| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Norwegian University of Science and Technology, Norwegian University of Science and Technology, Trondheim, Norway
Submitted 17 February 2005 ; accepted in final form 1 April 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
postural muscles; habitual muscle activity; long-term recording
Many studies in the occupational research tradition have noted the considerable individual differences in muscle activity patterns when performing apparently similar jobs or tasks (e.g., Refs. 1, 26). The variation in responses is usually perceived as a problem by masking the underlying, "true" muscle response and has triggered studies to determine, e.g., the number of subjects required to establish valid group mean responses (22). The interindividual differences in muscle usage are nevertheless of physiological interest as a general motor control issue and relating to muscle characteristics such as the percentage muscle fiber types and fiber size.
We have previously demonstrated considerable interindividual differences in upper trapezius activity pattern during controlled arm movement (36), contrasting consistent intraindividual responses in repeated recordings of the same arm movement (30). Large interindividual and consistent intraindividual responses in motor tasks point to considerable differences between individuals in their habitual control of the motor system. However, neither the interindividual variation in upper trapezius activity pattern nor the activity pattern of other postural muscles is well documented in long-term recordings. The aim of this study is to characterize the long-term, habitual activity patterns of trapezius and low back muscles of female subjects free of major work task-associated demands. Female subjects were selected because of our interest in shoulder pain development, which is more prominent for this gender. The results were compared with previous published results of muscle activity in upper and lower extremity muscles (15), with the anticipated outcome that the recordings of postural muscles, despite interindividual variation in responses, will show more sustained activity pattern than for the extremity muscles. Physiological implications of the different activity patterns of postural and extremity muscles are discussed.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
In some of the analyses, the material was supplemented by previous recordings of trapezius activity pattern for female university secretaries (n = 25), bank workers (n = 5), shop assistants in a small supermarket (n = 17), and health care workers (n = 19) to explore the full range of variation in sEMG activity patterns. The last two groups had a predominant standing posture at work. The age ranged from 19 to 64 yr (mean 44.0 ± 9.5 yr). Body mass ranged from 50 to 85 kg (mean 65.9 ± 7.9 kg), and body mass index ranged from 17.9 to 31.6 (mean 23.8 ± 2.8). The supplementary material has been analyzed previously to determine group results for median activity level and rest time (10); however, burst analysis (see sEMG analysis) and analyses with thresholds set for quantification of high-amplitude muscle activity were not performed. In the supplementary material, 26 subjects (health care workers, n = 18; shop assistants, n = 8) were recorded on two occasions, separated by 16–28 mo.
All subjects read and signed an informed consent form before inclusion. The study protocol was approved by the Regional Ethics Committee and carried out according to the Declaration of Helsinki.
Physiological recordings. Electrocardiographic (ECG) and bilateral sEMG activity from upper trapezius, lumbar multifidus, longissimus thoracis pars lumborum, and iliocostalis lumborum pars thoracis were recorded (Physiometer PHY-400, Premed, Norway) over 24 h. Inclinometers were placed on the arms and the back to record postural movements at work; however, these results are not reported here. Silver/silver chloride electrodes with diameter of 6 mm (Neuroline, Medicotest, Denmark) were used for ECG and sEMG recordings. EMG activity was sampled at 1,600 Hz and band-pass filtered at 20–800 Hz. EMG signals were thereafter analog-to-digital converted, and the root mean square (RMS) value was calculated and transmitted at 10 Hz to a palmtop personal computer carried by the subject (HP 200LX, Hewlett-Packard). A bipolar configuration with center-to-center distance of 20 mm was used for all sEMG recordings.
An artifact detection procedure, sensitive to sharp transients and slow deviations from baseline, was available and was used for checking sEMG recordings in the previous study (10). However, very few indications of artifacts were found, and these were most easily spotted as elevated, stand-alone events in time plots of the processed sEMG recordings. This simplified quality control procedure was therefore used in the present study. The recording system, utilizing active electrodes, was not sensitive to 50-Hz interference.
Electrodes were placed in standard positions across the chest for the ECG recordings. The QRS complex was detected, and the intervals between the R peaks (R-R intervals) were derived on a beat-by-beat basis. The processed ECG recording was transmitted at 10 Hz and stored synchronously with the EMG signal (Physiometer PHY-400, Premed, Norway). Instantaneous heart rate was determined by inverting the beat-to-beat intervals. A time resolution of 0.2 s was used in the statistical analyses of ECG and EMG signals.
