EMG Changes in Human Thenar Motor Units With Force Potentiation and Fatigue

Transcription

1 J Neurophysiol 95: , First published November 2, 2005; doi: /jn EMG Changes in Human Thenar Motor Units With Force Potentiation and Fatigue C. K. Thomas, 1,2 R. S. Johansson, 3 and B. Bigland-Ritchie 1 1 The Miami Project to Cure Paralysis, Department of Neurological Surgery, and 2 Department of Physiology and Biophysics, University of Miami Miller School of Medicine, Miami, Florida; and 3 Physiology Section, IMB, Umeå University, Umeå, Sweden Submitted 2 September 2005; accepted in final form 29 October 2005 Thomas, C. K., R. S. Johansson, and B. Bigland-Ritchie. EMG changes in human thenar motor units with force potentiation and fatigue. J Neurophysiol 95: , First published November 2, 2005; doi: /jn Few studies have analyzed activity-induced changes in EMG activity in individual human motor units. We studied the changes in human thenar motor unit EMG that accompany the potentiation of twitch force and fatigue of tetanic force. Single motor unit EMG and force were recorded in healthy subjects in response to selective stimulation of their motor axons within the median nerve just above the elbow. Twitches were recorded before and after a series of pulse trains delivered at frequencies that varied between 5 and 100 Hz. This stimulation induced significant increases in EMG amplitude, duration, and area. However, in relative terms, all of these EMG changes were substantially smaller than the potentiation of twitch force. Another 2 min of stimulation (13 pulses at 40 Hz each second) induced additional potentiation of EMG amplitude, duration, and area, but the tetanic force from every unit declined. Thus activity-induced changes in human thenar motor unit EMG do not indicate the alterations in force or vice versa. These data suggest that different processes underlie the changes in EMG and force that occur during human thenar motor unit activity. INTRODUCTION For a muscle to develop tension, its surface membranes have to be depolarized, a process that is reflected in the EMG activity. Studies in various mammals have shown that changes in EMG generally do not parallel changes in force. For example, the twitch force from most cat motor units increases significantly in response to brief periods of high-frequency stimulation, with little or no change in EMG potentials (Burke et al. 1973; Dum and Kennedy 1980; Kernell et al. 1975). Likewise, during muscle fatigue, the EMG may not change, may increase, or may decline for units that show large or intermediate force declines. In units that are more fatigue resistant, either no change or potentiation of EMG may accompany the decline in force (Burke et al. 1973; Celichowski et al. 1991; Enoka et al. 1992; Gardiner and Olha 1987; Hamm et al. 1989; Sandercock et al. 1985). Furthermore, changes in EMG potentials, when present, usually recover before the force is restored (Sandercock et al. 1985). In summary, these data suggest different processes account for the changes in motor unit EMG and force. Largely because of technical difficulties, few studies have examined activity-dependent changes in the relationships between EMG and force in human muscles at the level of single motor units. In particular, it is difficult to follow the activity of the same Address for reprint requests and other correspondence: C. K. Thomas, The Miami Project to Cure Paralysis, Univ. of Miami Miller School of Medicine, 1095 NW 14 Terrace, R48, Miami, FL ( cthomas@miami.edu). motor unit during prolonged contractions that involve strong forces. Thus most previous studies on long-term changes in EMG and force have been restricted to weak voluntary contractions and to changes in either the force or the EMG of low threshold units rather than on the relationships between these parameters (e.g., Carpentier et al. 2001; Christensen and Sjogaard 1999; Nordstrom and Miles 1990; Stephens and Usherwood 1977). However, simultaneous recording of single motor unit force and EMG during weak and strong contractions have been made in spinal cord injured individuals in whom only a few motor units in a muscle remained under voluntary control. In repeated voluntary contractions that elicited fatigue in these motor units, the unit EMG was either well maintained or declined (Thomas and del Valle 2001). Intraneural stimulation of individual motor axons in healthy humans (Westling et al. 1990) has been applied in one previous study to examine the force-emg relationship. Fuglevand et al. (1999) report that the EMG potentials of motor units in intrinsic and extrinsic hand muscles were unchanged despite depression of force. In contrast, using trains of pulses at 40 Hz and stimulation of the median nerve through the skin, Chan et al. (1998) found that the amplitude and area of the evoked potentials from thenar motor units more than doubled for fatigable units. Fatigue resistant units showed little change in EMG. Using intraneural stimulation, we have previously characterized single human thenar motor units with respect to their potentiation of twitch force and fatigue of tetanic force (Thomas et al. 1990, 1991a,b; Westling et al. 1990). In this study, we analyze the changes in unit EMG that accompany these changes in force. METHODS Single motor unit EMG and force were