Muscle weakness following dynamic exercise in humans

BURKE, R. E., D. N. LEVINE, P. TSAIRIS, AND F. E. ZAJAC. Physi- ological types and histochemical profiles in motor units of the cat gastrocnemius. J. Physiol.
1MB taille 3 téléchargements 315 vues
Muscle weakness following exercise in humans

dynamic

C. T. M. DAVIES AND M. J. WHITE Department of Physiology and Pharmacology, Medical School, Queen’s Medical Centre, Nottingham NG7 ZUH, United Kingdom

DAVIES, C. T. M., AND M. J. WHITE. Muscle weakness following dynamic exercise in humans. J. Appl. Physiol.: Respirat. Environ. Exercise Physiol. 53( 1): 236-241,1982.-Electrical stimulation of the triceps surae in five healthy male subjects showed that following l-2 h level running and uphill walking, at submaximal voltages of stimulation, exercise enhanced the twitch and tetanic responses, but the supramaximal time to peak tension (TPT), twitch (Pt ) and tetanic tensions (PO) at 10 and 20 Hz were reduced by 16 ms (-12.6%), 11 (-8.9%), 163 (- 17.5%)) and 230 N (- 18.1%)) respectively. High-frequency (50 and 100 Hz) tetanic stimulation produced qualitatively similar changes to the 20-Hz response, but the stimulus response curve for the two frequencies was different and the ratio of 20- to 50Hz response (20/50) (cf. Edwards et al., J. PhysioZ. London 272: 769-778, 1977) was voltage dependent. The reduction in P, at 100 Hz was associated with a decrease in maximal voluntary contraction (MVC). The effects of exercise on Pto and P, at 10, 20, 50, and 100 Hz were short lived and recovered within -2 h. In contrast box-stepping produced a greater fall in P, and P, at 10 and 20 Hz, which was long lasting (at least 22 h),Oand there was a consistent fall in the 20/50 ratio. A 2-min “fatigue” test showed that the muscles were weaker but not more fatigable after exercise. Our results seriously question the validity of using submaximal stimulation voltages and ratios for testing human muscle function and suggest that long-lasting muscle weakness is not associated with recovery from prolonged walking, running, and only observed after box-stepping exercise.

work; fatigue; electrically stimulated twitch and tetanic tensions

contractions;

strength;

HAVE BEEN SEVERAL STUDIES Of mu&! fatigue resulting from sustained isometric contractions (2, 12, P5, 16, 19), but only Edwards et al. (13) have presented observations on the effect of dynamic exercise on electrically stimulated muscle force generation in humans. They reported that box-stepping and in one case cycling produced profound effects on the tetanic muscle tension of the quadriceps at submaximal voltages. The ratio between 20- and 50-Hz tetanic tensions (20/50) was reduced by 50% following exercise and only recovered slowly over the next 24 h. They suggested that the reduction in the 20/50 tension ratio resulted from the loss of low-frequency (20 Hz) force due to excitationcontraction uncoupling, a phenomenon that (they suggest) is probably a common feature of recovery from most forms of moderate to intense physical activity. We have sought to examine and extend their observations by investigating three different forms of exercise, THERE

236

running, uphill walking, and stepping, using both submaximal and supramaximal twitch and tetanic stimulation of the triceps surae. MATERIALS

AND

METHODS

Five healthy male subjects were studied. Their physical details were as follows: age 25.0 t 5.5 yr, weight 78.4 t 5.5 kg, and height 181.5 t 4.4 cm. Each subject was investigated before and after 1 h of level running and uphill walking at a gradient of 25% on a motor-driven treadmill. In addition one subject box-stepped on and off a 0.5-m platform for 1 h. The experimental protocol was the same for each subject. Twitch responses of triceps surae were elicited by applying single electrical stimuli of 50 ,us duration controlled by a Digitimer through two pad electrodes, one placed over the tendocalcaneous and the other over the heads of the gastrocnemius. Tension development was measured using an isometric dynamometer (11). The subjects were seated and the leg was clamped in position with the thigh horizontal and the ankle at a set angle corresponding to a position that gave the “optimum” supramaximal twitch tension. This was determined in preliminary experiments and varied between 78 and 82” for the five subjects. The clamp transmitted the upward force to a transducer mounted on the frame of the dynamometer. The transducer was a thick steel bar on which two strain gauges were bonded to the lower and upper surfaces. The strain gauges formed one-half of a wheatstone bridge circuit, the output of which was amplified and displayed on a UV recorder and storage oscilloscope. Calibration of the dynamometer was by a 1O:l lever system mounted on the frame above the transducer. The transducer gave a linear response up to 2,500 N, a force greater than that which could be exerted maximally by our subjects. Three twitches at I-s intervals were recorded at a variety of intensities of stimulation until the tension developed showed no further rise with increasing voltage. The criterion for supramaximal twitch tension (Pt ) was that three “plateau” values should agree within’ &5%. Once Pt had been established tetanic stimulation was applied at successively higher frequencies of 10, 20, 50, and 100 Hz for 2 s at each frequency, forming a continuous 8-s stimulus train. Voltage increased in a stepwise manner, and 2 min were allowed between each train. The criterion for supramaximal lo- and 20-Hz (Poloand P,,)

