Volume 21, Issue 1 , Pages 59-66, February 2011
A study of forearm muscle activity and wrist kinematics in symptomatic office workers performing mouse-clicking tasks with different precision and speed demands
Article Outline
- Abstract
- 1. Introduction
- 2. Methods
- 3. Results
- 4. Discussion
- 5. Conclusions
- Conflict of interest statement
- Acknowledgements
- References
- Biography
- Copyright
Abstract
The present study examined various biomechanical parameters in symptomatic and asymptomatic computer users during mouse-clicking tasks with different speed and precision demands. Surface electromyography (EMG) of right wrist flexors and extensors were compared between individuals with computer-related wrist/hand symptoms (n
=9) and pain-free controls (n=8). Each subject performed four mouse tasks with high and low precision, constant and fastest speed of 5min each. Results showed that Case subjects recorded significantly lower EMG amplitudes during maximum voluntary contractions in three out of four forearm muscles (p=0.001–0.019). Normalised median amplitudes of extensor carpi radialis and flexor carpi ulnaris showed significant differences between groups in the speed conditions (p=0.01, 0.04, respectively). Case subjects reported significant discomforts in the wrist/hand region in all tasks, while recorded significantly longer response time and fewer mouse clicks compared to controls. Results suggested that forearm muscles in symptomatic individuals were inhibited in their maximal activation as well as during functional tasks, and this may imply altered motor control mechanisms in forearm muscles contributing to work-related musculoskeletal disorders.
Keywords: Computer, Mouse, Electromyography, Work-related musculoskeletal disorders, Motor control
1. Introduction
Nowadays, the computer has become an essential instrument of communication in the office and at home, and the mouse is the most commonly used input device next to the keyboard. Although manipulation of such a small device is not physically demanding, it has been documented as an important risk factor for work-related neck and upper limb disorders. Prevalence rates of about 35% in the wrist/hand region have been reported to be related directly to computer mouse use (Cook et al., 2000).
Past research examining computer mouse use had mainly focused on muscle activity in healthy pain-free subjects. Dennerlein and Johnson (2006) compared the forearm and shoulder muscle activities when healthy subjects performed a variety of common typing and mousing tasks; and reported fairly low levels of muscle activities for operating the mouse – generally below 10% maximum EMG (% MEMG). Laursen et al. (2001) also reported low levels of muscle activity associated with mouse-clicking tasks, with median activities at below 5% MEMG for extensor and flexor carpi radialis muscles. Bystrom et al. (2002) found low but continuous activities, as well as absence of muscle activity gaps in forearm extensors in graphic design tasks involving intensive mouse use. However, most of these studies are concerned with comparing different mousing tasks or use of different mouse devices in healthy individuals. Evidence of symptomatic–asymptomatic differences in motor control associated with computer mouse use has not been extensively investigated especially in comparing different muscles controlling wrist movements.
For mouse users, speed and precision stresses are common in daily tasks and has aroused much research interest. Birch et al. (2000) found that high time pressure combined with low precision and low mental demands resulted in higher muscle activities of finger flexors and extensors, as well as shoulder muscles in graphic design operators performing drawing tasks with the mouse. Visser et al. (2004) examined the effect of increasing speed and mental demand in a mouse aiming task and reported a significant increase in both neck and forearm muscle activities in the trapezius, extensor digitorum and flexor digitorum superficialis.
There is mounting evidence that altered muscle activation is an important motor control mechanism contributing to the development of musculoskeletal disorders (Falla et al., 2004, Szeto et al., 2005a). Such evidence seems to mainly come from spinal research involving postural stabilising muscles in the cervical spine (Falla et al., 2004, Szeto et al., 2005a). It is not clear whether the same altered motor control mechanisms also affect the muscles controlling peripheral joints such as the wrist, and how different stressful demands would induce or provoke such phenomena. Hence the present study aimed to examine the muscle activities in the wrist extensor and flexor muscles during different mousing tasks in a case–control study design.
2. Methods
2.1. Subjects
A total of 17 female office workers were recruited for the present study. All of them used the computer for a minimum of 4h per day and had computer experience for at least 2–3years. Subjects were excluded if they had any traumatic injuries in relevant regions, or suffered from medical conditions that may affect the spine or upper limbs. Nine subjects were allocated to the Case Group, and they all had symptoms such as pain, aching, burning, numbness, or tingling present in the right wrist/hand region related to mouse use. Their symptoms lasted for more than 3months in the past year, and were also present during the 7days prior to testing as well as on the day of testing. Physical examination procedures included resisted muscle testing, palpation, range of movement testing of the wrist, and specific tests such as the Finkelstein and Cozen test for detecting more specific syndromes. All subjects had full range of movements in the wrist and muscle strength testing were grade 5. Among the Case subjects, four of them showed a positive response towards the Finkelstein test, and symptoms were reproduced on resisted wrist extension and radial/ulnar deviation. Five subjects had their usual wrist/hand pain reproduced upon resisted fingers/thumb extension. Seven of them had received previous medical consultation and only one person was on present treatment. Eight subjects were recruited in the Control Group and all of them had no or minimal discomfort in wrist/hand region past or present.
All subjects in the study were asked to complete a questionnaire prior to their participation. Demographic data including age, gender, and computer use profile were recorded. Questions about musculoskeletal symptoms were adopted from the Standardised Nordic Questionnaire (Kuorinka et al., 1987) (Table 1). The experimental procedures were explained to each subject and informed consent was obtained before the experiment began. The study was approved by the Human Ethics Committee of the Hong Kong Polytechnic University.