recorded from the thenar muscles of 12 healthy subjects in response to intraneural stimulation of their motor axons in the median nerve at the level of the upper arm (Westling et al. 1990). All subjects gave their written informed consent to participate in these experiments, conducted in accordance with the Declaration of Helsinki. Experimental setup As described previously (Westling et al. 1990), each subject reclined in a dental chair with the right forearm supinated, extended, and resting in a vacuum cast supported by a platform (Fig. 1). Clay was molded to the shape of the hand to stabilize it, and U-shaped metal clamps were pressed into the clay to restrain the fingers. Three electrodes made of braided The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact /06 $8.00 Copyright 2006 The American Physiological Society

2 EMG CHANGES IN HUMAN MOTOR UNITS 1519 Baseline fluctuations from the pulse pressure and respiration were minimized in two ways (Westling et al. 1990). First, using signals from the pulse detector, all single stimuli and trains of stimuli were delivered to the thenar motor axons during periods when the force baseline remained relatively flat (typically ms after peak pulse pressure). Second, the force baseline was reset electronically to a defined level just before the delivery of single stimuli or just before the first stimulus in a train of pulses. FIG. 1. Experimental setup. The supinated arm was supported in a vacuum cast and clay. A tungsten electrode within the median nerve above the elbow was used to stimulate each single thenar motor axon in relation to the pulse pressure wave, detected with a sensor on the middle finger. EMG was recorded from distal and proximal muscle surfaces. Force was recorded in the abduction and flexion directions. strands of copper wire picked up EMG from the proximal and distal thenar muscle surfaces. A common electrode, 3 4 cm long, was placed transversely over the muscle belly. The other two electrodes were close to the distal and proximal tendons. One, 2 cm long, lay across the volar aspect of the metacarpophalangeal joint, and the other over 3cmofthe base of the thenar eminence. Although unconventional, these electrode placements permitted recording of the small EMG responses evoked by stimulation of just one thenar motor axon in the median nerve in the otherwise relaxed and electrically silent muscle. A ground electrode spanned the arm just proximal to the wrist. A custom-made transducer was aligned along the length of the thumb so that 0.5 N of resting tension was applied in both the abduction and flexion directions. Force was registered in both of these directions at right angles. An optical pulse detector attached around the middle finger recorded the pulse pressure wave. Protocol An insulated tungsten microelectrode (0.2 mm diam with a fine uninsulated tip; impedance: k ) was inserted into the median nerve 10 cm proximal to the elbow. It was moved in small steps until weak negative current pulses (0.2-ms duration) delivered through the electrode evoked EMG and force in a thenar motor unit in an all-or-none manner. Thus below threshold, neither EMG nor force was produced. With small increases in current over a particular range, thenar EMG and force signals of consistent magnitude and shape were evoked simultaneously, showing the excitation of a single motor unit. Further increments in current resulted in larger and more complex EMG and force signals because additional motor units were activated. After verifying that only one thenar motor axon was stimulated (for details, see Westling et al. 1990), the following stimuli were delivered to the axon: 1) 5 10 single pulses to elicit initial twitches (Thomas et al. 1990); 2) trains of stimuli at variable interpulse intervals to determine the series of pulses that maximized unit force (Thomas et al. 1999) and/or trains of pulses at 5, 8, and 10 Hz (each for 2 s) and 15, 20, 30, 50, and 100 Hz (each for 1 s) in ascending then descending order of frequency, or vice versa, to determine the force evoked at different stimulation frequencies (Thomas et al. 1991a); 3) 5 10 single pulses to elicit twitches after posttetanic potentiation (Thomas et al. 1990); 4) 2 min of stimulation at 40 Hz (13 pulses each second) to induce fatigue (Burke et al. 1973; Thomas et al. 1991b); 5) postfatigue responses including single pulses and trains of pulses at Hz as before fatigue. Data collection and analysis EMG recorded from the distal and proximal surfaces of the thenar muscles, and the abduction and flexion forces were each sampled on-line at 3,200 and 400 Hz, respectively, using SC/Zoom (Physiology Section, IMB, Umeå University, Sweden). All data analyses were done off-line. The vector sum of the measured abduction and flexion force components represented the magnitude of the evoked forces. Up to 10 twitches recorded before and after the series of pulse trains at different tetanic frequencies were averaged and are referred to as initial and potentiated twitches, respectively, because these tetanic stimuli increased the twitch force significantly (Thomas et al. 1990). We characterized EMG potentials obtained from both the distal and the proximal channels by amplitude, duration, and area. The sum of the durations and areas measured for the first and second phase