0161-7567/82/0000-0000$01.25

Copyright

0 1982 the American Physiological Society

C. T. M.

tetanic

DAVIES

tensions

AND

M.

237

J. WHITE

was the same as that

applied

to Pt . The

50- and 100”Hz stimulation were discontinued at a voltage level determined by the subject. This procedure invariably resulted in tetanic tensions at 50 and 100 Hz that approximated to the subject’s maximal voluntary contraction (MVC) but were not supramaximal in terms of the criterion applied to the twitch and lower frequency stimulation. Subjects were asked to perform three maximal voluntary contractions (MVC) of the triceps surae 2-3 min

after

the last set of tetani.

During

the recovery

17) and a significant

change

in the speed

of contraction

of the triceps surae. Following exercise, time to peak tension (TPT) decreased from 127 t 7 to 111 t 10 ms (P < 0.001)

but

half-relaxation

time

(%RT)

remained

un-

changed (P > 0.05; see Table 1). The fall in PtO was greater after uphill walking than after running and was related to the severity of the fatigue. There was no loss of maximal twitch force until the maximal (20 Hz) tetanic force had fallen by approximately 15% (Fig. 2). Posttetanic

potentiation

was observed

at rest

and

remained

period from exercise, an electrically stimulated test of muscular fatigue was used. The test consisted of trains of

unaffected by exercise. The relationship of maximal twitch tension (Pt max)t o initial twitch tension before and

stimuli

upon

at 20 Hz lasting

300 ms repeated

every second

for

the cessation

of work

was Ptmax (N) = -8.75

+ 1.52

2 min. The test was applied 30 min after the cessation of exercise; control measurements were performed in separate experiments. The speed of running and walking was adjusted so that the oxygen intake corresponded to approximately 70% of

P, (N); r = +0.70. The response to tetanic low- (20 Hz) and high-fre-

the predicted

after

maximal

aerobic

power

output

(vozmax)

for

each subject. The box-stepping experiment also corresponded to -70% VOW maxand was conducted at a fixed rate

of 20 steps/min

with

a box

height

of 0.5 m. The

twitch and tetanic stimulation procedure was applied to the triceps surae at 15 min, 1 h, and 2 h after the cessation of the running

and walking

exercise.

Following

box-step-

quency

(100

Hz)

stimulation

under

control

conditions

and following running and walking uphill are shown in Figs. 3 and 4. Supramaximal stimulation both before and exercise

was achieved

on all subjects

at 10 and 20

Hz; the mean data are summarized in Table 1. Prior to exercise, plateau values for the two frequencies of stimulation averaged 931 t 98 and 1,267 $- 157 N, respectively. At 20 Hz the twitch-to-tetanus ratio (Pt /p,,) was 0.09 before exercise commenced and rose to “0.11 (P < 0.05) 15 min

after

the

cessation

of work.

Following

exercise

ping additional measurements were made at 4 and 22 h.

low=-frequency tetanic stimulation at a submaximal volt-

RESULTS

TABLE

Running and walking. The curves relating twitch tension to stimulus voltage before and after uphill walking are given in Fig. 1, and the mean supramaximal twitch (Pt ) data are summarized in Table 1. At submaximal voltages PtO was enhanced, but PtOat supramaximal stimulation was depressed, following exercise. Before uphill walking and level running, Pt averaged 123 t 18 compared with 112 $- 18 N (P < 0.05) on the cessation of work. The fall in plateau tension was associated with a distinct inflection in the twitch curve response (cf. Ref. I

1

I

I

I

I

II

1. Contractile properties of the triceps surae before and after exercise TPT, ms

Before

127 +7

After

P t09 N

1/2RT9 ms

77 tll

76

lll$ HO

p

O20’

