Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2013 Jan 2;591(Pt 5):1373–1383. doi: 10.1113/jphysiol.2012.248989

Twitch interpolation: superimposed twitches decline progressively during a tetanic contraction of human adductor pollicis

S C Gandevia 1, C J McNeil 1,2, T J Carroll 3, J L Taylor 1
PMCID: PMC3607877  PMID: 23283762

Abstract

The assessment of voluntary activation of human muscles usually depends on measurement of the size of the twitch produced by an interpolated nerve or cortical stimulus. In many forms of fatiguing exercise the superimposed twitch increases and thus voluntary activation appears to decline. This is termed ‘central’ fatigue. Recent studies on isolated mouse muscle suggest that a peripheral mechanism related to intracellular calcium sensitivity increases interpolated twitches. To test whether this problem developed with human voluntary contractions we delivered maximal tetanic stimulation to the ulnar nerve (≥60 s at physiological motoneuronal frequencies, 30 and 15 Hz). During the tetani (at 30 Hz) in which the force declined by 42%, the absolute size of the twitches evoked by interpolated stimuli (delivered regularly or only in the last second of the tetanus) diminished progressively to less than 1%. With stimulation at 30 Hz, there was also a marked reduction in size and area of the interpolated compound muscle action potential (M wave). With a 15 Hz tetanus, a progressive decline in the interpolated twitch force also occurred (to ∼10%) but did so before the area of the interpolated M wave diminished. These results indicate that the increase in interpolated twitch size predicted from the mouse studies does not occur. Diminution in superimposed twitches occurred whether or not the M wave indicated marked impairment at sarcolemmal/t-tubular levels. Consequently, the increase in superimposed twitch, which is used to denote central fatigue in human fatiguing exercise, is likely to reflect low volitional drive to high-threshold motor units, which stop firing or are discharging at low frequencies.

Key points

  • The size of an interpolated muscle twitch during a voluntary muscle contraction is used to assess the extra force that the central nervous system can harness from the muscle with volition. During human exercise, this interpolated twitch commonly increases in size and this reduced voluntary activation of the muscle is termed ‘central’ fatigue.

  • Recent work on isolated mouse muscle fibres suggests an alternative ‘peripheral’ explanation for the increased twitch based on altered sensitivity to intracellular calcium. We tested this possibility with tetanic stimulation of the ulnar nerve while measuring thumb adductor force.

  • During maximal tetani lasting 1 min at 30 Hz (or 3 min at 15 Hz), muscle force declined (i.e. peripheral fatigue developed) but the relative and absolute size of superimposed twitches declined progressively. They did not increase as predicted from the mouse study.

  • Twitch interpolation can reveal central fatigue during voluntary muscle contractions.

Introduction

To understand the brain's motor output it is imperative to know how effectively this output generates force at the muscle. Those interested in human muscle performance have studied this critical question by measurement of the level of ‘voluntary drive’ to the muscle, particularly during maximal voluntary contractions when perceived effort and drive are maximized by the person performing the task. This measurement is not straightforward and various approaches have been tried. Simple measures of electromyographic activity and single motor unit firing have been used, but they are limited because their link to actual force production is not included. Comparisons of forces produced by volition and by tetanic stimulation at high frequency are problematic because maximal voluntary forces are not produced by single muscles or muscle groups that have a nerve supply which can be stimulated selectively (Marsden et al. 1983; Gandevia, 2001).

The technique of ‘twitch interpolation’ was introduced by Merton (1954) to measure voluntary activation. It relies on the twitch force added by a supramaximal motor nerve stimulus during a voluntary contraction, i.e. the superimposed twitch produced by an interpolated stimulus. It provides a way to quantify how much of a muscle's possible force is not engaged by voluntary drive and therefore, the extra force available to be driven by the brain in a voluntary effort. It compares the maximal twitch force produced by supramaximal nerve stimulation at rest with that produced by the same stimulus interpolated during a static voluntary contraction; the smaller the interpolated twitch the smaller the extra force available from the muscle, and hence the greater the level of ‘voluntary’ activation of the muscle. Voluntary activation is directly related to the size of the interpolated twitch. The original observations by Merton using adductor pollicis and ulnar nerve stimulation were followed later by Belanger and McComas (1981) using plantar and dorsiflexors of the ankle. Subsequently, the technique became widely used although caveats on its use were recognized (e.g. Allen et al. 1998; Kooistra et al. 2007; see also Gandevia, 2001; Taylor et al. 2009). Its reliability was established (Allen et al. 1995; Behm et al. 1996) and a realistic model provided theoretical underpinnings for its operation in adductor pollicis (Herbert & Gandevia 1999), the muscle originally tested by Merton (1954).

Apart from its use to assess fresh muscle, the principal application of the technique has been to study voluntary activation in muscle fatigue. Here, a progressive impairment in voluntary activation has been inferred because the superimposed twitch did not decline prominently during a maximal voluntary isometric contraction. Indeed, it often increased, and this apparent impairment of voluntary activation was defined as central fatigue (e.g. McKenzie et al. 1992, Gandevia et al. 1996; for review, Gandevia, 2001). This reflected a failure in obtaining optimal force from the muscle and must represent a failure in sustaining recruitment and discharge frequencies of motoneurones. Recordings of motor unit firing confirmed both declining firing rates (Bigland-Ritchie et al. 1983; Gandevia et al. 1990; see also Marsden et al. 1971) and the cessation of firing of some units (Peters & Fuglevand 1999; see also Bigland-Ritchie et al. 1978). Furthermore, this evidence for central fatigue has been bolstered by studies showing that motor cortical stimulation can be used to assess voluntary activation and this measure decreases with muscle fatigue and hyperthermia (e.g. Gandevia et al. 1996; Todd et al. 2003, 2005).

