This article was downloaded by: [Purdue University] On: 28 August 2014, At: 11:41 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Ergonomics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/terg20
Cycle to cycle variability in a repetitive upper extremity task a
b
Jimmy Tat , Michael W.R. Holmes & Peter J. Keir
a
a
Department of Kinesiology, Occupational Biomechanics Laboratory, McMaster University, Hamilton, Canada b
Faculty of Health Sciences, University of Ontario Institute of Technology, Oshawa, Canada Published online: 13 Jun 2014.
To cite this article: Jimmy Tat, Michael W.R. Holmes & Peter J. Keir (2014) Cycle to cycle variability in a repetitive upper extremity task, Ergonomics, 57:9, 1405-1415, DOI: 10.1080/00140139.2014.926396 To link to this article: http://dx.doi.org/10.1080/00140139.2014.926396
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Ergonomics, 2014 Vol. 57, No. 9, 1405–1415, http://dx.doi.org/10.1080/00140139.2014.926396
Cycle to cycle variability in a repetitive upper extremity task Jimmy Tata, Michael W.R. Holmesb and Peter J. Keira* a
Department of Kinesiology, Occupational Biomechanics Laboratory, McMaster University, Hamilton, Canada; b Faculty of Health Sciences, University of Ontario Institute of Technology, Oshawa, Canada
Downloaded by [Purdue University] at 11:41 28 August 2014
(Received 14 October 2013; accepted 13 May 2014) The purpose of this study was to examine the variability in muscle activity at rest and work during a repetitive task. A total of 20 participants performed a bimanual push task using three frequencies (4, 8, 16 pushes/min), three loads (1 kg, 2 kg, 4 kg) and two grip conditions (no grip, 30% maximum). The coefficient of variation (CoV) of muscle activity was determined for the anterior deltoid, biceps brachii, extensor digitorum and flexor digitorum superficialis. Faster push frequencies and heavier loads had lower work – rest ratio CoV and higher mean muscle activity ( p , 0.01). Sixteen pushes per minute produced the lowest CoV for the anterior deltoid ( p , 0.01), while the 1- kg load produced the lowest CoV for the extensor digitorum and flexor digitorum superficialis ( p , 0.01). Changes were driven by the rest phase rather than by the work phase, except for grip decreasing forearm muscle CoV. These findings underscore the importance of variability at rest and indicate that low variability of muscle activity is associated with ergonomic risk factors. Practitioner Summary: Decreased motor variability has been associated with pain and injury. A cyclical push task, evaluated in terms of work and rest phases, found that greater workloads increased variability primarily due to changes in the rest phase. Muscle variability, especially for the rest phase, may provide insight into injury risk. Keywords: electromyography; work– rest ratio; rest; upper extremity; motor variability
1.
Introduction
In the search to identify risk factors for occupational upper extremity injuries, many facets of the work task have been evaluated. For example, high repetition and high force, especially in combination, are known risk factors for the development of upper extremity disorders (Silverstein, Fine, and Armstrong 1986). These biomechanical measures of workplace exposure are often averaged or summated over the entire shift. To further examine components of the task that could relate to injury, cyclic workplace tasks have been evaluated and partitioned into work and rest phases (Kilbom 1994). The size of motor variability (coefficient of variation [CoV] or standard deviation [SD] of mean muscle activity), during work and rest, has been useful in elucidating conditions that may be associated with injury and pain (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Srinivasan and Mathiassen 2012; van Dieen et al. 1993). When performing repetitive cyclic tasks, such as those found in automotive manufacturing, analysing motor variability and muscular control changes between cycles has provided insight into workplace disorders. Madeleine, Mathiassen, and Arendt-Nielsen (2008) found that the magnitude of motor variability differed depending on acute and chronic pain. After experimentally inducing acute pain in healthy participants, they found increased variability in muscle activity and concluded that individuals respond to acute pain by seeking alternative motor solutions to reduce nociceptive influx. They defined motor variability as the SD of the ratio between work and rest muscle activity. With chronic pain, motor variability of muscles is diminished with exposure to a repetitive demand. This could lead to a cycle of degradation since performing a task with low motor variability continually stresses the same muscles, increasing fatigue and susceptibility to injury (Mathiassen, Moller, and Forsman 2003). Interestingly, in a study that examined butchers during a repetitive cutting task, experienced butchers had greater motor variability and a decreased tendency to report pain than inexperienced butchers (Madeleine et al. 2003a). Greater flexibility in motor patterns is likely a positive trait for performing cyclic workplace tasks and motor variability may have important implications for predicting muscle fatigue, chronic pain and injury in the workplace. Large motor variability has been linked to a reduction in muscular fatigue (Farina et al. 2008; Madeleine and Farina 2008; Madeleine, Mathiassen, and Arendt-Nielsen 2008; van Dieen et al. 1993) and shown to improve performance in a repetitive task (Farina et al. 2008). Conversely, muscle activity patterns characterised by low motor variability during work cycles are often identified as risk factors for the development of upper extremity disorders (Kilbom, Persson, and Jonsson 1986; Veiersted, Westgaard, and Andersen 1993). Pain can also induce a decrease in variability and continue to limit the
*Corresponding author. Email: [email protected] q 2014 Taylor & Francis
Downloaded by [Purdue University] at 11:41 28 August 2014
1406
J. Tat et al.
postural patterns after pain is removed (Moseley and Hodges 2006). Thus, motor variability appears to be an important characteristic of muscle activity for examining work-related musculoskeletal disorders (WMSD). Workers who perform tasks in a stereotyped manner demonstrate low variability and may increase their risk for injury as compared to individuals who make use of the redundancy in the muscular system (Mathiassen, Moller, and Forsman 2003). Elevated muscle activity has been well documented as a risk factor for WMSDs (Finsen, Christensen, and Bakke 1998; Madeleine et al. 2003b; Veiersted, Westgaard, and Andersen 1993). Muscle activity during both work and rest phases of a task is also linked to feelings of discomfort in the neck and shoulder (Madeleine et al. 2003b). For example, work cycles with short rest times often elevate resting muscle activity and can cause fatigue, pain and injury (Veiersted, Westgaard, and Andersen 1993). A number of factors can increase muscle activity in the upper extremity. Multitasking and high task complexity can increase concurrent mental effort and lead to increases in muscle activity (Au and Keir 2007; Laursen et al. 2002; MacDonnel and Keir 2005). The coupling of grip force with dynamic shoulder action can alter shoulder activity by decreasing motor variability (Hodder and Keir 2012). Though motor variability and muscle activity have been investigated previously, the inclusion of additional task demands could reduce motor variability due to imposed constraints, as with workers required to grasp an object while performing a push or pull task (Di Domizio and Keir 2010). This study used a cyclic push task, with and without a grip task, to elucidate motor control and variability strategies to further our understanding of muscular injuries of the shoulder and forearm. The purpose of this study was to investigate the effects of load and frequency on muscle activity and the relative size of variability using the CoV during work and rest phases of the movement for a repetitive pushing task with the constraint of a handgrip. 2. Methods 2.1 Participants A total of 10 men and 9 women participated in this study to represent a mixed working population. Mean age, mass and height can be found in Table 1. All participants were verbally screened for right-hand dominance and having no history of upper extremity injury or pain in the past year. The study was approved by the McMaster University Research Ethics Board and all participants provided written informed consent. 2.2 Experimental protocol Methods have been reported in detail previously (Keir and Brown 2012) but will briefly be described here. A bimanual pushing task was created using two parallel low friction tracks with handles mounted vertically on platforms that could slide independently (Figure 1). The tracks were 0.25-m long and the handles were 0.38-m apart. The position of each handle was monitored via linear potentiometers affixed to the platform of each handle. The handle mounted to the right track was instrumented with a grip dynamometer (MIE Medical Research Ltd. Leeds, UK; mass ¼ 0.45 kg; grip span ¼ 5.25 cm), while a matching wooden post was on the left side. The track was fixed to an adjustable table set at elbow height for each participant to allow a start position of 908 elbow flexion and forearms parallel to the table top. Participants performed 18 different bilateral pushing tasks, each 120 s in duration. The 18 tasks included all combinations of three push frequencies (4, 8, 16 pushes/min), three force levels (1 kg, 2 kg, 4 kg) and two grip conditions (no required grip, 30% maximum). Trials were randomised for each participant and 120 s of rest was given between each trial. Push loads were chosen to replicate those found in the automotive assembly plant, but frequencies were chosen to include common definitions of low and high repetition (Moore and Garg 1995). The participant was aware of the load, frequency and grip requirements prior to the start of each trial. A metronome indicated the start of each push cycle through headphones. The participant extended both arms to a comfortable distance and returned to the starting position at a self-selected speed, such that they would be ready for indication of the next cycle. When gripping was required, participants were instructed to grip to the target level first, prior to beginning the push cycle. A target grip force was provided by visual feedback that displayed the target force of 30 ^ 1.5% maximal voluntary grip force (MVG). Table 1.
Mean participant characteristics (^SD) including age, height, mass and maximum grip force.
Number of participants Age (years) Height (cm) Mass (kg) Maximum grip (N)
Males
Females
Combined
10 27.7 ^ 6.0 179.5 ^ 6.8 84.4 ^ 11.4 629.8 ^ 95.4
9 23.7 ^ 3.1 165.9 ^ 8.4 64.4 ^ 9.2 333.5 ^ 50.2
19 25.8 ^ 5.1 172.8 ^ 9.9 75.0 ^ 14.4 481.7 ^ 173.3
Ergonomics
Downloaded by [Purdue University] at 11:41 28 August 2014
Figure 1.
1407
Start (left) and finish (right) postures using dual pushing track with grip dynamometer on right-hand side (in foreground).
