Vol. 83, Issue 5, 1557-1565, 1997

Mechanical behavior of skeletal muscle during intermittent voluntary isometric contractions in humans

N. K. Vøllestad¹, I. Sejersted², and E. Saugen²

¹ Section for Postgraduate Studies in Health Science, and ² Department of Physiology, National Institute of Occupational Health and University of Oslo, 0316 Oslo, Norway

ABSTRACT

INTRODUCTION

METHODS

RESULTS

DISCUSSION

ACKNOWLEDGEMENTS

FOOTNOTES

REFERENCES

ABSTRACT

Vøllestad, N. K., I. Sejersted, and E. Saugen. Mechanical behavior of skeletal muscle during intermittent voluntary isometric contractions in humans. J. Appl. Physiol. 83(5): 1557-1565, 1997.Changes in contractile speed and force-fusion properties were examined during repetitive isometric contractions with the knee extensors at three different target force levels. Seven healthy subjects were studied at target force levels of 30, 45, and 60% of their maximal voluntary contraction (MVC) force. Repeated 6-s contractions followed by 4-s rest were continued until exhaustion. Contractile speed was determined for contractions elicited by electrical stimulation at 1-50 Hz given during exercise and a subsequent 27-min recovery period. Contraction time remained unchanged during exercise and recovery, except for an initial rapid shift in the twitch properties. Half relaxation time (RT_1/2) decreased gradually by 20-40% during exercise at 30 and 45% of MVC. In the recovery period, RT_1/2 values were not fully restored to preexercise levels. During exercise at 60% MVC, the RT_1/2 decreased for twitches and increased for the 50-Hz stimulation. In the recovery period after 60% MVC, RT_1/2 values declined toward those seen after the 30 and 45% MVC exercise. The force oscillation amplitude in unfused tetani relative to the mean force increased during exercise at 30 and 45% MVC but remained unaltered during the 60% MVC exercise. This altered force-fusion was closely associated with the changes in RT_1/2. The faster relaxation may at least partly explain the increased energy cost of contraction reported previously for the same type of exercise.

energy cost; fatigue; relaxation rate; skeletal muscle

INTRODUCTION

MUSCLE FATIGUE is often associated with slowing of contractile speed (see Refs. 16, 37 for review). This has been demonstrated by reduced rates of force generation and relaxation following intense sustained voluntary contractions and tetanic stimulation of animal muscle in vitro (5, 9, 12). Much less is known about the changes in contractile speed during more normal activation patterns of human skeletal muscle. It is well documented that submaximal voluntary contractions are generated by only a fraction of the motor units and that they are often activated by subtetanic excitation rates (2, 14, 25). Other studies demonstrate a temporal rise in the myoelectrical activity as assessed by the integrated electromyogram amplitude (2, 23). This increase is probably caused in part by a gradual recruitment of fresh motor units, but an increased rate of excitation in units already active could probably also contribute (2). Studies of physiological responses during fatigue from low-force voluntary isometric activity thus introduce additional degrees of freedom not present in other studies. Nevertheless, the interpretation of changes seen during this type of exercise is important to understand the physiological responses to prolonged voluntary muscle activity.

During exercise consisting of voluntary low-force repetitive contractions, Vøllestad and co-workers (35) demonstrated a gradual decline in force-generating capacity, without the concomitant chemical changes seen during sustained ischemic contractions (12, 18). Repetitive isometric contractions also induce a progressive rise in energy cost of contraction, demonstrated by an increased oxygen uptake in the muscle (27, 36) and a higher rate of metabolic heat production during contraction (29). Contractile speed is closely associated with the metabolic heat production and energy cost of contraction (16, 39, 41), suggesting that the rate of force generation and relaxation may increase with fatigue from repetitive isometric exercise. Thus the exercise-induced changes in muscle energetics and activation pattern occur in a direction opposite to that seen during sustained voluntary or electrically stimulated contractions (4, 5, 12, 18).

Prolonged submaximal exercise is associated with a fatigue-induced recruitment of the faster type II muscle fibers, which have a higher energy cost of contraction compared with type I (6, 7). To what extent this increased activation of type II fibers causes a temporal rise in energy turnover can be examined by comparing the responses during repetitive isometric exercise at different target force levels. With an increased target force, a larger proportion of type II fibers will be activated from start, and the rise in energy cost of contraction and, consequently, contractile speed is expected to be smaller.

The aim of the present study was to investigate the changes in muscle contractile speed during and after voluntary repetitive isometric contractions carried out until exhaustion and to compare the changes at different contraction levels. This was accomplished by examining time to peak force and half relaxation time (RT_1/2), together with the maximum rate of force change, when the muscle was stimulated electrically at stimulation frequencies from 1 to 50 Hz. Furthermore, we wanted to examine the effects of fatigue and altered contractile speed on the force-fusion properties during submaximal tetanic activation. The relative force oscillation amplitude was examined when the muscle was stimulated at rates similar to those recorded during voluntary contractions at submaximal force levels, i.e., 10-20 Hz (2, 25). Preliminary reports have been given elsewhere (31, 34).