Recently, an in vitro model using single mouse muscle fibres was used to explore the intracellular mechanisms which generate superimposed twitches (Place et al. 2008). During tetanic contraction at 70 Hz (350 ms duration), an extra stimulus evoked relatively more force during a fatigue run when the muscle was fatigued by repeated contractions. This was proposed to reflect a move to the steep part of the sigmoid force–[Ca2+]i relationship from near the plateau in fresh muscle. This means that the change in force for the extra Ca2+ released with the extra stimulus increased relative to the force produced by a steady rate of stimulation. Thus, a purely peripheral mechanism could contribute to, and potentially dominate, what had been proposed as a purely central mechanism.

Therefore, this present study was undertaken to determine whether in the human adductor pollicis in vivo, the size of a superimposed twitch increased, as predicted from the mouse in vitro study, during a sustained fatiguing contraction produced by a steady rate of nerve stimulation. We also examined changes in the compound muscle action potential.

Methods

Subjects

Six healthy subjects (39 ± 8 years, mean ± s.d.; one woman) participated in the main experiment. Subsets of these subjects and one other man (36 years) participated in additional experiments on separate days. All procedures were approved by the University of New South Wales Human Research Ethics Committee and conformed to the Declaration of Helsinki. Written consent was obtained from each of the participants.

Experimental set-up

Subjects were seated at a table with their right forearm resting in a neutral position on a padded L-shaped support that ran from the elbow to the fingertips (Fig. 1A). The posterior aspect of the forearm and hand was secured tightly to the support with a strap just distal to the elbow and a padded plate against the palmar surface of the fully extended fingers. A padded clamp on the superior (radial) aspect of the wrist prevented ulnar deviation during stimulation. The thumb was abducted and positioned in a ring which was mounted to a linear strain gauge to measure adduction force (250 N; Xtran, Melbourne, Australia). Electromyographic activity (EMG) of adductor pollicis was recorded via adhesive Ag–AgCl electrodes (10 mm diameter) positioned over the muscle belly and metacarpophalangeal joint of the thumb. A large ground electrode (10 × 5 cm) was placed on the anterior aspect of the wrist.

Figure 1. Details of experimental arrangements.

Figure 1

Open in a new tab

A, experimental set-up. Electromyographic activity was recorded from surface electrodes on the palm over the adductor pollicis and stimuli were delivered to the ulnar nerve at the elbow (not shown). B, timing of interpolated stimulus in the 33 ms interval between stimuli in the train at 30 Hz delivered to the ulnar nerve. Three stimuli in the ongoing train are represented by solid boxes. Broken boxes (with an arrow) represent the interpolated stimulus. Six delays of 5, 10, 15, 20, 25 and 30 ms after a stimulus in the train were tested in the 12 control tetani shown in D. C, amplitude of the superimposed twitch (SIT) for the six timings of interpolated stimulus in six individual subjects. D, protocol for the main experiment. Brief tetani (∼2.5 s, 30 Hz) with the interpolated stimulus at six different delays after a stimulus in the train (see B and C). Subsequently, there was a 60 s tetanus (30 Hz) with interpolated stimulus (15 ms delay) delivered at ∼4.3 s and every 5 s until ∼59 s. Recovery was tracked for 2 min. Filled arrows (downward pointing) indicate a stimulus interpolated during a tetanus and open arrows (upward pointing) indicate a stimulus delivered at rest.

Force and EMG data were recorded to computer using a 12-bit A/D converter (CED 1401 Plus; Cambridge Electronic Design Ltd, Cambridge, UK) and Spike2 software (version 6.16; Cambridge Electronic Design). The force and EMG data were sampled at 1000 and 2000 Hz, respectively. To improve resolution of superimposed twitches, the force signal was further amplified (×10) after subtraction of the background force for ∼500 ms after each stimulus (Hales & Gandevia 1988). The EMG signal was amplified (×100) and bandpass filtered (16–1000 Hz) using a CED 1902 amplifier (Cambridge Electronic Design).

Ulnar nerve stimulation

A constant-current stimulator (DS7AH, Digitimer, Welwyn Garden City, Hertfordshire, UK) was used to deliver electrical stimuli (500 μs pulse width) to the ulnar nerve at the elbow to activate the adductor pollicis. The cathode and anode were placed just proximal and distal to the medial epicondyle of the humerus, respectively. The precise location was chosen to minimize the stimulus current required to elicit the maximal compound muscle action potential (Mmax). To determine the stimulus current which evoked Mmax, single stimuli to the ulnar nerve were applied at increasing stimulus intensity until the compound muscle action potential showed no further increase in amplitude. To account for activity-dependent changes in motor axonal excitability (e.g. Vagg et al. 1998), all subsequent stimuli were delivered at 150% of this current (18–24 mA).

Experimental procedures

Tetanic contractions of the adductor pollicis muscle were produced by supramaximal ulnar nerve stimulation at a frequency of 30 Hz. Additional supramaximal stimuli were occasionally interpolated between stimuli in the 30 Hz train to mimic the addition of a peripheral nerve stimulus to a maximal voluntary contraction.