of an EMG potential (as defined by isoelectric crossings) represented the duration and area of the potential, respectively. EMG amplitude was calculated as the peak-topeak voltage (Fig. 2A). The corresponding EMG parameters and the peak twitch force of each unit obtained after stimulation at different frequencies were normalized to the respective initial values to show the changes in twitch EMG and force induced by this stimulation. During the fatigue protocol, the EMG and force responses from five trains of pulses were averaged every 20 s (i.e., trains at 0 4 s, s, etc.) to examine the changes in these signals over time. The amplitude, duration, and area of the first and last EMG potentials in these averaged responses were analyzed for both the distal and proximal EMG signals to evaluate changes across (long-term) and within trains (short-term) because decrements in EMG may result in force declines. Peak force was also measured. Time-dependent changes in the EMG amplitude, duration, and area for the first potential in the train were calculated by normalizing the values at 20, 40, 60, 80, 100, and 120 s to the respective values at 0 s. Changes in the parameters of the last EMG potential in these same trains were assessed by normalizing the values to the respective parameters for the first EMG potential at 0 s. Thus at 120 s, these data represent the changes in EMG parameters after the delivery of 1,560 pulses. Fatigue (force decline) was calculated by dividing the peak force measured every 20 s by the force measured at 0 s. Statistics Separate repeated measures ANOVAs with EMG channel (distal vs. proximal) and test point (initial vs. potentiated) as factors were used to compare EMG amplitude, duration, and area before and after tetanic stimulation at frequencies between 5 and 100 Hz. Likewise, repeated-measures ANOVAs with EMG channel (distal vs. proximal) and train number (7 levels corresponding to 7 times) were used to assess EMG changes during the fatigue protocol. One-way repeatedmeasures ANOVAs were used to examine effects of the tetanic stimulation and the fatigue protocol on twitch forces. Tukey HSD tests were used in posthoc analyses that involved pairwise comparisons of data. Relationships between parameters were analyzed using Pearson or product-moment correlations. Use of other statistical analyses is indicated in the Results section. Mean SD values are given, and the threshold for statistical significance was P 0.05.

3 1520 C. K. THOMAS, R. S. JOHANSSON, AND B. BIGLAND-RITCHIE RESULTS Presented here are EMG data from single human thenar motor units previously characterized with respect to potentiation of twitch force and fatigue of tetanic force (Thomas et al. 1990, 1991a,b; Westling et al. 1990). We also address the changes in unit EMG that accompany the force changes, focusing on the 23 units that were subjected to brief trains of stimulation at frequencies between 5 and 100 Hz (Thomas et al. 1991a) and to fatigue (Thomas et al. 1991b). Figure 2A shows the EMG and twitch force recorded from a single thenar motor unit when single pulses were first delivered to its motor axon. These pulses were followed by a series of trains of pulses delivered at frequencies from 100 to 5 Hz (Fig. 2B) and by more single pulses (Fig. 2C). For this unit, the TABLE 1. EMG and force data for 23 human thenar motor units Amplitude, V FIG. 2. EMG and force evoked by stimulation at different frequencies. Surface EMG (top traces) and resultant force (bottom traces) recorded from 1 single thenar motor unit in response to single pulses delivered at the beginning of the experiment (A, initial twitch) and after delivery of a series of trains of pulses designed to potentiate the twitch force (C, potentiated twitch). Initial twitch data are also overlaid (dotted traces) to show changes in EMG and force induced by stimuli at different frequencies. Trains of pulses were delivered to this unit at decreasing frequencies (B, Hz) and increasing frequencies (5 100 Hz, data not shown). Measurements of peak-to-peak amplitude, duration, and area of the 1st 2 phases of the potential are shown for the distal EMG of the initial twitch (A). twitch force increased by 94% after these trains of stimuli compared with the force of the initial twitches, whereas there were only small changes in the corresponding EMG potentials. Overall, the potentials recorded from the distal muscle surface were similar in shape to those from the proximal muscle surface, but typically larger in amplitude. Furthermore, the distal and proximal EMG was of opposite polarity for 22 of the 23 units studied (96%). This suggested that the end plates of most motor units lay near the center of the muscles. Changes in unit EMG with twitch force potentiation Brief periods of stimulation at frequencies between 5 and 100 Hz caused significant increases in EMG amplitude (P 0.05), Distal EMG Proximal EMG Force Duration, ms Area, Vs Amplitude, V Twitch potentiation Duration, ms Initial Potentiated Potentiated, % initial Fatigue First potential 40 Hz Force, mn at 0 s First potential at 120 s Last potential at 0 s Last potential at 120 s First potential First potential at 120 s, Force, % initial % initial Last potential at 120 s, % initial Area, Vs Twitch, mn Values are means SD. Numbers under Fatigue are 40 Hz force.