METHODS

Subjects. Five healthy men and two healthy women (age 22-30 yr, height 170-182 cm; weight 59-83 kg) volunteered to participate in the present experiments. They were all physically active students but not particularly trained for any sports activity. After a routine medical examination, an informed consent to participate was obtained. During the last 14 days before the experiment, the subjects were familiarized with the experimental setup and the exercise protocol. The study was approved by the Regional Ethics Committee for Medical Studies, Norway.

Protocol. All subjects participated in three separate experiments, with repetitive isometric contractions at target forces of 30, 45, and 60% of maximal voluntary contraction (MVC). The experiments were carried out on separate days with at least 1-wk interval. All subjects carried out the 30% MVC experiments first. The order of the experiments at the two highest force levels was randomized. The exercise was carried out while the subjects were sitting with the knee flexed at an angle of ~80° and with the back inclined backward at a hip angle of ~110°. The subjects' hips were strapped to the seat to ensure minimal movement. The force generated with each of the knee extensors was measured with a strain gauge connected to a padded loop around the ankle. Voluntary contractions were generated bilaterally, but electrically evoked contractions were studied in the right leg only. The force responses during maximal voluntary effort and during electrical stimulation at 1 (single twitch), 10, 15, 20, and 50 Hz were examined before the repetitive exercise. The corresponding resting values were similar at the three experimental days (Table 1). MVC force was determined as the highest force that could be held for 2 s in three consecutive maximal efforts. Unfatigued mean values ± SD for MVC force were 478 ± 143, 479 ± 163, and 466 ± 151 N for the experiments requiring 30, 45, and 60% MVC, respectively. The subjects thereafter rested in the chair for 5 min before control data for stimulated contractions were obtained. With an interval of 60 s, a twitch contraction and trains of 10 pulses at the different test stimulation frequencies were given. Two sets of force responses from the electrically stimulated contractions were obtained from the unfatigued muscle. Average values of the two contractions at each frequency were used as the preexercise control values.

Table  1.   Preexercise control values for force and contractile speed Stimulation Frequency Target Force, %MVC Force, N CT, ms MCR, ms1 RT1/2, ms MRR, ms1 Twitch 30 31 ± 3  41.4 ± 3.0  21.9 ± 0.9  50.9 ± 5.7  13.5 ± 1.5  45 35 ± 2  38.5 ± 0.7  22.1 ± 0.5  61.2 ± 9.8  12.0 ± 1.5  60 35 ± 4  40.9 ± 2.2  22.1 ± 0.7  53.9 ± 5.5  13.6 ± 1.6  10 Hz 30 62 ± 10  90.4 ± 7.2  7.2 ± 0.7  45 74 ± 8  90.4 ± 6.9  7.1 ± 0.6  60 72 ± 10  87.7 ± 8.2  7.5 ± 0.9  15 Hz 30 90 ± 15  81.9 ± 4.3  8.5 ± 0.6  45 100 ± 12  83.8 ± 3.2  8.6 ± 0.3  60 108 ± 14  83.0 ± 2.2  8.9 ± 0.4  20 Hz 30 108 ± 17  80.6 ± 3.9  9.7 ± 0.3  45 116 ± 14  80.7 ± 2.6  9.8 ± 0.4  60 135 ± 18  76.4 ± 2.4  10.4 ± 0.3  50 Hz 30 124 ± 17  138.9 ± 4.3  13.3 ± 0.8  95.8 ± 9.3  8.1 ± 0.4  45 132 ± 12  142.6 ± 4.1  13.1 ± 0.4  103.7 ± 7.6  8.3 ± 0.6  60 137 ± 17  141.1 ± 5.1  13.2 ± 0.6  80.4 ± 3.2  8.6 ± 0.4 Data are means ± SE; n = 7 subjects. CT, contraction time; RT1/2, half relaxation time; MCR, maximal rate of force generation; MRR, maximal rate of force relaxation.

The exercise consisted of repeated voluntary contractions at either 30, 45, or 60% MVC. The contractions were maintained for 6 s with a 4-s rest between them (Fig. 1) and were continued until exhaustion, defined as the point when the subjects were unable to maintain target force for the required 6 s. At given intervals during the exercise, two sequences of test contractions were performed. Test sequence A consisted of stimulation with a single pulse, 15 and 50 Hz, given in three consecutive 4-s rest periods, followed by a 3-s MVC instead of one target force contraction. Test sequence B consisted of stimulation at 10 and 20 Hz in two consecutive rest periods. Test sequence A was given after the first target force contraction and repeated after 3, 5, and 10 min and then every 5 min. In addition, this sequence was performed in the last 40 s before exhaustion. The test sequence B was given after 2, 6, and 11 min and repeated every 5 min thereafter.

After exhaustion, the subjects remained seated without moving their legs for a recovery period of ~27 min. The test sequence A was performed 3 and 5 min after cessation of exercise and then every 5 min. Before the twitch, a 6-s target force contraction was performed to facilitate twitch potentiation allowing comparison with the exercise data. The test sequence B was given after 2 and 6 min of recovery and every 5 min thereafter.