During a maximal voluntary contraction, motoneurones discharge asynchronously and over a range of discharge rates (e.g. Bigland-Ritchie et al. 1983; see also Gandevia et al. 1990; Seki et al. 2007). Hence, the peripheral nerve stimulus used to evoke a superimposed twitch will be delivered to axons which will span the complete range of responsiveness. That is, some axons will be absolutely refractory and, at the other end of the spectrum, some axons will be just about to discharge their next action potential. In contrast, during a tetanic contraction, all axons are activated synchronously at the stimulus site and will be in the same state when an additional stimulus is delivered. To investigate how the timing of the additional stimulus affects the size of the superimposed twitch during a 30 Hz tetanus, six different intervals were tested during brief control tetani (∼2.5 s duration). Two seconds after the onset of tetanus, the additional stimulus was delivered in the 33 ms interval between two pulses of the ongoing 30 Hz train. The interpolated stimulus was delivered at 5, 10, 15, 20, 25 or 30 ms after the first of these pulses (Fig. 1B). Each of these six conditions was tested twice and so there was a total of 12 control tetani. At ∼2 s after each tetanus, a single stimulus was delivered to evoke a twitch of the resting muscle. Superimposed twitch size showed a gradual increase with the interstimulus interval until a marked decrease between the 25 and 30 ms intervals. Data for all subjects are shown in Fig. 1C.

A second part of the study examined the responses to interpolated stimuli during fatigue and recovery. Fatigue was induced by a 60 s tetanus at 30 Hz (Fig. 1D). This stimulus frequency was selected as it represents the mean of motor unit firing rates early in maximal efforts for this muscle (Bigland-Ritchie et al. 1983). Beginning at ∼4.3 s, an additional stimulus was delivered every 5 s so that the superimposed twitch was tested 12 times. The additional stimulus was always delivered at the 15 ms interval because it was nearest to the middle of the two pulses of the ongoing tetanus (as used by Place et al. 2008). A single stimulus was delivered ∼2 s after the tetanus during relaxation to examine the effect of fatigue on the resting twitch. To assess recovery of the superimposed and resting twitches, brief ∼2.5 s trains of stimuli were delivered at 5, 20, 40, 60, 90 and 120 s after the end of the 60 s tetanus. An additional stimulus (15 ms interval) was delivered during each tetanus and a single stimulus evoked a resting twitch after each tetanus.

Additional experiments

The main study included 12 interpolated stimuli that produced instances of high-frequency stimulation (∼60 Hz) compared with the ongoing rate (30 Hz) during the 60 s tetanic contraction. Because interpolation of high-frequency stimuli early in a contraction can assist the maintenance of force in human muscles (e.g. Bigland-Ritchie et al. 2000) and mouse single fibres (e.g. Abbate et al. 2002), we repeated the study in three subjects with an additional stimulus delivered only in the last second of the 60 s tetanus. Only the two control tetani with the 15 ms interval were collected for this experiment.

The main study used a tetanic stimulus frequency of 30 Hz to mimic motor unit firing rates at the start of a sustained maximal effort of adductor pollicis. In three subjects, the basic protocol was repeated but the stimulus frequency was lowered to 15 Hz and duration of the fatiguing tetanus increased to 3 min. This was done to determine if any changes in the size of the superimposed twitch occurred when the frequency mimicked motor unit firing rates at the end of sustained contractions of the adductor pollicis (Bigland-Ritchie et al. 1983). To maintain the position of the additional stimulus relative to stimuli in the ongoing 15 Hz tetanus, the 15 ms interval used in the main study was doubled to 30 ms in this study. During the 3 min tetanus, an additional stimulus was delivered every 10 s so that the superimposed twitch was tested 18 times.

Data analysis and statistics

During off-line analysis, Signal software (version 4.07; Cambridge Electronic Design) was used to determine all measures. For statistical analysis, data were expressed as a percentage of the control value obtained during the brief tetani. The exceptions were tetanic force during fatigue and recovery, which was normalized to the peak force at the onset of the 60 s tetanus and M wave characteristics (amplitude, area and duration), which were normalized to the M wave at 2 s of the 60 s tetanus. This M wave was selected because it equates to the response before the superimposed stimulus during control and recovery tetani. All statistical analyses were conducted using IBM SPSS Statistics version 20. The effect of the stimulus interval on superimposed twitch force was tested with a one-way repeated-measure ANOVA. In post-hoc analysis, paired t-tests with a Bonferroni correction factor were used to determine which intervals differed from one another. One-way repeated measures ANOVAs were also used to investigate the effect of time (fatigue and recovery) on all other measures. Fatigue and recovery data were assessed separately. If the main effect of time was significant, results of paired t-tests were compared to a two-tailed Dunnett's table to control for multiple comparisons and determine time points which differed from the control or peak value. Data are reported in the text as mean ± s.d. The significance level was P < 0.05.

Results

Main experiment

For the six subjects the peak tetanic force across the 12 control tetani was 74.2 ± 9.4 N. Superimposed twitch force varied with the stimulus interval (P < 0.001; Fig. 1C) and ranged from a low of 1.2 ± 0.5 N at a 30 ms interval to a high of 2.2 ± 0.6 N at a 25 ms interval. For the group, the superimposed twitch at 30 ms was significantly smaller than all intervals but 10 ms, whereas the superimposed twitch at 25 ms was significantly larger than all but the 20 ms interval. In two control tetani at the 15 ms interval, the size of the superimposed twitch was 1.8 ± 0.5 N and this interval was used in the fatiguing tetanus.

During the 60 s tetanus, force decreased 42% from 71.7 ± 14.1 N at the onset to 41.8 ± 11.4 N at the end (P < 0.001; Figs 2A and 3B). Size of the superimposed twitch decreased rapidly during the prolonged tetanus (P < 0.001) and was only 20% of the control value at 30 s and <1% of control at the end (Figs 2B and 3). After 5 s of recovery, tetanic force rose to 77% of control and increased progressively thereafter. It remained significantly lower than control until the final tetanus (Fig. 3B). Superimposed twitch force was also restored to ∼20% of control at 5 s but continued to recover rapidly such that it was equivalent to the control value 90 s after the fatiguing contraction (Fig. 3B).

