Medicine & Science in Sports & Exercise:
APPLIED SCIENCES: Biodynamics
Decreased EMG median frequency during a second bout of eccentric contractions
WARREN, GORDON L.; HERMANN, KARL M.; INGALLS, CHRISTOPHER P.; MASSELLI, MARIA R.; ARMSTRONG, R. B.
Muscle Biology Laboratory, Department of Health and Kinesiology, Texas A&M University, College Station, TX 77843; and Muscle Biology Laboratory, Department of Physical Therapy, Georgia State University, Atlanta, GA 30303
Submitted for publication August 1998.
Accepted for publication January 1999.
Address for correspondence: Gordon L. Warren, Department of Physical Therapy, University Plaza, Georgia State University, Atlanta, GA 30303-3083; E-mail email@example.com.
WARREN, G. L., K. M. HERMANN, C. P. INGALLS, M. R. MASSELLI, and R. B. ARMSTRONG. Decreased EMG median frequency during a second bout of eccentric contractions. Med. Sci. Sports Exerc., Vol. 32, No. 4, pp. 820–829, 2000.
Purpose: Others have reported preferential recruitment of fast motor units in muscles during performance of eccentric contractions and there is evidence that fast muscle fibers are more susceptible to eccentric contraction-induced injury. We tested the hypothesis that during a second bout of maximal eccentric contractions 1 wk after the first, there would be a reduction in the electromyographic (EMG) median frequency (MF) with minimal change in the EMG root-mean-square (RMS), indicating greater reliance on slower motor units. This could provide an explanation for the enhanced resistance to eccentric contraction-induced injury after a single bout of eccentric exercise.
Methods: Human subjects performed 50 maximal voluntary eccentric (N = 10) or concentric (N = 10) contractions of the anterior crural muscles on two occasions separated by 1 wk. To determine whether MF changes during the second bout could be a consequence of injury to fibers in fast motor units, the anterior crural muscles of mice were electrically stimulated to perform 50 maximal eccentric (N = 10) or concentric (N = 9) contractions on two occasions separated by 1 wk. In both the humans and mice, torque production and tibialis anterior muscle RMS and MF were measured during the two exercise bouts.
Results: In human tibialis anterior muscle, MF was 30% lower (P < 0.01) during the second eccentric bout although RMS was the same. In the mice, RMS and MF were unchanged at any time after the first eccentric bout despite torque deficits similar to those observed in the humans.
Conclusions: The data indicate that with repetition of maximal voluntary eccentric contractions, there is an increased activation of slow motor units and a concomitant decrease in activation of fast units.
Asingle bout of exercise with eccentric contractions that induces injury is followed by a decreased susceptibility of the muscle to injury in subsequent bouts of eccentric exercise (6,28,29). The mechanism underlying this rapid adaptation has not been elucidated. Hypotheses proposed to explain the phenomenon include: 1) modification of susceptible structures within the muscles; and 2) alteration of patterns of fiber recruitment within or among muscles (6).
Motor unit recruitment with increasing force production normally progresses from slow (Type I fibers) to fast fatigue-resistant (Type IIa fibers) to fast fatiguable (Type IIb fibers) units (5). However, there is evidence that when muscles perform eccentric contractions, there is a greater reliance on fast motor units, both within muscles (16,26) and among synergistic muscles (27). In fact, Nardone and coworkers (26) have shown that some large, very high threshold (and presumably fast) motor units are recruited exclusively during eccentric contractions. In that study, ∼20% of the motor units studied in the gastrocnemius and soleus muscles discharged only during lengthening and not during maximal isometric or ballistic shortening contractions. In light of the evidence that fast muscle fibers are more susceptible to injury than slow fibers during performance of eccentric contractions (10,23,33), it is of interest that these motor units are preferentially recruited for these tasks.
It is not clear from the reported studies whether or not fast motor units continue to be preferentially recruited during eccentric contractions after the movement patterns are “learned.” In fact, there is evidence that in some stereotyped movements, slow fibers are primarily recruited during eccentric contractions. For example, during walking, cat soleus muscle (100% slow fibers) is active during the early stages of the stance phase when the muscle is stretched, with lesser activation of the faster gastrocnemius muscle (32). Also, during downhill walking in rats, most of the injury that occurs in the extensor muscles is in slow fibers (2). These various observations suggest one hypothesis that would explain the apparent anomaly: when confronted with unaccustomed eccentric contractions, the nervous system “naively” elicits an inappropriate recruitment pattern, but with continued performance of the movements, there is a shift to the more typical pattern of activation of motor unit types. Furthermore, a shift away from recruitment of faster motor units toward activation of slower motor units during eccentric contractions could provide a potential mechanism for the reported rapid adaptation that reduces susceptibility to injury in later bouts of eccentric exercise. Also, it may be the initial injury to fast fibers that necessitates greater activation of slower motor units.
The first objective of this study was to determine if electromyographic (EMG) median frequency (MF) was reduced during a second bout of eccentric contractions. Human subjects performed 50 maximal voluntary eccentric contractions of the anterior crural muscles on two occasions separated by one week. The initial EMG MF in tibialis anterior muscle was significantly decreased in the second bout, even though EMG root-mean-square (RMS) did not change. One interpretation of these results was that during the second bout, torque production was maintained through a greater contribution of slower motor units (11,12,22).
The second objective of this study was to determine whether the apparent change in motor unit recruitment in eccentric contractions between the two bouts was the result of injury to fast fibers. We sought to find out if mouse anterior crural muscles when electrically stimulated to perform 50 maximal eccentric contractions would result in injury to fast fibers and elicit a decreased EMG activity during subsequent contractions. The mouse data suggest that the reduction in MF observed in the human study was not a consequence of fibers in fast motor units becoming electrically unexcitable as a result of the injury.