Three subjects carried out control experiments, in which no voluntary exercise was performed. The subjects sat for 60 (n = 1) or 90 min (n = 2) in the chair apparatus, and the test sequences (twitch, 10-50 Hz stimulation, and MVC) were performed as in the exercise experiments.

Electrical stimulation and force measurements. The right quadriceps muscle was stimulated through surface electrodes (Tenzcare, 3M, St. Paul, MN) placed proximally and distally over the thigh. Electrical square-wave pulses (0.5-ms duration) were generated by a stimulator (Pulsar 6bp, FHC, Brunswick, ME), and the frequency and number of pulses were controlled by a computer running a custom-made program. In preliminary experiments, it was established that peak force was obtained by a stimulation of 70-80 V. In all experiments, 90-V stimulation was used to ensure activation of the largest possible muscle mass.

Force was measured by a strain gauge (HBM U2AC2, Darmstadt, Germany) connected via a steel rod and an inflexible loop around the ankle. The time constant of the strain gauge was 5 ms. The subjects exercised with both legs, and the target force required, together with that from each contraction, was continuously displayed on a screen in front of the subject. This enabled the subjects to generate a constant target force during the voluntary contractions. Force was continuously analog-to-digital converted (Metrabyte Das-16, Keithley, Cleveland, OH) and stored on a hard disk. The data-sampling frequency was 1 kHz during the electrically evoked test contractions and 50 Hz in the other periods.

Calculations. Contraction time (CT) was calculated as the time elapsed for force to increase from 10 to 90% of peak force. The maximal rate of force generation (MCR) was calculated as the peak value of dF/dt divided by peak force, where F is force and t is time (3, 5, 40). Peak force in unfused tetanic contractions elicited by 10- to 20-Hz stimulation appeared after a variable number of pulses (cf. Fig. 4). Hence, reliable assessment of CT could only be obtained from the twitch and the 50-Hz contractions that displayed smooth force increments until peak force was reached. RT_1/2 was calculated as the time elapsed for force to fall by 50% from peak force after the last stimulation pulse. The corresponding maximal relaxation rate (MRR) was calculated as the nadir value of dF/dt divided by peak force.

Contractions generated by stimulation frequencies of 10 and 15 Hz were unfused and showed clear oscillations. The average amplitude of these oscillations (F) was determined after the initial rapid rise in tension had subsided (cf. Fig. 4). The mean force (F_m) in the same period was also determined. Any changes in contractile speed would be expected to affect F. In addition, fatigue will influence both F and F_m. The combined effect of repetitive isometric exercise on force-fusion properties and force-generating capacity was monitored as the the change in the relative force oscillation amplitude (F/F_m).

Statistics. Values are means ± SE, unless otherwise stated. One-way analysis of variance for repeated measures was performed to evaluate the temporal changes, except for the changes in F/F_m, which were tested by paired t-tests.

RESULTS

Force and endurance. The endurance times were 45 ± 5, 15 ± 1, and 5.6 ± 0.3 min for the repetitive exercise at 30, 45, and 60% MVC, respectively. The MVC force decreased gradually by ~35-40% during exercise in all experiments (Table 2). Over the 27-min postexercise recovery period, MVC force increased to 86-88% of preexercise levels in all three protocols. The final values were significantly lower than preexercise control values (P < 0.04).

Table  2.   Force responses determined at exhaustion and after 27 min of recovery Exhaustion 27-min Recovery 30% MVC 45% MVC 60% MVC 30% MVC 45% MVC 60% MVC Force, %    MVC 66 ± 5  58 ± 5  63 ± 2  88 ± 4  86 ± 4  88 ± 3    10 Hz 43 ± 3  47 ± 10  55 ± 7  56 ± 4  53 ± 5  45 ± 5    20 Hz 52 ± 5  53 ± 10  45 ± 8  67 ± 6  71 ± 6  53 ± 6 Values are means ± SE; n = 7 subjects. Data given as %preexercise control in experiments consisting of repetitive isometric contractions at 30, 45, and 60% maximal voluntary contraction (MVC).

The force response to all stimulation frequencies declined during the exercise period in all experiments, as illustrated in Fig. 2 for twitches and 15- and 50-Hz stimulation (F_tw, F_15, and F₅₀, respectively). The reductions (relative to the preexercise values) were similar to, or larger than, those observed for MVC (Fig. 2, Table 2). The rate of force loss increased with increasing target force. A notable exception from the steady decline in force during exercise was the F_tw, which increased initially, before a gradual decline was seen. This initial rise probably reflected the degree of posttetanic potentiation that occurs predominantly in type II fibers (11). With increasing target force, a larger initial rise in F_tw was observed.

In the postexercise recovery period, the time course of force recovery varied, but none of the force responses was fully restored after 27 min (Table 2 and Fig. 2). After the 30% MVC exercise, F_tw remained stable, whereas a consistent increase in the peak force of the trains at 15-50 Hz was observed (P < 0.007). In the first 3 min, F₅₀ increased rapidly before it leveled off at 80-85% of control values. A similar rapid initial increase was observed after the 45% MVC exercise for twitch force and contractions elicited by 15 Hz or more (Fig. 2). In contrast, F₁₀ remained unchanged in the recovery period (P = 0.88). After exhaustion from exercise at 60% MVC, peak force at all stimulation frequencies recovered rapidly before it declined again during the remaining recovery period (P < 0.0001).

