Since the start of the industrial revolution in the middle of the 19th century, there have been huge social upheaval and massive technological advances, majorly impacting our way of life. This encompasses a more sedentary working life with extensive computer use [
Systematic reviews of prospective cohort studies show that gender (woman) and a prior history of neck pain are the strongest predictors for development of neck pain in computer workers [
Tension or activity of the neck/shoulder muscles may play an important role in the development of neck/shoulder pain and can be measured with electromyography during work. The type of activity patterns in the neck/shoulders muscles associated with computer work causes a selective activation of low-threshold motor units with type I muscle fibers. This causes both reduced local blood flow and an accumulation of calcium (Ca2+), which can lead to musculoskeletal pain [
Previous research has shown that physical exercise reduces musculoskeletal pain [
This study investigates the effect of brief daily resistance training on the acute and longitudinal changes in occupational electromyographic activity of the neck muscles (m. splenius and m. trapezius) in female office workers with neck/shoulder pain. We hypothesized that performing two minutes of daily neck/shoulder resistance training for 10 weeks will beneficially alter the muscular activity pattern and thereby reduce neck/shoulder pain. In detail, we hypothesized that the training group will experience (i) an enhanced frequency of EMG gaps, (ii) a prolonged duration of the EMG gaps, and (iii) have a larger percentage of time with a minimal muscular activity compared with the control group.
This study is nested in a larger randomized controlled trial performed in Copenhagen, Denmark. In the larger parallel-group single-blind randomized controlled trial, the participants were allocated to training groups of two or twelve minutes of daily resistance training or to a control group. For the present analyses, we were particularly interested in the mechanisms of pain reduction in the group performing a single set to failure and included a subsample of 2 × 15 participants. In the larger study, 198 office workers with frequent neck/shoulder pain, but without traumatic injuries or serious chronic disease participated. However, due to the time-consuming procedure of performing full-day EMG measurements, it was not possible in the present study to perform daily EMG measurements on all 198 participants. The detailed procedure of recruitment and concealed randomization of the 198 participants is described elsewhere [
Flow chart.
The outcomes in this nested study of the trial were change in (i) frequency of EMG gaps under 0.5% EMGmax (number per minute), (ii) duration per EMG gap under 0.5% EMGmax (length in seconds), and (iii) time spent under 0.5% EMGmax (percentage distribution). On an exploratory basis, the time spent under 1.0%, 1.5%, and 2.0% EMGmax was also investigated. These outcomes were assessed both acutely after a training session and longitudinally following the 10-week intervention. There were no changes made to either methods or study protocol after trial registration.
All participants were informed about the purpose and content of the study and gave their written informed consent prior to participating in the study, which conformed to The Declaration of Helsinki and was approved by the local ethical committee of Copenhagen and Frederiksberg (HC2008103).
The intervention has been described in detail elsewhere [
The adherence in both groups was monitored by weekly internet-based questionnaires. Adherence for the training group was defined as the number of completed training session expressed as a percentage of the total number of training sessions throughout the intervention period. The adherence for the control group was defined as the number of read informational emails expressed as a percentage of the total number of informational emails throughout the intervention period.
The EMG signal was recorded from m. trapezius and m. splenius of the dominant side. The recordings were collected using a bipolar surface EMG configuration (Ambu Blue Sensor N, N-00-S, Ambu A/S, Ballerup, Denmark) using an interelectrode distance of two cm [
Each pair of EMG electrodes was connected to a wireless probe (Velamed Medizintechnik GmbH) connected to the skin, serving as reference electrode. Furthermore, the probe preamplified the EMG signal (gain 400) before transmitting the data to 16-channel 16 bit PC-interface receiver in real-time (Noraxon Telemyo DTS Telemetry, Noraxon, AZ, USA). All data were collected using a sample rate of 1500 Hz within a bandwidth of 10–500 Hz. This wireless EMG-system has shown to be valid and reliable for collecting EMG-data from the neck/shoulder musculature [
All EMG recordings were performed during normal working hours while the participants performed their usual work. To obtain resting EMG at the beginning of the workday, participants performed 30 seconds of instructed seated rest with closed eyes and complete arm support while focusing on completely relaxing the shoulder and neck muscles. This was followed by the three reference tasks performed in accordance with outlined guidelines [
After another period of between 60 to 90 minutes just before terminating the measurement, the participants again conducted the reference task. This was followed by a resisted maximal voluntary contraction to obtain maximal EMG for normalization of the obtained EMG signals. The maximal contraction was conducted in the position of the reference tasks with the only addition of an opposing force provided by the test instructor. The participants then performed isometric maximal voluntary contractions two times for five seconds separated by rest periods of 30 seconds. For an overview of the sampling protocol see Figure
A schematic overview of the measurement period. Rest is equivalent to the resting period where the resting EMG amplitude was determined, REF is equivalent to the three reference tasks, and Max corresponds to the time of the maximal contraction.