For the upper trapezius, EMG electrodes were placed at a point two-thirds of the distance from the spinous process of the seventh cervical vertebra (C7) toward the lateral edge of the acromion (12). For the iliocostalis, the electrodes were placed 
2 cm medial to the lateral border of the muscle belly at the 12th rib, in line with the assumed muscle fiber direction (2). For the longissimus, the electrodes were placed 3–3.5 cm lateral to the L3 (20). For the multifidus, the electrodes was placed at the L5–S1 level in line with the assumed muscle fiber orientation, i.e., parallel and medial to a line between the posterior superior iliac spine and the L1–L2 interspinous space. At this position, the multifidus is a wide, superficially located muscle (21). The electrode positions for iliocostalis and multifidus were similar to those used in a study by Ng and coworkers (25).
Whole-day recordings (from start of work until bedtime) shorter than 9 h or recordings with work or leisure recording period shorter than 4.5 h were excluded from further analysis. EMG recording failure was due to involuntary removal of the electrodes or technical failure during data acquisition. EMG recordings failed on both right and left side for multifidus, longissimus, and iliocostalis in three, five, and eight subjects, respectively, whereas recordings failed on only the right or left side in six subjects for iliocostalis, two subjects for multifidus, and one subject for longissimus. Trapezius EMG recordings failed on both sides for four subjects, whereas three recordings failed on either the right (dominant) or left side. After removing incomplete recordings, the average recording time ranged from 13.3 to 13.8 h for the different muscles (Table 1), in which the work and leisure periods averaged 6.1 (range 4.9–6.9) and 7.7 h (range 4.6–10.6 h). Total recording time of the supplementary trapezius material ranged from 10.3 to 17.5 h (mean 14.0 h).
|
3 s, separated by 1-min pauses. Verbal encouragement was given by the experimenter for all calibration trials. Intraclass correlation analysis and calculation of mean percentage difference between EMGmax obtained at the start and end of the recording period were used to assess the reliability of the EMGmax responses. Intraclass correlation coefficients ranged from 0.91 to 0.97, whereas mean percentage difference ranged from 4.4 to 7.4% (Table 1), indicating high reproducibility of the EMGmax response and unchanged calibration during the recording period. Body mass index was inversely correlated to EMGmax, with correlation coefficients ranging from 0.36 to 0.77 (0.001 < P < 0.05) for the different muscles.
sEMG analysis. Median and mean sEMG activity (% EMGmax), and time with sEMG activity of >2, >10, >30, and >50% EMGmax were determined. sEMG activity with amplitude of >2% EMGmax was also quantified by burst analysis, as described by Kern and coworkers (15). Outcome variables were number of bursts (bursts/h), mean burst duration (s), and mean burst amplitude (% EMGmax), additional to burst time (% of recording period). Further EMG variables in the complete trapezius material were duration of sEMG activity of <0.5% EMGmax ("rest time"), 90% of the amplitude distribution and the highest sEMG response, using a time resolution of 0.2 s. The sEMG responses were checked by visual inspection of an amplitude-time display of the sEMG recordings to ensure that obvious artifacts (singular, stand-alone events) were excluded. Such events were eliminated in 12 recordings (8%). System noise (0.6–0.9 µV; Ref. 24) was subtracted before quantification of sEMG variables.
To allow use of independent-samples statistics, the sEMG responses are presented as the mean of the left and right side in the tables. If the recording on one side was lost, the successful recording was taken to represent the activity pattern of this muscle for that subject. The left and right sides are separately included in some of the figures to show the range of trapezius sEMG responses.
sEMG amplitude was also quantified during the calibration of inclinometers for posture recording ("uninformed rest"; Ref. 31). The subjects adopted a neutral standing posture with eyesight fixed at a far point at eye height for 45 s but received no instruction to relax their shoulders or otherwise modify their muscle activity level. This procedure was carried out both in the beginning and the end of the work period.
Subjective scoring of physical fatigue. The subjects scored their level of physical fatigue on Borg's scale (3) on an hourly basis throughout the daytime recording. Average and maximum scores during work and leisure were used to assess the level of subjective physical fatigue.
Statistical analyses. A Shapiro-Wilk W-test for normality was performed on all dependent variables before statistical analysis. Nineteen of 32 EMG variables (8 EMG variables in 4 muscles) were found to be nonnormally distributed. Nonparametric statistical methods were therefore used in the analyses. All comparisons were performed two tailed, and the significance level was set to P < 0.05. Data are reported as means ± SD unless otherwise stated.
One-way ANOVA on ranks (Kruskal-Wallis) with a Kruskal-Wallis z-score post hoc test was used to test the hypothesis that EMG activity did not differ between muscles during daytime activity. A two-way random intraclass correlation analysis was used to assess agreement between EMGmax obtained at the start and end of the recording period and to assess the agreement between repeated recordings of work and uninformed rest in upper trapezius. Linear regression coefficients and Spearman's rho was used for other correlation analyses.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Daytime sEMG activity of postural muscles. Table 2 shows daytime sEMG activity for trapezius and low back muscles. Group mean values for the six EMG variables, SD, and range of responses are presented. Similar activity levels were found for all muscles, except somewhat less activity was indicated for iliocostalis. This trend was significant for mean burst duration (compared with all other muscles) and mean activity (vs. multifidus). Burst time was indicated significant by the multivariate analysis, but post hoc analyses did not show significant differences between groups. However, unitary comparisons indicate that burst time for iliocostalis was less than for the other muscles (P < 0.05 for all unitary comparisons). Another noted feature in Table 2 is the large interindividual variation in sEMG responses, whatever sEMG variable is used. There was no association between sEMG variables and calibration responses (i.e., EMGmax).
|
12 s for a daytime period of 17 h for trapezius. The corresponding duration for multifidus was 11 s/h or 3 min for the full daytime period.
|
The distributions of median sEMG activity level, burst time, and rest time for the complete material are shown by histograms in Fig. 1, A–C, with left and right trapezius recordings separately included. Scatterplots of responses for the dominant vs. nondominant side are shown in Fig. 1, D–F. Regression line and line of identity are included in the scatterplots. Trapezius muscle activity was mostly symmetrical as measured by the three sEMG variables, except for a markedly higher median sEMG activity for the dominant trapezius in a few subjects. Interindividual variation in sEMG responses was large, regardless of method used for quantification. Finally, scatterplots showing interrelationships between the three sEMG variables are presented in Fig. 1, G–I. In case of median EMG level vs. burst time and rest time, the relationships are clearly curvilinear: median EMG level vary considerably for similar, high burst time. Burst time and rest time vary considerably for the same low median EMG level.
|
2 yr. The first recording included only the work period (37), and trapezius sEMG responses for the two work periods were compared (Fig. 3, A–C). The scatterplots show highly consistent responses, despite the separation in time and likely differences in work tasks performed: there was no attempt to control movements or work tasks during the recordings. The uninformed rest responses were also very consistent, similar to repeated recordings of uninformed rest the same day (Fig. 3D).
|
|
50% EMGmax (Fig. 4A). Group mean value for the 90th percentile of the amplitude distribution curve was 10.4% EMGmax (range 1.7–28.8% EMGmax; Fig. 4B). Corresponding mean values for the low back muscles were 11% (multifidus), 9.5% (longissimus), and 5.7% EMGmax (iliocostalis).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Figure 5 presents results from this study and presumed comparable results of recordings from upper (first dorsal interossius, biceps brachii) and lower extremity (medial and lateral vastus) muscles of moderately active female students (15). Burst time was substantially different between muscles: mean values were 40–50% of the recording time for trapezius, longissimus, and multifidus, 20% for upper extremity muscles, and 15% for lower extremity muscles. Mean burst amplitude was 4–5% EMGmax for postural, 7–8% EMGmax for upper extremity, and 20% EMGmax for lower extremity muscles. Motor units with relatively high recruitment threshold (e.g., >10% EMGmax) may thus be active for longer periods of time in case of the lower extremity muscles than for the postural low back muscles and trapezius. The mean burst amplitude of trapezius is markedly lower than for all the other muscles and quite invariant, indicating that a select population of motor units is active during bursts for this muscle. The activity patterns of the four postural muscles are clearly distinguished from upper and lower extremity muscles, thereby verifying the introductory assumption.
|
Rest time was quantified only for trapezius due to low calibration responses for the low back muscles. The detection level of 0.5% EMGmax corresponds to the sEMG contribution of two to three motor units underneath the sEMG electrode (33). Burst amplitude of 2–5% EMGmax is estimated to represent the sEMG contribution of 40–70 motor units (33), with numbers increasing if contributing motor units are located away from the sEMG electrode. Thus rest time represents periods of rest and recovery for the population of upper trapezius motor units detected by the sEMG electrode, whereas burst time represents duration of sustained activity of at least some of the low-threshold motor units. Periods of motor unit silence occur, but such periods seem to be of short duration if sEMG amplitude is substantially higher than the motor unit recruitment threshold (34).
Burst time ranged from <10 to >80% of the daytime recording period and was similar for trapezius, multifidus, and longissimus. This compares to type I motor units in rat soleus muscle, which were active for 22–35% of the recording period (8). Many motor units in postural muscles are likely to have a duty time substantially longer than those observed in the rat, even after adjusting for reduced activity during sleep (24). Thus conditions that may cause type I motor unit exhaustion and thereby development of "ragged red fibers" exist (7, 17).
Putative type IIB motor units in rat leg muscles were active for 0.04–0.22% of time over 24 h (8). Time periods with sEMG amplitude >50% EMGmax were of similar duration for the low back muscles but were shorter or nonexistent in case of trapezius. The question may thus be raised whether human muscle fibers tolerate day-long periods of inactivity. Studies of membrane properties in controlled stimulation of inactive and denervated rat and baboon muscle fibers provide a partial answer. Extrajunctional acetylcholine sensitivity in nerve-blocked baboon lumbrical muscle fibers was reduced after 1 wk of stimulation with 500 pulses (5 pulses/s stimuli for a total duration of 100 s) but not with 50 pulses (5 pulses/s for 10 s) once every 24 h (6). Denervated rat muscle fibers showed reduced acetylcholine sensitivity after 2 wk of stimulation with 100 pulses (10 Hz for 10 s) every 5.5 h but not with 100 pulses at 12-h intervals (19). Thus a few seconds of activity per 24 h seem sufficient for maintenance of muscle membrane structure both for the rat and the baboon. Few firings of high-threshold motor units in postural muscles of females may explain the smaller cross-sectional area of type II relative to type I muscle fibers (18, 29).
The few or missing periods of moderately high sEMG amplitude may seem at odds with motor unit recruitment thresholds up to 80% maximal voluntary contraction for proximal extremity muscles (5, 16) and for sEMG amplitudes exceeding 50% EMGmax in case of trapezius motor units (Ref. 35 and unpublished observations). However, a considerable increase in firing rates of trapezius motor units is observed even in relatively slow ramp contractions (35), whereas firing rates tend to stay low in sustained contractions (32). More motor units are therefore required to generate the same sEMG amplitude, resulting in an apparent reduction of recruitment threshold. This firing behavior is in contrast to first dorsal interossius motor units, which have firing patterns that closely mimic the sEMG activity profile both in sustained contractions and in contractions with rapid amplitude modulation (35).
Monster et al. (23) found that muscles with the highest proportion of type I fibers had the longest daily duration of muscle activity in most comparisons, but this finding was neither corroborated by Kern et al. (15) nor is evident from Fig. 5 and known fiber type proportions of these muscles. The discrepant results may be due to differences in calibration procedures since the duration of measurements by Monster and coworkers were based on muscle activity higher than 8% of the 90th percentile of the amplitude distribution. The effective threshold for detecting muscle activity by the Monster et al. criterion would vary between 0.14 and 2.3% EMGmax for trapezius. There will also be systematic variation between muscles.
The year-long consistency of the daytime sEMG responses was a surprise. We speculate that the long recording period average out short-term, task-based variations in activity patterns. Also, the consistency of responses is probably helped by this being work recordings of subjects in predominant standing postures with little or no demands of high force exertions or working in constrained postures. The long-term sEMG recordings may thus be a measure of the habitual muscle usage of subjects, their motor habit: a subject-specific muscle activity pattern that is also a characteristic when the standardized resting posture is adopted.
The profound differences in muscle activity pattern are likely to influence the ability of muscle fibers to sustain contractions, also within the same fiber-type population. The muscle metabolic apparatus is very adaptable, increasing the number and volume of muscle fiber mitochondria and capillary density with exercise and reducing metabolic capacity when the muscle returns to the nontrained state (27). The marked differences in muscle activity patterns seem stable over periods of several years; thus motor unit adaptation to exercise may represent an individual characteristic right through to adolescence. If so, group-based quantification of muscle activity patterns may not be well suited to show muscle fiber overexertion and risk of muscle pain when tasks are performed that require moderate physical activity.
In conclusion, the activity pattern of postural low back muscles and trapezius show considerable interindividual variation; however, they remain distinctly different from the activity patterns of upper and lower extremity muscles reported by others (15). We speculate that the observed interindividual variation in long-term recordings of muscle activity pattern represents habitual differences in postural motor programming: their motor habit.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
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.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. Grunheid, G.E.J. Langenbach, A. Zentner, and T.M.G.J. Van Eijden Duty Time of Rabbit Jaw Muscles Varies with the Number of Activity Bursts Journal of Dental Research, December 1, 2006; 85(12): 1112 - 1117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Mork and R. H. Westgaard Low-amplitude trapezius activity in work and leisure and the relation to shoulder and neck pain J Appl Physiol, April 1, 2006; 100(4): 1142 - 1149. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