CT and MCR. CT determined from the 50-Hz stimulation remained unaltered during both exercise and recovery at all three intensities (P > 0.23; Fig. 3). Twitch CT decreased in all experiments by ~20% after the first 6-s target force contraction. Thereafter, no further changes were seen in the fatiguing exercise or in recovery (P > 0.5). The initial decline in CT was associated with an increased MCR (Table 3). For all experiments, both CT and MCR remained virtually unchanged during the entire exercise and recovery period.

Table  3.   Contractile speed at exhaustion and after 27-min recovery in response to electrical stimulation at different frequencies Stimulation Frequency Exhaustion 27-min Recovery 30% MVC 45% MVC 60% MVC 30% MVC 45% MVC 60% MVC MRR, %    Twitch 171 ± 15  169 ± 8  132 ± 9  181 ± 36  152 ± 11  126 ± 12    10 Hz 268 ± 24  261 ± 25  124 ± 11  270 ± 32  240 ± 17  212 ± 21    15 Hz 175 ± 17  142 ± 16  105 ± 20  147 ± 16  133 ± 9  117 ± 13    20 Hz 149 ± 6  116 ± 8  71 ± 9  135 ± 4  124 ± 3  108 ± 7    50 Hz 151 ± 15  123 ± 15  85 ± 7  140 ± 15  128 ± 8  132 ± 8  MCR, %   Twitch 119 ± 5  119 ± 6  124 ± 6  126 ± 6  117 ± 3  126 ± 7    50 Hz 118 ± 8  109 ± 6  95 ± 5  104 ± 6  94 ± 1  101 ± 3 Values are means ± SE; n = 7 subjects. Data given as %preexecise control values in experiments consisting of repetitive isometric contractions at 30, 45, and 60% MVC. {/CAPT;;;left;stack}

DISCUSSION

The major finding of the present experiments was a gradual fall in RT_1/2 during fatigue from repetitive isometric exercise at 30 and 45% MVC target force while the CT remained unchanged. Furthermore, RT_1/2 values changed consistently in twitches and trains of stimuli from low to high frequency and always inversely to the MRR. These changes are in sharp contrast to those frequently observed during sustained voluntary contractions or tetanic stimulation (5, 9, 12) but in keeping with a small decline in twitch time recorded during repeated maximal contractions at low duty cycles (22). The amplitude of force oscillations F during tetani elicited by low-frequency stimulation (10 and 15 Hz) increased as the F_m declined. These changes were maintained in the recovery period. During exercise at 60% MVC, the changes in RT_1/2 values were more variable and depended on the stimulation frequency. Furthermore, repetitive isometric exercise at this highest target force induced a parallel decline in the F in unfused tetanic contractions and F_m. In the postexercise recovery period, however, the oscillation amplitude increased to values comparable to those seen after the exercise at 30 and 45% MVC.

The RT_1/2 either remained shorter than control (30 and 45% MVC exercise) in the postexercise recovery period or shifted from values longer than preexercise control values to shorter values (e.g., 50-Hz stimulation after 60% MVC exercise). The end points for RT_1/2 were thus remarkably similar among the three target force levels. A similar convergence in the recovery period was also observed for the F/F_m ratio. There was a close and inverse relationship between F/F_m and RT_1/2 under all conditions.

The methods used in this study to determine the contractile speed have been validated in numerous other studies that showed contractile slowing with fatigue of both human and animal muscle (5, 8, 12, 19). Here consistent measurements were made of both time measurements (RT_1/2 and CT) as well as rate measurements (MRR and MCR) in response to a wide range of stimulus frequencies. All seven subjects showed a gradual decline in RT_1/2 in response to five different stimulation regimes in each of two experiments carried out on separate days at different target force levels (30 and 45% MVC). Similar values were obtained from unfatigued muscle of each subject in three separate experiments on different days (Table 1), and when the subjects sat resting in the chair for 60-90 min. It is, therefore, reasonable to assume that both the constancy of CT as well as the marked changes in RT_1/2 values reflect the real mechanical properties of the muscle-tendon unit.

Mechanical responses in relation to muscle chemistry. Fatigue in human muscle in vivo is mostly studied by experimental protocols allowing a precise determination of the excitation pattern (electrical stimulation) or in situations where the excitation pattern is well known (MVCs). The results from these studies, together with data extrapolated from fatigue induced by electrical stimulation in vitro, almost unequivocally associate fatigue with a slowing of relaxation [see review by Fitts (16)]. This slowing is usually attributed to the large changes in high-energy substrate and metabolite concentrations (5, 9, 12, 18).