In the baseline screening questionnaire, the participants reported that they spend the vast majority of their working hours doing computer work, see Table
Baseline characteristics, Mean ± SD. No significant differences were observed.
Training |
Control |
|
---|---|---|
Age (years) |
|
|
Height (cm) |
|
|
Weight (kg) |
|
|
BMI (kg |
|
|
Pain intensity previous 3 weeks (Scale 0–10) |
|
|
Systolic BP (mmHg) |
|
|
Diastolic BP (mmHg) |
|
|
Isometric muscle strength (Nm) |
|
|
Computer use (% work time) |
|
|
Weekly working time (hours) |
|
|
Duration of office work (Years) |
|
|
Recording time corresponding to the average number of total minutes recorded. Effective recording time corresponding to the percentage time where the participants were present in the predefined data collection area, that is, percentage of the total minutes recorded included in the data analysis, median (interquartile range). No significant differences were observed.
Training |
Control |
||
---|---|---|---|
Recording time (min) | |||
Before daily training session | Week 0 | 60.0 (42.0–69.1) | 64.5 (49.8–79.2) |
Week 10 | 72.1 (65.3–77.4) | 75.5 (56.8–87.8) | |
After daily training session | Week 0 | 82.7 (70.1–96.2) | 60.9 (47.9–74.4) |
Week 10 | 62.3 (39.6–82.6) | 59.7 (46.7–73.2) | |
Effective recording time (%) | |||
Before daily training session | Week 0 | 79.7 (70.8–92.9) | 84.2 (71.9–95.1) |
Week 10 | 89.3 (84.3–92.5) | 93.6 (40.9–97.3) | |
After daily training session | Week 0 | 92.2 (73.4–97.0) | 93.3 (87.9–95.1) |
Week 10 | 93.7 (85.6– 98.5) | 94.3 (73.5–97.1) |
All data processing was performed in MatLab (MathWorks, version 7.5.0 342, R2007b). The first step in the data processing was to filter out the periods of work time were the participants were outside the predefined data collection area. In the measurements, this was visualized as a completely flat line without fluctuations of EMG amplitude, and the program therefore removed periods which assumed identical values over a period of minimum 100 ms.
There were no statistical differences regarding the total recording time and the computer work time between the two groups, see Table
Finally, the RMS for the working periods before (first 60–90 minutes of data sampling) and after the daily training session (last 60–90 minutes) was determined, using the same procedure as described above. This allowed the identification of periods where the EMG amplitude was below a predefined percentage of the normalized EMGmax, which was termed an EMGgap. In this study, the following percentages of the normalized EMGmax had a particular interest: 0.5%, 1.0%, 1.5%, and 2% EMGmax. According to previous studies, 0.5% EMGmax represents the boundary for total relaxation of a motor unit, whereas the remaining values represent different degrees of activation of the smallest motor units [
All statistical analyses were performed in SAS statistical software (SAS version 9.2, SAS Institute, Cary, NC) and were performed in accordance with the intention-to-treat principle by including data from all available participants regardless of actual adherence [
Table
During the intervention period, the training group performed an average of 4.3 of the 5 scheduled training sessions per week, which is equivalent to an 86.8% training adherence, while the control group had read on average 8.9 of the 10 informational emails corresponding to an adherence of 89%.
Overall, two participants were lost to follow-up, one participant in each intervention group, both due to lack of time. No adverse events were reported during the intervention or EMG measurements.
Table
Table
Frequency of EMG gaps (periods per minute below 0.5% EMGmax) for m. trapezius and m. splenius, median (interquartile range).
Training |
Control |
||
---|---|---|---|
Trapezius | |||
Before daily training session | Week 0 | 8.5 (4.4–11.6) | 4.2 (1.9–11.5) |
Week 10 | 8.2 (6.3–15.3) | 3.7 (1.1–13.2) | |
After daily training session | Week 0 | 7.4 (5.4–11.7) | 3.0 (2.0–12.2) |
Week 10 | 7.6 (4.4–15.2) | 2.2 (1.2–6.4) | |
Splenius | |||
Before daily training session | Week 0 | 3.1 (1.4–10.9) | 5.0 (1.6–10.9) |
Week 10 | 12.3 (4.8–15.2)b | 1.1 (0.5–5.8) | |
After daily training session | Week 0 | 5.0 (2.7–7.8) | 3.1 (1.3–11.7) |
Week 10 | 8.0 (3.5–14.5)d | 1.3 (0.5–6.5) |
Table
Table
Pain intensity and muscular strength at week 0 and week 10, Mean ± SD.
