INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

TECHNOLOGICAL ADVANCES HAVE MADE long-term recordings of muscle activity a viable strategy for quantifying the functional use of muscles during activities of daily living. Although chronic electromyographic (EMG) recordings have been obtained in cats and rats during daily cage activity (1, 2, 5, 13, 21), long-term EMG recordings from human muscles are scarce. In one of the first studies to use this technique in humans, muscle activity of 12 men was recorded continuously over 8-h periods in pairs of muscles that were functionally linked but comprised different proportions of muscle fiber types (23). The main finding from this study was that the duration of total muscle activity was highly correlated with the proportion of type I fibers in the muscle. A similar association has been observed among hindlimb muscles of experimental animals (1, 5, 16, 19, 32). These findings suggest that the fiber-type composition of a muscle can be reasonably predicted based on the total time a muscle is active during normal daily use.

The amount of daily activity in the muscle, however, does not appear to be solely determined by the proportion of type I fibers. Monster et al. (23), for example, found for 24% (8/33) of comparisons between pairs of muscles that the muscle with the greater proportion of type I fibers was active for a lesser percentage of the recording time. Although Monster et al. studied only men, one factor that might influence this association is the sex of the individual. This possibility was suggested by sex differences that were observed in long-term EMG recordings during baseline measurements in a study on the effects of limb immobilization (30). In this study, several features of EMG activity recorded in the elbow flexor muscles was greater in a group of subjects that was comprised mainly of women (6/7 subjects) compared with a group of five men. Because the statistical power of these comparisons was marginal and the study involved a single muscle group, we decided to expand our comparison of muscle activity.

The purpose of the study was to quantify and compare long-term muscle activity in upper- and lower-limb muscles of men and women. We recorded the concurrent activity of four muscles for 10 h and performed a detailed analysis of the EMG to assess differences between muscles and effects of an individual's sex on the level of muscle activity. Results of most previous studies (16, 19, 23, 32) predicted that we should find a significant relation between the amount of EMG activity and the fiber-type proportions of selected muscles but not between the sex of the individual. In contrast, we found no relation between the amount of EMG activity and the reported fiber-type proportions of the muscles. However, there were modest sex differences in EMG levels. A preliminary report of these data has been presented (18).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Muscles Examined

Experimental Arrangement

EMG. Long-term EMG activity was recorded from selected muscles on a subject's left (nondominant) side. Pairs of disposable surface electrodes (5-mm diameter Meditrace pellet electrodes, Graphic Controls, Buffalo, NY) were attached over the bellies of the four muscles. Electrodes over the first dorsal interosseus muscle were positioned ~1.5 cm apart, whereas those over the biceps brachii, vastus medialis, and vastus lateralis muscles were 2.0 cm apart. To stabilize electrode location during the recording period, subjects wore a modified breathable mesh glove (E-Force, San Diego, CA) covering the electrodes of first dorsal interosseus and a flexible wrapping (Co-Flex, Andover Coated Products, Salisbury, MA) over the electrodes of biceps brachii, vastus medialis, and vastus lateralis muscles. The electrodes were connected to a portable EMG device (ME-3000 Professional, Mega Electronics, Kuopio, Finland) that was protected in a hip sack and worn around a subject's waist. At no time was movement restricted by the EMG recording apparatus. EMG signals were originally sampled at 1,000 Hz, amplified (412×), band-pass filtered (20-500 Hz), then averaged every 0.1 s, and stored for subsequent analysis.

Maximum voluntary contractions. Subjects performed maximum voluntary contractions (MVCs) with the first dorsal interosseus muscle while seated in a chair with the forearm resting on a table. This task involved exerting an abduction force with the index finger. The hand was positioned in an apparatus containing a plastic restraint that conformed to the radial side of the index finger and a metal support aligned to the palmar side of the thumb. For MVCs performed with elbow flexor and knee extensor muscles, the subject was seated and secured with waist straps to the back of a chair. The elbow joint was flexed to 1.57 rad (90°) for elbow flexor MVCs with hand supinated, forearm strapped to an armrest, and elbow resting on a padded armrest. Knee extensor MVCs were performed on the same apparatus with the knee flexed to 1.57 rad and an adjustable strap that connected the ankle to the base of the chair.