It was previously demonstrated that only moderate initial changes in high-energy P_i, lactate, and H⁺ levels occur during repetitive isometric exercise at 30% MVC (27, 35). Furthermore, we have previously estimated that the rate of ATP turnover during this type of exercise is well below the limits for aerobic metabolism (29, 36). The low amounts of lactate production support the opinion that the ATP demand during this type of exercise is adequately met by oxidative ATP resynthesis (27, 36). The absence of large changes in substrate and metabolite levels may explain why contractile slowing was not found during repetitive isometric exercise at the two lowest force levels. Even though we have no direct evidence of changes in metabolites in the present study, it is reasonable to assume that the 60% MVC repetitive isometric exercise may have induced larger changes in high-energy P_i and other metabolites compared with the 30 and 45% MVC exercise. These chemical changes may have caused the slowing of relaxation demonstrated for the tetanic contractions during the 60% MVC exercise. The slowing of relaxation was reversed with a time course similar to recovery of metabolite changes after exercise (17, 30). However, no known metabolic factors can explain why RT_1/2 values decreased during the low-force contractions or in the recovery period.

A recent paper by Ferrington and co-workers (15) demonstrated that prolonged exercise in rats induced an increased proportion of functional Ca²⁺ pumps at the sarcoplasmic reticulum (SR) membrane. If this response occurs also in our human model involving isometric contractions, faster relaxation is predicted. Analysis of functional Ca²⁺ pumps during repetitive isometric exercise may thus be an important step toward understanding the mechanisms behind the increased contractile speed.

Changes in muscle fiber activation pattern. The sequence in which motor units are recruited may influence the contractile properties of muscle during voluntary contractions executed at different intensities (25, 40). It has previously been shown that low-force contractions involve only a fraction of the motor-unit pool and that these are predominantly type I fibers (2, 24). As exercise progresses, these fibers fatigue and thus produce less force when a given electrical stimulation is applied. Thus an increasing fraction of the force response elicited by external stimulation will be generated by the faster type II fibers. The muscle as a whole will thus appear with faster contractile properties, even in the absence of changes at the cellular level.

At higher target forces, the initially recruited fraction of type II fibers increases. The greater twitch potentiation after the first 45% compared with the first 30% contraction indicates that a substantially larger number of type II fibers were recruited initially during this more intense exercise regime (11). Thus, at the higher target force, fatigue will develop simultaneously in both type I and in the active type II fibers, and the dominance of unfatigued type II fibers on contractile speed will be decreased. In consequence, the temporal changes in muscle activation pattern would predict a lesser decline in RT_1/2 during exercise at higher target force. We did not observe this. Furthermore, one would expect that RT_1/2 and CT would be altered in a similar manner, as they are both heavily dependent on the fiber type (1, 20, 40). The lack of changes in CT, apart from an initial change, indicates that the changes in contractile speed observed in these experiments were not primarily due to selective fatigue of type I fibers. This interpretation is in keeping with a small decline in twitch time recorded when 10-s maximal contractions with the elbow flexors were repeated every 100 or 200 s (22). Therefore, we conclude that the gradual fall in RT_1/2 reflects a change in mechanical properties at the cellular level.

Muscle temperature. The contractile speed of skeletal muscle is sensitive to temperature, exhibiting a Q₁₀ of ~2 above 25°C (33, 39). We have recently reported a 3-4°C increase in muscle temperature during 30% MVC repetitive isometric exercise to exhaustion (29), and a similar increase has been seen during 60% MVC repetitive isometric exercise (unpublished observations). The expected increase in contractile speed due to increasing muscle temperature would then be ~30%. The relaxation rates increased by 50-170% (cf. Table 3), and, hence, part of this rise could possibly be a temperature effect. However, temperature changes are shown to affect CT and RT_1/2 equally (5, 33). In addition, muscle temperature falls linearly in the recovery period (29), predicting a linear rise in RT_1/2. Therefore, the constancy of CT presently observed while RT_1/2 decreased markedly during exercise, with only little normalization in the postexercise recovery period, does not indicate that the changes in RT_1/2 primarily are a consequence of the changes in muscle temperature.

Contractile changes and muscle energetics. At the cellular level, the relaxation process is regulated by enzymes controling the rate of Ca²⁺ reuptake in the SR and, possibly, also the rate of cross-bridge detachment (12, 16). Hence, the decline in RT_1/2 during the 30 and 45% MVC exercise is expected to coincide with an increase in the rate of ATP utilization. This contention agrees well with the observed increase in energy cost during 30% MVC repetitive isometric exercise, seen as a twofold increase in oxygen uptake in the working muscle (27, 36), and a 75% increase in metabolic heat production during contraction (29). Furthermore, the decrease in RT_1/2 and increase in energy turnover follow similar time courses during exercise and show a similar slow return toward control values during postexercise recovery. The present data, therefore, strongly indicate that the decline in RT_1/2 seen during the 30 and 45% MVC exercise is caused by a higher rate of ATP utilization either at the cross bridges or at the membrane of the SR.

The changes in RT_1/2 during the 60% MVC exercise clearly differ from those seen at the lower exercise intensities in the respect that RT_1/2 increased initially from all stimulation regimes and remained elevated for the higher frequencies throughout the exercise. At exhaustion, the change in RT_1/2 seemed to be "dose" dependent, with the slowest relaxation seen from the 50-Hz stimulation. The short endurance may indicate that the energy turnover during the 60% MVC exercise is at or beyond the limit for aerobic ATP replenishment during the hyperaemic rest period between contractions. If this is true, the test contractions may perturb the balance between ATP demand and resynthesis even further. Because RT_1/2 from twitch contraction became shorter and CT did not increase, it is likely that the slowing of RT_1/2 from 50-Hz stimulation was a specific effect of this regime. Therefore, we believe that this slowing of relaxation may reflect a transient larger rise in P_i and H⁺ levels due to a disturbance of the aerobic recovery by the high-frequency pulse train.