4 EMG CHANGES IN HUMAN MOTOR UNITS 1521 FIG. 3. EMG and force potentiation after stimulation at different frequencies. Distributions of distal EMG amplitude ( ), duration (Œ), area ( ), and twitch force ( ), expressed as a percentage of the respective initial values (n 23 units). Dotted line at 100 represents no change in a parameter as a result of stimulation at frequencies between 5 and 100 Hz. duration (P 0.001), and area (P 0.001; Table 1). Increases in each EMG parameter were seen in all but four units (83% of units; Fig. 3). The EMG potentials recorded from the distal muscle surface were similar in duration (P 0.31) to those recorded from the proximal muscle surface, but significantly larger in amplitude (P 0.01) and area (P 0.05). There was no interaction between EMG channel (distal vs. proximal) and test point (initial vs. potentiated) for any of the EMG variables. Initial and potentiated peak-to-peak EMG amplitudes for distal and proximal records were positively correlated, as were the respective EMG areas and durations (all r 0.96). The stimulation at frequencies between 5 and 100 Hz also increased twitch force (P 0.001). Twitch force potentiation occurred in all but two units (91% of units), but its magnitude varied widely between units. Potentiated twitch forces ranged from 93 to 290% of those recorded initially (Table 1; Thomas et al. 1990). There was a significant positive correlation between the initial and potentiated twitch forces (r 0.82, P 0.001). In relative terms, the potentiation of twitch force was greater than the increases in any of the EMG parameters (P 0.05 in all instances, paired t-test; Fig. 3). However, across units, no significant relationships were evident between the relative changes in any of the EMG parameters and twitch force (Fig. 4). The r values for the EMG amplitude, duration, and area for the 23 units analyzed in this study were 0.23, 0.37, and 0.15, respectively. The lack of correlations between the changes in EMG parameters and force were further emphasized by a weakening of the corresponding correlations (r 0.04, 0.26, and 0.14) when we included data from another seven units stimulated at frequencies between 5 and 100 Hz (Thomas et al. 1991a) but not subjected to the fatigue protocol. Changes in unit EMG during fatigue of tetanic force Fatigue was induced in each unit by stimulating its motor axon for 2 min with trains of 13 pulses at 40 Hz delivered once per second. Figure 5 exemplifies the changes in the EMG and force during this protocol for three motor units. For the more fatigable units (Fig. 5, A and B), the first EMG potentials of the trains increased in amplitude, duration, and area during the fatigue protocol. For the last train, the EMG amplitudes for these two units were 132 (Fig. 5A) and 110% (Fig. 5B) of that recorded for the first train. The corresponding values for the duration of the potentials were 114 and 118% initial. Area data were 144 and 111% initial. The evoked tetanic forces declined to 41 (Fig. 5A) and 49% of initial (Fig. 5B). In contrast, the force for the more fatigue resistant motor unit had only fallen by 8% after 2 min of stimulation (Fig. 5C). However, the amplitude, duration, and area of the first EMG potentials in the trains increased to 128, 103, and 136% initial, respectively. Figure 5 also shows changes in the last EMG potentials in the trains. For one of the more fatigable units, the amplitude of the last potential in the train at 120 s declined during the fatigue protocol, but its duration increased resulting in an increase in its area (Fig. 5A). The duration and area of the EMG also increased for the other fatigable unit, whereas the amplitude was maintained (Fig. 5B). The value of all three EMG parameters increased for the more fatigue-resistant unit (Fig. 5C). Considering the first EMG potentials in the trains, significant changes occurred in the average unit data for all three EMG parameters during the 2 min of intermittent stimulation at 40 Hz (all P 0.001). The values of all parameters gradually increased. The increase in EMG amplitude was significant at 20 s, duration at 60 s, and area at 40 s (Fig. 6, A C). The distal EMG amplitudes were larger than the proximal EMG ampli- FIG. 4. Correlations between changes in EMG and twitch force after stimulation at different frequencies. Changes in distal EMG amplitude (A), duration (B), and area (C) in response to brief trains of stimulation at frequencies between 5 and 100 Hz, plotted against the changes in twitch force (n 23 units; r 0.23, 0.37, and 0.15, respectively). Data are expressed as a percentage of the respective initial values. Dotted lines at 100 represent no change from initial values.