Twenty healthy men volunteered for this study. The mean (±SD) age, height, and weight of the subjects were 26.7 (±4.8) yr, 178.0 (±8.2) cm, and 80.6 (±14.7) kg, respectively. The subjects were randomly assigned to one of two groups, a group that performed concentric contractions (N = 10) or a group that performed eccentric contractions (N = 10). The subjects were screened to ensure that they possessed no contraindications to exercise or any pathology affecting the neuromusculoskeletal system of the lower extremities. All subjects gave written informed consent before initiation of the study. The experimental protocol was approved by the Texas A&M University Institutional Review Board.
The subjects performed either two bouts of 50 maximal voluntary eccentric or two bouts of 50 maximal voluntary concentric contractions of the right anterior crural muscles; the two exercise bouts were separated by 7 d. The anterior crural muscles were chosen because it is assumed they do not perform forceful eccentric contractions during normal daily activities. The intent was that the eccentric contractions would be a novel task for these muscles during the eccentric contraction group’s first bout. The 7-d interval between bouts was chosen because: 1) a decreased susceptibility to eccentric contraction induced-injury exists as early as 5 d after the first bout (6), and 2) we knew from preliminary work that muscle strength and EMG RMS were essentially recovered after 7 d.
The subjects were positioned sitting with the hip and knee in ∼90° and ∼72° flexion, respectively. A seatbelt was placed across the hips, and the subject’s arms were folded across the chest. The foot was strapped to a footplate that was attached to a computer-controlled servomotor (Omnitech Robotics model MC-1000; Englewood, CO). The footplate was adjusted so that each subject’s axis of rotation for ankle dorsiflexion/plantarflexion was aligned with the shaft of the servomotor. The axis of rotation was assumed to be that of a line perpendicular to the tibial long axis that passed through the distal tips of the malleoli. Before performing either the 50 eccentric or concentric contraction protocols, five maximal voluntary isometric contractions of the anterior crural muscles were performed with the ankle in 0° of dorsiflexion. Eccentric contractions were begun with the ankle dorsiflexed by 10°, and movement of the footplate was resisted using the anterior crural muscles until 50° of plantarflexion was attained. Conversely, concentric contractions were begun with the ankle plantarflexed by 50°, and the movement ended at 10° of dorsiflexion. The 60° movements were performed at an angular velocity of 45°·s−1, so the contractions took 1.3 s; there was a 30-s rest period between contractions. For both types of contractions, movement of the footplate was not initiated until the subject had exerted a torque equal to 50% of his maximal isometric torque measured before the first exercise bout. Before performing any isokinetic contractions, each subject was familiarized with the range of movement by performing three passive movements over the 60° range. Torque and ankle angular position were sampled by an 80486–66 MHz computer at 3000 Hz using an interface board (Keithley-Metrabyte DAS-1801 ST/DA, Cleveland, OH).
Surface EMG was used to assess muscle recruitment during the two exercise bouts. EMG signals were recorded from the tibialis anterior, medial gastrocnemius, and soleus muscles. The tibialis anterior muscle is the primary ankle dorsiflexor, and the medial gastrocnemius and soleus muscles are antagonistic muscles that could affect torque measured at the ankle during contraction of the anterior crural muscles. The skin was prepared by shaving, abrading, and cleansing with alcohol. Pairs of silver-silver chloride electrodes (8-mm diameter recording surface, 20-mm interelectrode distance) were applied to the skin overlying the three muscles. The distance from the fibular head to lateral malleolus was measured, and the soleus muscle electrodes were placed midway between the fibular head and lateral malleolus. At one-third of the distance from the fibular head to the lateral malleolus, the leg circumference was measured; the medial gastrocnemius muscle electrodes were placed at one-third of this leg circumferential distance moving medially from the tibial crest (21). The tibialis anterior muscle site was located over the muscle belly at a level equal to one-third of the distance from the fibular head to the lateral malleolus (26). The reference electrode was placed on the skin overlying the lateral malleolus. After the electrodes were in position, electrode impedance at 30 Hz was checked to ensure that it was below 5 kOhm. To ensure the same electrode placement in the two exercise bouts, the electrode sites were marked on the skin with indelible ink at the time of the first bout.
The EMG signals were amplified 200 times and passed through a bandpass filter with low and high cut-off frequency settings of 3 and 1000 Hz, respectively (Grass Instruments model P511K, Quincy, MA). The amplified EMG signals were sampled at 3000 Hz using the same computer and interface board used for acquisition of the torque and ankle angular position data. The EMG signal was subsequently passed through a high-pass (i.e., 10 Hz) digital filter (Keithley-Metrabyte VTX ver. 1.1). EMG root-mean-square (RMS) and median frequency (MF) were calculated for each of the six 10° angular segments during the 60° movement because of the RMS dependence on muscle length (7,20) and to better approximate a stationary EMG signal that is assumed in the calculation of MF. In calculation of MF, a Hamming window was first applied to the data (i.e., ∼667 data points per angular segment). The data were then zero-padded before running a 2048-point fast Fourier transform. MF was calculated from the power density spectrum as the frequency at which 50% of the points in the spectrum lay at or below.
Reliability of the torque and tibialis anterior EMG RMS measurements was checked in a group of seven men. These individuals performed nine maximal voluntary isometric contractions spread over a 30-min period; this protocol was then repeated 2 d later. For the two measures, intraclass correlation coefficients exceeded 0.96 and 0.93 for within and between day retests, respectively.