After cessation of the 60% MVC exercise, rapid shifts in RT_1/2 values were seen. Values for RT_1/2 from 50-Hz trains declined throughout recovery from ~40% above control to a level ~20% below. The time course of the initial rapid decline is similar to that for the recovery of phosphocreatine and P_i after sustained ischemic contraction (17). These final values converged with those recorded after 30 and 45% MVC exercise. Thus all three types of exercise lead to a generally faster muscle once substrate and metabolite changes had been reversed.

Consequences of faster relaxation. Relaxation rate is an important factor in determining the contractile response of the muscle fibers during isometric contraction. It was previously documented that the motor unit discharge rate is in the order of 10-15 Hz during submaximal voluntary contractions in humans (2, 25). As shown in Fig. 4 for unfused tetanic contraction of the whole muscle during electrical stimulation, these low excitation rates result in large oscillations of the force output. The F, together with F_m, is expected to be influenced by any changes in relaxation rate, as shown for the F/F_m ratio and RT_1/2 in Fig. 6. The same relationship between these indicators of mechanical behavior was observed independent of fatigue and target force level, despite quite different changes in RT_1/2 and F/F_m ratio during repetitive isometric exercise at 60% MVC compared with 30 and 45% MVC (cf. Figs. 3 and 5).

Oscillation of isometric tension is associated with a higher cross-bridge cycling rate than during a fully fused isometric contraction, in which cross-bridge turnover is relatively slow (41). The energy cost of contraction, expressed as the ratio between the rate of ATP turnover and force, is thus expected to be higher in unfused compared with fully fused isometric contractions. In keeping with this, Wiles and Edwards (38) found that when the unfatigued human adductor pollicis muscle was stimulated at a wide range of frequencies, the energy cost of contraction was linearly and closely correlated to the F/F_m ratio. In fact, at the lowest stimulation frequencies in which the highest values for both energy cost and F/F_m were found, the energy cost of contractions was six times higher than that observed during maximal stimulation. We have recently reported a similar sixfold difference in energy cost between submaximal (10% MVC) and MVC in the vastus lateralis muscle (28). The increase in energy cost during and after the repetitive isometric contractions may thus involve a two-step process. An enhanced rate of ATP turnover at the Ca²⁺-adenosinetriphosphatase (ATPase) causes a faster relaxation. Consequently, larger tension and possibly also length adjustments consequent to the faster contractile speed occur, and this altered mechanical behavior causes a larger rate of ATP turnover at the myosin ATPase.

This hypothesis may also explain why the relationship between a change in contractile properties and energetics seems to depend on the type of contractile activity. When exhaustive 30% MVC repetitive isometric exercise was performed between two bouts of submaximal bicycling, the oxygen uptake during bicycling remained unaltered, even though a 70% rise in energy cost during repetitive isometric exercise was observed (27). We have observed similar results with dynamic knee extension exercise before and after repetitive isometric exercise (unpublished observations). An explanation for the coupling between type of contraction and the effect of fatigue on contraction efficiency may be found in the different patterns of motor unit excitation. During repetitive isometric exercise, as in the present study, each contraction is held for several seconds, so that the energy utilized over a given time period depends primarily on the force-fusion properties of the active motor units and is less dependent on the initial force generation. Because the motor unit firing rate during repetitive isometric exercise at 30% MVC is low (2), an altered F/F_m ratio will have a large impact on the total energy utilization. In contrast, during dynamic exercise such as bicycling, each contraction is short (a fraction of a second), but the motor units are activated at high frequencies (10). Under these conditions, the F/F_m ratio is small and the overall energy turnover is dominated by that associated with the initial force generation. In addition, the shape of the force-frequency relationship predicts that changes in RT_1/2 will have minor effect on F/F_m during high-intensity stimulation. These lines of reasoning are in keeping with the observation of higher energy cost for isometric twitch and short intermittent tetani than for tetanic contractions of longer duration (21, 26). Other experiments showing that the slowing of relaxation often seen during sustained contraction is associated with a decreased energy cost (13, 32) is in keeping with a close connection between changes in mechanical and metabolic properties.

In conclusion, our results indicate that repetitive low-force isometric contractions induce a gradual fall in the RT_1/2 while the CT is unaltered. With higher target force levels, the faster relaxation may be masked during exercise, probably by the slowing effect of anaerobic metabolites. Our data further suggest that the reduced RT_1/2 occurs at the cellular level and is not primarily caused by temperature rise or selective muscle fiber recruitment and fatigue. The most likely cause is an increased turnover rate of the SR Ca²⁺-ATPase or the myosin ATPase. We argue that voluntary submaximal isometric contractions are associated with low motor-unit excitation rates and corresponding oscillations in force output. The faster relaxation may thus induce increased amplitudes of force oscillations in the active motor units and thereby increase the energy cost of contraction.