Experimental Procedures

Three MVCs were performed in the morning and three in the evening to measure the maximum EMG for each muscle and check the integrity of the recording electrodes. Data were discarded if there was a significant difference between the maximum EMG recorded in the morning and evening sessions. MVC comprised a ramp-and-hold protocol that involved gradually increasing force from baseline to maximum over 2-3 s and then holding the maximum force for an additional 2-3 s. Participants were verbally encouraged to achieve maximum force during each MVC trial. Subjects were given a 30-s rest between each MVC. For first dorsal interosseus MVCs, subjects were instructed to keep the palmar surface of their hands flat on the support surface and to abduct the index finger against the plastic restraint while minimizing activity in other hand and arm muscles. Similarly, subjects were encouraged to relax the hand, forearm, and shoulder muscles while pulling up with the wrist to perform an MVC with the elbow flexor muscles. For knee extensors MVCs, the task was to pull against the ankle restraint.

Data Analysis

View larger version (10K): [in this window] [in a new window]   Fig. 1.   An example of an electromyographic (EMG) recording and details for quantification of an individual burst of EMG. A: 1-h segment of EMG activity recorded from the first dorsal interosseus (FDI) muscle of a man during normal, everyday use. B: expanded view of EMG recording at the location of the arrow shown in A. EMG activity was based on the number of bursts identified, which was defined as a signal with an amplitude >2% of maximum EMG [obtained during maximum voluntary contractions (MVCs)] and a duration of >0.1s. For this subject, FDI was active for 18% of total recording time.

Statistical analysis. A paired t-test was performed to compare MVCs obtained in morning and evening sessions. An unpaired t-test was used to compare weekly energy expenditure reported from the Paffenbarger Physical Activity Questionnaire between men and women. A three-factor, repeated-measures ANOVA (session × muscle × sex) for each EMG-dependent variable was used to test differences between sessions. Absence of a session effect enabled us to collapse data across sessions and perform a two-factor, repeated-measures ANOVA (muscle × sex) to examine the effect of sex on EMG data. Fisher's post hoc test was performed as required. Significance was set at a P value <0.05. Data are reported as means ± SD unless otherwise stated.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

EMG activity was recorded from two upper-limb (first dorsal interosseus and biceps brachii) and two lower-limb (vastus medialis and vastus lateralis) muscles for a 10-h period on two separate occasions. All subjects, with one exception, were university students. Subjects were predominantly involved in educational activities for the duration of each recording session, which included attending lectures and performing computer work. All subjects performed some mild physical activity during the recording sessions, such as walking and climbing stairs. None of the subjects performed activities that damaged the equipment. Self reports from subjects indicated that the recording equipment did not limit their activities.

Average Energy Expenditure

MVC EMG

A paired t-test for each muscle determined that there was no significant difference (P values from 0.4 to 1.0) between peak EMG recorded during MVCs performed in the morning and evening sessions. This suggests that the integrity of the electrodes remained intact for the duration of the recording period. Peak EMG for MVCs performed in the two sessions was used to normalize EMG data obtained from each muscle in that recording session. Because there was no significant difference between peak EMG for each muscle between sessions 1 and 2 (two-factor ANOVA), these EMG data were averaged between sessions. Peak EMG averaged across sessions for each subject was largest for the biceps brachii muscle, intermediate for the first dorsal interosseus muscle, and least for the vastus medialis and vastus lateralis muscles (Table 1).

Sessions

EMG Activity in Upper and Lower Limbs

View larger version (36K): [in this window] [in a new window]   Fig. 2.   EMG activity recorded continuously for ~10 h in 1 woman. All data are represented as a percent of maximum EMG obtained during MVCs performed for each muscle. For this subject, FDI and biceps brachii were active for 22 and 15% of total recording time, respectively. The vastus medialis and vastus lateralis (VL) were both active for 9% of recording time.

Burst analysis data obtained during long-term EMG recordings in upper- and lower-limb muscles of all subjects are shown in Table 1. Activity in upper-limb muscles was significantly greater (P < 0.05) than in lower-limb muscles for number of bursts, burst rate, and total burst time. Furthermore, summed duration of activity for upper-limb muscles was significantly greater than the vastus medialis muscle, and summed duration for the biceps brachii muscle was greater than for the vastus lateralis muscle. Differences in summed duration of EMG activity between the first dorsal interosseus muscle and the vastus lateralis muscle just failed to reach statistical significance (P = 0.05). On the basis of the four significantly different outcome variables from burst analysis, upper-limb muscles were activated 67% (range = 41-90%) more often than lower-limb muscles. No significant differences in EMG activity were observed between the two upper-limb muscles (first dorsal interosseus compared with biceps brachii) and the two lower-limb muscles (vastus medialis compared with vastus lateralis).