5 1522 C. K. THOMAS, R. S. JOHANSSON, AND B. BIGLAND-RITCHIE tudes (P 0.02), but the durations (P 0.51) and areas were similar (P 0.44) for the two EMG channels. There was no interaction between EMG channel (distal and proximal) and train number for any of the EMG variables. The general potentiation of the first EMG potential in a train was accompanied by a force decline (P 0.001) that was manifest in every unit (Fig. 6D; Table 1). The force decline was significant at 60 s (Thomas et al. 1991b). However, as shown in Fig. 7 A C, there were no significant correlations between the relative changes in force and any of the EMG parameters (amplitude: P 0.15, duration: P 0.06, area: P 0.22 for distal EMG). As with the first EMG potentials in the trains, the last potentials in the trains also increased in amplitude with fatigue (P 0.05) but more significantly in duration (P 0.001) and area (P 0.001). Again, the distal EMG amplitudes were larger than the proximal EMG amplitudes (P 0.02), whereas the durations (P 0.62) and areas (P 0.36) were similar. No interaction was observed between EMG channel (distal and proximal) and the number of the train for any of the EMG variables. The increases in EMG duration and area were both significant at 40 s, whereas the increase in EMG amplitudes only reached significance at s when considering the distal and proximal EMGs separately. Figure 8, A C, shows the EMG changes that occurred during the fatigue protocol by comparing EMG parameters obtained for the first and last potential of the last train (at 120 s) with the first potential of the first train (at 0 s). For the first EMG potential of the last train, the amplitude, duration, and area remained higher than their initial values for 100, 83, and 96% of units, respectively. The corresponding percentages for the last potential of the trains were 61, 48, and 65%. In separate ANOVAs, we compared the first and last potentials in the first (0 s) and last (120 s) 13-pulse trains. Overall, the first potential had a longer duration than the last one at 0s (P 0.002), and the duration increased from the first to the last train (P 0.001). Furthermore, an interaction between the position of the potential in the train and train number (P 0.001) indicated that the longer duration of the first potential compared with the last in a train only occurred for the first train, whereas there was little effect on the duration in the last train (Fig. 8E). For the amplitude, there was no effect of the potential s position in the train (P 0.16), but overall, the amplitude increased from the first to the last train (P 0.05). However, an interaction between these factors (P 0.05) suggested that this increase in amplitude largely was accounted for by an increase in the amplitude of the first potential in the last train (Fig. 8D). Finally, the area of the potential increased with both the potential s position in the train (P 0.001) and from the first to the last stimulus train (P 0.001; Fig. 8F). There was no interaction between these two factors. Also in these ANOVAS, we included the EMG channel as a factor and, again, it only had an effect on the amplitude (P 0.02), which was larger for the distal pair of electrodes. Units whose EMG potentials had slowed the most by the end of the fatigue test had greater reductions in EMG amplitude (Fig. 9A; P 0.001), larger increases in EMG area (Fig. 9B; P 0.014), and were the more fatigable units (Fig. 9C; P 0.01). There were no significant correlations across units between the relative change in force and relative change in EMG amplitude or area. This together with the observation that the EMG area for most units exceeded the initial values at the end of the fatigue protocol (Fig. 8C) suggested that the force loss related to processes beyond the electrical excitation of the muscle fibers. DISCUSSION FIG. 5. Typical changes in EMG and force with fatigue. Trains of distal surface EMG potentials (top traces) recorded from 2 fatigable motor units (A and B) and a more fatigue resistant motor unit (C). Overlays of the 1st EMG potentials at 0 (solid traces) and 120 s (dotted traces) and the last EMG potentials at 120 s (dashed traces) show changes in EMG over time and within the trains (middle traces). Force (bottom traces) from each unit is overlaid to show how it declines and slows with time. These data show that thenar motor unit EMG potentials and twitch force change to different extents and largely independently after the delivery of brief trains of pulses at different frequencies between 5 and 100 Hz. Furthermore, when the tetanic forces were reduced by brief trains of stimuli at 40 Hz