Female ICR mice obtained from Harlan Laboratories (Indianapolis, IN) were used. The mice were 3–4 months old and weighed 35.0 (±3.4) g at the time of the first exercise bout. They were housed with a 12:12-h light-dark photoperiod at the American Association of Laboratory Animal Care-accredited Laboratory Animal Research and Resources facility at Texas A&M University. For surgical implantation of the stimulating nerve cuff on the common peroneal nerve, the mice were anesthetized with an intraperitoneal injection of sodium pentobarbital (100 mg·kg−1). For surgical implantation of the EMG electrodes on the tibialis anterior muscle, the mice were anesthetized with an intraperitoneal injection of fentanyl (0.33 mg·kg−1), droperidol (16.7 mg·kg−1), and diazepam (5 mg·kg−1). This latter anesthetic regimen was also used when the mice performed the exercise protocols because it has no adverse effects on contractility (18). All animal care and use procedures were approved by the Institutional Animal Care and Use Committee and were in conformance with the policy statement of the American College of Sports Medicine on research with experimental animals.
A stimulating nerve cuff was implanted on the left common peroneal nerve as described previously (34). Briefly, the nerve cuff was constructed from two Teflon-coated, multi-stranded 90% platinum (Pt):10% iridium (Ir) wires (0.15 mm dia.) (Medwire-Sigmund Cohn Corp. 10Ir9/49T, Mount Vernon, NY). An incision was made through the biceps femoris muscle and the two loops formed from 2.5-mm segments of deinsulated Pt:Ir wire were placed around the common peroneal nerve. The proximal end of the nerve cuff was externalized in the dorsal cervical region.
Eleven days after implantation of the stimulating nerve cuff, maximal isometric tetanic torque produced by the left anterior crural muscles was measured using the servomotor system described previously (18,25); stimulations consisted of 200 ms trains of 75-μs biphasic pulses at 300 Hz. The tibialis anterior and extensor digitorum longus muscles produce 89% and 11%, respectively, of the torque in this model (25). Tibialis anterior muscle EMG electrodes were implanted if torque production exceeded 2.54 N·mm; this minimum torque level was selected based on observations that 95% of all female ICR mice in the weight range studied produce this torque or greater when using the fentanyl/droperidol/diazepam anesthetic regimen (34).
The tibialis anterior muscle EMG electrodes were implanted as described previously (34). Briefly, after making a skin incision running longitudinally along the anterior lower leg, a 23-gauge needle was passed just beneath the fascial sheath covering the tibialis anterior muscle at its midbelly. The needle was passed in a direction perpendicular to the superficial fiber longitudinal orientation. The deinsulated end of a Pt:Ir wire was passed through the needle and the needle withdrawn, leaving a 3-mm segment of deinsulated wire beneath the fascia. A second length of wire was routed underneath the fascia in a direction parallel to the first wire but offset in the distal direction by 2 mm. The electrode wire spacing theoretically permitted sampling of EMG activity from the full thickness of the tibialis anterior muscle beneath the electrodes (3). The wires were secured to adjacent tissue and the proximal ends of the wires externalized in the dorsal cervical region. Using these surgical techniques, we have demonstrated good reliability for measurement of maximal isometric tetanic torque and RMS when measured six times over a 14-d period (34).
Exercise protocols were initiated five days after surgical implantation of the EMG electrodes. The anesthetized mice were positioned in the servomotor system as described previously (18,25). Peak torque production was then optimized by varying stimulation voltage in a series of isometric contractions (i.e., 5–8). Next, the relationships of stimulation frequency to torque, RMS, and MF were determined in a series of isometric contractions (data not shown).
The mice then performed a series of either 50 eccentric (N = 10) or 50 concentric (N = 9) contractions. The eccentric contractions were done as described previously (i.e., from 20° of ankle dorsiflexion to 20° of ankle plantarflexion at an angular velocity of 2000°·s−1; this movement was preceded by a 100-ms isometric stimulation, so the contraction lasted a total of 120 ms). From previous experience with this model (e.g., 18,25), we designed the eccentric contraction protocol to cause a similar decrease in torque as that observed in the human study. The concentric contractions were done similarly but in the reverse direction (i.e., from 20° of plantarflexion to 20° of dorsiflexion). There were 10 s between contractions except for every fifth contraction, which took an additional 12 s while the data were written to disk. Two minutes after the 50th eccentric or concentric contraction, maximal isometric tetanic torque was measured. The relationships of stimulation frequency to torque, RMS, and MF were determined again (data not shown). The 50-contraction protocol was repeated 7 d later. In addition, maximal isometric tetanic torque and the relationships of stimulation frequency to torque, RMS, and MF were determined at 1 and 3 d after the first protocol.
The proximal ends of the tibialis anterior muscle EMG electrodes were externalized and cleaned with acetone. A wire acutely implanted beneath the skin in the abdominal region served as the reference electrode. Electrode impedance at 30 Hz was then measured; impedance remained constant over the 7-d period at 10–15 kOhm. The EMG signal was amplified 1000 times and passed through a bandpass filter with low and high cut-off frequency settings of 10 and 3000 Hz, respectively (Grass Instruments model P-15). The amplified EMG signal was sampled at 5000 Hz by the computer used for acquisition of the torque and ankle angular position data. The EMG signal was subsequently passed through a low-pass (i.e., 2000 Hz) digital filter (Keithley-Metrabyte VTX ver. 1.1) to remove aliasing in the 2000–3000 Hz range; any DC offset was removed at this time as well. The methods for calculation of EMG RMS and MF were the same as for the human study except that MF was calculated on the averaged M-wave as in Kupa et al. (22). RMS and MF were only calculated for the 100-ms isometric contraction immediately preceding the 20-ms isokinetic contraction. It should be kept in mind that the human and mouse MFs are not directly comparable. Whereas the mouse MF is primarily determined by the action potential conducting properties of the fibers, the human MF is also affected by motor unit firing rates and synchronization. However, it is thought that these latter two factors have little effect on the EMG frequency spectrum (7,15).