Although lower-limb muscles were activated less often, EMG amplitude was greater in these muscles when activated (Table 1). Mean burst amplitude was 170% (range = 118-225%) greater in lower-limb muscles compared with upper-limb muscles. No significant difference in mean burst amplitude was observed within the two upper- and two lower-limb muscles.

A frequency distribution plot of mean burst amplitudes in the first dorsal interosseus and vastus lateralis muscles of all subjects is shown in Fig. 3. For the first dorsal interosseus muscle, 79% (7,558/9,527) of bursts had an amplitude <10% MVC EMG, whereas only 55% (3,338/6,051) of bursts were <10% MVC EMG in the vastus lateralis muscle (Fig. 3A). In contrast, the vastus lateralis muscle had more bursts above 90% MVC EMG than the first dorsal interosseus muscle (Fig. 3B). The small number of bursts above 100% MVC EMG in the vastus lateralis muscle is a reflection of the protocol used to measure MVC EMG. For experimental purposes, the MVC EMG was obtained during an isometric contraction that was performed in a controlled manner, whereas EMG during normal, everyday activities included anisometric contractions obtained at varying muscle lengths and contraction velocities. A comparison of burst amplitude distribution between the first dorsal interosseus and vastus lateralis muscles (Fig. 3) suggests that relative distribution is concentrated at higher burst amplitudes in lower-limb muscles and lower burst amplitudes in upper-limb muscles. The behavior of the two upper-limb muscles was quantitatively similar, as was the behavior of the two lower-limb muscles.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Distribution of burst amplitude in FDI and VL muscles. Data represent average burst amplitude from all subjects during 10-h recordings. Burst amplitude on the horizontal axis was plotted with a resolution of 1% MVC, whereas the vertical axis represents number of bursts on a linear scale (A) and on a log scale (B).

EMG Activity Between Muscles

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Correlated EMG activity in upper- and lower-limb muscles. Data show the relation between EMG activity of FDI and biceps brachii muscles (A) and for the vastus medialis and vastus lateralis muscles (B) for the data shown in Fig. 2. Each data point represents EMG activity from the two muscles recorded at the same time. There was a strong association between the EMG activity of lower-limb muscles (B; r2 = 0.83) but not of upper-limb muscles (A; r2 = 0.037).

EMG Activity in Men and Women

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The purpose of the study was to quantify muscle activity in upper- and lower-limb muscles of men and women during normal activities of daily living. We found that upper-limb muscles were more active during 10-h recording compared with lower-limb muscles. Although lower-limb muscles were activated less often, mean burst amplitude was greater for these muscles. A high temporal association existed between activity of the two thigh muscles, which was not observed in the two upper-limb muscles. Furthermore, we found only modest differences in 10-h EMG activity between men and women.

Muscle Activity and Fiber-Type Composition

A major application of chronic EMG recordings has been to quantify muscle activity during chronic interventions, such as hindlimb suspension, spinal cord transection, and limb immobilization. These studies have shown that hindlimb suspension significantly decreases the quantity of EMG activity (5, 28), but this occurs in some muscles and not in others (2). Although substantial changes in neuromuscular properties occur after hindlimb suspension, these changes do not seem to be related to the extent of muscle activity during the intervention (34). One factor that could be responsible for this discrepancy is the differences in fiber-type proportion of the muscles under examination. For example, transformation of slow-twitch fibers to fast-twitch fibers after spinal cord transection may be related to the total level of spontaneous daily activity, where reduction in EMG activity is greater in slow-twitch muscles (1). Unfortunately, we do not know how these data apply to humans because of the lack of long-term recordings of muscle activity in humans and the differences in fiber-type proportions and muscle morphology between animals and humans.