N

pt”/po,,

123

931

1,267

+18

t98

t157

to.02

1,037$ tl16

kO.02

112* 218

tll

P ~107N

768-f t,132

MVC,

0.09

1,979 _t445

1,807-t t478

0.11”

Values are means t SD, n = 10 observations on 5 subjects. TPT, time to peak tension; % RT, half relaxation time; Pt,,, maximal twitch tension; P,,,, PozO, tetanic tensions at 10 and 20 HZ, respectively; ratio; MVC, maximal voluntary contracP tpo20, twitch-to-tetanus tion. * P < 0.05. t P < 0.01. $I P < 0.001. 0

0

100

90

80 0

tir *

70

60

'

I

I

1

1

I

1

I

20

30

40

50

60

70

80

Stimulus

.

(VI

1. Twitch tension response curves before, and during recovery from uphil walking at 3.65 km for 2 h. SoLid circles, control; open circles, 15 min gZes, 2 h postexercise.

20

10

FIG.

l

immediately h-l, +25% postexercise;

after, gradient, trian-

N

%APo FIG.

to change

2. Relative

change in supramaximal

30

40

at 20 Hz

in maximal twitch tension (%I’, tetanic forces at 20 Hz (%AP,).

) in relation

238

C. T. M.

*

I

I

I

Low (20 Hz) frequency

I

I

1

60

70

before

and after

response

z .-g900 c” e 600

20

30

40

50

Stimulus

(VI

FIG. 3. Low-frequency (20 Hz) tetanic ning. Symbols as in Fig. 1.

2200

.

High WI0 Hz) frequency

tension

MVC (control\

response

80 run-

t+

2an

1800 2z ‘$iOO c 1400

1200

10

20

30

40

50

Stimulus FIG.

AND

M.

J. WHITE

ratios were voltage dependent (Table 2). Under control conditions the ratio varied from 0.41 at low voltage to 0.77 at the highest voltage used for high-frequency stimulation. Following exercise the 20/50 ratio was still voltage dependent but at the lower voltages gave consistently higher values; only at the highest voltages was a small fall in ratio observed. The control and postexercise maximal ZO-Hz forces were related to the MVC exerted by each subject but the association (r = 0.74; P c 0.001) accounted for ~50% of the total variance of the two variables. The association was improved if high-frequency (50 Hz) forces were considered (r = +0.95; P < 0.001) for the data available (17 observations on 3 subjects, Fig. 5). The decrease in MVC following exercise was of the order of 8.6%, the mean value being 1,979 t 445 before compared with 1,807 t 478 N after exercise. Box-stepping. The results from the box-stepping experiments performed on a single subject are summarized in Fig. 6. The twitch and tetanic responses (10 and 20 Hz) were enhanced at low-stimulation voltages and the plateau tensions were reduced following exercise in a manner similar to that observed for running and walking, but recovery was more gradual. The reductions inPto and P,, were -28 (19.4%) and -319 N (22.8%) in the right leg and -17 (12.2%) and -363 N (27.4%) in left leg. At 10 Hz there was evidence of “sag” [cf. Burke et al. (7,8)] in the supramaximal tension response (Fig. 7). At submaximal voltages although both the high- and low-frequency tetanic responses were enhanced, there was a consistent fall in the 20/50 ratio. In the 2-min fatigue test, which was applied 30 min after cessation of the three forms of exercise, the initial (20 Hz) tetanic forces reflected those given in Table 1 for running and walking and above for box-stepping. How2. Relationship between tetanic forces at 20 and 50 Hz (20/50 ratio) before and after exercise

TABLE

.‘

running.

DAVIES

4. High-frequency

Symbols

(100 Hz) as in Fig. 1. MVC,

60

70

80

(VI

tetanic maximal

tension before and after voluntary contraction.

Voltage

Before

25

30

35

40

0.41

0.54 0.15

0.64

kO.08

0.61* kO.12

0.64 to.11

to.13

age produced an enhanced response, but, as with the twitch, supramaximal stimulation produced plateau tensions that were reduced significantly (P < 0.01). The mean falls in lo- and ZO-Hz forces at supramaximal voltages were 163 (-17.5%) and 230 N (-18.1%), respectively (Table 1). The high-frequency (50 and 100 Hz) tension response at submaximal voltages 15 min after exercise in all subjects studied was increased by the same order of magnitude as that described for low-frequency stimulation. In the three subjects who were stimulated close to supramaximal levels at 50 and 100 Hz the forces attained at the higher stimulation voltages were reduced (Fig. 4). The changes in the forces developed in response to supramaximal tetani were relatively short lived and had returned to near normal levels after 2 h of recovery. The ratio between the low- (20 Hz) and high-frequency (50 Hz) supramaximally stimulated forces reflected the changes in absolute values, but because of the different stimulus-response curves for the two frequencies the

After

0.50* kO.21

Values are means value before exercise:

45

50

0.68 to.12

0.74 kO.13

0.76 to.07

0.66 kO.06

0.71-t to.03

0.72* to.02

t SD. Significantly different * P < 0.05. -/- P < 0.01.