The recorded EMG in this model is not contaminated with stimulation artifact. First, administration of the two neuromuscular junction blockers, succinylcholine chloride (3.0 mg·kg−1) and d-tubocurarine chloride (0.4 mg·kg−1), can reduce RMS by over 70%, an observation one would not expect if the EMG signal was composed predominately of stimulation artifact. Second, when stimulation voltage is increased stepwise to maximize torque production by the anterior crural muscles, RMS plateaus after a certain threshold voltage is attained. No effect on RMS is observed when stimulation voltage is increased to ≥ 3 times the threshold voltage (34).
For the human study, the effects of group (concentric vs eccentric), bout (bout 1 vs bout 2), contraction number (1, 5, 10, . . . , 50), and angle (−10°−0°, 0–10°, .., 40–50°) on torque, RMS, and MF were evaluated using a four-way ANOVA with repeated measures on bout, contraction number, and angle. For the animal study, the effects of group, bout, and contraction number on torque, RMS, and MF were evaluated using a three-way ANOVA with repeated measures on bout and contraction number. When significant main effect interactions were found, significant differences between means were determined using contrast statements. PC-SAS 6.08 (SAS Institute) and Systat 7.0 (SPSS Inc.) were used for the human and animal analyses, respectively. With one exception in the human study, the effects of group, bout, and contraction number were manifested across all angle segments, so for clarity of presentation, the data in the Results are presented as collapsed across angle segments. An α level of 0.05 was used for all analyses and the values reported in the Results are means ± standard errors.
Anterior crural muscle torque.
Maximal voluntary eccentric and concentric torques during the 50 contractions in bouts 1 and 2 are presented in Figure 1. Peak torque during the first eccentric contraction in bout 1 was 70% greater than the maximal isometric torque (i.e., 46.7 ± 0.7 N·m) measured immediately before the eccentric contraction protocol. Eccentric torques progressively decreased during bout 1 and by the 50th contraction, torque was 29% lower than on the initial contraction. In bout 2, the initial eccentric torque was 11% lower than in bout 1. In bout 2, torques again progressively declined, and by the 50th contraction had decreased by 18%.
Peak torque during the first concentric contraction in bout 1 was 35% lower than the maximal isometric torque measured immediately prior (i.e., 40.2 ± 1.4 N·m). Initial concentric torques in bouts 1 and 2 were about one-third those for the eccentric contractions. There were no differences in concentric torques between bouts 1 and 2, and no changes in either bout over the 50 contractions.
Tibialis anterior muscle RMS.
Human tibialis anterior muscle RMS data during the 50 maximal voluntary eccentric and concentric contractions in the two bouts are presented in Figure 2. Eccentric contraction RMS progressively decreased over the 50 contractions in both bout 1 (24%) and bout 2 (21%). There were no differences in RMS between bout 1 and bout 2.
Similarly, concentric contraction RMS in tibialis anterior muscle progressively decreased over the 50 contractions in both bouts (24% and 15% in bout 1 and bout 2, respectively). There were no differences in RMS between bouts 1 and 2.
Tibialis anterior muscle MF.
Human tibialis anterior muscle MF data during the 50 maximal voluntary eccentric or concentric contractions in bouts 1 and 2 are presented in Figure 3. Over the 50 eccentric contractions in bout 1, MF decreased 34%. In bout 2, MF during the initial eccentric contraction was 28% lower than in bout 1. MF decreased 42% over the course of the 50 eccentric contractions in bout 2. Thus, MF was lower throughout the eccentric contraction protocol in bout 2; on average over the contractions, MF was 30% lower in bout 2.
MF decreased 22% during the 50 concentric contractions in bout 1 and 14% in bout 2. In bout 2, MF was elevated compared to bout 1; on average over the contractions, MF was 20% higher in bout 2.
Soleus and medial gastrocnemius muscle RMS.
Initial RMS in soleus and medial gastrocnemius muscles was low (i.e., ≤20% of that in tibialis anterior muscle) in both eccentric and concentric protocols in both bouts. For both protocols, there was no difference in initial RMS between bouts 1 and 2 in either soleus or medial gastrocnemius muscles; soleus muscle RMS decreased over time during both contraction protocols in both bouts (∼10%). Medial gastrocnemius muscle RMS did not change in either protocol or in either bout.
Anterior crural muscle torque.
Mouse eccentric and concentric torques over the 50 contraction protocol in bouts 1 and 2 are presented in Figure 4. In bout 1, eccentric torques decreased progressively over the 50 contractions; by the 50th contraction, torque had decreased by 26% (similar to the 29% decrement in torque observed in the human study). Maximal isometric tetanic torque was decreased by 24% immediately after the eccentric contraction protocol, and remained depressed over the succeeding three days (Table 1). In bout 2, eccentric torques also decreased progressively, and during the first 20 contractions were similar to those in bout 1 (Fig. 4). By the 50th contraction, eccentric torque had decreased by 19% from the initial contraction.
