experiments. The example in Fig. 2 shows a decline in the force record immediately prior to the unit discharge and a short latency afterwards before the force record rises. Occasionally, and in one subject in particular, the records
1 00
UVI
20,uV
G
4g
50 msec
Fig. 2. The average data for 256 discharges of a single unit. The upper trace shows the signal from the needle electrode, with the trace initiated about 1O msec before the discharge of the unit. Below are shown the corresponding averages of the signal from the surface electrodes and the filtered force record. The unit had a force threshold of about 1000 g. A
I-m
100 U~V
20 pV
. D
I 50 msec
Fig. 3. The average of the surface e.m.g. The traces are: A, the average impulse from the needle electrode; B, the average of the surface e.m.g.; C, the average of the rectified surface e.m.g.; D, the zero reference for the rectified e.m.g. The record, which is from the recording of the same unit as shown in Fig. 2 and is the average of 256 sweeps, is typical of those obtained from other units and reveals no tendency for synchronization of other units (see Methods).
168 R. YEMM revealed a rising force level in the early part of the record (Fig. 4). In these cases, it was found that the firing rate of the unit had been very slow with intervals of up to 1 sec. On inspection of the average of the rectified surface e.m.g. it was deduced that the subject was voluntarily allowing the TABLE 1. Contractile properties of seven units determined from two separate recordings. The peak of an interval histogram obtained from each recording is also given, as an indication of the change in firing rate
Slow firing record A
Muscle Masseter
Temporal
1*8 0-6 0*7 40 pV
-
Contrac- Interspike tion time interval (msec) (msec) 31 85 72 120 52 90 64 105 30 80 37 75 55 105
Twitch tension (g) 09 2-4 2-6 4-6
Twitch tension (g) 1.1 2-6
3*2 3-5 1-4 2-1 1-3
Contrac- Interspike tion time interval (msec) (msec) 45 27 65 61 54 60 64 75 30 55 39 55 60 61
f
15g/
2FvI I~4i\|Rwglf
2 pV
100 msec
Fig. 4. Records from a unit which discharged very slowly. The upper record is the average of the needle electrode signal, derived from 64 single events. Below are the average force record and the average of the rectified e.m.g., the latter at high gain. The force level rises prior to the unit impulse and falls rapidly to its lowest level after reaching its peak.The rectified surface e.m.g. shows a fluctuation during the time course of the record, rising in the period immediately before the unit discharge and falling as the force record exhibits its peak.
169 ORDERLY RECRUITMENT OF MOTOR UNITS unit to stop firing, gradually raising the force of contraction to induce one or a few discharges and then relaxing (Fig. 4). This produced a gross increase in the force variation apparently associated with the unit discharge. Fig. 5 shows an example of the average force record for a unit firing in this slow, irregular fashion, compared with the record obtained for the same unit discharging more rapidly. Note that the apparent twitch measured
1 00 /IV
1 0 g|
1 g
.
75 msec
Fig. 5. The average unit impulse (upper record), and the average force records obtained (centre) from a recording where the unit was discharging at long intervals and irregularly, and (lower record) from a period where the unit was discharging at shorter intervals and more regularly. Each record is the average of 128 events.
from the moment of the unit discharge is some eight times larger in the record from the slow firing period. Results from fourteen units (nine from the masseter muscle) where the average force record showed a rise before the motor unit discharge, often with a sharp fall in force record after the apparent peak (Fig. 4), were eliminated from the study. From all other units, values for the twitch tension were obtained. Also, from the majority, the contraction time (time from onset of motor unit impulses to the peak of the force record) could be measured. Halfrelaxation-time, determined for about a third of the units detected in the first dorsal interosseous muscle in the earlier experiments (Milner-Brown et al. 1973b), was not measured. In too few units was the rate of discharge sufficiently slow and regular for the falling phase of the average force change to be reliable. Table 2A summarizes the results from four subjects for the masseter muscle and Table 2B for three subjects from the temporal muscle.
R. YEMM
170
Twitch tension Fig. 6A and B shows the relationship between twitch tension and threshold force for recruitment for one subject for units of the two muscles. Similar evidence of a relationship was found for all other subjects. Slopes were fitted to each set of data. The values, with corresponding coefficients of linear correlation, are given in Table 3 A, B. The finding is consistent, for both muscles, with that for the first dorsal interosseous muscle. There is a well organized orderly recruitment of units of increasing size during the rising levels of voluntary isometric contraction. TABLE 2. A, summary of data obtained for masseter motor units
Y)
Recruitment threshold (I ,
Subject
Range
R.L R.Y. H.F. H.G.