6 EMG CHANGES IN HUMAN MOTOR UNITS 1523 maximal evoked contractions (e.g., Duchateau and Hainaut 1985). In cat hind limb or foot muscles, however, twitch force potentiation also occurred in almost all motor units examined, but no changes were reported for the EMG waveforms (Burke et al. 1973; Dum and Kennedy 1980; Kernell et al. 1975). Potentiation of EMG amplitude with muscle activity can result from increases in Na K pump activity and from greater synchronization of muscle fiber potentials (Hicks and McComas 1989; McComas et al. 1994). Our results suggest that increased synchronization may not fully explain the potentiation of EMG amplitude in experiments on whole human thenar muscles since the duration of most of our unitary EMG potentials increased rather than decreased in response to brief stimulation at different frequencies (Figs. 3 and 4). Increases in EMG potential duration probably reflect the slowing of muscle fiber conduction velocity that occurs when extracellular K increases with exercise (Juel 1988). In comparison, potentiation of twitch force after repetitive stimulation is proposed to relate to phosphorylation of myosin regulatory light chains (Grange et al. 1998; Sweeney et al. 1993). That the EMG and force data changed to quite different extents further suggests Downloaded from FIG. 6. Activity-dependent changes in EMG (1st potential) and force. Relative changes (% initial) in peak-to-peak amplitude (A), duration (B), and area (C) of the distal surface EMG and force (D) for all motor units during fatigue. Each solid line shows data from 1 motor unit (n 23). Dashed lines at 100 indicate no change in a parameter compared with the value at 0 s. Circles show mean for each parameter. *Significant change from the mean value at 0 s. by on November 30, 2017 for another 2 min, the EMG potentials potentiated even further. These divergent changes in EMG and force (Figs. 4 and 7) suggest that different processes control increases in EMG, potentiation of twitch force and fatigue of tetanic force. These issues have not been explored previously in humans at the level of single motor units. Changes in unit EMG with twitch force potentiation Stimulation at various frequencies between 5 and 100 Hz induced significant increases in EMG amplitude, duration, and area, changes that were accompanied by even greater increases in twitch force (Fig. 3). Significant increases in evoked EMG have also been seen in whole human muscles after a brief maximal voluntary contraction (e.g., Hicks et al. 1989) or FIG. 7. Changes in EMG and force with fatigue. Change in the distal EMG amplitude (A), duration (B), and area (C) for the 1st EMG potential in the trains at 120 s, normalized to corresponding values for 1st EMG potential at 0s(n 23 units) vs. relative changes in 40-Hz force.

7 1524 C. K. THOMAS, R. S. JOHANSSON, AND B. BIGLAND-RITCHIE that different mechanisms underlie the potentiation of thenar unit EMG and twitch force in response to stimulation at different frequencies (Fig. 3). Likewise, there were no reliable correlations between the changes in unit EMG parameters and twitch force (Fig. 4). Changes in unit EMG with fatigue Two minutes of intermittent stimulation at 40 Hz resulted in significant potentiation of all EMG parameters, whereas the force of the motor units decreased. That the changes in EMG parameters were largely unrelated to the changes in force show that the changes in EMG signals do not predict motor unit fatigue and vice versa. These results also show that excitation of the sarcolemma and transmission across the neuromuscular junctions was effective and changes in these processes cannot explain the force loss. Similar conclusions have been reached in studies of single cat and rat motor units (Burke et al. 1973; Celichowski et al. 1991; Enoka et al. 1992; Hamm et al. 1989; Sandercock et al. 1985), where it has been suggested that force loss relates to changes in phosphate metabolism, intercellular Ca 2 handling and altered cross-bridge kinetics (Edman 1995; Westerblad and Allen 1991; Westerblad et al. 1998). As for motor units in cat or rat hind limb muscles, the duration and area of the EMG potentials increased with fatigue in human thenar units, consistent with data of Chan et al. (1998). However, for thenar units, there was usually an increase in the amplitude of the first EMG potential in each train of stimuli (Fig. 6A), whereas more variable amplitude effects FIG. 8. Distributions of relative changes in EMG and force with fatigue. Distal EMG amplitude (A), duration (B), and area (C) for the 1st EMG potential (Œ) and the last EMG potential in the train at 120 s ( ), normalized to the corresponding values for 1st EMG potential at 0s(n 23 units). Relative changes in tetanic (40 Hz) force are also shown ( ). Dotted lines at 100 represent no change in a parameter. Mean distal EMG amplitude (E), duration (F), and area (G) ofthe 1st and last EMG potentials from the 1st and last trains of pulses. Vertical bars denote 95% confidence intervals. have been observed in cat units exposed to the same fatigue protocol (Enoka et al. 1992; Sandercock et al. 1985). These results may reflect differences in relative contraction intensity. Trains of pulses at 40 Hz evoke near maximal force in human thenar units (Thomas et al. 1991a) but cause unfused contractions in most cat motor units (Botterman et al. 1986; Kernell et al. 1983). Thenar units are also relatively fatigue resistant compared with the units in many cat hind limb muscles (Thomas et al. 1991b), so the largest changes in EMG parameters do not necessarily occur in fatigable units (force 75% of the initial force after 2 min of stimulation), as is typical for motor units in cat and rat muscles (Celichowski et al. 1991; Enoka et al. 1992; Gardiner and Olha 1987; Hamm et al. 1989; Sandercock et al. 1985). Although Enoka et al. (1992) suggested that measurements of either EMG amplitude and duration or area are adequate to assess EMG changes, further comparison of results across studies is difficult because of the general lack of consensus on how to analyze EMG signals. Sometimes an entire train of potentials is averaged, which masks possible EMG changes occurring within the train (Figs. 5 and 8). The increases in unit EMG duration and area during fatiguing contractions may reflect slowing of muscle conduction velocity or fatigue-induced variability in conduction velocity across the different fibers of a motor unit (Gydikov et al. 1979; Stalberg 1966). Both these factors, in turn, will influence the amplitude of the unitary EMG potential. Furthermore, during high-frequency stimulation, a temporal overlap between succeeding EMG