Training |
Control |
||
---|---|---|---|
Pain intensity (scale 0–10) | Week 0 |
|
|
Week 10 |
|
|
|
Isometric muscle strength (Nm) | Week 0 |
|
|
Week 10 |
|
|
Tables
(a) Percentage time spent under given % EMGmax for m. trapezius, median (interquartile range). (b) Percentage time spent under given % EMGmax for m. splenius, median (interquartile range).
Training |
Control |
|||
---|---|---|---|---|
Trapezius | 0.5% EMGmax | |||
Before daily training session | Week 0 | 18.5% (5.1–39.1) | 11.4% (4.3–17.6) | |
Week 10 | 24.0% (16.0–34.6) | 4.1% (0.9–27.8) | ||
After daily training session | Week 0 | 15.1% (11.1–30.2) | 9.8% (3.1–12.7) | |
Week 10 | 25.8% (13.4–42.2) | 4.9% (1.5–7.9) | ||
1.0% EMGmax | ||||
Before daily training session | Week 0 | 39.0% (14.1–50.8) | 21.3% (7.6–34.1) | |
Week 10 | 32.3% (24.3–54.8) | 9.9% (2.9–35.5) | ||
After daily training session | Week 0 | 25.9% (21.2–41.2) | 12.6% (6.3–26.0) | |
Week 10 | 37.0% (20.7–54.3) | 6.6% (3.9–13.9) | ||
1.5% EMGmax | ||||
Before daily training session | Week 0 | 47.3% (24.3–57.3) | 28.3% (10.2–43.9) | |
Week 10 | 38.6% (32.3–66.1) | 19.5% (5.1–40.4) | ||
After daily training session | Week 0 | 36.1% (29.1–50.6) | 15.2% (10.5–33.3) | |
Week 10 | 46.2% (26.5–62.8) | 11.0% (5.7–18.8) | ||
2.0% EMGmax | ||||
Before daily training session | Week 0 | 55.0% (37.6–61.9) | 34.7% (15.6–52.3) | |
Week 10 | 44.5% (39.2–72.9) | 27.9% (7.3–47.7) | ||
After daily training session | Week 0 | 44.9% (35.8–58.3) | 19.5% (14.2–39.4) | |
Week 10 | 53.6% (32.9–74.7) | 14.4% (7.8–28.5) |
Training |
Control |
|||
---|---|---|---|---|
Splenius | 0.5% EMGmax | |||
Before daily training session | Week 0 | 2.3% (1.0–20.7) | 7.1% (3.8–11.1) | |
Week 10 | 15.6% (11.7–28.5)a | 0.8% (0.2–4.6) | ||
After daily training session | Week 0 | 5.0% (2.2–8.9) | 4.1% (1.3–10.4) | |
Week 10 | 14.3% (8.5–20.0) | 1.9% (0.3–5.9) | ||
1.0% EMGmax | ||||
Before daily training session | Week 0 | 7.6% (5.8–35.1) | 11.1% (6.4–24.7) | |
Week 10 | 26.0% (21.9–45.0)b | 3.5% (1.4–12.6) | ||
After daily training session | Week 0 | 11.9% (4.3–19.4) | 5.9% (2.0–22.5) | |
Week 10 | 23.9% (12.3–29.8) | 3.7% (1.0–12.4) | ||
1.5% EMGmax | ||||
Before daily training session | Week 0 | 18.2% (11.1–48.5) | 21.0% (7.6–29.8) | |
Week 10 | 37.2% (29.8–55.4) | 7.3% (2.8–20.8) | ||
After daily training session | Week 0 | 24.6% (10.1–37.6) | 8.2% (4.6–32.5) | |
Week 10 | 34.2% (19.2–40.8) | 5.7% (2.5–18.8) | ||
2.0% EMGmax | ||||
Before daily training session | Week 0 | 30.1% (19.8–57.9) | 31.0% (10.7–37.8) | |
Week 10 | 46.0% (36.2–62.6) | 10.6% (5.4–29.3) | ||
After daily training session | Week 0 | 35.1% (19.9–51.2) | 11.9% (8.8–35.5) | |
Week 10 | 41.0% (29.8–52.8) | 9.7% (4.3–26.5) |
Table
Duration of each EMG gap (seconds) under 0.5% EMGmax for m. trapezius and m. splenius, median (interquartile range).