I

I

I

I

1000

from

1

55 0.77 to.03

corresponding

1

zoo0

Tension (N)

5. Relationship of maximal high-frequency electrically stimulated FIG.

voluntary tension.

contraction

(MVC)

to

MUSCLE

WEAKNESS

2.a5

1200

gQ> 900 .-v c g 600

20

30

40

50

60

70

20 Stimulus

6. Twitch

FIG.

and 20-Hz

tetanic

tension

responses

before

30

40

50

60

Symbols

as in Fig. 1.

70

(VI

and after

1 h of box-stepping.

-500 (N) -0

60 bu

-

i, A

B

7. Traces of original tetanic stimulation records at supramaxima1 voltage before (A) and after (B) box-stepping exercise, showing depression of tension at 10, 20, 50, and 100 Hz and presence of “sag” at lowest frequency of stimulation. FIG.

ever, when expressed in relative terms (A%) the loss of tension during the 2-min test was comparable in the three forms of exercise and was similar to that found under resting conditions (Fig. 8). DISCUSSION

Our results are in agreement with Edwards et al. (13) for box-stepping but not for the other forms of exercise studied (Figs. 1, 3, and 4), though we doubt if it is correct to term the loss of twitch and tetanic tension observed as indicative of fatigue as stated in their studies. Indeed the classic signs of fatigue per se are absent following exercise in the present investigation. TPT actually decreases (rather than increases), %RT remains unchanged, and the rate of relative force loss in response to a standard fatigue test (7) is the same following exercise. The leg muscles become weaker but retain their ability to sustain a given force albeit at a lower level (Fig. 8). We would

50

:

.. I

I

30

60

I

90

I

.

120

Time (s) FIG. 8. Relationship of relative change in 20-Hz tetanic tension, expressed as percentage of initial value, to time during 2-min fatigue test modified from Burke et al. (7).

thus prefer the term muscle weakness rather than fatigue to describe our results. Under control conditions our TPT and %RT data agree closely with work from this laboratory (11) and other workers in the field for the triceps surae (5). The decrease in TPT of 16 ms (P < 0.001) following exercise is of the same order as we have found for passively heating the leg muscles by water immersion (11). An increase in the speed of contraction of the muscle was associated with a change in the shape of the twitchtension curve. After box-stepping, running, and uphill walking the twitch curve was no longer smooth but was characterized by a distinct inflection. A curve of this type has been noted by others (18), and the work of Biscoe and Taylor (3) suggests that it may be due to a relative change in force-generating capabilities of the different motor units within the muscle. A selective loss of force