To determine 30% MVG, the MVG was determined using the grip dynamometer in the start position. Participants were instructed to slowly increase grip force to maximum, hold for 2 s and relax over a 10-s trial. MVG was calculated over a 250- ms window centred about the peak force. Two maximal trials were performed and if the MVG were within 5%, the mean of the two trials was used to determine the 30% grip condition. If the two trials were not within 5%, additional trials were completed until the two highest calculated forces were within 5% (occasionally three trials were required, one participant required four trials). Surface EMG was collected from eight muscles of the right upper extremity (AMT-8, Bortec Biomedical, Calgary, AB, Canada; bandwidth 10-1000 Hz, CMRR . 115 dB at 60 Hz, input impedance < 10GV). The muscles that were monitored included: posterior deltoid, anterior deltoid (AD), biceps brachii (BB), triceps brachii, flexor digitorum superficialis (FDS), flexor carpi radialis, extensor digitorum (ED) and extensor carpi radialis. The muscles were identified using anatomical landmarks and muscle specific isometric contractions (Perotto 1994). For the four muscles of primary interest (AD, BB, FDS and ED), the electrode positions included: below the anterior margin of the acromion for AD; BB was at mid-humerus level; FDS was distal to the medial epicondyle with the hand fully supinated; and ED was distal the lateral epicondyle with the hand fully pronated. Note that the current study focussed on only four of the eight collected muscles to simplify the document and highlight the muscles previously shown to be of sufficient activation (. 5% maximum) and interest (Keir and Brown 2012). Each electrode site was prepared by shaving and scrubbing with isopropyl alcohol before attaching disposable Ag-AgCl surface electrodes (Meditrace, Kendall, MA, USA) over the muscle belly parallel to muscle fibre orientation with an inter-electrode distance of 2.5 cm. Next, the participant performed a series of muscle specific maximal isometric contractions for 10 s each. Maximal voluntary excitation (MVE) for each muscle followed a similar protocol to our previous studies (Au and Keir 2007; Di Domizio and Keir 2010). For the shoulder muscles, participants performed resisted arm raises in flexion and for biceps, participants performed resisted elbow flexion in supination. For the forearm muscles, the tasks included resisted wrist flexion and extension. Grip force, linear potentiometer and EMG data were sampled at 2048 Hz and stored for analysis using a custom programme (Labview v. 7.1, National Instruments, TX, US).
2.3
Data analysis
For the current study, four muscles were selected: AD, BB, ED and FDS. Raw EMG was full wave rectified and dual Butterworth filtered using a 3 Hz low pass filter. All EMG signals were normalised to the muscle specific MVE that was calculated as the mean amplitude over a 500- ms window about the largest peak of each muscle. Grip force was filtered at 10 Hz and normalised to MVG. The linear potentiometer data were used to determine the start, peak and end position of each individual cycle and to determine cycle time in seconds. Each cycle was divided into work and rest phases of the movement (Figure 2). The phases were defined as (1) work – including push and return phases and (2) rest – time between work phases. Average EMG (AEMG), grip force and timing information were determined for each phase using a custom programme (MatLab v.7.6, The MathWorks, Natick, MA, USA). AEMG was used to determine the CoV, defined as the SD divided by the mean, and represented the relative size of EMG variability for each work phase (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen 2008). We used CoV because it is self-scaling and can assess differences in variability between muscle groups. CoV has previously been used to describe motor variability in relation to
Downloaded by [Purdue University] at 11:41 28 August 2014
1408
Figure 2.
J. Tat et al.
Sample potentiometer data illustrating individual components of the movement that make up each cycle.
work experience and discomfort (Madeleine and Madsen 2009). The result of the work phase AEMG divided by the rest phase EMG represented the work– rest (WR) ratio. The WR ratio CoV was also calculated for AEMG (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen 2008). 2.4 Statistical analysis For each cycle, AEMG was determined for work and rest phases, and the WR ratio calculated. The mean and CoV of each parameter were calculated for each trial. A three-way repeated measures analysis of variance was used to determine the effect of load, frequency and grip on grip force, cycle times, AEMG and CoV for each muscle at work and rest, and WR ratio. Significant main effects and interactions were further analysed using Tukey’s honestly significant difference test (SPSS v 13.0, SPSS Inc, Chicago, IL). Significance was set at a ¼ 0.01 and all values are reported as means ^ SD. 3.
Results
3.1 Grip force For the rest phase, a grip £ frequency interaction (F2,36 ¼ 32.38, p , 0.01) affected grip force. Faster push frequencies increased grip force during all gripping trials ( p , 0.01); however, the effect was lost without grip (Figure 3). A main effect
Figure 3. Grip force at rest. During gripping trials, the grip force was significantly different for all push/pull frequencies. Error bars represent one standard error of mean. *Denotes significance ( p , 0.01).
Ergonomics
1409
of grip was found in the work phase (F1,18 ¼ 229.61, p , 0.01), where the mean grip force was consistent but lower than the target 30% MVG at 21.4 ^ 5.8% of MVG, while the no grip condition produced a grip force of 1.1 ^ 0.8% of MVG. 3.2
Cycle times
Downloaded by [Purdue University] at 11:41 28 August 2014
Frequency resulted in faster work times (F2,36 ¼ 11.98, p , 0.01) with the greatest difference in the 16/min condition compared to the 4/min and 8/min push frequencies ( p , 0.01). The work times for each frequency were 2.18 ^ 0.45 s (4/min), 2.13 ^ 0.51 s (8/min) and 1.90 ^ 0.27 s (16/min). The remainder of the cycle was allocated to rest and consisted of rest times 10.23 ^ 0.49 s (4/min), 5.03 ^ 0.47 s (8/min) and 1.80 ^ 0.27 s (16/min). 3.3 Muscle activity For AEMG, a frequency £ load interaction was found in the rest phase (AD, F4,72 ¼ 4.27, p , 0.01 and BB, F4,72 ¼ 2.75, p ¼ 0.035) but not during the work phase. A frequency £ load interaction was also found for the AD WR ratio (F4,72 ¼ 17.85, p , 0.01) and BB WR ratio (F4,72 ¼ 17.20, p , 0.01). For both AD and BB, a faster push frequency increased AEMG at rest and decreased WR ratio in all push/pull loads, with the greatest difference in the fastest frequency (Table 2). For the forearm muscle activity, there was a significant frequency £ grip interaction for ED (F2,36 ¼ 21.97, p , 0.01) and FDS (F2,36 ¼ 24.83, p , 0.01) in the rest phase but not in the work phase. AEMG at rest had the greatest increase with higher push frequencies and with grip (Table 2). There was also a frequency £ grip interaction for WR ratio for these muscles (ED, F2,36 ¼ 26.77, p , 0.01; FDS, F2,36 ¼ 29.41, p , 0.01). Likewise, WR ratio decreased with higher push frequencies and with grip (Table 3). In the work phase, grip independently increased muscle activity in ED by 13.1 ^ 1.3% MVE (F1,18 ¼ 50.39, p , 0.01) and FDS by 15.9 ^ 2.0% MVE (F1,18 ¼ 67.45, p , 0.01) over trials without grip. For FDS activity, a frequency effect was found for the work phase (F2,36 ¼ 6.06, p , 0.01) that was primarily seen between the 16/min (11.7% MVE) and 4/min (10.5% MVE) push frequencies. 3.4
EMG variability
AD and BB CoV were significantly affected by task frequency (Figure 4). For AD, faster push frequencies increased the AEMG CoV in the rest phase (F2,36 ¼ 10.55, p , 0.01) but not during the work phase. The WR ratio CoV also decreased Table 2.
Muscle activity in % MVE (mean ^ SD) for rest and work phases. 4/min
Rest
Load
Muscle
1 kg
AD BB ED FDS AD BB ED FDS AD BB ED FDS AD BB ED FDS AD BB ED FDS AD BB ED FDS
2 kg
4 kg
Work
1 kg
2 kg
4 kg
8/min
16/min
No grip
Grip
No grip
Grip
No grip
Grip
1.3 ^ 1.4 0.8 ^ 0.5 2.1 ^ 1.8 1.3 ^ 0.8 1.4 ^ 1.7 0.9 ^ 0.6 3.1 ^ 3.8 1.5 ^ 1.2 1.8 ^ 2.1 1.1 ^ 1.2 3.5 ^ 3.0 1.7 ^ 1.0 11.2 ^ 6.6 1.9 ^ 1.3 4.4 ^ 4.1 2.2 ^ 1.4 14.2 ^ 8.0 2.4 ^ 1.8 5.2 ^ 4.8 2.7 ^ 2.0 24.8 ^ 13.9 5.7 ^ 4.3 7.9 ^ 5.9 3.6 ^ 2.3
1.3 ^ 1.4 0.9 ^ 0.5 5.0 ^ 2.4 3.7 ^ 2.0 1.4 ^ 1.4 0.9 ^ 0.6 5.4 ^ 2.3 3.9 ^ 2.0 1.9 ^ 2.1 1.2 ^ 0.7 6.7 ^ 4.2 3.9 ^ 1.9 9.6 ^ 6.3 2.6 ^ 1.9 16.7 ^ 9.3 17.5 ^ 9.4 14.2 ^ 10.3 3.5 ^ 2.9 18.2 ^ 8.6 18.4 ^ 8.2 25.5 ^ 15.6 7.6 ^ 6.3 21.5 ^ 10.0 18.4 ^ 8.7
1.5 ^ 1.9 0.8 ^ 0.5 2.6 ^ 3.1 1.5 ^ 1.3 1.8 ^ 2.1 0.9 ^ 0.5 3.1 ^ 3.3 1.7 ^ 1.0 2.4 ^ 2.8 1.3 ^ 1.0 4.7 ^ 4.7 2.1 ^ 1.4 9.9 ^ 5.9 1.6 ^ 1.1 4.5 ^ 3.4 2.6 ^ 2.2 13.6 ^ 8.6 1.9 ^ 1.4 6.0 ^ 5.9 3.0 ^ 2.2 25.0 ^ 15.7 6.0 ^ 6.0 9.7 ^ 6.9 3.9 ^ 2.7
1.5 ^ 1.8 1.0 ^ 0.7 9.0 ^ 5.2 5.8 ^ 4.1 1.9 ^ 2.0 1.3 ^ 0.9 9.5 ^ 3.9 6.0 ^ 4.9 2.5 ^ 2.9 1.7 ^ 1.1 10.1 ^ 3.9 7.0 ^ 5.1 9.0 ^ 6.1 2.3 ^ 1.8 17.4 ^ 9.6 19.2 ^ 9.1 14.5 ^ 9.6 3.8 ^ 2.5 18.0 ^ 9.5 17.6 ^ 7.9 25.9 ^ 16.4 7.5 ^ 5.0 22.3 ^ 9.4 19.4 ^ 8.9
2.2 ^ 2.7 1.1 ^ 0.7 3.7 ^ 2.2 1.6 ^ 1.1 2.9 ^ 3.0 1.3 ^ 0.9 4.0 ^ 2.6 2.0 ^ 1.5 4.5 ^ 5.2 2.1 ^ 1.4 4.0 ^ 1.8 2.6 ^ 1.9 9.8 ^ 6.1 1.8 ^ 1.4 4.5 ^ 4.1 2.7 ^ 1.8 13.6 ^ 8.5 2.2 ^ 1.6 5.3 ^ 4.3 3.1 ^ 2.1 26.1 ^ 14.3 5.1 ^ 4.4 8.9 ^ 5.9 4.1 ^ 2.6
3.2 ^ 5.0 2.0 ^ 1.5 2.0 ^ 1.9 10.3 ^ 7.8 3.7 ^ 4.9 2.0 ^ 1.1 2.3 ^ 2.1 10.3 ^ 7.1 4.5 ^ 4.7 2.8 ^ 1.6 4.5 ^ 4.7 10.6 ^ 6.8 13.2 ^ 10.8 4.0 ^ 3.7 19.2 ^ 8.4 21.2 ^ 9.9 15.1 ^ 11.0 3.4 ^ 1.8 19.4 ^ 8.8 18.8 ^ 8.2 25.6 ^ 14.1 6.8 ^ 5.1 21.4 ^ 9.5 20.1 ^ 9.1