Figure 2. Data from the main protocol from a typical subject.

Figure 2

Open in a new tab

A, above: force recorded during a brief tetanus with interpolated stimulus and resting twitch. Below: force recorded during a 60 s tetanus with interpolated stimuli and resting twitch. B, responses to the interpolated stimuli at high gain. Times in the 60 s stimulus train are indicated. C, resting twitch before (larger trace) and after (smaller trace) the 60 s tetanus. The filled arrows indicate a stimulus interpolated during a tetanus and the open arrows (upward pointing) indicate a stimulus delivered at rest.

Figure 3. Changes in tetanic and superimposed twitch force during and after the 60 s tetanus.

Figure 3

Open in a new tab

A, amplitude of the superimposed twitch (SIT) during the 60 s tetanus. Data from six subjects joined by single lines. The size of the superimposed twitch dropped to below half its initial value within 30 s in all subjects. B, group data (mean ± s.e.m., n = 6) for tetanic (open circles) and superimposed twitch force (black circles) during and after the 60 s tetanus. Timing of the tetanus is indicated by the shaded box. Tetanic force was measured 10 ms before each superimposed stimulus. The superimposed twitch declined during the fatiguing tetanus and recovered within 3 min.

Before fatigue, the resting twitch force was 11.5 ± 1.5 N and time-to-peak tension and half-relaxation time were 63.3 ± 5.6 and 57.9 ± 8.6 ms, respectively. After the 60 s tetanus, twitch force was reduced (by 41.9 ± 24.6%), with prolongations of both the time-to-peak tension (by 32.5 ± 4.7 %) and half-relaxation time (by 74.9 ± 26.1%) (all P < 0.001; Fig. 2C). Twitch force and time-to-peak tension were equivalent to control values 90 s after the fatiguing contraction, whereas half-relaxation time did not recover until the final twitch (data not shown).

During the 60 s tetanus the M wave stimuli delivered at 30 Hz showed the expected changes: the area increased as the compound muscle potential diminished in amplitude but became broader (all P < 0.001). Pooled data for the M waves produced by the stimuli immediately preceding and following the interpolated stimulus are shown in Fig. 4A and C respectively, with raw traces in Fig. 4D. Despite a >40% reduction in amplitude at the end of the tetanus, M wave area remained significantly elevated until 55 s and non-significantly greater than the initial value over the final 5 s. In contrast, an important finding was that the M wave produced by the superimposed stimulus was well maintained for only 25 s and then diminished rapidly until it was fully occluded during the final 10 s of the tetanus (Fig. 4B and D). However, 10 s into the tetanus, when the area of the M wave was above control levels and its amplitude had not declined, the superimposed twitch had already declined to 60% of control.

Figure 4. Changes in M wave parameters during and after the 60 s tetanus.

Figure 4

Open in a new tab

A, amplitude (open circle), area (black triangle) and duration (open diamond) of the M wave associated with the last stimulus of the 30 Hz train before the superimposed stimulus (mean ± s.e.m., n = 6). B, amplitude, area and duration of the M wave associated with the superimposed stimulus (mean ± s.e.m., n = 6). C, amplitude, area and duration of the M wave associated with the first stimulus immediately after the interpolated stimulus (mean ± s.e.m., n = 6). For A–C, timing of the tetanus is indicated by the shaded box. Data are normalized to the M wave collected at 2 s into the 60 s tetanus. D, raw traces of the M waves associated with the superimposed stimulus and the stimuli from the 30 Hz tetanus, which preceded and followed the additional stimulus. Timing within the 60 s tetanus is indicated to the left of each trace. M waves during the recovery period are also shown. Solid vertical lines indicate the time of delivery of ulnar nerve stimuli in the train. Arrow and broken vertical line indicate the time of delivery of the interpolated stimulus.

Additional experiments

Additional studies were performed in which the superimposed stimulus was not delivered until the last second of the 60 s tetanus. The same result occurred as in the main study: the superimposed twitch response (and the associated M wave) was virtually abolished. Protocols performed at 15 rather than 30 Hz also yielded results similar to the main study for the superimposed twitch but not for the associated M wave. All subjects behaved similarly and data for a single subject are shown in Fig. 5. The superimposed twitch force reached a stable nadir after 100 s (<10% of the control value) but the area of the M wave at 100 s was still greater than the initial value.

Figure 5. Changes in tetanic and SIT force and M wave area during 3 min tetanus at 15 Hz.

Figure 5

Open in a new tab

A, data from a single subject. Timing of tetanus indicated by the shaded box whereas the darker area indicates the timing of fatigued raw traces seen in B. Tetanic force is normalized to the peak value at the onset of the contraction whereas SIT force is normalized to the mean value recorded during the control tetani. M wave area is normalized to the M wave collected at 4 s into the 3 min tetanus. B, raw traces of SIT and M waves from the same subject. For SIT, the nearly identical traces from the two control tetani are overlaid with the trace collected at 100 s. For the M waves, the response collected 4 s into the 3 min tetanus is overlaid with the response to the superimposed stimulus at 100 s. SIT, superimposed twitch.