ACKNOWLEDGEMENTS

The authors thank Brenda Bigland-Ritchie for helpful suggestions and comments on a preliminary draft of this paper and Zhaopeng Li for extensive technical assistance.

FOOTNOTES

Address for reprint requests: N. K. Vøllestad, Section for Postgraduate Studies in Health Science, Univ. of Oslo, Box 1153 Blindern, 0316 Oslo, Norway (E-mail: nina.vollestad{at}helsefag.uio.no).

Received 2 January 1997; accepted in final form 15 July 1997.

REFERENCES

1.	Bakels, R., and D. Kernell. Matching between motoneurone and muscle unit properties in rat medial gastrocnemius. J. Physiol. (Lond.) 463: 307-324, 1993[Abstract/Free Full Text].
2.	Bigland-Ritchie, B., E. Cafarelli, and N. K. Vøllestad. Fatigue of submaximal static contractions. Acta Physiol. Scand. 128: 137-148, 1986.
3.	Bigland-Ritchie, B., R. Johansson, O. C. J. Lippold, and J. Woods. Contractile speed and EMG changes during fatigue of sustained maximal voluntary contractions. J. Neurophysiol. 50: 313-324, 1983[Abstract/Free Full Text].
4.	Bigland-Ritchie, B., D. A. Jones, and J. J. Woods. Excitation frequency and muscle fatigue: electrical responses during human voluntary and stimulated contractions. Exp. Neurol. 64: 414-427, 1979[Medline].
5.	Bigland-Ritchie, B., C. K. Thomas, C. L. Rice, J. V. Howarth, and J. J. Woods. Muscle temperature, contractile speed, and motoneuron firing rates during human voluntary contractions. J. Appl. Physiol. 73: 2457-2461, 1992[Abstract/Free Full Text].
6.	Bolstad, G., and A. Ersland. Energy metabolism in different human skeletal muscles during voluntary isometric contractions. Eur. J. Appl. Physiol. 38: 171-179, 1978.
7.	Burke, R. E. Motor units: anatomy, physiology, and functional organization. In: Handbook of Physiology. The Nervous System. Motor Control. Bethesda, MD: Am. Physiol. Soc., 1981, sect. 1, vol. II, pt. 1, chapt. 10, p. 345-422.
8.	Cady, E. B., H. Elshove, D. A. Jones, and A. Moll. The metabolic causes of slow relaxation in fatigued human skeletal muscle. J. Physiol. (Lond.) 418: 327-337, 1989[Abstract/Free Full Text].
9.	Dawson, M. J., D. G. Gadian, and D. R. Wilkie. Mechanical relaxation rate and metabolism studied in fatiguing muscle by phosphorus nuclear magnetic resonance. J. Physiol. (Lond.) 299: 465-484, 1980[Abstract/Free Full Text].
10.	Desmedt, J. E., and E. Godaux. Ballistic contractions in fast or slow human muscles: discharge patterns of single motor units. J. Physiol. (Lond.) 285: 185-196, 1978[Abstract/Free Full Text].
11.	Duchateau, J., and K. Hainaut. Nonlinear summation of contractions in striated muscle. I. Twitch potentiation in human muscle. J. Muscle Res. Cell Motil. 7: 11-17, 1986[Medline].
12.	Edwards, R. H. T., D. H. Hill, and D. A. Jones. Metabolic changes associated with the slowing of relaxation in fatigued mouse muscle. J. Physiol. (Lond.) 251: 287-301, 1975. [Abstract/Free Full Text]
13.	Edwards, R. H. T., D. K. Hill, and D. A. Jones. Heat production and chemical changes during isometric contractions of the human quadriceps muscle. J. Physiol. (Lond.) 251: 303-315, 1975. [Abstract/Free Full Text]
14.	Fallentin, N., K. Jørgensen, and E. B. Simonsen. Motor unit recruitment during prolonged isometric contractions. Eur. J. Appl. Physiol. 67: 335-341, 1993.
15.	Ferrington, D. A., J. C. Reijneveld, P. R. Bär, and D. J. Bigelow. Activation of the sarcoplasmic reticulum Ca2+-ATPase induced by exercise. Biochim. Biophys. Acta 1279: 203-213, 1996[Medline].
16.	Fitts, R. H. Cellular mechanisms of muscle fatigue. Physiol. Rev. 74: 49-94, 1994[Abstract/Free Full Text].
17.	Harris, R. C., R. H. T. Edwards, E. Hultman, L.-O. Nordesjö, B. Nylind, and K. Sahlin. The time course of phosphorylcreatine resynthesis during recovery of the quadriceps muscle in man. Pflügers Arch. 367: 137-142, 1976[Medline].
18.	Hultman, E., and H. Sjöholm. Electromyogram, force and relaxation time during and after continuous electrical stimulation of human skeletal muscle in situ. J. Physiol. (Lond.) 339: 33-40, 1983[Abstract/Free Full Text].
19.	Jewell, B. R., and D. R. Wilkie. An analysis of the mechanical components in frog's striated muscle. J. Physiol. (Lond.) 143: 515-540, 1958.