1410 Table 3.
J. Tat et al. Work– rest ratio (mean ^ SD) for all conditions. 4/min
WR ratio
Load
Muscle
1 kg
AD BB ED FDS AD BB ED FDS AD BB ED FDS
2 kg
Downloaded by [Purdue University] at 11:41 28 August 2014
4 kg
8/min
16/min
No grip
Grip
No grip
Grip
No grip
Grip
11.4 ^ 6.6 2.6 ^ 1.2 2.4 ^ 0.9 1.7 ^ 0.5 14.0 ^ 6.8 3.2 ^ 2.0 2.8 ^ 1.4 1.9 ^ 0.8 21.3 ^ 11.2 6.4 ^ 3.5 3.7 ^ 1.7 2.4 ^ 1.6
9.5 ^ 5.2 3.0 ^ 1.6 5.2 ^ 2.5 5.5 ^ 3.0 13.3 ^ 7.8 4.3 ^ 3.4 5.2 ^ 2.0 5.4 ^ 2.6 19.6 ^ 9.7 6.9 ^ 4.8 5.8 ^ 2.9 5.0 ^ 2.0
9.3 ^ 4.5 2.1 ^ 0.9 2.4 ^ 1.3 1.8 ^ 0.6 10.6 ^ 5.1 2.6 ^ 1.5 2.3 ^ 1.5 1.8 ^ 0.7 14.4 ^ 6.1 4.7 ^ 2.6 2.9 ^ 0.6 1.8 ^ 0.5
8.0 ^ 4.2 3.0 ^ 1.6 3.8 ^ 1.7 4.2 ^ 2.1 10.1 ^ 5.2 3.2 ^ 2.0 3.5 ^ 1.5 3.7 ^ 1.7 14.3 ^ 5.9 4.9 ^ 2.2 3.9 ^ 1.6 3.4 ^ 1.5
6.1 ^ 3.2 1.7 ^ 1.0 1.9 ^ 0.7 1.8 ^ 0.5 6.4 ^ 2.7 1.9 ^ 1.2 1.9 ^ 0.7 1.7 ^ 0.6 8.6 ^ 3.6 2.6 ^ 1.4 2.1 ^ 0.6 1.7 ^ 0.6
6.1 ^ 3.2 2.3 ^ 1.3 2.5 ^ 1.3 2.8 ^ 1.4 6.3 ^ 2.7 1.8 ^ 0.9 2.2 ^ 1.1 2.5 ^ 1.5 7.9 ^ 3.9 2.6 ^ 1.4 2.2 ^ 0.9 2.3 ^ 1.2
with task frequency (F2,36 ¼ 13.95, p , 0.01). For BB, there was a frequency effect on AEMG CoV at rest (F2,36 ¼ 20.52, p , 0.01) and for WR ratio CoV (F2,36 ¼ 14.07, p , 0.01). The rest BB CoV decreased and the WR ratio CoV increased, both linearly with frequency ( p , 0.01). No significant changes were found for work variability in either muscle.
Figure 4. Frequency effects for AD muscle variability during (a) work and rest phases, (b) WR ratio and BB muscle variability at (c) work and rest phases, (d) WR ratio. Error bars represent one standard error of mean. *Denotes conditions in AD that are significantly different from 16/min push frequency ( p , 0.01). **Denotes BB significant differences with 16/min frequency ( p , 0.01).
Downloaded by [Purdue University] at 11:41 28 August 2014
Ergonomics
1411
Figure 5. Push/Pull load effects in the no grip condition for ED muscle variability at (a) work and rest, (b) WR Ratio and FDS muscle variability at (c) work and rest phases, (d) WR ratio. Error bars represent one standard error of mean. *Significantly different from 4- kg load in FDS ( p , 0.01) and **significantly different from 4- kg load in ED ( p , 0.01).
Forearm muscle EMG variability (ED and FDS) was significantly affected by the work conditions in WR ratio. Post hoc comparisons showed that this was due to changes in the rest phase and not in the work phase (Figure 5). For ED, there was a load £ grip interaction on CoV at rest (F2,36 ¼ 7.89, p , 0.01) and WR ratio CoV (F2,36 ¼ 5.49, p , 0.01) which increased ED CoV at rest and significantly decreased WR ratio CoV for the 4- kg load, no grip condition ( p , 0.01). This load £ grip interaction was not found at work; however, there was a grip main effect (F1,18 ¼ 51.94, p , 0.01). The addition of 30% MVG decreased ED CoV at work by an average of 0.13 ^ 0.05 across all conditions. For FDS, there was also a significant load £ grip interaction on CoV at rest (F2,36 ¼ 14.92, p , 0.01) and WR ratio CoV (F2,36 ¼ 6.20, p , 0.01). Similarly, the load £ grip interaction was seen in the no grip condition, which increased FDS CoV at rest and decreased FDS WR ratio CoV with the largest differences being in the 4- kg load. Frequency also affected EMG variability in FDS at rest and WR ratio. Faster push frequencies resulted in a linear increase in FDS CoV in the rest phase (F2,36 ¼ 19.63, p , 0.01). For FDS, the WR ratio CoV had a frequency £ load interaction (F4,72 ¼ 4.48, p , 0.01) that was mainly seen in the 2- kg load, where 4/min (CoV 2.07) was significantly larger than 8/min (1.13) and 16/min (1.08). There was also a frequency £ grip interaction (F2,36 ¼ 5.69, p , 0.01) such that the effect of frequency was amplified in the trials without grip. There were no significant effects in COV at work. 4.
Discussion
This study examined cycle-to-cycle EMG variability (CoV in AEMG) in selected upper extremity muscles during a repetitive pushing task. There were no significant effects of load and frequency on EMG variability during work phases of the task; however, WR ratio variability was predominantly influenced by changes in variability in the rest phase. For AD and BB, a higher push frequency increased CoV at rest and decreased WR ratio CoV. For the forearm muscles (ED and
J. Tat et al.
Downloaded by [Purdue University] at 11:41 28 August 2014
1412
Figure 6. EMG data replotted versus workload illustrating frequency, load and grip effects in FDS at (a) rest, (b) work, (c) WR ratio. Error bars represent one standard error of mean.
FDS), higher loads increased the CoV at rest and decreased the WR ratio CoV. In this paradigm, greater resting variability suggests the muscle may be unable to relax between efforts that could result in the overuse of the same motor units and their muscle fibres. In the workplace, this could lead to muscular fatigue and injury to the muscle. This is supported by epidemiological studies, which indicate workers who perform repetitive tasks with reduced pauses are more likely to develop injury (Veiersted, Westgaard, and Andersen 1990). Our findings highlight motor variability changes in the rest
1413
Downloaded by [Purdue University] at 11:41 28 August 2014
Ergonomics
Figure 7. Variability (CoV) presented versus workload illustrating frequency, load and grip effects in FDS at (a) rest, (b) work, (c) WR ratio. Error bars represent standard error.
phase but not in the work phase of a cyclic task and underscore the importance of muscle activity at rest. Muscle activity at rest might provide a more sensitive measure of physical exposure to frequency and loads, which is contrary to the traditional emphasis on work components (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen 2008; Mathiassen, Moller, and Forsman 2003; Veiersted, Westgaard, and Andersen 1993). EMG has been used to identify potential warning signs for repetitive workplace tasks that pose risk to the development of WMSD (Madeleine et al. 2003b; Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen
Downloaded by [Purdue University] at 11:41 28 August 2014
1414
J. Tat et al.
2008; Sandsjo et al. 2000; Veiersted 1994). In rest and work phases, there was an increase in activity that corresponded with larger loads and faster push frequencies (Table 2); increased AEMG during a work task has been linked to pain (Sandsjo et al. 2000; Veiersted 1994). Chronic pain in the upper extremity has also been characterised by a decrease in WR ratio EMG variability (Madeleine, Voigt, and Mathiassen 2008). In the present study, we have shown a change in the direction (decrease WR ratio CoV) and magnitude of variability (smaller WR ratio CoV) with exposure to more demanding workloads. The smallest CoV for the AD and BB was found in the highest frequency (16/min), and the smallest forearm CoV was found with the largest load (4 kg). Madeleine et al. (2003a) demonstrated that novice workers had lower variability and a lower WR EMG ratio in the shoulder muscles (including AD) when performing a repetitive task. A strategy that involves reduced variability may help explain the tendency for novice workers to experience fatigue, pain or injury as it has been described as less efficient due to its failure to take advantage of the redundant upper extremity muscular system (Mathiassen, Moller, and Forsman 2003). Our findings in AEMG and EMG variability support the hypothesis that increased frequency and load may lead to muscle fatigue and chronic pain (Farina et al. 2008; Madeleine, Mathiassen, and ArendtNielsen 2008). Simultaneous gripping and pushing are common in the workplace and grip forces have been found to alter muscle activity patterns in the shoulder (Au and Keir 2007; Hodder and Keir 2012; MacDonell and Keir 2005; Smets, Potvin, and Keir 2009). The effect of grip force in this study had a significant impact on the work components of the repetitive task. The addition of a 30% MVG increased muscle activity in the forearm (ED by 13.1 ^ 1.3%; FDS by 15.9 ^ 2.0%) and significantly decreased EMG variability (ED CoV by 0.13 ^ 0.05; FDS CoV by 0.07 ^ 0.06), which are both risk factors for the development of injury and chronic pain (Madeleine, Mathiassen, and Arendt-Nielsen 2008). It is also noteworthy that participants increased resting grip force with faster push frequencies, suggesting that participants had difficulty returning to resting conditions between pushes. The mean work grip force was substantially below the target of 30% because we defined the work phase to include both the push (28.1 ^ 4.1% MVG) and return (15.5 ^ 10.4% MVG) components. Participants were not asked to maintain the target grip force during the return. Keir and Brown (2012) evaluated the 120-s trial as a whole and noted that the complex interactions between load and frequency might be more easily explained by workload, the product of load and frequency. We have replotted the data by categorising the 18 work conditions into five workloads (4, 8, 16, 32, 64 kg/min) (Figures 6 and 7). For example, conditions in the 8 kg/min workload included the conditions low load –med frequency (1 kg £ 8/min) and med load – low frequency (2 kg £ 4/min). For FDS, the relationship between muscle activity and workload is driven by the AEMG in the rest phase (Figure 6). With gripping, the linear decline of the WR ratio is mirrored by the linear increase in rest AEMG but the relationship is static across workload for gripping. Workload variability (CoV) provides a different perspective as the changes in the rest phase dictate the relationship when a grip was not required (Figure 7). We can see that the variability with a grip appears relatively constant across workload but the variability in rest EMG increases with workload. This indicates the workload method might be helpful to simplify the interpretation of EMG results to identify risk in cyclical workplace tasks. There are a few limitations to this study. Participants were healthy young adults who were not employed in manual labour. Madeleine, Mathiassen, and Arendt-Nielsen (2008) showed that experienced workers have greater motor pattern flexibility. Thus, generalising our findings to more skilled and advanced assembly workers may be limited. While we relate our findings to the work of Madeleine, Mathiassen, and Arendt-Nielsen (2008) who related skill and pain to variability, we had healthy participants without pain. Thus, this study is limited to the effects of load and frequency on muscle activity. However, it is unclear whether the large mean muscle activity at rest and greater variability in muscle activity at rest were a result of the task demands or an anticipatory effect due to the subject knowing the condition. Faster push frequencies may result in early muscle activation for the next cycle that would increase muscle activity and/or variability. However, a metronome was used to indicate the start of each cycle, and we noted with work times that subjects voluntarily moved faster for faster frequencies to meet the task demand. Nonetheless, an increase in muscle activity at rest is a relevant risk factor for developing pain (Sandsjo et al. 2000; Veiersted 1994). 5. Conclusions This study found that the rest phase is important in the assessment of muscle activity in a repetitive cyclic task. Frequency and push load significantly altered the muscle activity during the rest phase (but not the work phase), which drove changes in WR ratio. The relative size of variability in muscle activity and WR ratio was reduced with higher frequency and load, suggesting a potential role in the development of upper extremity WMSD. Further research is needed to examine whether the effects of frequency, load and grip on motor variability contribute to acute or chronic pain during both rest and work phases. Understanding the relationship between these workload variables and pain can lead to a better comprehension of injury mechanisms.
Ergonomics
1415
Funding This study was funded by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada [#217382].
Downloaded by [Purdue University] at 11:41 28 August 2014
References Au, A. K., and P. J. Keir. 2007. “Interfering Effects of Multitasking on Muscle Activity in the Upper Extremity.” Journal of Electromyography & Kinesiology 17 (5): 578– 586. Di Domizio, J., and P. J. Keir. 2010. “The Effects of Posture and Force Coupling on Upper Extremity Muscle Activity.” Ergonomics 53 (3): 336– 343. Farina, D., F. Leclerc, L. Arendt-Nielsen, O. Butteli, and P. Madeleine. 2008. “The Change in Spatial Distribution of Upper Trapezius Muscle Activity is Correlated to Contraction Duration.” Journal of Electromyography & Kinesiology 18 (1): 16 – 25. Finsen, L., H. Christensen, and M. Bakke. 1998. “Musculoskeletal Disorders among Dentists and Variation in Dental Work.” Applied Ergonomics 29 (2): 119– 125. Hodder, J. N., and P. J. Keir. 2012. “Targeted Gripping Reduces Shoulder Muscle Activity and Variability.” Journal of Electromyography & Kinesiology 22 (2): 186– 190. Keir, P. J., and M. M. Brown. 2012. “Force, Frequency and Gripping Alter Upper Extremity Muscle Activity during a Cyclic Push Task.” Ergonomics 55 (7): 813– 824. Kilbom, A. 1994. “Repetitive Work of the Upper Extremity: Part II – The Scientific Basis for the Guide.” International Journal of Industrial Ergonomics 14 (2): 59 –86. Kilbom, A., J. Persson, and B. G. Jonsson. 1986. “Disorders of the Cervicobrachial Region among Female Workers in Electronics Industry.” International Journal of Industrial Ergonomics 1 (1): 37 –47. Laursen, B., B. R. Jensen, A. H. Garde, and A. H. Jorgensen. 2002. “Effect of Mental and Physical Demands on Muscular Activity during the Use of a Computer Mouse and Keyboard.” Scandinavian Journal of Work Environment & Health 28 (4): 215– 221. MacDonell, C. W., and P. J. Keir. 2005. “Interfering Effects of the Task Demands of Grip Force and Mental Processing on Isometric Shoulder Strength and Muscle Activity.” Ergonomics 48 (15): 1749– 1769. Madeleine, P., and D. Farina. 2008. “Time to Task Failure in Shoulder Elevation is Associated to Increase in Amplitude and to Spatial Heterogeneity of Upper Trapezius Mechanomyographic Signals.” European Journal of Applied Physiology 102 (3): 325– 333. Madeleine, P., B. Lundager, M. Voigt, and L. Arendt-Nielsen. 2003a. “Standardized Low-Load Repetitive Work: Evidence of Different Motor Control Strategies between Experienced Workers and a Reference Group.” Applied Erognomics 34 (6): 533– 542. Madeleine, P., B. Lundager, M. Voigt, and L. Arendt-Nielsen. 2003b. “The Effects of Neck-Shoulder Pain Development on SensoryMotor Interactions among Female Workers in Poultry and Fish Industries. A Prospective Study.” International Archives of Occupational and Environmental Health 76 (1): 39 – 49. Madeleine, P., and T. M. T. Madsen. 2009. “Changes in the Amount and Structure of Motor Variability during a Deboning Process are Associated with Work Experience and Neck-Shoulder Discomfort.” Applied Ergonomics 40 (5): 887– 894. Madeleine, P., S. E. Mathiassen, and L. Arendt-Nielsen. 2008. “Changes in the Degree of Motor Variability Associated with Experimental and Chronic Pain Neck-Shoulder Pain during Standardised Repetitive Arm Movement.” Experimental Brain Research 185 (4): 689– 698. Madeleine, P., M. Voigt, and S. E. Mathiassen. 2008. “The Size of Cycle-to-Cycle Variability in Biomechanical Exposure among Butchers Performing a Standardised Cutting Task.” Ergonomics 51 (7): 1078– 1095. Mathiassen, S. E., T. Moller, and M. Forsman. 2003. “Variability in Mechanical Exposure within and between Individuals Performing a Highly Constrained Industrial Work Task.” Ergonomics 46 (8): 800– 824. Moore, S. J., and A. Garg. 1995. “The Strain Index: A Proposed Method to Analyze Jobs for Risk of Distal Upper Extremity Disorders.” American Industrial Hygiene Association Journal 56 (5): 443–458. Moseley, G. L., and P. W. Hodges. 2006. “Reduced Variability of Postural Strategy Prevents Normalization of Motor Changes Induced by Back Pain: A Risk Factor for Chronic Trouble?” Behavioural Neuroscience 120 (2): 474– 476. Perotto, A. 1994. Anatomical Guide for the Electromyographer: The Limbs and the Trunk. Springfield, IL: CC Thomas. Sandsjo, L., B. Melin, D. Rissen, D. Ingela, and U. Lundberg. 2000. “Trapezius Muscle Activity, Neck and Shoulder Pain, and Subjective Experiences during Monotonous Work in Women.” European Journal of Applied Physiology 83 (2): 235– 238. Silverstein, B. A., L. J. Fine, and T. J. Armstrong. 1986. “Hand Wrist Cumulative Trauma Disorders in Industry.” British Journal of Industrial Medicine 43 (11): 779– 784. Smets, M. P. H., J. R. Potvin, and P. J. Keir. 2009. “The Effect of Constrained Hand Gripping on Arm Strength and Muscle Activation of the Upper Extremities.” Ergonomics 52 (9): 1144– 1152. Srinivasan, D., and S. E. Mathiassen. 2012. “Motor Variability in Occupational Health and Performance.” Clinical Biomechanics 27 (10): 979– 993. van Dieen, J. H., H. H. E. O. Vrielink, A. F. Housheer, F. B. J. Lotters, and H. M. Toussaint. 1993. “Trunk Extensor Endurance and Its Relationship to Electromyogram.” European Journal of Applied Physiology 66 (5): 388– 396. Veiersted, K. B. 1994. “Sustained Muscle Tension as a Risk Factor for Trapezius Myalgia.” International Journal of Industrial Ergonomics 14 (4): 333– 339. Veiersted, K. B., R. H. Westgaard, and P. Andersen. 1990. “Pattern of Muscle Activity during Stereotyped Work and Its Relation to Muscle Pain.” International Archives of Occupational and Environmental Health 62 (1): 31 – 41. Veiersted, K. B., R. H. Westgaard, and P. Andersen. 1993. “Electromyographic Evaluation of Muscular Work Pattern as a Predictor of Trapezius Myalgia.” Scandinavian Journal of Work Environment & Health 19 (4): 284– 290.