It has been shown in both animal (19) and human (23) studies that the percentage of type I fibers is typically higher for muscles with greater levels of daily activity compared with muscles that are more quiescent. More importantly, these studies have suggested that the fiber-type proportion of a muscle can be reasonably predicted based on the duration of daily muscle activity. We examined this issue by matching recorded EMG activity from four muscles during normal activities of daily living with estimates of fiber-type compositions as reported in the literature. As indicated in Fig. 5, there was no association between the amount of EMG activity in a 10-h period and the approximate proportion of type I muscle fibers.

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Relation between EMG activity (total duration, Tables 2 and 3) and proportion of type I fibers as reported in the literature. Mean proportion of type I fibers for men is 57% for FDI (17), 41% (range = 39-43%) for biceps brachii (3, 22, 24, 29), 44% for vastus medialis (17), and 44% (range = 36-56%) for VL (9, 11, 20, 22, 31, 33). Mean proportion of type I fibers for women is 50% for FDI (8), 44% (range = 39-50%) for biceps brachii (3, 22, 24, 29), and 46% (range = 34-66%) for VL (9, 11, 20, 22, 31, 33).

EMG Activity in Upper and Lower Limbs

The most striking difference in muscle activity between upper and lower limbs was the greater mean burst amplitude in the vastus lateralis and vastus medialis muscles (Fig. 3). Although activated less often, we found that mean burst amplitudes of lower-limb muscles were two to three times greater than those of upper-limb muscles. These data were not reported by Monster et al. (23) because they were unable to compare EMG amplitude between muscles when they were recorded at different times. Between-muscle and between-subject comparisons of EMG amplitude was possible in the present study because we used a procedure of normalizing EMG activity to the value obtained during a controlled MVC performed at the beginning and end of each recording period. As such, this is the first study to report greater EMG amplitude in lower-limb muscles during normal, everyday use. This finding is to be expected, as lower-limb muscles are used to support body weight during postural control and are often subjected to high-impact forces during locomotion.

On the basis of a qualitative examination of long-term EMG patterns, it has been suggested that there is a strong temporal similarity between muscles of the lower limb but not between muscles of the upper limb. For example, Monster et al. (23) noted a strong functional linkage in concurrent recordings of muscle activity in four leg muscles, with the strongest correlation being observed between the biceps femoris and tibialis anterior muscles. In contrast, no strong temporal association was observed between four muscles of the wrist and arm (wrist flexor and extensor, biceps and triceps brachii). We used a correlational technique to quantify functional coupling between muscles and found a high temporal similarity between the vastus lateralis and vastus medialis muscles of the leg (Fig. 4B) but not for the first dorsal interosseus and biceps brachii muscles of the upper limb (Fig. 4A).

Sex Differences in Muscle Activity

The largest differences between men and women were detected in the biceps brachii muscle (Table 2) where the summed duration and total burst time were ~70% greater in women. No significant differences were observed for any other dependent variable from EMG between men and women. Based on this finding and our previous observation (30), we now have convincing evidence that sex differences in muscle activity exist for the biceps brachii muscle, at least in the nondominant arm of university students. Although no significant sex differences were observed for the other muscles, many EMG-dependent variables just failed to reach statistical significance. A larger sample size may provide additional support for sex differences in muscle activity, particularly in lower-limb muscles.

There are conflicting reports in the literature regarding sex differences in fiber-type composition. The lack of consensus on this issue is presumably due to the inherent variability in obtaining a biopsy sample from different sites within the same muscle. Furthermore, the sex differences in fiber-type composition appear to exist in some muscles but not in others. For example, there is substantial evidence suggesting that fiber-type proportions are similar in the biceps brachii muscle for men and women (3, 22, 24, 29). Interestingly, it was in the biceps brachii muscle where we found the largest sex differences in long-term EMG activity. This dissociation suggests that long-term EMG activity in different muscles is related more to other physiological characteristics of each muscle, such as muscle strength, muscle mass, or endurance capacity, rather than to fiber-type proportions. However, it is possible that the amount of long-term EMG activity for men and women could be different on the dominant side, especially in upper-limb muscles.

In summary, we found greater average amplitudes but less frequent bursts of EMG in lower-limb muscles compared with upper-limb muscles of nondominant limbs of young men and women. The duration of long-term muscle activity was not related to the reported fiber-type composition of the muscle. Furthermore, modest sex differences were observed in EMG activity of the biceps brachii muscle. It appears that factors other than fiber-type distribution influence the extent of EMG activity in a muscle during normal activities of daily living performed by young men and women.