Discussion

This study was done primarily to determine the change in size of the force evoked by a twitch superimposed during a sustained electrically evoked contraction in vivo. We chose to examine sustained stimulated contractions because sustained maximal voluntary contractions are a common human fatigue paradigm. Increases in superimposed twitches are reported during such contractions of various muscles and are attributed to central fatigue. The muscle selected here, the adductor pollicis, has been used commonly for studies of human fatigue mechanisms (e.g. Merton, 1954; Jones et al. 1979; Bigland-Ritchie et al. 1983; Fuglevand & Keen, 2003). During sustained maximal voluntary contractions requiring thumb adduction, the firing rate of adductor pollicis motor units declines from ∼30 Hz to ∼15 Hz over 60 s (Bigland-Ritchie et al. 1983; see also Gandevia et al. 1990). This decline in discharge rate is too great to maintain full force-generating capacity by the muscle (Fuglevand & Keen, 2003). Our new finding is that during a 60 s tetanus at 30 Hz in which the tetanic (and twitch) force declines by ∼40%, the size of a superimposed twitch drops more rapidly and is less than 10% of control values after ∼40 s. This decline in the superimposed twitch could not be explained by a history of repeated high-frequency intervals as it occurred if an interpolated twitch was not delivered until the last second of the tetanus. A progressive decline also occurred when stimuli were interpolated during a tetanus at only 15 Hz for 3 min. Hence, the main conclusion is that the force produced by a supramaximal interpolated stimulus during tetanic contraction of a whole human muscle does not increase with time when all the muscle fibres are activated at constant frequencies in the range reported for human maximal voluntary contractions. This finding contrasts with the increase in force produced by interpolated stimuli in fatigued single mouse muscle fibres stimulated at 70 Hz (Place et al. 2008).

The major implication of our result is that in studies of human voluntary muscle fatigue produced by sustained efforts when the relative size of the superimposed twitch rises (to motor nerve or cortical stimulation, see Introduction), this increase is not likely to represent a purely peripheral mechanism. Rather the additional force is likely produced by recruitment of some motor units that have ceased to fire and others that have reduced their firing frequency compared to before fatigue. For these motor units, the lower the firing frequency the greater the superimposed twitch evoked by the interpolated stimulus.

Our studies were not designed to assess the mechanism for the reduction in the interpolated twitch amplitude during the prolonged tetani at 15 Hz or 30 Hz. Particularly with the 30 Hz tetanus, it is likely that the muscle action potential fails, as shown in both animal (e.g. Lännergren & Westerblad, 1987) and human studies (e.g. Bigland-Ritchie et al. 1979; see also Fuglevand et al. 1993; Farina et al. 2006). However, after ∼25 s of the 30 Hz tetanus, when tetanic force had declined to ∼80% initial value and the area of Mmax was still 100% initial value, the interpolated twitch had declined to ∼20% of its initial value. Rather than a failure of neuromuscular transmission, excitability may have declined distal to the sarcolemma, probably within the t-tubule. At this site there is likely to be rapid recovery of the M wave after the interpolated stimulus (Fig. 4B and C). While the precise mechanisms of action potential failure will differ with different protocols (e.g. Duty & Allen, 1994), a shift in the force–[Ca2+]i relationship may be of little relevance when any such failure occurs. Our findings indicate that this may be the case whenever fatigue is caused by sustained maximal contractions.

It is possible that a role for intracellular mechanisms in growth of the superimposed twitch with fatigue could be revealed in human muscle by the use of specific stimulation paradigms that resemble more closely those used for mouse muscle fibres, i.e. brief intermittent near-maximal (∼85% maximum) contractions with a low duty cycle (350 ms contraction with 2–3 s rest; Place et al. 2008). This specific paradigm would not transfer well to human isometric maximal voluntary contractions, as even brief efforts usually last 2–3 s and, with fatigue, subjects take longer to reach their maximal force. In our study, force only just reached a plateau in the 2.5 s trains (30 Hz), which were delivered during the recovery period. Thus, shorter trains are unlikely to allow interpolation of a stimulus during maximal force production. For 2.5 s trains, delivery every 15–20 s would match the duty cycle of the mouse muscle paradigm. In our study, brief trains were delivered 15–30 s apart in the recovery period, and in the final brief tetanus when tetanic force had recovered to ∼90% of its initial level, the superimposed twitch was relatively increased (∼15%; see Fig. 3B). This may reflect an effect of the altered force–[Ca2+]i relationship described by Place et al. (2008), although both the level of fatigue and relative increase in the superimposed twitch were small at that point in our protocol.

The behaviour of the superimposed twitch at the beginning of the sustained contraction may also be relevant in setting a limit for when a peripheral mechanism may increase the superimposed twitch. In the 30 Hz sustained tetanus, the superimposed twitch and tetanic force were both marginally decreased at the first interpolated stimulus at 4.3 s but by 10 s the superimposed twitch was considerably reduced (Fig. 3B). This suggests that in maximal efforts of >4 s duration peripheral mechanisms are unlikely to increase the superimposed twitch directly if firing rates remain maximal. On the other hand, at 10 s into the 15 Hz tetanus the superimposed twitch was increased by a small amount relative to tetanic force (∼15%), although this increase had disappeared after another 10 s (Fig. 5). This suggests that intracellular mechanisms may sometimes increase the superimposed twitch when firing rates are less than optimal. Thus, when superimposed twitches increase during a voluntary sustained contraction, we propose that a decrease in firing rates [Ca2+]i is required but that, given low enough firing rates, the size of the increase may be enhanced by a change in the force–[Ca2+]i relationship.

Implications for quantitation of voluntary activation

To extrapolate the results of these studies to the quantification of levels of voluntary activation is not straightforward. Although the frequencies of stimulation used here are in the physiological range measured in maximal voluntary contractions, sustained firing at a single stable frequency is unlikely to describe the behaviour of any individual motor unit in a voluntary contraction. However, some arguments can be put forward about the implications of the findings.