Ergonomics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/terg20
Cycle to cycle variability in a repetitive upper extremity task a
b
Jimmy Tat , Michael W.R. Holmes & Peter J. Keir
a
a
Department of Kinesiology, Occupational Biomechanics Laboratory, McMaster University, Hamilton, Canada b
Faculty of Health Sciences, University of Ontario Institute of Technology, Oshawa, Canada Published online: 13 Jun 2014.
To cite this article: Jimmy Tat, Michael W.R. Holmes & Peter J. Keir (2014) Cycle to cycle variability in a repetitive upper extremity task, Ergonomics, 57:9, 1405-1415, DOI: 10.1080/00140139.2014.926396 To link to this article: http://dx.doi.org/10.1080/00140139.2014.926396
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Ergonomics, 2014 Vol. 57, No. 9, 1405–1415, http://dx.doi.org/10.1080/00140139.2014.926396
Cycle to cycle variability in a repetitive upper extremity task Jimmy Tata, Michael W.R. Holmesb and Peter J. Keira* a
Department of Kinesiology, Occupational Biomechanics Laboratory, McMaster University, Hamilton, Canada; b Faculty of Health Sciences, University of Ontario Institute of Technology, Oshawa, Canada
Downloaded by [Purdue University] at 11:41 28 August 2014
(Received 14 October 2013; accepted 13 May 2014) The purpose of this study was to examine the variability in muscle activity at rest and work during a repetitive task. A total of 20 participants performed a bimanual push task using three frequencies (4, 8, 16 pushes/min), three loads (1 kg, 2 kg, 4 kg) and two grip conditions (no grip, 30% maximum). The coefficient of variation (CoV) of muscle activity was determined for the anterior deltoid, biceps brachii, extensor digitorum and flexor digitorum superficialis. Faster push frequencies and heavier loads had lower work – rest ratio CoV and higher mean muscle activity ( p , 0.01). Sixteen pushes per minute produced the lowest CoV for the anterior deltoid ( p , 0.01), while the 1- kg load produced the lowest CoV for the extensor digitorum and flexor digitorum superficialis ( p , 0.01). Changes were driven by the rest phase rather than by the work phase, except for grip decreasing forearm muscle CoV. These findings underscore the importance of variability at rest and indicate that low variability of muscle activity is associated with ergonomic risk factors. Practitioner Summary: Decreased motor variability has been associated with pain and injury. A cyclical push task, evaluated in terms of work and rest phases, found that greater workloads increased variability primarily due to changes in the rest phase. Muscle variability, especially for the rest phase, may provide insight into injury risk. Keywords: electromyography; work– rest ratio; rest; upper extremity; motor variability
1.
Introduction
In the search to identify risk factors for occupational upper extremity injuries, many facets of the work task have been evaluated. For example, high repetition and high force, especially in combination, are known risk factors for the development of upper extremity disorders (Silverstein, Fine, and Armstrong 1986). These biomechanical measures of workplace exposure are often averaged or summated over the entire shift. To further examine components of the task that could relate to injury, cyclic workplace tasks have been evaluated and partitioned into work and rest phases (Kilbom 1994). The size of motor variability (coefficient of variation [CoV] or standard deviation [SD] of mean muscle activity), during work and rest, has been useful in elucidating conditions that may be associated with injury and pain (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Srinivasan and Mathiassen 2012; van Dieen et al. 1993). When performing repetitive cyclic tasks, such as those found in automotive manufacturing, analysing motor variability and muscular control changes between cycles has provided insight into workplace disorders. Madeleine, Mathiassen, and Arendt-Nielsen (2008) found that the magnitude of motor variability differed depending on acute and chronic pain. After experimentally inducing acute pain in healthy participants, they found increased variability in muscle activity and concluded that individuals respond to acute pain by seeking alternative motor solutions to reduce nociceptive influx. They defined motor variability as the SD of the ratio between work and rest muscle activity. With chronic pain, motor variability of muscles is diminished with exposure to a repetitive demand. This could lead to a cycle of degradation since performing a task with low motor variability continually stresses the same muscles, increasing fatigue and susceptibility to injury (Mathiassen, Moller, and Forsman 2003). Interestingly, in a study that examined butchers during a repetitive cutting task, experienced butchers had greater motor variability and a decreased tendency to report pain than inexperienced butchers (Madeleine et al. 2003a). Greater flexibility in motor patterns is likely a positive trait for performing cyclic workplace tasks and motor variability may have important implications for predicting muscle fatigue, chronic pain and injury in the workplace. Large motor variability has been linked to a reduction in muscular fatigue (Farina et al. 2008; Madeleine and Farina 2008; Madeleine, Mathiassen, and Arendt-Nielsen 2008; van Dieen et al. 1993) and shown to improve performance in a repetitive task (Farina et al. 2008). Conversely, muscle activity patterns characterised by low motor variability during work cycles are often identified as risk factors for the development of upper extremity disorders (Kilbom, Persson, and Jonsson 1986; Veiersted, Westgaard, and Andersen 1993). Pain can also induce a decrease in variability and continue to limit the
*Corresponding author. Email: [email protected] q 2014 Taylor & Francis
Downloaded by [Purdue University] at 11:41 28 August 2014
1406
J. Tat et al.
postural patterns after pain is removed (Moseley and Hodges 2006). Thus, motor variability appears to be an important characteristic of muscle activity for examining work-related musculoskeletal disorders (WMSD). Workers who perform tasks in a stereotyped manner demonstrate low variability and may increase their risk for injury as compared to individuals who make use of the redundancy in the muscular system (Mathiassen, Moller, and Forsman 2003). Elevated muscle activity has been well documented as a risk factor for WMSDs (Finsen, Christensen, and Bakke 1998; Madeleine et al. 2003b; Veiersted, Westgaard, and Andersen 1993). Muscle activity during both work and rest phases of a task is also linked to feelings of discomfort in the neck and shoulder (Madeleine et al. 2003b). For example, work cycles with short rest times often elevate resting muscle activity and can cause fatigue, pain and injury (Veiersted, Westgaard, and Andersen 1993). A number of factors can increase muscle activity in the upper extremity. Multitasking and high task complexity can increase concurrent mental effort and lead to increases in muscle activity (Au and Keir 2007; Laursen et al. 2002; MacDonnel and Keir 2005). The coupling of grip force with dynamic shoulder action can alter shoulder activity by decreasing motor variability (Hodder and Keir 2012). Though motor variability and muscle activity have been investigated previously, the inclusion of additional task demands could reduce motor variability due to imposed constraints, as with workers required to grasp an object while performing a push or pull task (Di Domizio and Keir 2010). This study used a cyclic push task, with and without a grip task, to elucidate motor control and variability strategies to further our understanding of muscular injuries of the shoulder and forearm. The purpose of this study was to investigate the effects of load and frequency on muscle activity and the relative size of variability using the CoV during work and rest phases of the movement for a repetitive pushing task with the constraint of a handgrip. 2. Methods 2.1 Participants A total of 10 men and 9 women participated in this study to represent a mixed working population. Mean age, mass and height can be found in Table 1. All participants were verbally screened for right-hand dominance and having no history of upper extremity injury or pain in the past year. The study was approved by the McMaster University Research Ethics Board and all participants provided written informed consent. 2.2 Experimental protocol Methods have been reported in detail previously (Keir and Brown 2012) but will briefly be described here. A bimanual pushing task was created using two parallel low friction tracks with handles mounted vertically on platforms that could slide independently (Figure 1). The tracks were 0.25-m long and the handles were 0.38-m apart. The position of each handle was monitored via linear potentiometers affixed to the platform of each handle. The handle mounted to the right track was instrumented with a grip dynamometer (MIE Medical Research Ltd. Leeds, UK; mass ¼ 0.45 kg; grip span ¼ 5.25 cm), while a matching wooden post was on the left side. The track was fixed to an adjustable table set at elbow height for each participant to allow a start position of 908 elbow flexion and forearms parallel to the table top. Participants performed 18 different bilateral pushing tasks, each 120 s in duration. The 18 tasks included all combinations of three push frequencies (4, 8, 16 pushes/min), three force levels (1 kg, 2 kg, 4 kg) and two grip conditions (no required grip, 30% maximum). Trials were randomised for each participant and 120 s of rest was given between each trial. Push loads were chosen to replicate those found in the automotive assembly plant, but frequencies were chosen to include common definitions of low and high repetition (Moore and Garg 1995). The participant was aware of the load, frequency and grip requirements prior to the start of each trial. A metronome indicated the start of each push cycle through headphones. The participant extended both arms to a comfortable distance and returned to the starting position at a self-selected speed, such that they would be ready for indication of the next cycle. When gripping was required, participants were instructed to grip to the target level first, prior to beginning the push cycle. A target grip force was provided by visual feedback that displayed the target force of 30 ^ 1.5% maximal voluntary grip force (MVG). Table 1.
Mean participant characteristics (^SD) including age, height, mass and maximum grip force.
Number of participants Age (years) Height (cm) Mass (kg) Maximum grip (N)
Males
Females
Combined
10 27.7 ^ 6.0 179.5 ^ 6.8 84.4 ^ 11.4 629.8 ^ 95.4
9 23.7 ^ 3.1 165.9 ^ 8.4 64.4 ^ 9.2 333.5 ^ 50.2
19 25.8 ^ 5.1 172.8 ^ 9.9 75.0 ^ 14.4 481.7 ^ 173.3
Ergonomics
Downloaded by [Purdue University] at 11:41 28 August 2014
Figure 1.
1407
Start (left) and finish (right) postures using dual pushing track with grip dynamometer on right-hand side (in foreground).