8 EMG CHANGES IN HUMAN MOTOR UNITS 1525 FIG. 9. Correlations between within train EMG (last potential) and force. Changes in the distal EMG duration for the last potential in the train at 120 s relative to the 1st potential at 0 s (i.e., change in EMG potential duration after delivery of 1,560 pulses at 40 Hz over 2 min) vs. corresponding changes in EMG amplitude (A), EMG area (B), and 40-Hz force (C). potentials may result in marked signal attenuation (Day and Hulliger 2001; Fuglevand 1995; Keenan et al. 2005). Indeed, during stimulation at 50 and 100 Hz, we observed a decrease in the recorded EMG amplitude for thenar units (Fig. 2), although the force generally increased. Thus rather than reflecting a reduced efficacy of electrical transmission, the decrements in unit EMG amplitude observed within a train of potentials during the fatigue protocol (Figs. 5A and 9) and in fatigable motor units in cat or rat muscles (Enoka et al. 1992; Gardiner and Olha 1987; Hamm et al. 1989) may be an artifact caused by signal attenuation. General implications These data show that thenar unit EMG signals usually became larger both when twitch force potentiated and when additional stimulation induced declines in tetanic force. These activity-dependent dissociations between EMG and force must underlie some of the discrepancies in EMG force relationships seen at the whole muscle level. For example, early increases in M-wave amplitude accompany force declines that result from repeated brief maximal voluntary contractions of thenar muscles (Hicks et al. 1989), whereas M-waves are maintained at initial levels as force declines during sustained maximal voluntary contractions (Bigland-Ritchie et al. 1982). After small daily amounts of chronic stimulation, the force of cat muscles declines in response to repeated stimulation but the EMG is maintained well (Kernell et al. 1987). These motor unit and whole muscle data emphasize that changes in EMG signals are poor indicators of motor unit and muscle force or vice versa. Thus inappropriate adjustments in force may occur when electrical stimulation is used to restore behaviors in paralyzed muscles if changes in EMG are used to control the stimulation parameters. Furthermore, when interpreting surface EMG data obtained during voluntary muscle contractions, it is important to consider that the magnitude of motor unit potentials may change in addition to alterations in motor unit recruitment and firing rate. ACKNOWLEDGMENTS The authors thank C. Kostov for help with some data analysis and B. Mas for help with the figures. GRANTS This research was funded by National Institutes of Health Grants NS , NS-14756, and HL-30026, the Swedish Research Council (08667), and the 6th Framework Program of the EU (project IST ). REFERENCES Bigland-Ritchie B, Kukulka CG, Lippold OC, and Woods JJ. The absence of neuromuscular transmission failure in sustained maximal voluntary contractions. J Physiol 330: , Botterman BR, Iwamoto GA, and Gonyea WJ. Gradation of isometric tension by different activation rates in motor units of cat flexor carpi radialis muscle. J Neurophysiol 56: , Burke RE, Levine DN, Tsairis P, and Zajac FE III. Physiological types and histochemical profiles in motor units of the cat gastrocnemius. J Physiol 234: , Carpentier A, Duchateau J, and Hainaut K. Motor unit behaviour and contractile changes during fatigue in the human first dorsal interosseus. J Physiol 534: , Celichowski J, Grottel K, and Rakowska A. Changes in motor unit action potentials during the fatigue test. Acta Neurobiol Exp (Warsz) 51: , Chan KM, Andres LP, Polykovskaya Y, and Brown WF. Dissociation of the electrical and contractile properties in single human motor units during fatigue. Muscle Nerve 21: , Christensen H and Sjogaard G. Muscular Disorders in Computer Users: Mechanisms and Models. Copenhagen: National Institute of Occupational Health, 1999, p Day SJ and Hulliger M. Experimental simulation of cat electromyogram: evidence for algebraic summation of motor-unit action potential trains. J Neurophysiol 86: , Duchateau J and Hainaut K. Electrical and mechanical failures during sustained and intermittent contractions in humans. J Appl Physiol 58: , Dum RP and Kennedy TT. Physiological and histochemical characteristics of motor units in cat tibialis anterior and extensor digitorum longus muscles. J Neurophysiol 43: , Edman KA. Myofibrillar fatigue versus failure of activation. Adv Exp Med Biol 384: 29 43, Enoka RM, Trayanova N, Laouris Y, Bevan L, Reinking RM, and Stuart DG. Fatigue-related changes in motor unit action potentials of adult cats. Muscle Nerve 15: , Fuglevand AJ. The role of the sarcolemma action potential in fatigue. Adv Exp Med Biol 384: , Fuglevand AJ, Macefield VG, and Bigland-Ritchie B. Force-frequency and fatigue properties of motor units in muscles that control digits of the human hand. J Neurophysiol 81: , Gardiner PF and Olha AE. Contractile and electromyographic characteristics of rat plantaris motor unit types during fatigue in situ. J Physiol 385: 13 34, 1987.