We argue that an increase in the interpolated twitch in a sustained maximal contraction cannot be due only to peripheral mechanisms. If this were true, it would mean that the interpolated twitch would increase if motor units continued to fire at the same frequency in a fatigued maximal voluntary contraction as they do when the muscle is fresh. Our results show that if motor units fired at between 15 and 30 Hz and there were no change in this rate with sustained firing and fatigue, then an interpolated stimulus would produce less force from these units when fatigued. As indicated above, our study cannot identify the mechanism for the reduction in the superimposed twitch in such circumstances, although the progressive reduction and eventual disappearance of the interpolated M wave suggests that failure in the sarcolemma and t-tubules is likely to contribute. However, regardless of the mechanism, if there is no reduction in drive to the muscle (i.e. a reduction in firing rate) we find no increase in the superimposed twitch.

As the peak firing rates of motor units in the adductor pollicis during maximal voluntary contractions are distributed from to ∼12 to 50 Hz (Bellemare et al. 1983), some fall outside the frequencies we have tested. Given the ultimate abolition of the M wave from the interpolated stimulus during the 30 Hz tetanus, it is likely that units that fire faster than 30 Hz would also have problems with muscle excitation by an interpolated stimulus if they continued at these high firing rates. About 2% of units are reported to have peak firing rates slower than 15 Hz and it is possible that these units behave as suggested by Place et al. (2008) and produce extra force in response to the interpolated stimulus when they are fatigued.

Of course, during a sustained maximal voluntary contraction, motor units do not maintain their peak firing frequency. Rather, neural drive to the muscle decreases. At the same time, despite the slowing of muscle contraction and relaxation, the motor unit firing frequency required to produce half-maximal force increases (Fuglevand et al. 1999). Thus, presumably, firing rates during fatigue are below the optimum for full force production. Then, one would expect the interpolated twitch of fatigued muscle fibres firing at a given suboptimal frequency to produce more force (relative to the force produced by ongoing firing) than that evoked from fresh muscle fibres firing at the same frequency. However, it is difficult to identify what this means in terms of central fatigue as such rates would be far too low to produce high levels of activation in a fresh muscle, which are typically observed in the first few seconds of an maximal voluntary contraction. Even with all the muscle fibres firing at 30 Hz, comparison of the interpolated twitch with the resting twitch gives an apparent activation of ∼85%, which is low compared to typical maximal voluntary drive. Thus, it seems that if motor units fire at frequencies that produce reasonable activation in fresh muscle, the interpolated twitch gets smaller. In contrast, firing frequencies that are low enough to show extra force with fatigue through the intracellular mechanism described by Place et al. (2008), are likely to be much lower than required for near-maximal activation of fresh muscle. While we may see ‘extra’ force from the periphery in a fatigued muscle, it may only occur if neural drive to the muscle is inadequate.

Realistically, both the output of the motoneurone pool and the contribution of motor units to the superimposed twitch are non-uniform. In particular, the high-threshold motor units are likely to be de-recruited and re-recruited with the small fluctuations in a maximal voluntary contraction and to have a very low average firing rate. The activation of their muscle fibres (at sarcolemmal and t-tubular levels) is unlikely to be compromised and so they will be available to generate force in response to interpolated stimuli during voluntary contractions. Furthermore, these units generate the largest forces and thus, are likely to contribute a large fraction of the superimposed twitch.

Implications for motor unit firing in human muscle fatigue

These results have implications for the understanding of human muscle fatigue. First, when increases in absolute or relative interpolated twitch force are evoked from muscles during fatiguing exercise, this is likely to be dominated by the response of those motoneurones that have a high threshold to voluntary effort and have discharged relatively few action potentials. Muscle fibres innervated by low-threshold motoneurones are known to show activity-dependent changes in action potential morphology likely produced by local K+ accumulation (Farina et al. 2009; for review Sejersted & Sjøgaard 2000; Allen et al. 2008). During sustained contractions, their capacity to respond to additional neural drive inputs may be limited by peripheral factors at the muscle membrane level. Second, these results suggest that, unlike when the muscle is fresh, attempts to relate the size of interpolated twitches to the exact level of voluntary activation of the muscle should be made with caution. Finally, our findings highlight the tight nexus that must exist between the firing rate of motoneurones and the driving of the intracellular contractile machinery. Important central changes accompanying sustained muscle contractions, including spike-frequency adaptation (e.g. Sawczuk et al. 1995), reduced ‘gain’ at the motoneurone pool (e.g. Butler et al. 2003; Johnson et al. 2004; McNeil et al. 2009; Khan et al. 2012; Rossi et al. 2012), and impaired volitional drive to the motor cortex (Gandevia et al. 1996) may act to preserve force-generating capacity in the periphery.

Acknowledgments

The National Health and Medical Research Council of Australia supported this work.

Glossary

EMG

electromyographic activity

Mmax

maximal compound muscle action potential

SIT

superimposed twitch

Author contributions

All authors contributed to the design of the study, which was then undertaken at Neuroscience Research Australia in Sydney. All authors were involved in data collection and interpretation. They contributed to drafting the manuscript and all approved the final version of the text.