To determine 30% MVG, the MVG was determined using the grip dynamometer in the start position. Participants were instructed to slowly increase grip force to maximum, hold for 2 s and relax over a 10-s trial. MVG was calculated over a 250- ms window centred about the peak force. Two maximal trials were performed and if the MVG were within 5%, the mean of the two trials was used to determine the 30% grip condition. If the two trials were not within 5%, additional trials were completed until the two highest calculated forces were within 5% (occasionally three trials were required, one participant required four trials). Surface EMG was collected from eight muscles of the right upper extremity (AMT-8, Bortec Biomedical, Calgary, AB, Canada; bandwidth 10-1000 Hz, CMRR . 115 dB at 60 Hz, input impedance < 10GV). The muscles that were monitored included: posterior deltoid, anterior deltoid (AD), biceps brachii (BB), triceps brachii, flexor digitorum superficialis (FDS), flexor carpi radialis, extensor digitorum (ED) and extensor carpi radialis. The muscles were identified using anatomical landmarks and muscle specific isometric contractions (Perotto 1994). For the four muscles of primary interest (AD, BB, FDS and ED), the electrode positions included: below the anterior margin of the acromion for AD; BB was at mid-humerus level; FDS was distal to the medial epicondyle with the hand fully supinated; and ED was distal the lateral epicondyle with the hand fully pronated. Note that the current study focussed on only four of the eight collected muscles to simplify the document and highlight the muscles previously shown to be of sufficient activation (. 5% maximum) and interest (Keir and Brown 2012). Each electrode site was prepared by shaving and scrubbing with isopropyl alcohol before attaching disposable Ag-AgCl surface electrodes (Meditrace, Kendall, MA, USA) over the muscle belly parallel to muscle fibre orientation with an inter-electrode distance of 2.5 cm. Next, the participant performed a series of muscle specific maximal isometric contractions for 10 s each. Maximal voluntary excitation (MVE) for each muscle followed a similar protocol to our previous studies (Au and Keir 2007; Di Domizio and Keir 2010). For the shoulder muscles, participants performed resisted arm raises in flexion and for biceps, participants performed resisted elbow flexion in supination. For the forearm muscles, the tasks included resisted wrist flexion and extension. Grip force, linear potentiometer and EMG data were sampled at 2048 Hz and stored for analysis using a custom programme (Labview v. 7.1, National Instruments, TX, US).
2.3
Data analysis
For the current study, four muscles were selected: AD, BB, ED and FDS. Raw EMG was full wave rectified and dual Butterworth filtered using a 3 Hz low pass filter. All EMG signals were normalised to the muscle specific MVE that was calculated as the mean amplitude over a 500- ms window about the largest peak of each muscle. Grip force was filtered at 10 Hz and normalised to MVG. The linear potentiometer data were used to determine the start, peak and end position of each individual cycle and to determine cycle time in seconds. Each cycle was divided into work and rest phases of the movement (Figure 2). The phases were defined as (1) work – including push and return phases and (2) rest – time between work phases. Average EMG (AEMG), grip force and timing information were determined for each phase using a custom programme (MatLab v.7.6, The MathWorks, Natick, MA, USA). AEMG was used to determine the CoV, defined as the SD divided by the mean, and represented the relative size of EMG variability for each work phase (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen 2008). We used CoV because it is self-scaling and can assess differences in variability between muscle groups. CoV has previously been used to describe motor variability in relation to
Downloaded by [Purdue University] at 11:41 28 August 2014
1408
Figure 2.
J. Tat et al.
Sample potentiometer data illustrating individual components of the movement that make up each cycle.
work experience and discomfort (Madeleine and Madsen 2009). The result of the work phase AEMG divided by the rest phase EMG represented the work– rest (WR) ratio. The WR ratio CoV was also calculated for AEMG (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen 2008). 2.4 Statistical analysis For each cycle, AEMG was determined for work and rest phases, and the WR ratio calculated. The mean and CoV of each parameter were calculated for each trial. A three-way repeated measures analysis of variance was used to determine the effect of load, frequency and grip on grip force, cycle times, AEMG and CoV for each muscle at work and rest, and WR ratio. Significant main effects and interactions were further analysed using Tukey’s honestly significant difference test (SPSS v 13.0, SPSS Inc, Chicago, IL). Significance was set at a ¼ 0.01 and all values are reported as means ^ SD. 3.
Results
3.1 Grip force For the rest phase, a grip £ frequency interaction (F2,36 ¼ 32.38, p , 0.01) affected grip force. Faster push frequencies increased grip force during all gripping trials ( p , 0.01); however, the effect was lost without grip (Figure 3). A main effect
Figure 3. Grip force at rest. During gripping trials, the grip force was significantly different for all push/pull frequencies. Error bars represent one standard error of mean. *Denotes significance ( p , 0.01).
Ergonomics
1409
of grip was found in the work phase (F1,18 ¼ 229.61, p , 0.01), where the mean grip force was consistent but lower than the target 30% MVG at 21.4 ^ 5.8% of MVG, while the no grip condition produced a grip force of 1.1 ^ 0.8% of MVG. 3.2
Cycle times
Downloaded by [Purdue University] at 11:41 28 August 2014
Frequency resulted in faster work times (F2,36 ¼ 11.98, p , 0.01) with the greatest difference in the 16/min condition compared to the 4/min and 8/min push frequencies ( p , 0.01). The work times for each frequency were 2.18 ^ 0.45 s (4/min), 2.13 ^ 0.51 s (8/min) and 1.90 ^ 0.27 s (16/min). The remainder of the cycle was allocated to rest and consisted of rest times 10.23 ^ 0.49 s (4/min), 5.03 ^ 0.47 s (8/min) and 1.80 ^ 0.27 s (16/min). 3.3 Muscle activity For AEMG, a frequency £ load interaction was found in the rest phase (AD, F4,72 ¼ 4.27, p , 0.01 and BB, F4,72 ¼ 2.75, p ¼ 0.035) but not during the work phase. A frequency £ load interaction was also found for the AD WR ratio (F4,72 ¼ 17.85, p , 0.01) and BB WR ratio (F4,72 ¼ 17.20, p , 0.01). For both AD and BB, a faster push frequency increased AEMG at rest and decreased WR ratio in all push/pull loads, with the greatest difference in the fastest frequency (Table 2). For the forearm muscle activity, there was a significant frequency £ grip interaction for ED (F2,36 ¼ 21.97, p , 0.01) and FDS (F2,36 ¼ 24.83, p , 0.01) in the rest phase but not in the work phase. AEMG at rest had the greatest increase with higher push frequencies and with grip (Table 2). There was also a frequency £ grip interaction for WR ratio for these muscles (ED, F2,36 ¼ 26.77, p , 0.01; FDS, F2,36 ¼ 29.41, p , 0.01). Likewise, WR ratio decreased with higher push frequencies and with grip (Table 3). In the work phase, grip independently increased muscle activity in ED by 13.1 ^ 1.3% MVE (F1,18 ¼ 50.39, p , 0.01) and FDS by 15.9 ^ 2.0% MVE (F1,18 ¼ 67.45, p , 0.01) over trials without grip. For FDS activity, a frequency effect was found for the work phase (F2,36 ¼ 6.06, p , 0.01) that was primarily seen between the 16/min (11.7% MVE) and 4/min (10.5% MVE) push frequencies. 3.4
EMG variability
AD and BB CoV were significantly affected by task frequency (Figure 4). For AD, faster push frequencies increased the AEMG CoV in the rest phase (F2,36 ¼ 10.55, p , 0.01) but not during the work phase. The WR ratio CoV also decreased Table 2.
Muscle activity in % MVE (mean ^ SD) for rest and work phases. 4/min
Rest
Load
Muscle
1 kg
AD BB ED FDS AD BB ED FDS AD BB ED FDS AD BB ED FDS AD BB ED FDS AD BB ED FDS
2 kg
4 kg
Work
1 kg
2 kg
4 kg
8/min
16/min
No grip
Grip
No grip
Grip
No grip
Grip
1.3 ^ 1.4 0.8 ^ 0.5 2.1 ^ 1.8 1.3 ^ 0.8 1.4 ^ 1.7 0.9 ^ 0.6 3.1 ^ 3.8 1.5 ^ 1.2 1.8 ^ 2.1 1.1 ^ 1.2 3.5 ^ 3.0 1.7 ^ 1.0 11.2 ^ 6.6 1.9 ^ 1.3 4.4 ^ 4.1 2.2 ^ 1.4 14.2 ^ 8.0 2.4 ^ 1.8 5.2 ^ 4.8 2.7 ^ 2.0 24.8 ^ 13.9 5.7 ^ 4.3 7.9 ^ 5.9 3.6 ^ 2.3
1.3 ^ 1.4 0.9 ^ 0.5 5.0 ^ 2.4 3.7 ^ 2.0 1.4 ^ 1.4 0.9 ^ 0.6 5.4 ^ 2.3 3.9 ^ 2.0 1.9 ^ 2.1 1.2 ^ 0.7 6.7 ^ 4.2 3.9 ^ 1.9 9.6 ^ 6.3 2.6 ^ 1.9 16.7 ^ 9.3 17.5 ^ 9.4 14.2 ^ 10.3 3.5 ^ 2.9 18.2 ^ 8.6 18.4 ^ 8.2 25.5 ^ 15.6 7.6 ^ 6.3 21.5 ^ 10.0 18.4 ^ 8.7
1.5 ^ 1.9 0.8 ^ 0.5 2.6 ^ 3.1 1.5 ^ 1.3 1.8 ^ 2.1 0.9 ^ 0.5 3.1 ^ 3.3 1.7 ^ 1.0 2.4 ^ 2.8 1.3 ^ 1.0 4.7 ^ 4.7 2.1 ^ 1.4 9.9 ^ 5.9 1.6 ^ 1.1 4.5 ^ 3.4 2.6 ^ 2.2 13.6 ^ 8.6 1.9 ^ 1.4 6.0 ^ 5.9 3.0 ^ 2.2 25.0 ^ 15.7 6.0 ^ 6.0 9.7 ^ 6.9 3.9 ^ 2.7
1.5 ^ 1.8 1.0 ^ 0.7 9.0 ^ 5.2 5.8 ^ 4.1 1.9 ^ 2.0 1.3 ^ 0.9 9.5 ^ 3.9 6.0 ^ 4.9 2.5 ^ 2.9 1.7 ^ 1.1 10.1 ^ 3.9 7.0 ^ 5.1 9.0 ^ 6.1 2.3 ^ 1.8 17.4 ^ 9.6 19.2 ^ 9.1 14.5 ^ 9.6 3.8 ^ 2.5 18.0 ^ 9.5 17.6 ^ 7.9 25.9 ^ 16.4 7.5 ^ 5.0 22.3 ^ 9.4 19.4 ^ 8.9
2.2 ^ 2.7 1.1 ^ 0.7 3.7 ^ 2.2 1.6 ^ 1.1 2.9 ^ 3.0 1.3 ^ 0.9 4.0 ^ 2.6 2.0 ^ 1.5 4.5 ^ 5.2 2.1 ^ 1.4 4.0 ^ 1.8 2.6 ^ 1.9 9.8 ^ 6.1 1.8 ^ 1.4 4.5 ^ 4.1 2.7 ^ 1.8 13.6 ^ 8.5 2.2 ^ 1.6 5.3 ^ 4.3 3.1 ^ 2.1 26.1 ^ 14.3 5.1 ^ 4.4 8.9 ^ 5.9 4.1 ^ 2.6
3.2 ^ 5.0 2.0 ^ 1.5 2.0 ^ 1.9 10.3 ^ 7.8 3.7 ^ 4.9 2.0 ^ 1.1 2.3 ^ 2.1 10.3 ^ 7.1 4.5 ^ 4.7 2.8 ^ 1.6 4.5 ^ 4.7 10.6 ^ 6.8 13.2 ^ 10.8 4.0 ^ 3.7 19.2 ^ 8.4 21.2 ^ 9.9 15.1 ^ 11.0 3.4 ^ 1.8 19.4 ^ 8.8 18.8 ^ 8.2 25.6 ^ 14.1 6.8 ^ 5.1 21.4 ^ 9.5 20.1 ^ 9.1