20.	Kernell, D., O. Eerbeek, and B. A. Verhey. Motor unit categorization on basis of contractile properties: an experimental analysis of the composition of the cat's m. peroneus longus. Exp. Brain Res. 50: 211-219, 1983[Medline].
21.	Loiselle, D. S., and B. Walmsley. Cost of force development as a function of stimulus rate in rat soleus muscle. Am. J. Physiol. 243 ((Cell Physiol. 12): C242-C246, 1982[Abstract/Free Full Text].
22.	McKenzie, D. K., and S. C. Gandevia. Recovery from fatigue of human diaphragm and limb muscles. Respir. Physiol. 84: 49-60, 1991[Medline].
23.	Mengshoel, A. M., E. Saugen, Ø. Førre, and N. K. Vøllestad. Muscle fatigue in early fibromyalgia. J. Rheumatol. 22: 143-150, 1995[Medline].
24.	Milner-Brown, H. S., R. B. Stein, and R. Yemm. The orderly recruitment of human motor units during voluntary isometric contractions. J. Physiol. (Lond.) 230: 359-370, 1973[Abstract/Free Full Text].
25.	Monster, A. W., and H. Chan. Isometric force production by motor units of extensor digitorum communis muscle in man. J. Neurophysiol. 40: 1432-1443, 1977[Free Full Text].
26.	Newham, D. J., D. A. Jones, D. L. Turner, and D. McIntyre. The metabolic cost of different types of contractile activitity of the human adductor pollicis muscle. J. Physiol. (Lond.) 488: 815-819, 1995[Abstract/Free Full Text].
27.	Sahlin, K., S. Cizinsky, M. Warholm, and J. Höberg. Repetitive static muscle contractions in humansa trigger of metabolic and oxidative stress? Eur. J. Appl. Physiol. 64: 228-236, 1992.
28.	Saugen, E., and N. K. Vøllestad. Nonlinear relationship between heat production and force during voluntary contractions in humans. J. Appl. Physiol. 79: 2043-2049, 1995[Abstract/Free Full Text].
29.	Saugen, E., and N. K. Vøllestad. Metabolic heat production during fatigue from voluntary repetitive isometric contractions in humans. J. Appl. Physiol. 81: 1323-1330, 1996[Abstract/Free Full Text].
30.	Saugen, E., N. K. Vøllestad, H. Gibson, P. A. Martin, and R. H. T. Edwards. Dissociation between metabolic and contractile responses during intermittent isometric intermittent exercise. Exp. Physiol. 82: 213-226, 1997[Abstract].
31.	Sejersted, I., E. Saugen, and N. K. Vøllestad. Muscle fatigue is associated with increased contractile speed during repeated submaximal isometric contractions (Abstract). Acta Physiol. Scand. 149: 35A, 1993.
32.	Spriet, L. L., K. Söderlund, and E. Hultman. Energy cost and metabolic regulation during intermittent and continuous tetanic contractions in human skeletal muscle. Can. J. Physiol. Pharmacol. 66: 134-139, 1987.
33.	Stein, R. B., T. Gordon, and J. Shriver. Temperature dependence of mammalian muscle contractions and ATPase activities. Biophys. J. 40: 97-107, 1982[Medline].
34.	Vøllestad, N. K., I. Sejersted, and E. Saugen. Increased relaxation rates during and following intermittent submaximal isometric contractions (Abstract). Biochem. Exercise 9th Int. Conf. Aberdeen Scotland 1994, p. 105-106.
35.	Vøllestad, N. K., O. M. Sejersted, R. Bahr, J. J. Woods, and B. Bigland-Ritchie. Motor drive and metabolic responses during repeated submaximal contractions in humans. J. Appl. Physiol. 64: 1421-1427, 1988[Abstract/Free Full Text].
36.	Vøllestad, N. K., J. Wesche, and O. M. Sejersted. Gradual increase in leg oxygen uptake during repeated submaximal contractions in humans. J. Appl. Physiol. 68: 1150-1156, 1990[Abstract/Free Full Text].
37.	Westerblad, H., J. A. Lee, J. Lännergren, and D. G. Allen. Cellular mechanisms of fatigue in skeletal muscle. Am. J. Physiol. 261 ((Cell Physiol. 30): C195-C209, 1991[Abstract/Free Full Text].
38.	Wiles, C. M., and R. H. T. Edwards. Metabolic heat production in isometric ischaemic contractions of human adductor pollicis. Clin. Physiol. 2: 499-512, 1982[Medline].
39.	Wiles, C. M., and R. H. T. Edwards. The effect of temperature, ischemia and contractile activity on the relaxation rate of human muscle. Clin. Physiol. 2: 485-497, 1982[Medline].
40.	Wiles, C. M., A. Young, D. Jones, and R. H. T. Edwards. Relaxation rate of constituent muscle-fibre types in human quadriceps. Clin. Sci. (Lond.) 56: 47-52, 1979[Medline].
41.	Woledge, R. C., N. A. Curtin, and E. Homsher. Energetic Aspects of Muscle Contraction. London: Academic, 1985.