440-5900 50-5700 300-3700 20-3840
~~~~A
Mean + S.D. of observation 1770 + 1550 1630 ± 1310 1435 + 795 1060 + 1050
Twitch tension (g)
Contraction time (msec)
No. of unitIs 25 22 18 20
Range 0-2-15-9 0*4-33-6 1-8-24-6
0*1-10-7
A
A,%
,
-_1
Mean + No. Mean + No. S.D. of of S.D. of of observation units Range observation units 25 28-84 3-5 + 4 0 55 + 15 23 9*0 + 10-4 22 24-91 61 + 18 21 18 42-80 7.5 ± 5.7 63 + 13 16 4-2 + 3-6 20 29-67 48 + 10 19
B, summary of data obtained for temporal motor units Twitch tension (g) Contraction time (msec) Recruitment threshold (g) 11
Mean + SD. of Subject Range observation R.L. 120-2760 1210+ 1020 40-3800 1610+ 1135 R.Y. 40-2300 565+ 640 S.L.
No. of units 7 25 18
-A
5
Mean + No. of S.D. of Range observation units 7 0*6-8-7 3-3 ± 3-1 1*9-30 3 12-5 + 8*9 25 0*3-9*1 3-0+3±1 18
Range 30-75 35-66 30-76
Mean + No. S.D. of of observation units 48+ 16 6 50 + 9 25 49+ 12 17
Contraction time Fig. 7 shows the relationship between contraction time and force threshold for units of both muscles of one subject. In contrast to the finding in the hand muscle, there was in some cases a tendency for some of the units recruited at low forces to exhibit faster contraction times. Table 4 summarizes the slopes of the fitted lines for the data for all subjects. For two subjects, this relationship reached statistical significance. With twelve exceptions, the units were found to have contraction times of less than 75 msec and the mean values were similar to those for the units of the hand muscle (Milner-Brown et al. 1973b). In none of the subjects was there evidence of a discontinuity in the distribution of contraction times to suggest a division into fast and slow twitch units.
ORDERLY RECRUITMENT OF MOTOR UNITS
171
TABLE 3. A, the relationship between twitch tension and force threshold for units of the temporal muscle
Subject R.L R.Y H.F H.G.
No. of units 25 22 18 20
Slope of fitted line 1-07 0*97 0 75 0-79
s.E. of mean
±0.15 ± 0*18 +0.19 +0*10
Correlation coefficient 0*83 0*77 0 70 0 88
For each subject, the slope, s.E. of mean of the slope, and correlation coefficient are given for the line fitted to a plot of the type shown in Fig. 5. In each case the correlation coefficient was significant at the 1 % level.
B, the relationship between twitch tension and force threshold for units of the temporal muscle Correlation No. of Slope of units fitted line coefficient s.E. of mean Subject 0 93 1-04 R.L. +0-18 7 + 0 08 R.Y. 25 0*39 0*70 +0*14 18 0*71 S.L. 0*59 The data are as for A. The correlation coefficients were all significant at the 1 % level. TABLE 4. A, the relationship between force threshold and contraction time for units of the masseter muscle No. of units 23 21 16
Correlation Slope of fitted line coefficient 0*57 25-6 R.L. - 0*4 0*01 R.Y. - 10-7 0-22 H.F 19 9-2 0*58 H.G. Data are for lines fitted to semilogarithmic plots as shown in Fig. 7. The correlation coefficient is significantly different from zero for both cases where the slope of the line is positive (R.L. and H.G.) at the 1 % level.
Subject
B, the relationship between force threshold and contraction time for units of the temporal muscle No. of Correlation Slope of units coefficient fitted line Subject R.L. 0-18 6 5*8 - 4.7 R.Y. 25 0 34 S.L. 0-43 17 9-7 Data are as shown in A. The slopes were not significantly different from zero at the 5 % level.
172
R. YEMM A
o
0
10OB
10 0
@10 _
oW
0~~~~~
1.0 0
00
6
0.1 100
/
0
0 1010, Threshold force (g)
00
100
1000 10,000 Threshold force (g)
0 U twitch tensions and force thresholds for single units of the Fig. 6. A, masseter and B. temporal muscles of one subject (R.L.). The scales are logarithmic and the fitted straight line has a slope close to unity (see
Table 3). 00
100
Thrshod frce(g) 0 0 EJ 100 Maue
80
Threshold force (g)
0 Fig. 6. oe U 00 ~ muscles ~ massleter and Btemporal (0ildcrls oesbcth(R.L.) of0 0 ofthescalesr e logaithmct and tho ofFitte line s.Tragh hhwas afsloed lostemsto munt(scee daaol n Ca 20lp infcnl ifrn rmzr seTable 3). . 0000 1000 1 thresold 010,000 tichotnsrionsnoc for units singlemsee of the
0
U 0
100
1000 Threshold force (g)
10,000
Fig. 7. Measured contraction times of motor units of the masseter (open circles) and temporal muscles (filled circles) of the units of the same subjects as those of Fig. 6. The line shown was fitted to the masseter muscle data only and has a slope significantly different from zero (see Table 4). DISCUSSION
Twitch tension The force fluctuation determined for each unit may be an underestimate of the true twitch tension, because of the partially fused nature of the unit contraction under the conditions of voluntary activation. The observation that an increase in firing rate, and hence a greater degree of fusion, does not result in a consistent diminution in the force fluctuation, suggests that the underestimate is not large. The correlation observed between the twitch tension of the motor units of both masseter and temporal muscles and the force at which the units