240 generation from slow fibers would be consistent with the observed decreased TPT of the triceps surae following exercise (Table 1). It is also worth noting that in this context, a change in the motor unit population contributing to force generation following exercise may be indicated by the presence of sag in unfused tetani at a frequency of 10 Hz. Sag has been reported in whole cat gastrocnemius ( though not soleus) by Cooper and Eccles (10) and more recently in single motor units of cat gastrocnemius by Burke et al. (7,8). The presence of sag at certain stimulus frequencies is used by Burke to classify motor units, and its appearance, particularly following box-stepping exercise (Fig. 7), may be indicative of the increased speed of contraction of the whole muscle and relatively greater contribution of the faster contracting fibers to force generation following exercise. It should be noted that the twitch/tetanus ratio only increases slightly after exercise (Table I), and posttetanic potentiation, which appears to be a characteristic of the human triceps surae [Davies et al. (ll)] in contrast to animal slow-twitch muscle (6), is unaffected by work. Both these observations underline the unchanging nature of the fiber composition of the muscle before and after exercise and the fact that it is merely the force contributions of the different fiber types that change. Tetanic tensions showed changes similar to those outlined for the twitch. At supramaximal voltages of stimulation, lo- and 20-Hz plateau forces were reduced by approximately 18% immediately after exercise (Table 1) but recovered rapidly within 2 h (Fig. 3). At submaximal voltages tetanic tensions were enhanced; at a given voltage of 40 V forces at 10 and 20 Hz were increased by 172 and 167 N (P < O.OOl), respectively. This pattern of enhancement of low voltages and depression at maximaland supramaximal-voltage stimulation was also apparent in the high-frequency response (Fig. 4). These findings underline the difficulty of using submaximal forces and low-to-high frequency (20/50) ratios (13) to assess muscle function following exercise in humans. Under control conditions the 20/50 ratio is voltage dependent (Table 2), which probably reflects a changing pattern of fiber recruitment with increasing voltage of stimulation. It is known from animal experiments that the larger the diameter of an axon the lower its threshold to an electrical stimulus applied through distant external electrodes and the higher its conduction velocity. Further, the higher the conduction velocity of an axon the faster the contraction time of the muscle fibers it innervates. The faster a motor unit contracts the higher will be the forces produced by tetanic stimulation at 50 relative to 20 Hz. Thus, since exercise lowers the tetanic stimulation voltage threshold (Fig. 4), one might expect the 20/50 ratio not only to be voltage dependent but to actually increase in magnitude at low voltages following work, which is the case (Table 2). One could conclude that the muscles did or did not show low-frequency fatigue [as defined by Edwards et al. (13)] depending on the voltage chosen. Equally, if one based one’s findings on absolute submaximal tension one would have to conclude that exercise effected an increased capacity of the muscle to generate force. Clearly this is not so; exercise of moderate intensity (-70%00 2max) for prolonged (l-2 h) periods results in a

C. T. M.

DAVIES

AND

M.

J. WHITE

decline in supramaximal tetanic tensions in the immediate postwork period and, from the limited data we have, is not limited to the lower frequencies of stimulation following exercise. The leg muscles become weaker across the frequency spectrum, the decline in high-frequency tensions being reflected in the loss of MVC (Fig. 5). The use of submaximal tensions and ratios serves, in our view, only to misguide, and the reporting of results in these terms should be discouraged particularly if dynamic exercise is involved. On the basis of our present results we would seriously question the validity of the technique of Edwards et al. (14) for testing muscle function in normal subjects and patients. In our view, investigations into muscle weakness and fatigue in humans should be based, as in animal experimentation, on supramaximal twitch and tetanic stimulation, which ensures the complete activation of the whole muscle and allows results to be expressed in absolute tensions. Box-stepping. Clearly box-stepping has an equally profound effect on twitch and low-frequency tetanic tensions as observed in uphill walking (cf. Figs. 1 and 6). The reduction of PtO (-23 N) and PozO (-341 N) is approximately the same in both forms of work, the major difference being in the time course of the recovery processes. Whereas in uphill walking (and running) control values are regained within 2 h, in box-stepping the recovery of tension extends over a period of at least 22 h (Fig. 7). Indeed, the subject reported muscle soreness for 5-7 days following box-stepping, which suggests that the aftereffects of the exercise were qualitatively more severe than even the loss of muscle tension suggest. The actual energy expenditures in the three forms of exercise were similar and subjects perceived (on a rated scale) (4) their exertion in running uphill, walking, and box-stepping similarly. The major difference among the three forms of work was the extent to which the muscles of the leg were required to contract eccentrically and concentrically. In uphill walking, Margaria (17) has shown that beyond a gradient of +15% the work performed by the body is wholly positive, but in running the center of gravity of the body is raised and lowered and the leg muscles are required to perform negative work during the final phase of each stride. However, the muscle tensions involved at the speeds we used are relatively small (9). In box-stepping the work performed by the muscles is similar to level running but the height (0.5 m) to which the body weight has to raise and lower is greater with a subsequent large increase in the forces involved. It is known that a muscle generates more force when it is stretched during the contracted phase of movement, and it is this factor that may have contributed to the pattern of tension loss and slow recovery observed. We are currently investigating this possibility in our laboratory. The pattern of highfrequency force enhancement at submaximal voltages and possible loss (as judged by the fall in MVC, Fig. 5) at high stimulation voltages is similar to that described for running and uphill walking. The relative increase of highfrequency tension at lower voltages in box-stepping is greater than the enhancement of both lo- and 20-Hz forces, which causes a fall in the 20/50 ratio. This is in accord with Edwards et al’s (13) observations as is the recovery time-course data, though again the 20/50 ratios