References

  1. Abbate F, Bruton JD, De Haan A, Westerblad H. Prolonged force increase following a high-frequency burst is not due to a sustained elevation of [Ca2+]i. Am J Physiol Cell Physiol. 2002;283:C42–C47. doi: 10.1152/ajpcell.00416.2001. [DOI] [PubMed] [Google Scholar]
  2. Allen DG, Lamb GD, Westerblad H. Skeletal muscle fatigue: cellular mechanisms. Physiol Rev. 2008;88:287–332. doi: 10.1152/physrev.00015.2007. [DOI] [PubMed] [Google Scholar]
  3. Allen GM, Gandevia SC, McKenzie DK. Reliability of measurements of muscle strength and voluntary activation using twitch interpolation. Muscle Nerve. 1995;18:593–600. doi: 10.1002/mus.880180605. [DOI] [PubMed] [Google Scholar]
  4. Allen GM, McKenzie DK, Gandevia SC. Twitch interpolation of the elbow flexor muscles at high forces. Muscle Nerve. 1998;21:318–328. doi: 10.1002/(sici)1097-4598(199803)21:3<318::aid-mus5>3.0.co;2-d. [DOI] [PubMed] [Google Scholar]
  5. Behm DG, St-Pierre DM, Perez D. Muscle inactivation: assessment of interpolated twitch technique. J Appl Physiol. 1996;81:2267–2273. doi: 10.1152/jappl.1996.81.5.2267. [DOI] [PubMed] [Google Scholar]
  6. Belanger AY, McComas AJ. Extent of motor unit activation during effort. J Appl Physiol. 1981;51:1131–1135. doi: 10.1152/jappl.1981.51.5.1131. [DOI] [PubMed] [Google Scholar]
  7. Bellemare F, Woods JJ, Johansson R, Bigland-Ritchie B. Motor-unit discharge rates in maximal voluntary contractions of three human muscles. J Neurophysiol. 1983;50:1380–1392. doi: 10.1152/jn.1983.50.6.1380. [DOI] [PubMed] [Google Scholar]
  8. Bigland-Ritchie B, Jones DA, Hosking GP, Edwards RHT. Central and peripheral fatigue in sustained maximum voluntary contractions of human quadriceps muscle. Clin Sci Mol Med. 1978;54:609–614. doi: 10.1042/cs0540609. [DOI] [PubMed] [Google Scholar]
  9. Bigland-Ritchie B, Jones DA, Woods JJ. Excitation frequency and muscle fatigue: electrical responses during human voluntary and stimulated contractions. Exp Neurol. 1979;64:414–427. doi: 10.1016/0014-4886(79)90280-2. [DOI] [PubMed] [Google Scholar]
  10. Bigland-Ritchie B, Johansson R, Lippold OC, Smith S, Woods JJ. Changes in motoneurone firing rates during sustained maximal voluntary contractions. J Physiol. 1983;340:335–346. doi: 10.1113/jphysiol.1983.sp014765. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Bigland-Ritchie B, Zijdewind I, Thomas CK. Muscle fatigue induced by stimulation with and without doublets. Muscle Nerve. 2000;23:1348–1355. doi: 10.1002/1097-4598(200009)23:9<1348::aid-mus5>3.0.co;2-0. [DOI] [PubMed] [Google Scholar]
  12. Butler JE, Taylor JL, Gandevia SC. Responses of human motoneurons to corticospinal stimulation during maximal voluntary contractions and ischemia. J Neurosci. 2003;23:10224–10230. doi: 10.1523/JNEUROSCI.23-32-10224.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Duty S, Allen DG. The distribution of intracellular calcium concentration in isolated single fibres of mouse skeletal muscle during fatiguing stimulation. Pflügers Arch. 1994;427:102–109. doi: 10.1007/BF00585948. [DOI] [PubMed] [Google Scholar]
  14. Farina D, Zennaro D, Pozzo M, Merletti R, Laubli T. Single motor unit and spectral surface EMG analysis during low-force, sustained contractions of the upper trapezius muscle. Eur J Appl Physiol. 2006;96:157–164. doi: 10.1007/s00421-004-1261-8. [DOI] [PubMed] [Google Scholar]
  15. Farina D, Holobar A, Gazzoni M, Zazula D, Merletti R, Enoka RM. Adjustments differ among low-threshold motor units during intermittent, isometric contractions. J Neurophysiol. 2009;101:350–359. doi: 10.1152/jn.90968.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fuglevand AJ, Keen DA. Re-evaluation of muscle wisdom in the human adductor pollicis using physiological rates of stimulation. J Physiol. 2003;549:865–875. doi: 10.1113/jphysiol.2003.038836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fuglevand AJ, Zackowski KM, Huey KA, Enoka RM. Impairment of neuromuscular propagation during human fatiguing contractions at submaximal forces. J Physiol. 1993;460:549–572. doi: 10.1113/jphysiol.1993.sp019486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Fuglevand AJ, Macefield VG, Bigland-Ritchie B. Force-frequency and fatigue properties of motor units in muscles that control digits of the human hand. J Neurophysiol. 1999;81:1718–1729. doi: 10.1152/jn.1999.81.4.1718. [DOI] [PubMed] [Google Scholar]
  19. Gandevia SC. Spinal and supraspinal factors in human muscle fatigue. Physiol Rev. 2001;81:1725–1789. doi: 10.1152/physrev.2001.81.4.1725. [DOI] [PubMed] [Google Scholar]
  20. Gandevia SC, Macefield G, Burke D, McKenzie DK. Voluntary activation of human motor axons in the absence of muscle afferent feedback: the control of the deafferented hand. Brain. 1990;113:1563–1581. doi: 10.1093/brain/113.5.1563. [DOI] [PubMed] [Google Scholar]
  21. Gandevia SC, Allen GM, Butler JE, Taylor JL. Supraspinal factors in human muscle fatigue: evidence for suboptimal output from the motor cortex. J Physiol. 1996;490:529–536. doi: 10.1113/jphysiol.1996.sp021164. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hales JP, Gandevia SC. Assessment of maximal voluntary contraction with twitch interpolation: an instrument to measure twitch responses. J Neurosci Methods. 1988;25:97–102. doi: 10.1016/0165-0270(88)90145-8. [DOI] [PubMed] [Google Scholar]