Training |
Control |
||
---|---|---|---|
Trapezius | |||
Before daily training session | Week 0 | 0.72 (0.6–1.74) | 0.9 (0.54–1.68) |
Week 10 | 1.26 (0.6–1.98)b | 0.6 (0.42–1.02) | |
After daily training session | Week 0 | 1.08 (0.66–1.8) | 0.96 (0.6–1.38) |
Week 10 | 1.56 (0.72–2.76) | 0.78 (0.36–1.26) | |
Splenius | |||
Before daily training session | Week 0 | 0.42 (0.36–0.48) | 0.6 (0.48–0.96) |
Week 10 | 0.72 (0.54–0.78)a | 0.36 (0.3–0.48) | |
After daily training session | Week 0 | 0.54 (0.42–0.6) | 0.54 (0.42–0.72) |
Week 10 | 0.72 (0.54–1.02) | 0.48 (0.42–0.54) |
There was no change in the average EMG amplitude during the reference contraction (i.e., arms 90 degree abducted) from before to after the daily training session, showing that the EMG measurements were stable throughout the day.
The main finding of the present study was the change in occupational neck muscle activity in response to brief daily resistance training. These alterations were shown both acutely in response to a single training session and longitudinally following the 10 week intervention—however with opposite impact on the muscle activity pattern. While the single training session acutely altered the muscle activity pattern so that less frequent periods of muscular relaxation were observed, the longitudinal change in muscle activity led to both longer and more frequent periods of complete muscular relaxation. The longitudinal changes were observed concurrently with increased muscle strength and reduced pain of the neck muscles.
The frequency of EMG gaps decreased immediately after the training session in the splenius muscle, which may lead to increased muscle tension and perceived discomfort. Although we did not measure acute changes in pain in the present study, previous research has reported an acute increase in muscular pain immediately after high-intensity resistance training in women with trapezius myalgia [
The 10-week training period led to decreased pain and increased muscular strength in the neck/shoulder muscles. This is in accordance with the main study including all 198 participants [
Our study showed increased frequency of EMG gaps, that is, periods with complete muscular relaxation, defined as muscular activity below 0.5% EMGmax, following 10 weeks of resistance training. This more relaxed activity pattern in the neck muscles is likely to reduce fatigue and pain. Additionally, increased duration of EMG gaps in both m. splenius and m. trapezius was found. Prolonged duration of EMG gaps leads to longer episodes of complete muscular relaxation, which potentially can reduce the pain in the neck/shoulder muscles. A possible explanation for this relationship between the length of the EMG gap and the level of pain intensity can be that shorter EMG gaps, compared to longer EMG gaps, cause a higher average work strain [
Rosendal and coworkers have shown that women suffering from chronic neck muscle pain experience increased levels of both lactate and pyruvate in the interstitium as a result of low-force repetitive work [
In general, the findings of the present study suggest that the splenius muscle compared with trapezius is the primary site for pain sensation in the neck/shoulder muscles due to the fact that EMG alterations primarily appear in the splenius. This is supported by findings of a higher prevalence of severe tenderness in the neck extensors compared with trapezius [
A limitation to the present study is that participants could not be blinded due to the general design with a designated training group. This introduces multiple risks of nonspecific effects including possible placebo effects in respect to changes in perceived pain [
The relatively small sample size increases the risk for statistical type II errors, that is, not finding a significant difference when there is in fact a difference. On the other hand, the lack of Bonferroni correction will increase the risk for statistical type I errors. However, performing a Bonferroni correction will increase the risk of type II errors [
The use of surface EMG to determine the muscular activity patterns is sensitive to a number of different parameters including electrode placement [
The primary objective of this study was to investigate whether a brief daily resistance training session would have an effect on the muscular activity pattern of the neck/shoulder muscles. In respect to our hypothesis, we reported beneficial long-term changes in both the frequency and duration of the EMG gaps alongside with alterations in the time with minimal muscular activation. In summary, the acute response to a single session of resistance training appeared to generate an unfavourable muscle activity pattern. By contrast, the longitudinal changes were beneficial in terms of longer and more frequent periods of complete muscular relaxation and reduced pain; however, these findings were more pronounced in m. splenius compared to m. trapezius. Future studies on neck/shoulder pain should consider focusing also on the splenius rather than the trapezius alone.
The authors thank senior researcher Jørgen Skotte for providing the Matlab script for the EMG analysis. They also thank the students from the Metropolitan University College and the Institute of Exercise and Sports Sciences, University of Copenhagen, for their practical help during the project. The author Lars L. Andersen received a grant from the Danish Rheumatism Association (Grant R68-A993) for this study. The Hygenic Corporation (Akron, OH) provided elastic tubing for this study but no monetary funding.