MUSCLE

241

WEAKNESS

were voltage dependent, and one cannot necessarily claim as they do that the exercise effect is specific to the low-frequency forces. Further, the long-lasting loss of tetanic tension is certainly not to be regarded as a common feature (as they claim) of exercise; in this study it is limited to box-stepping. Running and walking are characterized by rapid recovery of both low- and high-frequency tetanic tensions (Figs. 1, 3, and 4). This is an important conclusion, which at least accords with common experience and observation. The majority of normal

healthy people are able to participate in most forms of rhythmic dynamic exercise without experiencing, as Edwards et al.‘s (13) study suggests, a 50% loss of muscle force the following day, and certainly within the confines of a laboratory it is possible to subject people to concentric exercise and electrical stimulation on successive days and gain reproducible results (11). Received

20 April

1981; accepted

in final

form

21 January

1982.

REFERENCES 1. BAWA, P., AND R. B. STEIN. Frequency response of human soleus muscle. J. NeurophysioZ. 39: 788-793, 1976. 2. BIGLAND-RITCHIE, B., D. A. JONES, G. P. HOSKING, AND R. H. T. EDWARDS. Central and peripheral fatigue in sustained maximum voluntary contractions of human quadriceps muscle. CZin. Sci. MOL. Med. 54: 609-614, 1978. 3. BISCOE, T. J., AND A. TAYLOR. The effect of admixture of fast and slow muscle in determining the form of the muscle twitch. 1Med. BioZ. Eng. 5: 473-479, 1967. 4. BORG, G. Perceived exertion as an indicator of somatic stress. Stand. J. Rehab. Med. 2: 92-98, 1970. 5. BUCHTHAL, F., AND H. SCHMALBRUCH. Contraction times and fibre types in intact human muscle. Acta. Physiol. Stand. 79: 435-452, 1970. 6. BULLER, A. J., 6. J. KEAN, K. W. RANATUNGA, AND J. M. SMITH. Characteristics of post-tetanic depression of isometric twitch response in cat slow twitch muscle. J. Physiol. London 217: 62-63, 1971. 7. BURKE, R. E., D. N. LEVINE, P. TSAIRIS, AND F. E. ZAJAC. Physiological types and histochemical profiles in motor units of the cat gastrocnemius. J. Physiol. London 234: 723-748,1973. 8. BURKE, R. E., P. RUDOMIN, AND F. E. ZAJAC. The effect of activation history on tension production by individual muscle units. Brain Res. 109: 515-529, 1976. 9. CAVAGNA, C. A., F. P. SAIBENE, AND R. MARGARIA. Mechanical work in running. J. AppZ. PhysioZ. 19: 249-256, 1964. 10. COOPER, S., AND J. C. ECCLES. The isometric responses of mam1930. malian muscles. ,J- PhvsioZ. London 69: 377-385.

11. DAVIES, C. T. M., I. K. MECROW, AND M. J. WHITE. Contractile properties of the human triceps surae with some observations on the effects of temperature and exercise. Eur. J. AppZ. PhysioZ. Occup. PhysioZ. In press. 12. EDWARDS, R. H. T., D. K. HILL, AND D. A. JONES. Effect of fatigue on the time course of relaxation from isometric contractions of skeletal muscle in man. J. PhysioZ. London 227: 26-27, 1972. 13. EDWARDS, R. H. T., D. K. HILL, D. A. JONES, AND P. A. MERTON. Fatigue of long duration in human skeletal muscle after exercise. J. Physiol. London 272: 769-778, 1977. 14. EDWARDS, R. H. T., A. YOUNG, G. P. HOSKING, AND D. A. JONES. Human skeletal muscle function: description of tests and normal values. CZin. Sci. MOL. Med. 52: 283-290, 1977. 15. JONES, D. A., B. BIGLAND-RITCHIE, AND R. H. T. EDWARDS. Excitation frequency and muscle fatigue: mechanical responses during voluntary and stimulated contractions. Exp. NeuroZ. 64: 401-413, 1979. 16. LIND, A. R. Muscle fatigue and recovery from fatigue induced by sustained contractions. J. PhysioZ. London 127: 162-171, 1959. 17. MARGARIA, R. Positive and negative work performances and their efficiencies in human locomotion. Int. 2. Angew. PhysioZ. 25: 339351, 1968. C. D., AND J. C. MEADOWS. Effect of adrenaline on the 18. MARSDEN, contraction of human muscle. J. PhysioZ. London 207: 429-448, 1969. voluntary 19. STEPHENS, J. A., AND A. TAYLOR. Fatigue of maintained muscle contraction in man. J. Physiol. London 220: 1-18, 1972.