  23. Herbert RD, Gandevia SC. Twitch interpolation in human muscles: mechanisms and implications for measurement of voluntary activation. J Neurophysiol. 1999;82:2271–2283. doi: 10.1152/jn.1999.82.5.2271. [DOI] [PubMed] [Google Scholar]
  24. Johnson KV, Edwards SC, Van Tongeren C, Bawa P. Properties of human motor units after prolonged activity at a constant firing rate. Exp Brain Res. 2004;154:479–487. doi: 10.1007/s00221-003-1678-z. [DOI] [PubMed] [Google Scholar]
  25. Jones DA, Bigland-Ritchie B, Edwards RHT. Excitation frequency and muscle fatigue: mechanical responses during voluntary and stimulated contractions. Exp Neurol. 1979;64:401–413. doi: 10.1016/0014-4886(79)90279-6. [DOI] [PubMed] [Google Scholar]
  26. Khan SI, Giesebrecht S, Gandevia SC, Taylor JL. Activity-dependent depression of the recurrent discharge of human motoneurones after maximal voluntary contractions. J Physiol. 2012;590:4957–4969. doi: 10.1113/jphysiol.2012.235697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kooistra RD, de Ruiter CJ, de Haan A. Conventionally assessed voluntary activation does not represent relative voluntary torque production. Eur J Appl Physiol. 2007;100:309–320. doi: 10.1007/s00421-007-0425-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Lännergren J, Westerblad H. Action potential fatigue in single skeletal muscle fibres of Xenopus. Acta Physiol Scand. 1987;129:311–318. doi: 10.1111/j.1748-1716.1987.tb08074.x. [DOI] [PubMed] [Google Scholar]
  29. Marsden CD, Meadows JC, Merton PA. Isolated single motor units in human muscle and their rate of discharge during maximal voluntary effort. J Physiol. 1971;217:12–13P. [PubMed] [Google Scholar]
  30. Marsden CD, Meadows JC, Merton PA. ‘Muscular wisdom’ that minimizes fatigue during prolonged effort in man: peak rates of motoneuron discharge and slowing of discharge during fatigue. Adv Neurol. 1983;39:169–211. [PubMed] [Google Scholar]
  31. McKenzie DK, Bigland-Ritchie B, Gorman RB, Gandevia SC. Central and peripheral fatigue of human diaphragm and limb muscles assessed by twitch interpolation. J Physiol. 1992;454:643–656. doi: 10.1113/jphysiol.1992.sp019284. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. McNeil CJ, Martin PG, Gandevia SC, Taylor JL. The response to paired motor cortical stimuli is abolished at a spinal level during human muscle fatigue. J Physiol. 2009;587:5601–5612. doi: 10.1113/jphysiol.2009.180968. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Merton PA. Voluntary strength and fatigue. J Physiol. 1954;123:553–564. doi: 10.1113/jphysiol.1954.sp005070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Peters EJD, Fuglevand AJ. Cessation of human motor unit discharge during sustained maximal voluntary contraction. Neurosi Lett. 1999;274:66–70. doi: 10.1016/s0304-3940(99)00666-7. [DOI] [PubMed] [Google Scholar]
  35. Place N, Yamada T, Bruton JD, Westerblad H. Interpolated twitches in fatiguing single mouse muscle fibres: implications for the assessment of central fatigue. J Physiol. 2008;586:2799–2805. doi: 10.1113/jphysiol.2008.151910. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Rossi A, Rossi S, Ginanneschi F. Activity-dependent changes in intrinsic excitability of human spinal motoneurones produced by natural activity. J Neurophysiol. 2012;108:2473–2480. doi: 10.1152/jn.00477.2012. [DOI] [PubMed] [Google Scholar]
  37. Sawczuk A, Powers RK, Binder MD. Spike frequency adaptation studied in hypoglossal motoneurons of the rat. J Neurophysiol. 1995;73:1799–1810. doi: 10.1152/jn.1995.73.5.1799. [DOI] [PubMed] [Google Scholar]
  38. Sejersted OM, Sjøgaard G. Dynamics and consequences of potassium shifts in skeletal muscle and heart during exercise. Physiol Rev. 2000;80:1411–1481. doi: 10.1152/physrev.2000.80.4.1411. [DOI] [PubMed] [Google Scholar]
  39. Seki K, Kizuka T, Yamada H. Reduction in maximal firing rate of motoneurons after 1-week immobilization of finger muscle in human subjects. J Electromyogr Kinesiol. 2007;17:113–20. doi: 10.1016/j.jelekin.2005.10.008. [DOI] [PubMed] [Google Scholar]
  40. Taylor JL, de Haan A, Gerrits KHL, de Ruiter CJ. Point:Counterpoint: The interpolated twitch does/does not provide a valid measure of the voluntary activation of muscle. J Appl Physiol. 2009;107:354–368. doi: 10.1152/japplphysiol.91220.2008a. [DOI] [PubMed] [Google Scholar]
  41. Todd G, Taylor JL, Gandevia SC. Measurement of voluntary activation of fresh and fatigued human muscles using transcranial magnetic stimulation. J Physiol. 2003;551:661–671. doi: 10.1113/jphysiol.2003.044099. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Todd G, Butler JE, Taylor JL, Gandevia SC. Hyperthermia: a failure of the motor cortex and the muscle. J Physiol. 2005;563:621–631. doi: 10.1113/jphysiol.2004.077115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Vagg R, Mogyoros I, Kiernan MC, Burke D. Activity-dependent hyperpolarization of human motor axons produced by natural activity. J Physiol. 1998;507:919–925. doi: 10.1111/j.1469-7793.1998.919bs.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES