Work-Related Musculoskeletal Discomfort in the Shoulder due to Computer Use

The National Institute for Occupational Safety and Health (NIOSH, 1997) in the USA defines Musculoskeletal Disorder (MSD) as a disorder that affects a part of the body’s musculoskeletal system, which includes bones, nerves, tendons, ligaments, joints, cartilage, blood vessels and spinal discs. These are the injuries that result from repeated motions, vibrations and forces placed on human bodies while performing various job actions. The individual factors that can contribute to musculoskeletal symptoms include heredity, physical condition, previous injury, pregnancy, poor diet, and lifestyle.

This chapter presents the risk factors that contribute to musculoskeletal disorders in shoulders resulting from intensive use of computers in the workplaces.The risk factors of musculoskeletal disorders were revealed by assessing and analyzing workplace ergonomics, worker attitudes and experiences on the use of the computer keyboard and mouse.This was followed by an experimental data collection of muscle load, muscle force and muscular fatigue from the shoulder by Surface electromyogram (sEMG) to validate and verify the developed mathematical model.
Epidemiological studies in the literature confirmed that the work which is related with computer use brings higher risk for the development of musculoskeletal symptoms.Evans and Patterson (2000) tested the hypothesis that poor typing skill, hours of computer use, tension score and poor workstation setup are associated with neck and shoulder complaints, and they found out that tension score and gender were the only factors to predict neck and shoulder pain.Jensen et al. (2002) found that the duration of computer work is associated with neck and shoulder symptoms in women, and hand symptoms in men.Additionally, the use of mouse was observed to have an increase in hand/wrist and shoulder region symptoms among the intensive users of computers.Moreover, Karlqvist et al. (2002) concluded that for both genders the duration of computer work was associated with the musculoskeletal disorder symptoms, and women are at more risk of exposure to such disorder as they have less variability in work tasks.Fogleman and Lewis (2002) studied the risk factors associated with the self-reported musculoskeletal discomfort in a population of video display terminal (VDT) operators, where their results indicated that there is a statistically significant increased risk of discomfort on each of the body regions (head and eyes, neck and upper back, lower back, shoulders, elbows and forearms, and hands and wrists) as the number of hour of keyboard use increases.Blatter and Bongers (2002) studied the association of the effect of the gender differences with physical work factors as well as with the psychosocial factors.However their results showed that psychosocial factors were not related with the duration of computer use, whereas computer work of more than 6 hours per day was associated with musculoskeletal symptoms in all body regions of men, and computer work of more than 4 hours per day entailed the association with musculoskeletal disorders in women.Intensive computer use is associated with an increased risk of neck, shoulder, elbow, wrist and hand pain, paresthesias and numbness.Repetition, forceful exertions, awkward positions and localized contact stress are associated with the development of upper limb cumulative trauma in computer users.Ming and Zaproudina (2003) showed that the repetitive computer use causes cumulative trauma on neck, shoulder, arm and hand muscles and joints.
In their model, Carayon et al. (1999) stipulated that psychosocial work factors (e.g.difficulty of job, working with deadlines, supervisor's pressure, lack of control), which can cause stress, might also influence or be related to ergonomic factors such as force, repetition, and posture that have been identified as risk factor for WRMSDs.Peper et al. (2003) reviewed the ergonomic and psychosocial factors that affect musculoskeletal disorders at the workstation, and their results showed that there was a significant difference in right forearm extensor-flexor muscle tension and in right middle trapezius muscle tension between type tasks and rest.Shuval and Donchin (2005) examined the relationship between ergonomic risk factors and upper extremity musculoskeletal symptoms in VDT workers, by taking into account individual and work organizational factors, and stress.Their results of RULA (Rapid Upper Limb Assessment) observations indicated that there were no acceptable postures of the employees whom were exposed to excessive postural loadings.

Objectives
This research addresses worker perception and attitudes towards computer use, and their experiences with musculoskeletal symptoms in the shoulder and their diagnoses.The primary aim of this chapter is to present an in-debt assessment of the relationship between work-related musculoskeletal disorders in the shoulder and computer use.This study illustrates the idea of understanding how demographic structure (gender, age, height, and weight) physical and psychosocial job characteristics, office ergonomics, perceived musculoskeletal discomfort types and their frequencies may affect formation of musculoskeletal disorders in the shoulder.It then provides the evidence on the symptoms of musculoskeletal discomfort types and the frequency of these discomforts which are significant in the development of WRMSDs in the shoulder due to computer use.
The relevance of this study to the industry is to reduce the work-related musculoskeletal disorders associated with the intensive, repetitive and long period computer use that affect the shoulder.The developed risk assessment model also provides guidance for solving problems related to costly health problems (direct cost), lost productivity (indirect cost), and relieving the imposed economic burden.
As a summary, the research objectives of this study are:


To assess and analyze workplace ergonomics, worker attitudes and experiences on computer use, and musculoskeletal symptoms in the shoulder developed by computer use,


To determine a meaningful and statistically significant relationship between workrelated musculoskeletal disorders in the shoulder and computer use, and develop a risk assessment model,  To validate and verify the developed mathematical model through analysis of the data collected by the sEMG recordings.

Questionnaire
A questionnaire (see appendix) was developed based on the U.S. National Institute for Occupational Safety and Health (NIOSH) Symptoms Survey (NIOSH, 2011) and the Nordic Musculoskeletal Questionnaire (Dickinson et al., 1992).The questionnaire included questions in 7 modules according to the type of the questions.The questions were related with the demographic structure of the participant, physical job characteristics, psychosocial job characteristics, office ergonomics (workstation setup), types of musculoskeletal discomforts experienced at the shoulder, frequency of the musculoskeletal discomforts in the shoulder, and personal medical history.
The instrument was designed specifically for the current work.We are not aware of such an instrument being used for this purpose.In order to prevent any misunderstanding, the respondents were assisted at the time of answering the questionnaire.

Risk assessment model
In order to determine a meaningful and statistically significant relationship between workrelated musculoskeletal disorders and computer use, a risk assessment model needs to be developed.
Logistic Regression Analysis was used to determine a meaningful and statistically significant relationship between shoulder discomfort and computer use, as a risk assessment model.The Logistic Regression was used since many of the independent variables were qualitative and the normality of residuals could not be guaranteed.
Our dependent variable was the WRMSD diagnosis made by a medical doctor (dichotomous dependent variable), and the independent variables were the rest of the variables in the questionnaire.

Experimentation
The respondents of the questionnaire, who have experienced musculoskeletal symptoms, were invited to a lab experiment, where surface electromyogram (sEMG) was used to record muscle load, muscle force and muscular fatigue.This test took place in two phases; i. interrupting the work and performing test contractions of known force in a predetermined body posture and, ii.comparing situations connected with a certain reference activity.
Before conducting the sEMG experiment, those respondents who were under high risk of having WRMSDs in the shoulder were identified using logistic regression.The significance level in logistic regression analysis was chosen to be 5% in order to minimize the possibility of making a Type I error.An independent variable with a regression coefficient not significantly different from 0 (p>0.05)can be removed from the regression model.If p<0.05 then the variable contributes significantly to the prediction of the outcome variable (Pampel, 2000).
Odd ratios of the significant factors for each respondent were calculated to find out respondents who were at the risk of having WRMSDs in the shoulder, as given below:

    
In order to determine the respondents under high risk of having WRMSDs in the shoulder, the odds ratios of the significant factors for each respondent were calculated and those respondents who reflected maximum levels of odds ratios for each significant factor were invited for further investigation through electromyography.
Surface electromyogram was used to collect data from the shoulder.The procedure for the experimentation was as follows; twenty minutes typing exercise was given to each respondent at a time.Each respondent was asked to type a given standard text.Data were collected at 5th, 10th, 15th, and 20th minutes of the experiment.The mean value of the data collected for 30 seconds was then calculated and taken into consideration.
Analysis of Variance (ANOVA) and Factorial Analysis were applied at the end to the data collected by sEMG recordings, to validate and verify the significant risk factors of WRMSDs in the shoulder which were determined by logistic regression.

Respondents
A questionnaire was given to 130 people, who worked intensively with the computers for work/business purposes, such as; staff, research assistants and faculty members of Eastern Mediterranean University (EMU), web page designers, computer programmers, engineers, government officers, public relation officers, marketing officers, bank officers, customer representatives, commissioners, consultants, travel agents and translators.The reason for targeting such diverse disciplines was that the target population is expected to use computers intensively especially for work/business purposes and several other auxiliary purposes including personal and communication.Thus, the results were guaranteed not to be task-related, instead work-related.
The keyboard and mouse were reported to be the most popular (90.77%) input devices, whereas only 12 (9.23%) of the 130 respondents were using touchpad, keypad and trackball as primary input devices.Moreover, 88.43% of the respondents were using regular (Q-type) keyboards, 3.31% were using F-type keyboards, and 4.96% were using ergonomic (with wrist support) keyboards.Additionally, 72.31% of the respondents were using desktop and 27.69% of the respondents were using laptop computers.
Regarding the keyboard use, it was found that 55.04% of the respondents have been using keyboard for 10 or more years, and 37.98% have been using keyboard for at least 5 years.
Around 24.62% of the respondents reported their daily keyboard use as 5-6 hours per day, 23.85% of them as 7-8 hours per day, and 36.15% of them as more than 8 hours per day.
The results of the questionnaire indicated that 79.84% of the respondents found their job interesting, where 20.16% of the respondents indicated that they did not find their job interesting.Additionally, 74.42% of the respondents mentioned that their job gives them personal satisfaction; however 25.58% of the respondents mentioned that they were not having personal satisfaction from their job.A very high majority of the respondents (90.62%) reported that they have "good" relationship with their supervisor/advisor, where 9.38% reported that they have "not good" relationship with their supervisor/advisor.
More than two thirds of the respondents indicated that they share their office, where 35.66% share the office with more than three people, and 33.33% share the office with three or less people.On the other hand, 31.01%reported that they have their own office.
Majority of the respondents (84.38%) reported that they like their office environment, whereas 15.62% of the respondents reported that they do not like their office environment.Addition to this, a very high majority (94.57%) of the respondents indicated that they like working with computers, however only 5.43% of the respondents indicated that they do not like working with computers.
Most of the respondents (64.04%) reported that they have a stressful job, but 35.94% of the respondents reported that they do not have a stressful job.It was observed that, 48.84% of the respondents think that they have enough rest breaks, and 51.16% of the respondents do not think that they have enough rest breaks.Additionally, 46.88% of the respondents have repetitive (static) jobs, whereas 53.12% of the respondents have non-repetitive (dynamic) jobs.
Only 18.60% of the respondents were smokers when they answered the questionnaire, and 81.40% of the respondents were not smokers.More than half of the smoker respondents (63.41%) reported that they were smokers during the previous year, and 36.59% of the respondents were not smokers during the previous year.
Table 2 shows the results obtained on the workstation ergonomics.The results show that 56.92% of the respondents lean back to support their vertebrae, 67.69% reported that their feet were comfortable in the front, 81.54% stated that their seat and hands were centered on the keyboard, more than half (50.77%) of the respondents sit symmetrically, 79.23% use keyboard at the fingertips, 77.69% have the keyboard and the mouse at the same level, 80.62% of the respondents' screens were about arm length away from their eyes, 65.38% had the monitors at the eye level, 72.87% had sufficient lighting without glare, 78.46% had neutral wrist position, and 64.62% had neutral head and neck position.
However, the majority of the respondents didn't take into consideration of having 90 0 angle between the shoulders and the elbows.They did not care about sitting symmetrically at all, and they usually (73.64%) talked on the phone by having the handset between the head and the shoulder.Elbow, arm or leg supports also were not available in the respondents' workstations.Moreover, the majority of the respondents (78.46%) were not trained in posture (table 1).

Office Ergonomics Yes (%) No (%)
Lean Table 2 shows that the most prevalent discomfort experienced was having ache in the shoulder (46.15%).Discomfort (feeling of pain) was observed to be the next prevalent discomfort after ache.It was reported by the respondents that 34.62% of them were experiencing pain in the shoulder.Heaviness was reported by 17.69% of the respondents in the shoulder, and 9.23% of the respondents stated that they have a tightness in their shoulder.Having weakness was reported by 8.46% of the respondents in the shoulder, and having cramp in the shoulder was reported by 6.15%.Feeling of numbness was reported by 3.85% of the respondents and 3.08% of them reported tingling in their shoulder.Feeling hot and cold in the shoulder was reported by 2.31% of the respondents, and only 1.54% reported swelling in their shoulder.Therefore, the discomfort feelings of ache and pain were the most common types of discomforts which are experienced at the shoulder.