Table 1. Demographic data and computer use profile of subjects.
| Case group (n=9) | Control group (n=8) | Group difference statistics | |
|---|---|---|---|
| Age (year) [mean (SD; range)] | Mean=36.23 (5.40; 30–43) | Mean=26.50 (3.02; 23–30) | t=−4.642 p <0.001⁎ |
| Body height (cm) [mean (SD; range)] | Mean=152.88 (4.11; 150–160) | Mean=163.88 (4.97; 155–170) | t=4.911 p <0.001⁎ |
| Body weight (kg) [mean (SD; range)] | Mean=58.60 (17.41; 44.91–99.79) | Mean=49.06 (3.28; 42.92–52.27) | t=−1.613 p =0.143 |
| Experience in computer use [mode (range)] | Mode=>3years (>3 years – >3years) | Mode=>3years (>3 years – >3years) | Z=0.000 p =1.000 |
| Computer use in hours/day [mode (range)] | Mode=6–8 (6 to 8 – 10) | Mode=6–8 (4 to 6 – 10 to 12) | Z=−1.533 p =0.125 |
| Mouse use in hours/day [mode (range)] | Mode=2–4 (2 to 4 – 6 to 8) | Mode=2–4 (2 to 4 – 6 to 8) | Z<0.001 p =1.000 |
| Keyboard use in hours/day [mode (range)] | Mode=2–4 (2 to 4 – 6 to 8) | Mode=2–4 (2 to 4 – 4 to 6) | Z=−1.016 p =0.309 |
⁎Significant level at p |
2.2. Surface electromyography
Four muscles in the forearm region were selected for the electromyography (EMG) study: extensor carpi ulnaris (ECU), extensor carpi radialis (ECR), flexor carpi ulnaris (FCU) and flexor carpi radialis (FCR). Only the right forearm muscles were measured as all subjects use their right hands to operate the computer mouse. The Noraxon TeleMyo 2400T G2 system (Noraxon USA Inc., USA) was used to capture the surface EMG data, with a bandwidth of 10–500Hz and a common mode rejection ratio of 100dB. The EMG signals received from the transmitter underwent a 16bit analogue to digital (A/D) conversion at a sampling frequency of 1500Hz.
Bipolar Ag–AgCl surface electrodes (3M™ Infant Red Dot™ electrodes, 3M Limited, Hong Kong) of 15mm diameter were used. The precise locations of EMG electrodes were adopted from Perotto et al. (2004) and also based on past research studies (Dennerlein and Johnson, 2006, Visser et al., 2004). Before attaching electrodes, the skin was carefully prepared by cleansing with water, sand paper and alcohol (and shaved if necessary). The skin impedance was checked with an impedance meter and below 5K ohm was considered acceptable. The inter-electrode distance was fixed at 20mm, and the ground electrode was placed over the right medial styloid process of the wrist.
Prior to starting the experiment, subjects were asked to perform two trials of resisted isometric maximum voluntary contractions (MVC) of 5s each for each muscle, with 1min rest in between. The subject’s right forearm was stabilized by an inelastic strap on a wooden bench surface and the wrist was positioned at the edge of the bench. A load cell was connected to the hand by an inelastic strap wrapped around the metacarpal heads, and it was fixated to a board with a metal bar. Resisted muscle actions for the MVC trials and starting positions are summarised in Table 2.
Table 2. Results of group means (SD) of EMG (μV) at maximum voluntary contractions (MVC) of four muscles.
| Muscle | Resisted muscle action and starting position | Max EMG (RMS) mean (SD) | Within-subject reliability: ICC (3.1), (95% CI) | Between-group comparisons (indep. t-tests: t, p values) | MVC (N) mean (SD) | Between-group comparisons (indep. t-tests: t, p values) | ||
|---|---|---|---|---|---|---|---|---|
| Case | Control | Case | Control | |||||
| ECU | Extension of wrist with ulnar deviation, wrist in 0° flexion and 90° pronation | 258.51 (92.69) | 418.94 (150.90) | Case: ICC=0.956 (0.818–0.990) Control: ICC =0. 977 (0.853–0.994) | t15=−2.622, p=0.019⁎ | 37.47 (12.21) | 48.48 (34.64) | t9=0.786, p=0.452 |
| ECR | Extension of wrist with radial deviation, wrist in 0° flexion and 90° pronation | 195.83 (60.71) | 464.20 (194.38) | Case: ICC=0.913 (0.664–0.998) Control: ICC =0.786 (0.432–0.977) | t8.3=−3.816, p=0.005⁎ | 43.02 (8.07) | 56.55 (13.77) | t9=2.244, p=0.052 |
| FCU | Flexion of wrist with ulnar deviation, wrist in 0° flexion and 90° supination | 101.43 (60.77) | 279.33 (85.11) | Case: ICC=0.959 (0.818–0.991) Control: ICC =0. 945 (0.725–0.989) | t15=−5.048, p<0.001⁎ | 43.83 (7.17) | 48.80 (10.91) | t9=0.922, p=0.381 |
| FCR | Flexion of the wrist with radial deviation, wrist in 0° flexion and 90° supination | 201.89 (122.22) | 297.42 (129.42) | Case: ICC=0.952 (0.789–0.989) Control: ICC =0. 966 (0.831–0.993) | t15=−1.547, p=0.143 | 39.83 (12.56) | 54.40 (13.77) | t9=1.792, p=0.107 |
⁎Significance level at p |
2.3. EMG data processing
The Myoresearch XP master edition (Noraxon USA Inc., USA) software was used to process the EMG data during the mousing tasks. Data processing involved full-wave rectification and smoothing with root-mean-square (RMS) of 100ms window. These data were then exported to compute the three levels of Amplitude Probability Distribution Function (APDF) – 10th%, 50th% and 90th%. 50th% APDF was considered an indicator of median muscle activity and APDF range represented the difference between the static (10th%) and peak (90th%) levels of APDF (Szeto et al., 2009). The APDF parameters were expressed both in terms of “non-normalised” data (RMS) as well as “normalised” data – percentage of maximum voluntary exertion (% MVE), and the two types of data were compared in the results.