9 1526 C. K. THOMAS, R. S. JOHANSSON, AND B. BIGLAND-RITCHIE Grange RW, Vandenboom R, Xeni J, and Houston ME. Potentiation of in vitro concentric work in mouse fast muscle. J Appl Physiol 84: , Gydikov A, Kosarov D, and Dimitrov GV. Length of the summated depolarized area and duration of the depolarizing and repolarizing processes in the motor unit under different conditions. Electromyogr Clin Neurophysiol 19: , Hamm TM, Reinking RM, and Stuart DG. Electromyographic responses of mammalian motor units to a fatigue test. Electromyogr Clin Neurophysiol 29: , Hicks A, Fenton J, Garner S, and McComas AJ. M wave potentiation during and after muscle activity. J Appl Physiol 66: , Hicks A and McComas AJ. Increased sodium pump activity following repetitive stimulation of rat soleus muscles. J Physiol 414: , Juel C. Muscle action potential propagation velocity changes during activity. Muscle Nerve 11: , Keenan KG, Farina D, Maluf KS, Merletti R, and Enoka RM. Influence of amplitude cancellation on the simulated surface electromyogram. J Appl Physiol 98: , Kernell D, Donselaar Y, and Eerbeek O. Effects of physiological amounts of high- and low-rate chronic stimulation on fast-twitch muscle of the cat hindlimb. II. Endurance-related properties. J Neurophysiol 58: , Kernell D, Ducati A, and Sjoholm H. Properties of motor units in the first deep lumbrical muscle of the cat s foot. Brain Res 98: 37 55, Kernell D, Eerbeek O, and Verhey BA. Relation between isometric force and stimulus rate in cat s hindlimb motor units of different twitch contraction time. Exp Brain Res 50: , McComas AJ, Galea V, and Einhorn RW. Pseudofacilitation: a misleading term. Muscle Nerve 17: , Nordstrom MA and Miles TS. Fatigue of single motor units in human masseter. J Appl Physiol 68: 26 34, Sandercock TG, Faulkner JA, Albers JW, and Abbrecht PH. Single motor unit and fiber action potentials during fatigue. J Appl Physiol 58: , Stalberg E. Propagation velocity in human muscle fibres in situ. Acta Physiol Scand 287: 1 112, Stephens JA and Usherwood TP. The mechanical properties of human motor units with special reference to their fatiguability and recruitment threshold. Brain Res 125: 91 97, Sweeney HL, Bowman BF, and Stull JT. Myosin light chain phosphorylation in vertebrate striated muscle: regulation and function. Am J Physiol 264: C1085 C1095, Thomas CK, Bigland-Ritchie B, and Johansson RS. Force-frequency relationships of human thenar motor units. J Neurophysiol 65: , 1991a. Thomas CK and del Valle A. The role of motor unit rate modulation versus recruitment in repeated submaximal voluntary contractions performed by control and spinal cord injured subjects. J Electromyogr Kinesiol 11: , Thomas CK, Johansson RS, and Bigland-Ritchie B. Attempts to physiologically classify human thenar motor units. J Neurophysiol 65: , 1991b. Thomas CK, Johansson RS, and Bigland-Ritchie B. Pattern of pulses that maximize force output from single human thenar motor units. J Neurophysiol 82: , Thomas CK, Johansson RS, Westling G, and Bigland-Ritchie B. Twitch properties of human thenar motor units measured in response to intraneural motor-axon stimulation. J Neurophysiol 64: , Westerblad H and Allen DG. Changes of myoplasmic calcium concentration during fatigue in single mouse muscle fibers. J Gen Physiol 98: , Westerblad H, Allen DG, Bruton JD, Andrade FH, and Lannergren J. Mechanisms underlying the reduction of isometric force in skeletal muscle fatigue. Acta Physiol Scand 162: , Westling G, Johansson RS, Thomas CK, and Bigland-Ritchie B. Measurement of contractile and electrical properties of single human thenar motor units in response to intraneural motor-axon stimulation. J Neurophysiol 64: , Downloaded from by on November 30, 2017