Initial concentric torques in both bout 1 and bout 2 were about half of the initial eccentric torques (Fig. 4). Concentric torques showed very small but significant decreases (i.e., 0.3–2%) over the 50 contraction protocol in the two bouts, but there was no difference in concentric torques between bouts 1 and 2. Maximal isometric tetanic torque was elevated immediately after the concentric protocol (Table 1), but returned to baseline levels in the succeeding days.
Tibialis anterior muscle RMS.
Mouse tibialis anterior muscle RMS data for eccentric and concentric contractions in bouts 1 and 2 are presented in Figure 5. There were no changes in RMS over the 50 contractions for either eccentric or concentric protocols in either bout 1 or bout 2 and there were no differences between bouts for either protocol and no differences between protocols. Also, in the days after the protocols, there were no changes in tibialis anterior muscle RMS during maximal isometric tetanic contractions in either eccentric or concentric groups (Table 1).
Tibialis anterior muscle MF.
Mouse MF data for tibialis anterior muscle are presented in Figure 6. MF showed small but significant (i.e., 1–5%) decreases over the 50 contractions in both eccentric and concentric protocols in both bouts. There were no differences between bouts for either protocol and no differences between protocols. Also, in the days after the protocols, there were no changes in tibialis anterior muscle MF during maximal isometric tetanic contractions in either group (Table 1).
The MF data for the human subjects support the hypothesis that with experience, there is an increased reliance on slower motor units during performance of maximal voluntary eccentric contractions. The strongest evidence for this change is the observation that tibialis anterior muscle MF was lower during eccentric contractions in bout 2 than in bout 1 without a concomitant decrease in RMS. The similarity in RMS between bouts 1 and 2 during the eccentric contraction protocols suggests that total neural activation of the muscle was similar between the two bouts. Therefore, the decrease in MF between the two bouts may be interpreted as an increased activation of slower motor units and a decreased activation of fast units.
Our interpretation of the MF data is based on observations that MF values during muscle contractions are related to the fiber type populations in the muscles (12,22,40). Kupa et al. (22) recently reported high correlations between initial MF and the populations of slow oxidative (SO), fast oxidative glycolytic (FOG), and fast glycolytic (FG) fibers in three rat muscles (soleus, extensor digitorum longus, and diaphragm) during isometric tetanic contractions in vitro. For example, initial MF in soleus muscle (80% slow fibers) was 20% lower than in extensor digitorum longus muscle (3% slow fibers) during tetanic stimulation. These findings are important, because with the in vitro preparation one is reasonably confident that most if not all fibers are active, whereas in human studies, populations of active fast and slow motor units may not reflect fiber composition of the whole muscle, even during maximal voluntary contractions (26).
Our interpretation of the MF data is also contingent on the assumption that all motor units are not recruited during maximal eccentric contractions. Though there is no evidence to definitively support or refute this assumption, the available data do generally support the assumption. First, in recent work, it has been shown that the maximum eccentric torque predicted from quick stretches imposed upon different isometric force levels is twice the eccentric torque measured during maximal voluntary efforts (38). This observation indicates that motor unit activation is considerably lower during maximal eccentric contractions than during maximal isometric tetanic contractions. Second, RMS has been reported to be 15–30% lower during maximal eccentric contractions compared with maximal concentric contractions (31,39). In the present study, RMS was 16% lower on average during the eccentric contractions. Though a decrease in motor unit discharge rate without a change in the number of recruited units could account for the decreased RMS, it seems unlikely that this could account for all of the reduction in RMS. Third, in the work of Nardone et al (26)., a large percentage of the motor units (i.e., 26–54%) studied in the gastrocnemius and soleus muscles discharged only during shortening and/or maximal isometric tetanic contractions and not during eccentric contractions. However, the eccentric contractions in this study were not maximal and it is likely that at least some of these motor units would have been recruited during higher force eccentric contractions. Finally, there are data indicating that the excitability of small motoneurons is reduced during eccentric contractions (1,27). These data collectively argue that recruitment is incomplete during maximal voluntary eccentric contractions.
An alternative explanation for the decrease in human tibialis anterior muscle MF in the second bout is that the same motor units were active but were discharging at a lower frequency. Support for this possibility is provided by Nardone et al. (26) and Søgaard and coworkers (30). Both groups of investigators found that a substantial population of motor units are active during both concentric and eccentric contractions, but that during the eccentric phase, they fire at a lower rate. However, the frequency content of the EMG signal is determined primarily by the waveform of the action potentials of the active units. Discharge rates and synchronization contribute little to the EMG frequency spectrum (7,15) and should have minimal if any influence on MF. Furthermore, motor unit synchronization in the human tibialis anterior muscle has been shown to occur sporadically and in less than 8% of motor unit firings (8). Moreover, the fact that human tibialis anterior muscle RMS was not lower during eccentric contractions in bout 2 in the present study argues against this explanation for the decrease in MF. If the same motor units were active during the two bouts, but the firing rates were reduced in the second bout, RMS and torque should have been lower during bout 2.
Attenuation of human tibialis anterior muscle MF during the 50 eccentric contractions in both bout 1 and 2 presumably was not a result of a shift from faster to slower motor units, since the decreases in MF during the bouts were associated with decreases in RMS and torque. Changes in muscle pH, temperature, and blood flow can influence MF (22), and it is also possible that changes in these variables contributed to the progressive decreases in MF that occurred over the 50 contraction protocols in the two bouts. The percent decline in eccentric MF in the first bout (34%) was greater than the decrease in the concentric MF (22%); theoretically, the concentric contractions should have required greater energy expenditure (4,17), therefore producing lower pH, and higher temperature and blood flow. This argues against these metabolic factors as the primary mechanism of the decrease in tibialis anterior muscle MF during the bout of eccentric contractions but does not eliminate their potential contribution. It seems even less likely that these factors would explain the downward shift in bout 2 compared with bout 1, although this possibility also cannot be completely discounted.