2.4. Wrist kinematics
A biaxial electrogoniometer (Model 308, Noraxon USA Inc., USA) was used to measure radio-ulnar deviation angles of the right wrist, with the same sampling frequency of 1500Hz. During pilot trials, the extent of flexion/extension movement was found to be very limited which was probably due to the wrist being supported on the desk surface. Therefore, only the movement data of radio-ulnar deviation was examined in the present study. The joint angle data were also processed in terms of APDF, and angular accelerations were also examined.
2.5. Performance and musculoskeletal symptom measure
Performance or productivity of mouse clicks was determined by recording how many targets the subject had clicked within the 5min task time. The response time per mouse click was also measured and compared between tasks and between groups.
Subjective discomfort rating was assessed using a numerical rating score of 0–10 (0=no discomfort, l=minimal discomfort and 10=extreme or intolerable discomfort). The subject was asked to rate their discomfort in the right wrist/hand region, since this is the focus of the present study. The discomfort ratings were recorded before the start of the experiment and after each mousing task. The discomfort scores were compared as within-subject factor (pre- and post-task) and compared between tasks.
2.6. The workstation and experimental tasks
The experimental workstation included a computer desk with a slide-out tray for mouse and keyboard at a fixed height of 60cm. An optical mouse (Microsoft comfort optical mouse 3000, Microsoft Corporation, USA) with 1000dpi was used for all subjects. The mouse was placed on a mouse pad with wrist support on the right side of keyboard. The subject’s forearm is partially supported on the keyboard tray at the same height. A desktop computer with 15′′ visible diagonal size LCD screen was used, and the height of the screen was adjusted so that the subject could maintain a reasonably erect head posture. This workstation also included a height-adjustable swivel chair (seat height ranged from 34 to 45cm) with backrest but no armrest (Fig. 1).
Fig. 1.
Photograph showing the subject with EMG electrodes attached to the forearm muscles and performing the mousing task.
In the present study, subjects were asked to perform multidirectional mouse-clicking tasks under four conditions:
Each mouse task lasted for 5min with a 3min rest in between. Task sequence was randomized for each subject by drawing lots. The multidirectional clicking task was referenced to the ISO guideline on tests for evaluation of non-keyboard input devices (ISO, 2000), and involved clicking square targets around a circle that covers most part of the screen. The precision requirements in the mouse-clicking tasks were guided by the effective index of difficulty (ISO, 2000). In the FS condition, subjects were asked to click at the target as fast as possible. In the CS condition, they were asked to click at the target appearing at a constant rate of 3s. In the HP and LP conditions, subjects were asked to click at their usual work pace. For FS and CS conditions, precision requirements were set at medium level.
2.7. Data analysis
EMG data of the maximum voluntary contractions were examined for within-subject reliability using Intra-Class Correlation (model 3) to ensure that they were truly maximal efforts. The EMG values in terms of median activity (50th% APDF) and the APDF range were analysed with mixed model MANOVA to examine the effects of condition (within-subject factor) and group (between-subject factor). Speed and precision were analysed separately as within-subject factors with two levels (speed: fast and constant; precision: low and high). Since only Case Group subjects experienced discomforts during the experiment and parametric assumption could not be met, subjective discomfort was analysed using Mann–Whitney test to examine the between-group differences. The task performance variables of response time and productivity were also compared between task conditions and between groups. SPSS version 16.0 was used for all statistical analysis and level of significance was set at 0.05.
3. Results
3.1. Subject characteristics
Demographic data and computer use profiles of the subjects are summarised in Table 1. There were no significant differences in all the demographic and workload variables between Case and Control Groups, except for age and height. When age and height were included as covariates in the MANOVA analysis, these factors did not produce any statistically significant differences for both variables – median amplitudes and APDF range. Past studies have also found a similar result with symptomatic subjects having a higher mean age as compared with pain-free controls, but it did not have a significant effect on their muscle activity measures (Szeto et al., 2005a).
3.2. Maximum voluntary contractions
Initially it was intended to use maximum voluntary contractions (MVC) in the normalisation process of muscle activities in the experimental tasks. There was high within-subject reliability indicated by the ICC results of the two trials of MVC and these were consistent for both groups (Table 2). However, there were significant differences in the normalised maximum EMG values between two groups in the ECU, ECR and FCU muscles. These results suggested that the forearm muscles appeared to be somehow inhibited in maximal activation in the symptomatic subjects. Hence the EMG data in the experimental tasks were examined both as RMS values as well as % MVE and the two types of data were compared.