1410 Table 3.
J. Tat et al. Work– rest ratio (mean ^ SD) for all conditions. 4/min
WR ratio
Load
Muscle
1 kg
AD BB ED FDS AD BB ED FDS AD BB ED FDS
2 kg
Downloaded by [Purdue University] at 11:41 28 August 2014
4 kg
8/min
16/min
No grip
Grip
No grip
Grip
No grip
Grip
11.4 ^ 6.6 2.6 ^ 1.2 2.4 ^ 0.9 1.7 ^ 0.5 14.0 ^ 6.8 3.2 ^ 2.0 2.8 ^ 1.4 1.9 ^ 0.8 21.3 ^ 11.2 6.4 ^ 3.5 3.7 ^ 1.7 2.4 ^ 1.6
9.5 ^ 5.2 3.0 ^ 1.6 5.2 ^ 2.5 5.5 ^ 3.0 13.3 ^ 7.8 4.3 ^ 3.4 5.2 ^ 2.0 5.4 ^ 2.6 19.6 ^ 9.7 6.9 ^ 4.8 5.8 ^ 2.9 5.0 ^ 2.0
9.3 ^ 4.5 2.1 ^ 0.9 2.4 ^ 1.3 1.8 ^ 0.6 10.6 ^ 5.1 2.6 ^ 1.5 2.3 ^ 1.5 1.8 ^ 0.7 14.4 ^ 6.1 4.7 ^ 2.6 2.9 ^ 0.6 1.8 ^ 0.5
8.0 ^ 4.2 3.0 ^ 1.6 3.8 ^ 1.7 4.2 ^ 2.1 10.1 ^ 5.2 3.2 ^ 2.0 3.5 ^ 1.5 3.7 ^ 1.7 14.3 ^ 5.9 4.9 ^ 2.2 3.9 ^ 1.6 3.4 ^ 1.5
6.1 ^ 3.2 1.7 ^ 1.0 1.9 ^ 0.7 1.8 ^ 0.5 6.4 ^ 2.7 1.9 ^ 1.2 1.9 ^ 0.7 1.7 ^ 0.6 8.6 ^ 3.6 2.6 ^ 1.4 2.1 ^ 0.6 1.7 ^ 0.6
6.1 ^ 3.2 2.3 ^ 1.3 2.5 ^ 1.3 2.8 ^ 1.4 6.3 ^ 2.7 1.8 ^ 0.9 2.2 ^ 1.1 2.5 ^ 1.5 7.9 ^ 3.9 2.6 ^ 1.4 2.2 ^ 0.9 2.3 ^ 1.2
with task frequency (F2,36 ¼ 13.95, p , 0.01). For BB, there was a frequency effect on AEMG CoV at rest (F2,36 ¼ 20.52, p , 0.01) and for WR ratio CoV (F2,36 ¼ 14.07, p , 0.01). The rest BB CoV decreased and the WR ratio CoV increased, both linearly with frequency ( p , 0.01). No significant changes were found for work variability in either muscle.
Figure 4. Frequency effects for AD muscle variability during (a) work and rest phases, (b) WR ratio and BB muscle variability at (c) work and rest phases, (d) WR ratio. Error bars represent one standard error of mean. *Denotes conditions in AD that are significantly different from 16/min push frequency ( p , 0.01). **Denotes BB significant differences with 16/min frequency ( p , 0.01).
Downloaded by [Purdue University] at 11:41 28 August 2014
Ergonomics
1411
Figure 5. Push/Pull load effects in the no grip condition for ED muscle variability at (a) work and rest, (b) WR Ratio and FDS muscle variability at (c) work and rest phases, (d) WR ratio. Error bars represent one standard error of mean. *Significantly different from 4- kg load in FDS ( p , 0.01) and **significantly different from 4- kg load in ED ( p , 0.01).
Forearm muscle EMG variability (ED and FDS) was significantly affected by the work conditions in WR ratio. Post hoc comparisons showed that this was due to changes in the rest phase and not in the work phase (Figure 5). For ED, there was a load £ grip interaction on CoV at rest (F2,36 ¼ 7.89, p , 0.01) and WR ratio CoV (F2,36 ¼ 5.49, p , 0.01) which increased ED CoV at rest and significantly decreased WR ratio CoV for the 4- kg load, no grip condition ( p , 0.01). This load £ grip interaction was not found at work; however, there was a grip main effect (F1,18 ¼ 51.94, p , 0.01). The addition of 30% MVG decreased ED CoV at work by an average of 0.13 ^ 0.05 across all conditions. For FDS, there was also a significant load £ grip interaction on CoV at rest (F2,36 ¼ 14.92, p , 0.01) and WR ratio CoV (F2,36 ¼ 6.20, p , 0.01). Similarly, the load £ grip interaction was seen in the no grip condition, which increased FDS CoV at rest and decreased FDS WR ratio CoV with the largest differences being in the 4- kg load. Frequency also affected EMG variability in FDS at rest and WR ratio. Faster push frequencies resulted in a linear increase in FDS CoV in the rest phase (F2,36 ¼ 19.63, p , 0.01). For FDS, the WR ratio CoV had a frequency £ load interaction (F4,72 ¼ 4.48, p , 0.01) that was mainly seen in the 2- kg load, where 4/min (CoV 2.07) was significantly larger than 8/min (1.13) and 16/min (1.08). There was also a frequency £ grip interaction (F2,36 ¼ 5.69, p , 0.01) such that the effect of frequency was amplified in the trials without grip. There were no significant effects in COV at work. 4.
Discussion
This study examined cycle-to-cycle EMG variability (CoV in AEMG) in selected upper extremity muscles during a repetitive pushing task. There were no significant effects of load and frequency on EMG variability during work phases of the task; however, WR ratio variability was predominantly influenced by changes in variability in the rest phase. For AD and BB, a higher push frequency increased CoV at rest and decreased WR ratio CoV. For the forearm muscles (ED and
J. Tat et al.
Downloaded by [Purdue University] at 11:41 28 August 2014
1412
Figure 6. EMG data replotted versus workload illustrating frequency, load and grip effects in FDS at (a) rest, (b) work, (c) WR ratio. Error bars represent one standard error of mean.
FDS), higher loads increased the CoV at rest and decreased the WR ratio CoV. In this paradigm, greater resting variability suggests the muscle may be unable to relax between efforts that could result in the overuse of the same motor units and their muscle fibres. In the workplace, this could lead to muscular fatigue and injury to the muscle. This is supported by epidemiological studies, which indicate workers who perform repetitive tasks with reduced pauses are more likely to develop injury (Veiersted, Westgaard, and Andersen 1990). Our findings highlight motor variability changes in the rest
1413
Downloaded by [Purdue University] at 11:41 28 August 2014
Ergonomics
Figure 7. Variability (CoV) presented versus workload illustrating frequency, load and grip effects in FDS at (a) rest, (b) work, (c) WR ratio. Error bars represent standard error.
phase but not in the work phase of a cyclic task and underscore the importance of muscle activity at rest. Muscle activity at rest might provide a more sensitive measure of physical exposure to frequency and loads, which is contrary to the traditional emphasis on work components (Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen 2008; Mathiassen, Moller, and Forsman 2003; Veiersted, Westgaard, and Andersen 1993). EMG has been used to identify potential warning signs for repetitive workplace tasks that pose risk to the development of WMSD (Madeleine et al. 2003b; Madeleine, Mathiassen, and Arendt-Nielsen 2008; Madeleine, Voigt, and Mathiassen