Percent
Table 3 shows the frequency of the discomforts experienced by the respondents.Among the 130 respondents, 17 had a recent accident and 6 of those had this accident within 12 months (4.62% of the whole population).Also, 23 respondents reported that they had diagnosed with a work-related musculoskeletal disorder by a medical doctor, and 11 (8.46% of the whole respondents) of the sufferers reported this diagnosis had been made within the last 12 months.
It was reported by the respondents that, 41 (31.54%) of them exercise never/rarely, 57 (43.85%) sometimes, 25 (19.23%)often, and only 7 (5.38%) of them exercise very often.Moreover, 91 of the respondents (70%) stated that they were involved in sport activities, and 39 of them (30%) reported that they were not involved in any kind of sport activities.More than half of the respondents (76, 58.46%) reported that they were involved in walking as sport activity, 17 of the respondents (13.08%) did jogging, 15 (11.54%) of them played football, 4 (3.08%) of them played basketball, 5 (3.85%) of them played volleyball, 10 (7.69%) of them played tennis, 26 (20.00%) did swimming, and 27 (20.77%)involved in other sport activities.

Data analysis
Table 4 shows that only one of the above ergonomics factors, using keyboard and mouse at the same level (p=0.038<0.05)was found to be significant predictors of WRMSDs in the shoulder for the collected data.Table 5 shows that ache in the shoulder (p=0.024<0.05),pain in the shoulder (p=0.019<0.05),and having tightness in the shoulder (p=0.038<0.05)were found to be significant predictors of WRMSDs in the shoulder for the collected data.Table 6 shows that often in the shoulder (p=0.022<0.05)was found to be significant predictors of WRMSDs for the collected data.

Experimental results
After developing the risk assessment model, the model should be validated and be verified.Towards this end, we have to first identify those respondents under risk.Then, the data analysis of the surface EMG recordings is supposed to provide the validation and verification.
Odds ratios for each significant factor determined by the logistic regression analysis were calculated and those respondents who have higher odds ratios for each factor were identified.
It was observed that fifteen respondents were under risk of having WRMSDs according to the results of odds ratio analysis.However, only six of the fifteen respondents were able to be contacted and invited to the sEMG data collection experiment.That group of six respondents formed the test group, and among the non-risk respondent group, six more respondents were invited to form the control group.
In the sEMG experiment, muscular activity in the shoulder (posterior deltoid) was recorded by using sEMG device (MyoTrac Infiniti, model SA9800).The procedure for the experimentation is as follows; 20 minutes typing exercise was given to each respondent at a time.Data were collected at 5th, 10th, 15th, and 20th minutes of the experimentation.The mean value of the collected data for 30 seconds is then calculated and taken into consideration.

Test group experimental results
The readings from sEMG provides the information about the muscle activity in the shoulder over time.Table 7 illustrates the mean value for each 30 seconds interval readings for each test group respondent.
The muscle activity is converted to μV by sEMG and is shown on the vertical axis, and time is shown on the horizontal axis in minutes (figure 1). Figure 1    Where; H 0 = mean musculoskeletal strain (in time) of the 6 respondents does not differ, and H 1 = mean musculoskeletal strain (in time) of the 6 respondents does not differs.
The results obtained by ANOVA indicate that, the risk assessment model developed has been validated and verified with the data collected through sEMG recordings.

Control group experimental results
The control group respondents were selected among the group of respondents who were not under risk according to the odds ratios.Figure 2 illustrates that control group respondents' muscle activities do not significantly differ from each other and these readings were not at high levels.Moreover, the muscle activities of the control group respondents 4 and 5 showed slight but not significant increase throughout the experiment.Where; H 0 = mean musculoskeletal strain (in time) of the 6 respondents does not differ, and H 1 = mean musculoskeletal strain (in time) of the 6 respondents does not differs.
The results of the ANOVA for each control group respondent indicate that, the mean musculoskeletal strain that they experience does not differ in time.That is, the musculoskeletal strain at their shoulder do not differ as those in the test group respondents.
ANOVA results for the control group respondents support the risk assessment model developed to determine the risk factors of WRMSDs.