The force data recorded during the MVC trials also showed trends for higher values achieved by the Control Group compared to the Case Group (see Table 2). However, the between-group differences in force data was only marginally significant (p=0.052) in ECR and not statistically significant in the other muscles. This may be due to large variations in force exertion capacities between different individuals. Differences in age and body build between groups may also account for some of the differences in EMG amplitudes.
3.3. Muscular activity – median amplitudes and APDF range
When the speed and precision conditions were examined in univariate analyses, the normalised ECR amplitude showed a statistically significant effect for the group factor in both analysis (precision: F=4.48, p=0.05; speed: F=11.14, p=0.01). The FCU also recorded a statistically significant effect for group in the speed conditions analysis (F=5.37, p=0.04). These significant differences in median amplitudes between groups were found in the analysis for the MVE-normalised median amplitude data but not in the non-normalised RMS data.
Examining the normalised and non-normalised data in Table 3 produced interesting results. Comparisons of the Case and Control Group data as non-normalised (RMS) values showed trends for generally higher amplitudes in the Control Group over Case Group in mostly the ECU and ECR muscles and more so in the more stressful conditions. The flexor muscles seemed to be more similar in activities in various task conditions, and most of the statistical analyses for the non-normalised median amplitudes were not statistically significant. When the data were examined as normalised values, the results suggested that the Case Group subjects had trends for higher amplitudes (in terms of % MVE) compared to Control Group in all precision and speed conditions, but only the ECR muscle recorded significant group differences in both precision and speed condition analyses, and FCU in speed condition analysis.
Table 3. Results of median muscle activities (50th% APDF) in two groups (mean, SD) and statistical analysis results.
| Muscle | HP | LP | FS | CS | Univariate analysis: 2 precision conditions (F1,13, p) | Univariate analysis: 2 speed conditions (F1,13, p) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Case | Control | Case | Control | Case | Control | Case | Control | Cond. | Group | Cond.×Group | Cond. | Group | Cond.×Group | |
| Non-normalised (μV) | ||||||||||||||
| ECU | 29.78 (11.35) | 45.72 (13.56) | 30.53 (15.15) | 40.99 (9.97) | 36.46 (13.71) | 49.03 (13.27) | 22.05 (8.09) | 24.67 (12.75) | 0.96, 0.34 | 0.277, 0.61 | 0.56, 0.47 | 0.02, 0.96 | 0.77, 0.40 | 2.55, 0.13 |
| ECR | 21.53 (10.10) | 33.97 (3.48) | 22.12 (10.45) | 32.57 (13.16) | 25.54 (9.92) | 35.26 (13.77) | 17.76 (8.10) | 20.37 (14.02) | 0.68, 0.43 | 0.11, 0.75 | 0.38, 0.55 | 0.20, 0.66 | 0.59, 0.46 | 2.90, 0.11 |
| FCU | 12.11 (6.95) | 12.34 (5.30) | 9.93 (5.84) | 9.53 (6.30) | 12.86 (6.85) | 12.50 (5.36) | 9.74 (5.86) | 7.30 (5.22) | 0.13, 0.73 | 0.04, 0.85 | 0.66, 0.43 | 0.45, 0.51 | 0.21, 0.65 | 2.70, 0.13 |
| FCR | 13.84 (8.60) | 16.23 (10.67) | 10.75 (7.48) | 10.68 (5.42) | 14.75 (9.17) | 17.64 (11.72) | 8.76 (4.47) | 7.08 (3.24) | 0.27, 0.61 | 0.01, 0.94 | 0.56, 0.55 | 1.48, 0.25 | 0.51, 0.49 | 0.07, 0.79 |
| MVE-normalised (% MVE) | ||||||||||||||
| ECU | 10.451 (2.67) | 9.66 (1.57) | 10.63 (3.41) | 8.29 (1.59) | 12.92 (2.27) | 10.37 (3.73) | 7.42 (2.01) | 4.56 (2.18) | 0.20, 0.66 | <0.01, 1.00 | 0.09, 0.77 | <0.01, 1.00 | 1.81, 0.20 | 0.09, 0.77 |
| ECR | 8.15 (2.52) | 5.57 (3.079) | 8.42 (2.38) | 5.28 (2.81) | 9.99 (2.75) | 5.83 (3.19) | 6.51 (2.22) | 2.65 (1.65) | 0.07, 0.80 | 4.48, 0.05⁎ | 0.72, 0.41 | 0.86, 0.37 | 11.14, 0.01⁎ | 1.01, 0.33 |
| FCU | 7.97 (4.32) | 2.82 (2.02) | 6.36 (3.33) | 1.74 (1.19) | 8.26 (3.73) | 2.66 (1.48) | 6.11 (3.23) | 1.23 (0.89) | 0.08, 0.78 | 1.94, 0.19 | 1.06, 0.32 | 0.66, 0.43 | 5.37, 0.04⁎ | 1.31, 0.27 |
| FCR | 5.41 (2.06) | 3.96 (2.92) | 4.38 (2.05) | 2.36 (1.78) | 6.17 (3.41) | 4.22 (2.92) | 3.35 (1.63) | 1.22 (0.89) | 1.00, 0.34 | 0.02, 0.90 | 0.33, 0.58 | 0.01, 0.91 | 0.28, 0.61 | 0.19, 0.67 |
⁎Significant level at p |
The APDF range is defined as the difference between the 90th% and 10th% APDF (Szeto et al., 2009). This variable is an indicator of the extent of variation in EMG amplitudes from the high end to the low end of the muscle load during various tasks. In the non-normalised data, the ECR muscle showed a statistically significant interaction which was only observed between speed condition and group (p=0.04), and not between precision condition and group (Table 4). Also, differences in the APDF range of the normalised ECR were only significant between the speed conditions (p=0.02).
Table 4. Results of APDF range (90th%–10th% APDF) comparing Case and Control Groups (mean, SD) and statistical analysis in precision and speed comparisons.
| Muscle | HP | LP | FS | CS | Univariate analysis: 2 precision conditions (F1,13, p) | Univariate analysis: 2 speed conditions (F1,13, p) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Case | Control | Case | Control | Case | Control | Case | Control | Cond. | Group | Cond.×Group | Cond. | Group | Cond.×Group | |
| Non-normalised (μV) | ||||||||||||||
| ECU | 25.55 (9.04) | 24.67 (9.23) | 27.30 (11.56) | 35.60 (11.89) | 28.47 (10.39) | 36.96 (10.20) | 22.44 (7.97) | 34.56 (9.68) | 3.99, 0.07 | 0.44, 0.52 | 0.13, 0.72 | 4.05, 0.07 | 1.56, 0.23 | 4.67, 0.05⁎ |
| ECR | 14.50 (6.13) | 22.12 (7.79) | 16.55 (8.22) | 20.94 (7.36) | 18.29 (7.89) | 21.43 (8.08) | 13.59 (5.60) | 22.08 (7.85) | 6.96, 0.20 | 0.14, 0.72 | 3.26, 0.09 | 3.26, 0.09 | 0.03, 0.86 | 5.44, 0.04⁎ |
| FCU | 8.68 (5.69) | 14.00 (7.90) | 9.16 (5.80) | 12.28 (4.93) | 11.88 (11.23) | 14.99 (5.90) | 5.34 (3.17) | 9.13 (4.09) | 0.89, 0.36 | 0.74, 0.41 | 0.41, 0.53 | <0.01, 0.99 | 1.55, 0.24 | 0.06, 0.80 |
| FCR | 14.44 (15.68) | 16.03 (15.97) | 13.52 (17.33) | 20.49 (10.37) | 17.29 (17.53) | 22.68 (14.80) | 6.51 (5.09) | 14.38 (8.92) | 0.02, 0.89 | 0.06, 0.81 | 1.01, 0.33 | 3.10, 0.10 | 0.02, 0.89 | 2.37, 0.15 |
| MVE-normalised (% MVE) | ||||||||||||||
| ECU | 10.29 (3.51) | 8.52 (2.63) | 11.43 (5.62) | 8.21 (2.86) | 11.31 (3.87) | 8.50 (2.83) | 9.28 (4.06) | 7.88 (1.74) | 0.82, 0.38 | 0.36, 0.56 | 0.09, 0.77 | 2.88, 0.11 | 0.54, 0.48 | 1.44, 0.24 |
| ECR | 6.43 (2.06) | 4.41 (1.63) | 7.24 (2.30) | 4.22 (1.60) | 8.21 (3.03) | 4.19 (1.59) | 6.20 (2.52) | 4.48 (2.11) | 1.38, 0.26 | 2.71, 0.12 | 0.86, 0.37 | 6.75, 0.02⁎ | 4.01, 0.07 | 4.04, 0.07 |
| FCU | 6.53 (4.44) | 4.63 (3.08) | 6.13 (2.85) | 3.46 (1.53) | 7.82 (4.44) | 4.34 (2.40) | 4.32 (2.76) | 2.65 (1.54) | 2.89, 0.11 | 0.33, 0.57 | 0.98, 0.34 | <0.01, 0.95 | 0.05, 0.82 | 0.30, 0.59 |
| FCR | 6.25 (5.64) | 5.64 (3.26) | 5.87 (5.39) | 5.76 (2.37) | 7.57 (7.84) | 6.54 (3.54) | (3.59) (3.02) | 3.99 (2.14) | 0.11, 0.75 | 0.52, 0.48 | 1.64, 0.22 | 1.306, 0.27 | 0.87, 0.37 | 0.07, 0.80 |
⁎Significant level at p |
The median amplitudes of the four muscles were also examined as co-contraction ratios for the control of the radial and ulnar deviation movements. The “ulnar ratio” denotes the ratio of ECU amplitude over FCU, and radial ratio represents ECR over FCR (see Table 5).
Table 5. Co-contraction ratios of forearm flexors and extensors comparing groups and task conditions.
| HP | LP | FS | CS | Univariate analysis: 2 precision conditions (F1,13, p) | Univariate analysis: 2 speed conditions (F1,13, p) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Case | Control | Case | Control | Case | Control | Case | Control | Cond. | Group | Cond.×Group | Cond. | Group | Cond.×Group | |
| Non-normalised (μV) | ||||||||||||||
| Ulnar ratio | 3.32 | 4.59 | 3.71 | 5.85 | 3.93 | 4.92 | 3.19 | 4.88 | <0.00, 0.99 | 0.97, 0.34 | 0.65, 0.43 | 1.91, 0.19 | 1.62, 0.23 | 0.11, 0.74 |
| (ECU/FCU) | (2.34) | (3.33) | (1.99) | (3.54) | (2.98) | (3.04) | (2.30) | (4.03) | ||||||
| Radial ratio | 2.04 | 2.45 | 2.61 | 3.35 | 2.48 | 2.34 | 2.54 | 2.76 | 0.12, 0.74 | 0.10, 0.75 | 0.02, 0.90 | 2.51, 0.14 | 1.44, 0.25 | 0.04, 0.84 |
| (ECR/FCR) | (1.38) | (0.93) | (1.56) | (1.06) | (1.92) | (0.83) | (1.47) | (1.07) | ||||||
| MVE-normalised (% MVE) data | ||||||||||||||
| Ulnar ratio | 2.00 | 5.92 | 2.52 | 7.48 | 2.20 | 6.67 | 2.11 | 5.54 | 1.11, 0.31 | 0.41, 0.53 | 0.64, 0.44 | 0.08, 0.78 | 0.35, 0.57 | 0.09, 0.77 |
| (ECU/FCU) | (1.72) | (5.54) | (2.47) | (5.18) | (2.03) | (6.37) | (2.48) | (4.62) | ||||||
| Radial ratio | 1.63 | 2.93 | 2.63 | 4.03 | 1.95 | 2.82 | 2.30 | 3.25 | 0.28, 0.61 | 0.29, 0.60 | 1.60, 0.23 | 1.49, 0.24 | 0.73, 0.41 | 0.24, 0.63 |
| (ECR/FCR) | (0.59) | (3.50) | (2.24) | (3.50) | (0.96) | (3.94) | (1.10) | (2.69) | ||||||
3.4. Wrist kinematics
Kinematic variables were examined as median joint angles (50th% APDF) and the range (90th%–10th% APDF) of radio-ulnar deviation and these were compared between groups and between tasks (Table 6). Both groups had their median wrist positions predominantly in the direction of ulnar deviation, and both groups showed more apparent differences in median angles in the low precision and constant speed conditions. However, statistical analysis generally revealed no significant differences between groups and between task conditions.
Table 6. Wrist joint kinematics comparing groups and task conditions.
| HP | LP | FS | CS | Univariate analysis: 2 precision conditions (F1,13, p) | Univariate analysis: 2 speed conditions (F1,13, p) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Case | Control | Case | Control | Case | Control | Case | Control | Cond. | Group | Cond.×Group | Cond. | Group | Cond.×Group | |
| Joint position (°) | ||||||||||||||
| Median | −10.36 (11.31) | −9.25 (8.20) | −12.17 (8.68) | −5.96 (6.60) | −13.80 (8.01) | −13.00 (10.89) | −12.70 (10.78) | −5.57 (7.50) | 0.12, 0.74 | 0.35, 0.56 | 1.15, 0.30 | 0.58, 0.46 | 0.16, 0.67 | 2.43, 0.14 |
| Range | 9.83 (5.41) | 11.89 (3.67) | 7.50 (2.30) | 8.48 (5.04) | 8.36 (4.11) | 9.93 (5.37) | 6.40 (2.81) | 7.07 (2.16) | 0.01, 0.94 | 1.84, 0.20 | 0.20, 0.66 | 0.46, 0.51 | 1.38, 0.26 | <0.01, 0.96 |
| Acceleration magnitude (°/s2) | ||||||||||||||
| Towards Radial | 34.67 (10.83) | 46.61 (13.63) | 38.14 (6.64) | 44.02 (9.86) | 36.15 (11.07) | 42.71 (8.84) | 34.11 (15.97) | 42.90 (11.26) | 0.12, 0.74 | 0.34, 0.57 | 0.25, 0.63 | 1.33, 0.27 | 0.23, 0.64 | 0.70, 0.42 |
| Towards Ulnar | 26.62 (4.15) | 32.25 (14.35) | 31.80 (5.50) | 37.60 (11.27) | 29.80 (6.2) | 35.30 (13.70) | 26.15 (13.90) | 23.58 (7.44) | 0.50, 0.83 | 0.37, 0.55 | <0.01, 1.00 | 0.74, 0.41 | 0.03, 0.88 | 1.86, 0.20 |
When the wrist acceleration magnitudes towards ulnar and radial side were compared in the four tasks, these data revealed more apparent differences between groups in both speed and precision conditions. There was a trend for Control Group to show higher acceleration magnitude than Case Group towards both radial and ulnar sides. Nevertheless the repeated measures ANOVA showed no significant differences.
3.5. Productivity and discomfort scores
Control Group generally performed better with more mouse clicks and shorter reaction time than Case Group in all conditions, but the difference in productivity was only significant for the high precision condition. The differences in response time were also significantly higher in the Case Group in HP and FS conditions (Table 7). Only the Case Group subjects reported discomfort scores in the wrist/hand region and Control subjects reported no discomfort during the experiment.
Table 7. Response time, productivity and discomfort scores.
| Task | Response time (second) mean (SD) | Independent t-tests between-group comparisons | Productivity (number of clicks) mean (SD) | Independent t-tests between-group comparisons | Discomfort score (0–10) mean (SD) | Mann–Whitney U test between-group comparisons | |||
|---|---|---|---|---|---|---|---|---|---|
| Case | Control | Case | Control | Case | Control | ||||
| HP | 2.49 (0.35) | 2.07 (0.27) | t15=−2.75, p=0.02⁎ | 122.0 (15.64) | 145.5 (21.23) | t15=2.62, p=0.02⁎ | 4.44 (1.42) | 0 | U<0.001, p<0.001⁎ |
| LP | 0.98 (0.24) | 0.80 (0.19) | t15=−1.66, p=0.12 | 262.1 (25.75) | 271.5 (22.37) | t15=0.80, p=0.44 | 3.11 (1.76) | 0 | U=4.00, p=0.001⁎ |
| FS | 1.14 (0.17) | 0.97 (0.12) | t15=−2.04, p=0.04⁎ | 246.0 (25.06) | 256.8 (14.94) | t15=1.06, p=0.31 | 3.89 (1.76) | 0 | U<0.001, p<0.001⁎ |
| CS | 1.42 (0.20) | 1.27 (0.22) | t15=−1.51, p=0.15 | 91.0 (0.71) | 91.6 (0.92) | t15=1.59, p=0.13 | 3.11 (1.76) | 0 | U=4.00, p=0.001⁎ |
⁎Significant level at p |
4. Discussion
The study attempted to examine differences in muscle activation and movement control in the forearm and wrist region of symptomatic and asymptomatic individuals, when they performed standardised mousing tasks with different physical demands. Past research has mainly focused on either the nature of the mouse task, task performance or the equipment used while EMG is recorded in the forearm muscles in healthy pain-free subjects (Dennerlein and Johnson, 2006, Thorn et al., 2007, Visser et al., 2004). The present study has demonstrated that it may be necessary to examine both normalised and non-normalised data when the EMG signals and MVC force exertions were not the same between-subject groups. In the following sections, the results will be discussed and compared with what past research studies have found.
4.1. Case–control differences in muscle activation
In the present study, the main findings showed that the Case Group had a significant reduction in the EMG amplitudes during maximum voluntary contractions of the forearm muscles, namely the ECU, ECR and FCU muscles. The present results showing significant decreases in EMG amplitudes during maximum voluntary exertions suggest that the muscles may be inhibited from maximum active recruitment, according to the pain inhibition theory proposed by Johansson and Sojka (1991). This theory proposed that muscle contractivity may be inhibited when the muscle is in an acute pain situation, and the present study seemed to support that the forearm muscles may also respond in this way.
The issue of normalisation has been a controversial topic in EMG research. It is easier to compare EMG amplitudes between different muscles in different individuals and in different tasks, if normalisation is used. However, if there are differences in baseline MVC values between different subjects, then some researchers consider it more appropriate to compare raw EMG data without normalisation. Nordander et al. (2004) also examined the EMG amplitudes as both normalised and non-normalised RMS values when they compared exposure variability of different manual handling job tasks. It was suggested that normalisation procedure is still currently the most reliable method to reduce between-subject variability in examining EMG data compared to non-normalised registrations (Nordander et al., 2004). In the present study, when EMG was examined as non-normalised values, variations across individuals make it difficult to detect distinct differences between groups, and the statistical comparisons failed to reveal significant differences even though some trends of case–control differences could be observed. When the data were compared as normalised values, it was easier to observe case–control differences but the baseline discrepancies have to be taken into consideration.
The comparison of median amplitudes data between the symptomatic and asymptomatic groups seemed to suggest that the speed demand elicited greater differences between groups. This was supported by the significant differences in the ECR and FCU muscles (normalised amplitudes) in the speed condition analysis. These two muscles were likely to be most active in directing the wrist movements towards the radial and ulnar sides. The APDF range data revealed fairly similar patterns between Control and Case Groups under the four task conditions. Higher values of APDF range would suggest that a muscle was working with greater variations in muscle load during a given task. In another study that compared typing and mousing tasks of 15min each, neck–shoulder muscles showed higher median amplitudes (normalised) as well as higher APDF range in the neck–shoulder postural muscle activities of office workers with neck pain (Szeto et al., 2009).
4.2. Motor control and motor performance
The present study has shown no major differences in average joint angles of the wrist in terms of radio-ulnar deviations between groups and between tasks. The wrist joint acceleration also showed no significant differences between groups and between tasks. However, there were significant differences between groups in terms of response time in the high precision and fastest speed conditions. The significant group differences in response time and productivity indicated that the symptomatic group had poorer motor performance in the more stressful tasks and this was associated with higher discomfort scores.
In addition to the physical stress, speed and precision demands may also lead to increased mental pressure which may have a direct effect on biomechanical loading. Past studies have reported that mentally stressful tasks can contribute to increased muscle loading in both the neck–shoulder as well as forearm muscles (Wahlstrom et al., 2002, Visser et al., 2004). Future study should explore further the interactive effects of physical and psychological stress factors and how these may affect the coordination of movements and muscular efforts in spinal and peripheral joints in symptomatic individuals. In addition, different individuals may have different levels of sensitivity or reactivity to stress, which may be affected by their psychological disposition, and the pathology involved. (Galen et al., 2002, Szeto et al., 2005b).
4.3. Limitations of study
The present study has revealed some interesting differences between symptomatic and asymptomatic individuals in their forearm muscle control and wrist movements when performing certain mouse tasks. Although we attempted to standardise the task conditions and the environment as much as possible, individual variations and small sample size may have contributed to the lack of significant differences in many of the outcome variables. Initial power calculations based on ECR non-normalised data under HP condition (assuming 80% power and 5% Type I error), revealed an effect size for group at 1.647 and the sample size required per group is 7. However, if other EMG variables were used in the calculations, the sample size required ranges from 7 (ECR in HP condition) to over 11,000 for FCU in HP condition (effect size=0.037). For the speed conditions, sample size required range from 20 to ⩾4000 for the various muscles. Other factors contributing to the lack of significant differences in the results may be due to differences in personal characteristics such as age and body build of the subjects in the two groups. It is also possible that task demands were not stressful enough to trigger the motor control deviations expected in the symptomatic group.