Downloaded by [Purdue University] at 11:41 28 August 2014
1414
J. Tat et al.
2008; Sandsjo et al. 2000; Veiersted 1994). In rest and work phases, there was an increase in activity that corresponded with larger loads and faster push frequencies (Table 2); increased AEMG during a work task has been linked to pain (Sandsjo et al. 2000; Veiersted 1994). Chronic pain in the upper extremity has also been characterised by a decrease in WR ratio EMG variability (Madeleine, Voigt, and Mathiassen 2008). In the present study, we have shown a change in the direction (decrease WR ratio CoV) and magnitude of variability (smaller WR ratio CoV) with exposure to more demanding workloads. The smallest CoV for the AD and BB was found in the highest frequency (16/min), and the smallest forearm CoV was found with the largest load (4 kg). Madeleine et al. (2003a) demonstrated that novice workers had lower variability and a lower WR EMG ratio in the shoulder muscles (including AD) when performing a repetitive task. A strategy that involves reduced variability may help explain the tendency for novice workers to experience fatigue, pain or injury as it has been described as less efficient due to its failure to take advantage of the redundant upper extremity muscular system (Mathiassen, Moller, and Forsman 2003). Our findings in AEMG and EMG variability support the hypothesis that increased frequency and load may lead to muscle fatigue and chronic pain (Farina et al. 2008; Madeleine, Mathiassen, and ArendtNielsen 2008). Simultaneous gripping and pushing are common in the workplace and grip forces have been found to alter muscle activity patterns in the shoulder (Au and Keir 2007; Hodder and Keir 2012; MacDonell and Keir 2005; Smets, Potvin, and Keir 2009). The effect of grip force in this study had a significant impact on the work components of the repetitive task. The addition of a 30% MVG increased muscle activity in the forearm (ED by 13.1 ^ 1.3%; FDS by 15.9 ^ 2.0%) and significantly decreased EMG variability (ED CoV by 0.13 ^ 0.05; FDS CoV by 0.07 ^ 0.06), which are both risk factors for the development of injury and chronic pain (Madeleine, Mathiassen, and Arendt-Nielsen 2008). It is also noteworthy that participants increased resting grip force with faster push frequencies, suggesting that participants had difficulty returning to resting conditions between pushes. The mean work grip force was substantially below the target of 30% because we defined the work phase to include both the push (28.1 ^ 4.1% MVG) and return (15.5 ^ 10.4% MVG) components. Participants were not asked to maintain the target grip force during the return. Keir and Brown (2012) evaluated the 120-s trial as a whole and noted that the complex interactions between load and frequency might be more easily explained by workload, the product of load and frequency. We have replotted the data by categorising the 18 work conditions into five workloads (4, 8, 16, 32, 64 kg/min) (Figures 6 and 7). For example, conditions in the 8 kg/min workload included the conditions low load –med frequency (1 kg £ 8/min) and med load – low frequency (2 kg £ 4/min). For FDS, the relationship between muscle activity and workload is driven by the AEMG in the rest phase (Figure 6). With gripping, the linear decline of the WR ratio is mirrored by the linear increase in rest AEMG but the relationship is static across workload for gripping. Workload variability (CoV) provides a different perspective as the changes in the rest phase dictate the relationship when a grip was not required (Figure 7). We can see that the variability with a grip appears relatively constant across workload but the variability in rest EMG increases with workload. This indicates the workload method might be helpful to simplify the interpretation of EMG results to identify risk in cyclical workplace tasks. There are a few limitations to this study. Participants were healthy young adults who were not employed in manual labour. Madeleine, Mathiassen, and Arendt-Nielsen (2008) showed that experienced workers have greater motor pattern flexibility. Thus, generalising our findings to more skilled and advanced assembly workers may be limited. While we relate our findings to the work of Madeleine, Mathiassen, and Arendt-Nielsen (2008) who related skill and pain to variability, we had healthy participants without pain. Thus, this study is limited to the effects of load and frequency on muscle activity. However, it is unclear whether the large mean muscle activity at rest and greater variability in muscle activity at rest were a result of the task demands or an anticipatory effect due to the subject knowing the condition. Faster push frequencies may result in early muscle activation for the next cycle that would increase muscle activity and/or variability. However, a metronome was used to indicate the start of each cycle, and we noted with work times that subjects voluntarily moved faster for faster frequencies to meet the task demand. Nonetheless, an increase in muscle activity at rest is a relevant risk factor for developing pain (Sandsjo et al. 2000; Veiersted 1994). 5. Conclusions This study found that the rest phase is important in the assessment of muscle activity in a repetitive cyclic task. Frequency and push load significantly altered the muscle activity during the rest phase (but not the work phase), which drove changes in WR ratio. The relative size of variability in muscle activity and WR ratio was reduced with higher frequency and load, suggesting a potential role in the development of upper extremity WMSD. Further research is needed to examine whether the effects of frequency, load and grip on motor variability contribute to acute or chronic pain during both rest and work phases. Understanding the relationship between these workload variables and pain can lead to a better comprehension of injury mechanisms.
Ergonomics
1415
Funding This study was funded by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada [#217382].
Downloaded by [Purdue University] at 11:41 28 August 2014
References Au, A. K., and P. J. Keir. 2007. “Interfering Effects of Multitasking on Muscle Activity in the Upper Extremity.” Journal of Electromyography & Kinesiology 17 (5): 578– 586. Di Domizio, J., and P. J. Keir. 2010. “The Effects of Posture and Force Coupling on Upper Extremity Muscle Activity.” Ergonomics 53 (3): 336– 343. Farina, D., F. Leclerc, L. Arendt-Nielsen, O. Butteli, and P. Madeleine. 2008. “The Change in Spatial Distribution of Upper Trapezius Muscle Activity is Correlated to Contraction Duration.” Journal of Electromyography & Kinesiology 18 (1): 16 – 25. Finsen, L., H. Christensen, and M. Bakke. 1998. “Musculoskeletal Disorders among Dentists and Variation in Dental Work.” Applied Ergonomics 29 (2): 119– 125. Hodder, J. N., and P. J. Keir. 2012. “Targeted Gripping Reduces Shoulder Muscle Activity and Variability.” Journal of Electromyography & Kinesiology 22 (2): 186– 190. Keir, P. J., and M. M. Brown. 2012. “Force, Frequency and Gripping Alter Upper Extremity Muscle Activity during a Cyclic Push Task.” Ergonomics 55 (7): 813– 824. Kilbom, A. 1994. “Repetitive Work of the Upper Extremity: Part II – The Scientific Basis for the Guide.” International Journal of Industrial Ergonomics 14 (2): 59 –86. Kilbom, A., J. Persson, and B. G. Jonsson. 1986. “Disorders of the Cervicobrachial Region among Female Workers in Electronics Industry.” International Journal of Industrial Ergonomics 1 (1): 37 –47. Laursen, B., B. R. Jensen, A. H. Garde, and A. H. Jorgensen. 2002. “Effect of Mental and Physical Demands on Muscular Activity during the Use of a Computer Mouse and Keyboard.” Scandinavian Journal of Work Environment & Health 28 (4): 215– 221. MacDonell, C. W., and P. J. Keir. 2005. “Interfering Effects of the Task Demands of Grip Force and Mental Processing on Isometric Shoulder Strength and Muscle Activity.” Ergonomics 48 (15): 1749– 1769. Madeleine, P., and D. Farina. 2008. “Time to Task Failure in Shoulder Elevation is Associated to Increase in Amplitude and to Spatial Heterogeneity of Upper Trapezius Mechanomyographic Signals.” European Journal of Applied Physiology 102 (3): 325– 333. Madeleine, P., B. Lundager, M. Voigt, and L. Arendt-Nielsen. 2003a. “Standardized Low-Load Repetitive Work: Evidence of Different Motor Control Strategies between Experienced Workers and a Reference Group.” Applied Erognomics 34 (6): 533– 542. Madeleine, P., B. Lundager, M. Voigt, and L. Arendt-Nielsen. 2003b. “The Effects of Neck-Shoulder Pain Development on SensoryMotor Interactions among Female Workers in Poultry and Fish Industries. A Prospective Study.” International Archives of Occupational and Environmental Health 76 (1): 39 – 49. Madeleine, P., and T. M. T. Madsen. 2009. “Changes in the Amount and Structure of Motor Variability during a Deboning Process are Associated with Work Experience and Neck-Shoulder Discomfort.” Applied Ergonomics 40 (5): 887– 894. Madeleine, P., S. E. Mathiassen, and L. Arendt-Nielsen. 2008. “Changes in the Degree of Motor Variability Associated with Experimental and Chronic Pain Neck-Shoulder Pain during Standardised Repetitive Arm Movement.” Experimental Brain Research 185 (4): 689– 698. Madeleine, P., M. Voigt, and S. E. Mathiassen. 2008. “The Size of Cycle-to-Cycle Variability in Biomechanical Exposure among Butchers Performing a Standardised Cutting Task.” Ergonomics 51 (7): 1078– 1095. Mathiassen, S. E., T. Moller, and M. Forsman. 2003. “Variability in Mechanical Exposure within and between Individuals Performing a Highly Constrained Industrial Work Task.” Ergonomics 46 (8): 800– 824. Moore, S. J., and A. Garg. 1995. “The Strain Index: A Proposed Method to Analyze Jobs for Risk of Distal Upper Extremity Disorders.” American Industrial Hygiene Association Journal 56 (5): 443–458. Moseley, G. L., and P. W. Hodges. 2006. “Reduced Variability of Postural Strategy Prevents Normalization of Motor Changes Induced by Back Pain: A Risk Factor for Chronic Trouble?” Behavioural Neuroscience 120 (2): 474– 476. Perotto, A. 1994. Anatomical Guide for the Electromyographer: The Limbs and the Trunk. Springfield, IL: CC Thomas. Sandsjo, L., B. Melin, D. Rissen, D. Ingela, and U. Lundberg. 2000. “Trapezius Muscle Activity, Neck and Shoulder Pain, and Subjective Experiences during Monotonous Work in Women.” European Journal of Applied Physiology 83 (2): 235– 238. Silverstein, B. A., L. J. Fine, and T. J. Armstrong. 1986. “Hand Wrist Cumulative Trauma Disorders in Industry.” British Journal of Industrial Medicine 43 (11): 779– 784. Smets, M. P. H., J. R. Potvin, and P. J. Keir. 2009. “The Effect of Constrained Hand Gripping on Arm Strength and Muscle Activation of the Upper Extremities.” Ergonomics 52 (9): 1144– 1152. Srinivasan, D., and S. E. Mathiassen. 2012. “Motor Variability in Occupational Health and Performance.” Clinical Biomechanics 27 (10): 979– 993. van Dieen, J. H., H. H. E. O. Vrielink, A. F. Housheer, F. B. J. Lotters, and H. M. Toussaint. 1993. “Trunk Extensor Endurance and Its Relationship to Electromyogram.” European Journal of Applied Physiology 66 (5): 388– 396. Veiersted, K. B. 1994. “Sustained Muscle Tension as a Risk Factor for Trapezius Myalgia.” International Journal of Industrial Ergonomics 14 (4): 333– 339. Veiersted, K. B., R. H. Westgaard, and P. Andersen. 1990. “Pattern of Muscle Activity during Stereotyped Work and Its Relation to Muscle Pain.” International Archives of Occupational and Environmental Health 62 (1): 31 – 41. Veiersted, K. B., R. H. Westgaard, and P. Andersen. 1993. “Electromyographic Evaluation of Muscular Work Pattern as a Predictor of Trapezius Myalgia.” Scandinavian Journal of Work Environment & Health 19 (4): 284– 290.