The mechanisms underlying the apparent alteration in motor unit recruitment with repetition of maximal voluntary eccentric contractions are of interest. Two contrasting explanations for the shift can be proposed. First, repeated performance of the eccentric contraction protocol may have resulted in “learning” in which the nervous system purposely altered its pattern of recruitment and activated more slow motor units and fewer fast units. This explanation is based on the underlying assumption that it would be biologically “preferable” to recruit slow units during the eccentric contractions, which may not be the case. For example, Nardone et al. (26) argued that recruitment of fast units during eccentric contractions provides better control of the lengthening movement because of the more rapid relaxation times of fast fibers. Howell et al. (16) suggested that the alternate recruitment pattern simply reflects differences in synaptic input related to central neural organization, which may deter rapid adaptation. However, as pointed out in the introductory section, there is evidence that during slow locomotion, slow motor units are preferentially recruited during lengthening contractions (32). Also, it could be beneficial to primarily rely on slow units for force production during eccentric contractions because of their greater resistance to injury (10,23,33). Finally, previous observations in our laboratory suggest there are functional advantages to recruiting slow units during eccentric contractions. First, mouse slow soleus muscle can produce a higher percentage of maximal isometric tetanic force (Po) than the fast extensor digitorum longus (EDL) muscle during maximal eccentric contractions (i.e., 189% vs 167% Po) (24,33). Second, maximal eccentric force is 13% higher in mouse EDL muscle at 30°C than at 37°C (35). Both of these findings suggest that slower cross-bridge kinetics permit higher force production during eccentric contractions. Conversely, force production by fast motor units would be preferable during concentric contractions because of their ability to produce greater force at a given shortening velocity, which could explain the increase in tibialis anterior muscle MF during the second bout of concentric contractions. Thus, there is a teleological basis for expecting that repetition of maximal voluntary eccentric contractions may be accompanied by a purposeful shift in recruitment toward slower motor units.
A second explanation for the apparent alteration in recruitment in bout 2 is that in the first bout fast motor units were recruited during the eccentric protocol but that constituent fibers were injured, resulting in greater activation of slower units in the second bout. In support of this hypothesis, eccentric torque produced by the anterior crural muscles at the ankle was lower during the first contraction in bout 2 in spite of similar RMS, i.e., there was a reduction in torque/RMS ratio. This could be explained by increased activation of antagonistic muscles (i.e., soleus and gastrocnemius), but there were no differences in RMS between bouts 1 and 2 in either of the posterior crural muscles. It is therefore reasonable to suggest that some fibers were being activated but were unable to produce normal force, i.e., the reduction in performance was a manifestation of muscle fiber injury resulting from the eccentric contractions. Felici et al. (9) found that performance of 70 eccentric contractions of biceps brachii muscle resulted in a decrease in MF during isometric contractions at both 80% and 50% MVC that persisted over the succeeding 4 d. These authors interpreted their data to indicate that “selective or prevalent” injury occurred in fast-twitch fibers during the eccentric contraction protocol.
We attempted to address the fiber injury hypothesis with the mouse study. In these experiments, the anterior crural muscles were maximally activated with electrical stimulation of the common peroneal nerve during the eccentric protocol. Although maximal isometric tetanic torque was depressed for several days after the first eccentric contraction protocol, there was no change in either tibialis anterior muscle RMS or MF at any time, suggesting the injury that occurred did not affect plasmalemmal transmission of action potentials. Furthermore, performance of 150 eccentric contractions leads to prolonged (i.e., ≥2 wk) torque deficits that are two-fold greater than those observed in the present study, but MF is unaffected and RMS is only transiently depressed to a small extent, returning to baseline one day after the protocol (36). Thus, because mouse tibialis anterior muscles are >95% fast twitch (unpublished observations), one can conclude that when injured fast-twitch fibers are recruited, normal EMG activity results. This conclusion is consistent with our previous observations in injured muscles of an excitation-contraction coupling failure explaining most of the force loss, with the site of failure being distal to the t-tubule in the excitation pathway (19,33,36,37). Thus, extrapolating to the human study, our data to date suggest that the reduction in the human EMG MF during the second eccentric bout was not a result of injured fast twitch fibers becoming electrically unexcitable.
How confidently the mouse data can be used to interpret the human results is of course open to question. It is not possible to determine the equivalency of the effects of the contractile protocols on the muscles in the respective species. It could be argued that, if anything, injury should be greater in the mouse muscles. First, there is a marked difference in the fiber type composition of the tibialis anterior muscle in the two species. Several reports indicate that human tibialis anterior muscle is a predominantly slow muscle similar to soleus muscle. Henriksson-Larsén et al. (14) and Helliwell et al. (13) reported means of 66% and 77% Type I fibers, respectively. On the other hand, tibialis anterior muscle in the mouse is composed of >95% fast-twitch fibers. Thus, the higher population of fast fibers in the mouse tibialis anterior muscle would suggest it might be relatively more susceptible to injury than the slower muscle in human subjects. Second, one would expect more injury in the mouse fibers because the contractions were true maximal efforts through optimized electrical stimulation of all motor units.
Another problem in comparing results from humans and mice is that it is not possible to assess the relative susceptibilities of the muscles in the two species to injury from eccentric contractions. Of perhaps even greater importance than the fiber type composition of the muscles, susceptibility to injury has been shown to be related to the activity history of the muscle (33); chronically active muscles are more resistant to injury than less-active muscles independent of fiber type. The relative activity levels of the mouse and human anterior crural muscles are unknown. Nonetheless, the findings in the animal study are consistent with the interpretation that the human subjects did not show altered EMG MF in bout 2 because of injury-induced inactivation of fast fibers.