Conclusion
Most of the studies on the formation of WRMSDs during computer use have been focused on the gender differences, physical and psychological aspects of the user and have not yet considered extra-rational factors such as the perceived musculoskeletal discomfort types and their frequencies.This study presents the idea of understanding how office ergonomics, perceived musculoskeletal discomfort types and their frequencies may affect formation of musculoskeletal disorders at the shoulder.
After collecting data from 130 respondents, the significant findings related with discomfort in shoulder during computer use were: This study provided the evidence that, for the study groups tested and for the given computer use activity, ache and pain are the most common types of the discomforts in the shoulder.Also, this study showed that the mean musculoskeletal strain at the shoulder of test group respondents differ in time, whereas for each control group respondent, the mean musculoskeletal strain that they experience, does not differ in time.
www.intechopen.comIf "Yes", have you been smoking in the last year?

Yes No
The questions in the table below are related with your working posture.Mark with "Yes" if the statement is applicable, mark with "No", if it is not applicable with your working posture.During the last 12 months, have you experienced, while using a keyboard or a mouse, the following symptoms in the following body regions?(mark all that apply)

Fig. 1 .
Fig. 1.Muscle activity recordings in the shoulder of test group respondents.

Fig. 2 .
Fig. 2. Muscle activity recordings in the shoulder of control group respondents.

5 .
What type of computer do you mostly use? (mark only one) type of computer input devices do you mostly use? (mark only one) , how much time daily in total you spend typing on a computer keyboard or using your current job give you personal satisfaction?Yes No 12.How do you define your relationship with your current supervisor/advisor?Good Not good 13.In what kind of office environment you work?I share the office with more than 3 people I share the office with 3 or less people I have my own office 14.Do you like your office environment?

NO 20 .
Lean back in chair to support your vertebrae 21.Elbows form a 90 degree angle while hanging at sides from the shoulders 22. Feet are comfortable on the floor in front of you 23.Your seat and your hands are centered on the keyboard 24.Sit symmetrically (not bending either sides) 25.The keyboard and the mouse are at the fingertips 26.The keyboard and the mouse are on the same level (side by side) 27.The screen is about an arm's length away from the eyes 28.The top of the monitor is at the eye level 29.Sufficient lightening available without glare from lights, windows, surfaces, and etc… 30.Frequent use of telephone between head and shoulder 31.Neutral position of the wrist (straight from fingers to the elbow) 32.Neutral position of the head and the neck 33.Elbow/arm support provided for intensive/long durations 34.Leg support provided for intensive/long durations 35.Change sitting position at least every 15 minutes 36.Take active breaks (phone call, file paper, drink water, etc…) every 30 minutes 37. Take frequent microbreaks (while seated on your workstation) 38.Trained in proper posture 39.
then the odds ratio is defined as

Table 2 .
Type of discomfort and percent occurrence in the shoulder.

Table 3 .
Frequency of discomfort.

Table 4 .
Logistic Regression of Ergonomic Factors that affects the Shoulder.

Table 5 .
Logistic Regression of Feelings of Discomforts in the Shoulder.

Table 6 .
Logistic Regression of Frequency of Discomforts in the Shoulder.

Table 7 .
illustrates that test group respondents have significantly high levels of muscle activities.There was a very significant Muscle activities (µV) of the test group respondents at the shoulder.increase in the shoulder muscle activity of the test group respondent 3 throughout the experiment.Test group respondent 5 has been suffering from discomforts at the shoulder very significantly more than that of the other 5 respondents.Test group respondent 1 was experiencing almost a constant shoulder muscle activity during the experiment.The test group respondent 6 was observed to have a decreasing muscle activity during the experiment.

Table 8
shows the ANOVA results for the test group respondents.

Table 9
illustrates the mean value for each 30 seconds interval sEMG readings for each control group respondent.

Table 9 .
Muscle activities (µV) of the control group respondents at the shoulder.

Table 10
shows the ANOVA results for the control group respondents.

Table 10 .
ANOVA results for the control group respondents.