5. Conclusions
Present results have shown that symptomatic individuals had significantly decreased muscle activity amplitudes during maximum voluntary contractions in their forearm muscles compared to asymptomatic controls. Mousing tasks with greater speed demand produced significantly increased median amplitudes in the normalised activity of ECR and FCU muscles in the symptomatic group. Other comparisons of normalised and non-normalised EMG amplitudes, as well as wrist joint kinematics did not show significant differences between Case and Control Groups. The asymptomatic controls performed the various mouse tasks with significantly shorter response time and produced significantly higher number of mouse clicks in the high precision condition. The symptomatic subjects had moderate degrees of symptoms in their wrist-hand region while control group had no symptoms at all. On the whole, the symptomatic subjects seemed to have an inhibited capacity to generate maximum muscle activation and their motor performance in terms of speed and precision control of the mousing tasks were also inferior compared to the healthy control subjects. Further study should examine the muscle performance in a larger sample of computer workers and the task demands may need to impose even greater stress levels in order to produce meaningful comparisons between symptomatic and asymptomatic individuals.
Conflict of interest statement
The authors of this paper declare that there is no conflict of interest with the funding body for this project, and all authors have no financial ties with any private company to disclose.
Acknowledgements
The authors would like to acknowledge funding support from the Department of Rehabilitation Sciences, the Hong Kong Polytechnic University for this project. We would like to thank Mr. Man Cheung for technical support and Mr. Raymond Chung for statistical advice.
References
- . Acute response to precision, time pressure and mental demand during simulated computer work. Scand J Work Environ Health. 2000;26(4):299–305
- . Physical workload on neck and upper limb using two CAD applications. Appl Ergon. 2002;33(1):63–74
- . The prevalence of neck and upper extremity musculoskeletal symptoms in computer mouse users. Int J Ind Ergon. 2000;26:347–356
- . Different computer tasks affect the exposure of the upper extremity to biomechanical risk factors. Ergonomics. 2006;49(1):45–61
- . Patients with chronic neck pain demonstrate altered patterns of muscle activation during performance of a functional upper limb task. Spine. 2004;29(13):1436–1440
- . Forearm EMG response activity during motor performance in individuals prone to increased stress reactivity. Am J Ind Med. 2002;41(5):406–419
- ISO. Non-keyboard Input Devices. In: ISO 9241 ergonomic requirements of office work with visual display terminals (VDT). Geneva: International Organization for Standardization; 2000. p. 1–47.
- . Pathophysiological mechanisms involved in genesis and spread of muscular tension in occupational muscle pain syndromes: a hypothesis. Med Hypothesis. 1991;35:196–203
- Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms. Appl Ergon. 1987;18(3):233–237
- . Performance and muscle activity during computer mouse tasks in young and elderly adults. Euro J Appl Physiol. 2001;84(4):329–336
- . Precision of measurements of physical workload during standardised manual handling. Part I: surface electromyography of M. trapezius, M. infraspinatus and the forearm extensors. J Electromyo Kinesiol. 2004;14:443–454
- . Anatomical guide for the electromyographer: the limbs and trunk. 4th ed.. Springfield, Illinois: Charles C Thomas Publisher Ltd; 2004;
- . A comparison of symptomatic and asymptomatic office workers performing monotonous keyboard work – 1: neck and shoulder muscle recruitment patterns. Man Ther. 2005;10(4):270–280
- . A comparison of symptomatic and asymptomatic office workers performing monotonous keyboard work – 2: neck and shoulder kinematics. Man Ther. 2005;10(4):281–291
- . Examining the low, high and range measures of muscle activity amplitudes in symptomatic and asymptomatic computer users performing typing and mousing tasks. Eur J Appl Physiol. 2009;106:243–251
- Trapezius muscle rest time during standardised computer work – a comparison of female computer users with and without self-reported neck/shoulder complaints. J Electromyo Kinesiol. 2007;17(4):420–427
- . Effects of precision demands and mental pressure on muscle activation and hand forces in computer mouse tasks. Ergonomics. 2004;47(2):202–217
- . Influence of time pressure and verbal provocation on physiological and psychological reactions during work with a computer mouse. Eur J Appl Physiol. 2002;87(3):257–263
Dr. Grace Szeto works as Assistant Professor in the Department of Rehabilitation Sciences, at the Hong Kong Polytechnic University. She has worked as a physiotherapist in Hong Kong, Canada and Australia since her graduation from University of Toronto in 1982. She obtained her PhD from the School of Physiotherapy, Curtin University of Technology, Western Australia in 2003. Her PhD thesis was awarded the David Ferguson Award for the Best Post-graduate thesis by the Australian Ergonomics Society. Her research interest is on work-related musculoskeletal disorders and motor control mechanisms, using surface electromyography and motion analysis. In particular, the problems of neck and upper limb pain associated with computer use is her major focus. Grace is currently the chairperson of the OSH and Rehabilitation Specialty Group of the Hong Kong Physiotherapy Association, and has been actively promoting the concepts of good ergonomic practices related to computer use in Hong Kong.
Joseph Lin received his M.Sc. degree in Healthcare (Physiotherapy) from The Hong Kong Polytechnic University in 2009. He is currently a physiotherapist working in clinical field. His research interest is in work-related musculoskeletal disorders, biomechanics and surface EMG.
PII: S1050-6411(10)00095-7
doi:10.1016/j.jelekin.2010.06.006
© 2010 Elsevier Ltd. All rights reserved.
Volume 21, Issue 1 , Pages 59-66, February 2011