If repetition of bouts of eccentric contractions is in fact associated with a greater activation of slower motor units, this could provide an explanation for the enhanced resistance to injury in repeated bouts. It has been noted by various investigators that even a single bout of eccentric contractions that induces injury results in greater resistance to injury in subsequent bouts (6,28). It has never been clear whether this decreased susceptibility to injury stems from structural adaptations in the initially injured fibers or from alterations in the patterns of fiber recruitment. The animal study reported by Sacco and Jones (28) indicated that the training effect resulted from adaptation within the muscles, because they used electrical stimulation to elicit contractions; the smaller reduction in mouse anterior crural muscle torque in bout 2 compared with bout 1 in the present study agrees with that report. On the other hand, the human data in the present study are consistent with the altered recruitment hypothesis.
In conclusion, the results of this study support the hypothesis that with repetition of a second bout of maximal voluntary eccentric contractions, there is an increased activation of slow motor units and a concomitant decrease in activation of fast units. The data also indicate that this change is not due to fibers in fast units becoming electrically unexcitable as a result of injury from high force eccentric contractions. This phenomenon may provide an explanation for the rapid adaptation in which muscles become less susceptible to injury from eccentric contractions with repetition of the movement.
This study was supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant AR-42761 and the Omar Smith Chair.
1. Abbruzzese, G., M. Morena, L. Spadavecchia, and M. Schieppati. Response of arm flexor muscles to magnetic and electrical brain stimulation during shortening and lengthening tasks in man. J. Physiol. (Lond.) 481:499–507, 1994.
2. Armstrong, R. B., R. W. Ogilvie, and J. A. Schwane. Eccentric exercise-induced injury to rat skeletal muscle. J. Appl. Physiol. 54:80–93, 1983.
3. Basmajian, J. V., and C. J. De Luca. Muscles Alive: Their Functions Revealed by Electromyography (5th Ed). Baltimore: Williams & Wilkins, 1985, pp. 42–45.
4. Bonde-Peterson, F., H. G. Knuttgen, and J. Henriksson. Muscle metabolism during exercise with concentric and eccentric contractions. J. Appl. Physiol. 33:792–795, 1972.
5. Burke, R. E. Motor units: anatomy, physiology, and functional organization. In:Handbook of Physiology,
Sect. 1, The Nervous System, Vol. II, Motor Control, Pt. 1. J. M. Brookhart, V. B. Mountcastle, and V. B. Brooks (Eds.). Bethesda, MD: American Physiological Society, 1981, pp. 343–422.
6. Clarkson, P. M., K. Nosaka, and B. Braun. Muscle function after exercise-induced muscle damage and rapid adaptation. Med. Sci. Sports Exerc. 24:512–520, 1992.
7. De Luca, C. J. The use of surface electromyography in biomechanics. J. Appl. Biomech. 13:135–163, 1997.
8. De Luca, C. J., A. M. Roy, and Z. Erim. Synchronization of motor unit firing in several human muscles. J. Neurophysiol. 70:2010–2023, 1993.
9. Felici, F., L. Colace, and P. Sbriccoli. Surface EMG modifications after eccentric exercise. J. Electromyogr. Kinesiol. 7:193–202, 1997.
10. Fridén, J., M. Sjöström, and B. Ekblom. Myofibrillar damage following intense eccentric exercise in man. Int. J. Sports Med. 4:170–176, 1983.
11. Gerdle, B., K. Henriksson-Larsén, R. Lorentzon, and M.-L. Wretling. Dependence of the mean power frequency of the electromyogram on muscle force and fibre type. Acta Physiol. Scand. 142:457–465, 1991.
12. Gerdle, B., M-L. Wretling, and K. Henriksson-Larsén. Do the fibre-type proportion and the angular velocity influence the mean power frequency of the electromyogram? Acta Physiol. Scand. 134:341–346, 1988.
13. Helliwell, T. R., J. Coakley, P. E. M. Smith, and R. H. T. Edwards. The morphology and morphometry of the normal human tibialis anterior muscle. Neuropathol. Appl. Neurobiol. 13:297–307, 1987.
14. Henriksson-Larsén, K., J. Fridén, and M-L. Wretling. Distribution of fibre sizes in human skeletal muscle: an enzyme histochemical study in m tibialis anterior. Acta Physiol. Scand. 123:171–177, 1985.
15. Hermens, H. J., T. A. M. Bruggen, C. T. M. Baten, W. L. C. Rutten, and H. B. K. Boom. The median frequency of the surface EMG power spectrum in relation to motor unit firing and action potential properties. J. Electromyogr. Kinesiol. 2:15–25, 1992.
16. Howell, J. N., A. J. Fuglevand, M. L. Walsh, and B. Bigland-Ritchie. Motor unit activity during isometric and concentric-eccentric contractions of the human first dorsal interosseus muscle. J. Neurophysiol. 74:901–904, 1995.
17. Infante, A. A., D. Klaupiks, and R. E. Davies. Adenosine triphosphate changes in muscles doing negative work. Science 144:1577–1578, 1964.
18. Ingalls, C. P., G. L. Warren, D. A. Lowe, D. B. Boorstein, and R. B. Armstrong. Differential effects of anesthetics on in vivo skeletal muscle contractile function in the mouse. J. Appl. Physiol. 80:332–340, 1996.