173 ORDERLY RECRUITMENT OF MOTOR UNITS were recruited during voluntary isometric contraction seems to confirm the mechanism of orderly recruitment reported for the first dorsal interosseous muscle of the hand. As suggested previously, the arrangement may be considered appropriate to provide optimum precision of control of light forces but may not apply under other conditions, such as rapid movement. Some variation was found in the slopes of the straight lines fitted to the logarithmic plots of twitch tension against force threshold (Table 3) with the two lowest values obtained for the temporal muscle data of two subjects. It is possible that such variations indicate a real difference between the muscles or between subjects. However, they may also arise from variations in experimental situation, where the geometry of the muscles and of the site of force recording cannot be controlled. For these reasons also, no valid conclusions can be drawn on the absolute values obtained for the twitch tensions measured for units of different muscles. No evidence was obtained to indicate regional differences in the distribution of units within the relatively large muscles studied. For all but one muscle, the data were accumulated in several recording sessions; the site of entry of the needle electrode was varied, to provide a wide sample of unit locations. The only region from which no units were sampled was the extreme posterior part of the temporal muscle. Contraction times The measured contraction times for the units of the masseter and temporal muscles ranged from 24 to 91 msec (masseter), and 30 to 75 msec (temporal muscle). For the hand muscle, the corresponding range was 30-100 msec (Milner-Brown et al. 1973b). In the earlier report it was suggested that the units might therefore be classified as being composed of fast twitch fibres. Such a tentative conclusion may be equally justified for the units of the muscles examined in the present study. No evidence was obtained to suggest that there was a difference in the contraction times of the units of the masseter and temporal muscles. For the hand muscle, a weak relationship was found between contraction time and force threshold of the units. Units recruited at high forces tended to reach peak tension more rapidly. The units from the jaw muscles did not show this. Indeed, for four of the seven muscles tested, the data suggested the reverse relationship, with the smaller, low threshold units showing the faster contraction (Table 4 fitted lines with positive slopes). The relationship was statistically significant for two of these muscles (masseter muscle units, R.L., H.G.). This finding was not expected; it conflicts with that for the hand muscle and with the results obtained in animal experiments. McPhedran, Wuerker & Henneman (1965), Wuerker, McPhedran & Henneman (1965) and Burke (1967) have shown a tendency for small units
174
R. YEMM
to exhibit longer contraction times than large units. However, Stephens & Stuart (1975) have made similar observations but show that the relationships between unit size and speed of contraction depend largely on the presence of a mixed population of slow and fast twitch units in the muscle studied (cat medical gastrocnemius). Examination of the groups separately was shown to eliminate or weaken the correlations. In a few experiments, they observed a just significant increase in contraction time of units with higher conduction velocity of the axon of the motor neurone (it is generally accepted that axon conduction velocity increases with motor unit size). In any event, it seems that the relationship between force threshold or unit size and contraction time is weak, at least in those human muscles studied so far (Sica & McComas, 1971; Milner-Brown et al. 1973b). Thanks are due to Professor D. J. Anderson and Professor A. I. Darling for provision of research facilities. Dr R. J. Linden rendered great assistance, both as the first subject and during recordings of my own units. REFERENCES
BuRKE, R. E. (1967). Motor unit types of cat triceps surae muscle. J. Phy8iol. 193, 141-160. LIPPOLD, 0. C. J. (1952). The relation between integrated action potentials in a human muscle and its isometric tension. J. Phy8iol. 117, 492-499. MCPHEDRAN, A. M., WUERKER, R. B. & HENNEMAN, E. (1965). Properties of motor units in a homogeneous red muscle (soleus) of the cat. J. Neurophygiol. 28, 71-84. MILNER-BROWN, H. S., STEIN, R. B. & YEMM, R. (1973a). The contractile properties of human motor units during voluntary isometric contractions. J. Phygiol. 228, 285-306. MILNER-BROWN, H. S., STEIN, R. B. & YEMM, R. (1973 b). The orderly recruitment of human motor units during voluntary isometric contractions. J. Phy-giol. 230, 359-370. SICA, R. E. P. & MCCOMAS, A. J. (1971). Fast and slow twitch units in a human muscle. J. Neurol. P.ychiat. Lond. 34, 113-120. STEIN, R. B., FRENCH, A. S., MANNARD, A. & YEMM, R. (1972). New methods for analyzing motor function in man and animals. Brain Re8. 40, 187-192. STEPHENS, J. A. & STUART, D. G. (1975). The motor units of cat medial gastrocnemius: speed-size relations and their significance for the recruitment order of motor units. Brain Res. 91, 177-195. WUERKER, R. B., MCPHEDRAN, A. M. & HENNEMAN, E. (1965). Properties of motor units in a heterogeneous pale muscle (m. gastrocnemius) of the cat. J. Neurophysiol. 28, 85-99. YEMM, R. (1976). The muscles of mastication: the properties of their motor units, and length-tension relationships of the muscles. In Mastication, ed. ANDERSON, D. J. & MATTHEWS, B. pp. 25-32. Bristol: Wright.