19. Ingalls, C. P., G. L. Warren, J. H. Williams, C. W. Ward, and R. B. Armstrong. E-C coupling failure in mouse EDL muscle after in vivo
eccentric contractions. J. Appl. Physiol. 85:58–67, 1998.
20. Kamen, G., and G. E. Caldwell. Physiology and interpretation of the electromyogram. J. Clin. Neurophysiol. 13:366–384, 1996.
21. Knutson, L. M., G. L. Soderberg, B. T. Ballantyne, and W. R. Clarke. A study of various normalization procedures for within day electromyographic data. J. Electromyogr. Kinesiol. 4:47–59, 1994.
22. Kupa, E. J., S. H. Roy, S. C. Kandarian, and C. J. De Luca. Effects of muscle fiber type and size on EMG median frequency and conduction velocity. J. Appl. Physiol. 79:23–32, 1995.
23. Lieber, R. L., and J. Fridén. Selective damage of fast glycolytic muscle fibres with eccentric contraction of the rabbit tibialis anterior. Acta Physiol. Scand. 133:587–588, 1988.
24. Lowe, D. A., G. L. Warren, D. A. Hayes, M. A. Farmer, and R. B. Armstrong. Eccentric contraction-induced injury of mouse soleus muscle: effect of varying [Ca2+
. J. Appl. Physiol. 76:1445–1453, 1994.
25. Lowe, D. A., G. L. Warren, C. P. Ingalls, D. B. Boorstein, and R. B. Armstrong. Muscle function and protein metabolism after initiation of eccentric contraction-induced injury. J. Appl. Physiol.:1260–1270, 1995.
26. Nardone, A., C. Romanò, and M. Schieppati. Selective recruitment of high-threshold human motor units during voluntary isotonic lengthening of active muscles. J. Physiol. (Lond.) 409:451–471, 1989.
27. Nardone, A., and M. Schieppati. Shift of activity from slow to fast muscle during voluntary lengthening contractions of the triceps surae muscles in humans. J. Physiol. (Lond.) 395:363–381, 1988.
28. Sacco, P., and D. A. Jones. The protective effect of damaging eccentric exercise against repeated bouts of exercise in the mouse tibialis anterior muscle. Exp. Physiol. 77:757–760, 1992.
29. Schwane, J. A., and R. B. Armstrong. Effect of training on skeletal muscle injury from downhill running. J. Appl. Physiol. 55:969–975, 1983.
30. Søgaard, K., H. Christensen, B. R. Jensen, L. Finsen, and G. Søgaard. Motor control and kinetics during low level concentric and eccentric contractions in man. Electroenceph. Clin. Neurophysiol. 101:453–460, 1996.
31. Tesch, P. A., G. A. Dudley, M. R. Duvoisin, B. M. Hather, and R. T. Harris. Force and EMG signal patterns during repeated bouts of concentric or eccentric muscle actions. Acta Physiol. Scand. 138:263–271, 1990.
32. Walmsley, B., J. A. Hodgson, and R. E. Burke. Forces produced by medial gastrocnemius and soleus muscles during locomotion in freely moving cats. J. Neurophysiol. 41:1203–1216, 1978.
33. Warren G. L., D. A. Hayes, D. A. Lowe, J. H. Williams, and R. B. Armstrong. Eccentric contraction-induced injury in normal and hindlimb-suspended mouse soleus and EDL muscles. J. Appl. Physiol. 77:1421–1430, 1994.
34. Warren, G. L., C. P. Ingalls, and R. B. Armstrong. A stimulating nerve cuff for chronic in vivo
measurements of torque produced about the ankle. J. Appl. Physiol. 84:2171–2176, 1998.
35. Warren, G. L., C. P. Ingalls, and R. B. Armstrong. Marked effect of temperature on the eccentric contraction-induced force deficit (Abstract). Med. Sci. Sports Exerc. 29:S102, 1998.
36. Warren, G. L., C. P. Ingalls, S. J. Shah, and R. B. Armstrong. Uncoupling of in vivo
torque production from EMG in mouse muscles injured by eccentric contractions. J. Physiol. (Lond.) 515:609–619, 1999.
37. Warren, G. L., D. A. Lowe, D. A. Hayes, C. J. Karwoski, B. M. Prior, and R. B. Armstrong. Excitation failure in eccentric contraction-induced injury of mouse soleus muscle. J. Physiol. (Lond.) 468:487–499, 1993.
38. Webber, S., and D. Kriellaars. Neuromuscular factors contributing to in vivo eccentric moment generation. J. Appl. Physiol. 83:40–45, 1997.
39. Westing, S. H., A. G. Cresswell, and A. Thorstensson. Muscle activation during maximal voluntary eccentric and concentric knee extension. Eur. J. Appl. Physiol. 62:104–108, 1991.
40. Wretling, M. L., B. Gerdle, and K. Henriksson-Larsén. EMG: a non-invasive method for determination of fibre type proportion. Acta Physiol. Scand. 131:627–628, 1987.
MUSCLE; MUSCLE INJURY; MOTOR UNIT RECRUITMENT; MUSCLE FORCE
©2000The American College of Sports Medicine
What does "Remember me" mean?
By checking this box, you'll stay logged in until you logout. You'll get easier access to your articles, collections,
media, and all your other content, even if you close your browser or shut down your
To protect your most sensitive data and activities (like changing your password),
we'll ask you to re-enter your password when you access these services.
What if I'm on a computer that I share with others?
If you're using a public computer or you share this computer with others, we recommend
that you uncheck the "Remember me" box.
Highlight selected keywords in the article text.
Data is temporarily unavailable. Please try again soon.