ARTICLE IN PRESS

International Journal of Industrial Ergonomics 35 (2005) 569581 www.elsevier.com/locate/ergon

Prevalence of upper extremity musculoskeletal symptoms and ergonomic risk factors at a Hi-Tech company in Israel
Kerem Shuvala,, Milka Donchinb
b a School of Public Health, University of Haifa, Mount Carmel, Haifa 31905, Israel School of Public Health and Community Medicine, Hadassah and the Hebrew University, P.O.B 12000, Jerusalem 91120, Israel

Received 24 June 2004; received in revised form 10 September 2004; accepted 4 January 2005 Available online 2 March 2005

Abstract This cross-sectional study examines the relationship between ergonomic risk factors and upper extremity musculoskeletal symptoms (UEMSS) in VDT workers at a Hi-Tech company, while taking into account individual and work organizational factors, and stress. The study population of 84 workers (92% response rate) is comprised of computer programmers, managers, administrators, and marketing specialists. Data on UEMSS, individual and organizational factors, and stress were derived from a questionnaire, while ergonomic data were collected through two direct observations via the rapid upper limb assessment (RULA) method. Results of the RULA observations indicate excessive postural loading with no employee in acceptable postures (all scores X3). Hand/wrist/nger symptoms were related to the RULA arm/wrist score (in a logistic regression model) as well as working with a VDT between 7.1 and 9 h a day, and working in Hi-Tech companies for more than 2 years. Neck/shoulder symptoms were related to: gender (female), working 410 h a day, working at the Hi-Tech company for more than 2 years, and being uncomfortable at the workstation. The results underline the need for implementing an intervention program focusing on arm/wrist posture, and for taking into account the special needs of subgroups: gender, working 10 h a day, working 7.19 h a day with a VDT, and employees experiencing discomfort at workstations. Relevance to industry: Upper extremity musculoskeletal disorders and ergonomic research have not focused enough on the Hi-Tech industry. Even with the .com crash, this industry remains a major force in the world economy. VDT work in the Hi-Tech industry has unique risk factors, which might lead to specic needs for intervention. r 2005 Elsevier B.V. All rights reserved.
Keywords: UEMSD; Hi-Tech; VDT; Work posture; RULA

1. Introduction
Corresponding author. Tel.: +972 4 6011603; fax: +972 4 8288637. E-mail addresses: kshuval@study.haifa.ac.il (K. Shuval), milka@hadassah.org.il (M. Donchin).

Ample research has been conducted on the prevalence of upper extremity musculoskeletal disorders (UEMSD) in VDT workers. Such

0169-8141/$ - see front matter r 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.ergon.2005.01.004

ARTICLE IN PRESS
570 K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581

research has focused on the multi-factorial etiology of work-related UEMSD (Bernard, 1997; WHO, 1989). Different aspects of VDT work i.e., physical, work organizational and psychosocialinteract and lead to UEMSD in VDT users (Bergqvist et al., 1995b; Punnett and Bergqvist, 1997). For example, hand/wrist disorders are caused by extensive use of the keyboard, which poses a combination of ergonomic risk factors (mechanical pressure on the soft tissue of the forearm and wrists) and work organizational factors (hours of keying a day). Additionally, in an extensive review by Bongers et al. (2002), high-perceived job stress was consistently associated with UEMSD in cross-sectional studies. High-perceived job stress might, in turn, cause physiological changes leading to musculoskeletal problems. Most of the research done has focused on the telecommunication industry, banks, and newspapers, all VDT jobs that involve data entry, word processing, computertelephone tasks, and CAD (Cook et al., 2000; Hales et al., 1994; Karlqvist et al., 1994, 1996; Ferreira et al., 1997; Polanyi et al., 1997; Sauter et al., 1991; Yun et al., 2001). To the best of our knowledge, no studies examining work-related UEMSD, using Hi-Tech companies as the setting, have been published. In addition, scant research has centered on computer programmers. In a study by Jensen et al. (2002), programmers comprised 3% of the sample, and in a study by Karlqvist et al. (2002), 9.7% of the sample were programmers. Neither study focused on specic risk factors and prevalence rates of UEMSD among computer programmers. Computer programmers and other Hi-Tech workers might encounter specic risk factors compared with those found in other VDT work, especially when taking into account the numerous hours worked a week, the high job stress setting, the young age of employees and high education level. Some studies have found that working with a VDT for 2 h a day is considered a risk factor for musculoskeletal disorders (MSD), so what is the effect of working 10, 11, or even 12 h a day with a VDT (Faucett and Rempel, 1994; Karlqvist et al., 2002; Oxenburgh et al., 1985)?

This study examines musculoskeletal symptoms in a Hi-Tech communication company in northern Israel. The aims of the study are: (i) To determine the prevalence of upper extremity musculoskeletal symptoms (UEMSS) stratied by the following risk factors: individual, occupational characteristics, and job stress. (ii) To examine the association between ergonomic workplace risk factors and UEMSS.

2. Methods 2.1. Subjects The entire software communication Hi-Tech company was included in the study. Out of 91 employees, three from management declined, three were on vacation, and one employee was on maternal leave. Thus, seven employees did not participate due to the above-mentioned reasons, resulting in a 92% response rate. The study examined a relatively small Hi-Tech company, which agreed to participate in this study. More than 10 medium and large Hi-Tech companies from northern and central Israel declined to participate in the research. 2.2. Methods of collecting data and analysis Data were gleaned from a questionnaire and direct observation. This questionnaire was based on the Nordic Musculoskeletal Questionnaire (Kourinka et al., 1987). Subjects were asked to report musculoskeletal symptoms both in the last year and the last week in the neck, shoulder, hand/ wrist, and ngers. The neck and shoulder symptoms were placed into one category. Therefore, an employee reporting having trouble (ache, pain, discomfort) in the neck and/or the shoulder was dened as suffering from UEMSS in the neck/ shoulder region. Additionally, hand/wrist and nger symptoms were also brought into one category. Hence, an employee reporting having trouble in the hand/wrist and/or the ngers was

ARTICLE IN PRESS
K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 571

regarded as suffering from UEMSS in the hand/ wrist/shoulder region. These grouping were made because: (a) statistically signicant associations were found between neck and shoulder symptoms (OR 4.5; 95% CI 1.414.6), and hand/wrist and nger symptoms (OR 5.0; 95% CI 1.122.9); (b) the number of subjects in each category was relatively small; (c) these categories are acceptable in the literature (Bergqvist et al., 1995a; Marcus and Gerr, 1996; Gerr et al., 2002). Subjects were also asked if they had been injured in an upper extremity accident unrelated to their job at the Hi-Tech company. Those injured were excluded from the group, and were not dened as suffering from UEMSS in the specic region. UEMSS in the last year and the last week were analyzed separately both in univariate analysis and logistic regression models. A logistic regression model of hand/wrist/nger symptoms in the last week was not performed due to the low prevalence of this problem in the study population (4.8%). In addition, the questionnaire included questions regarding personal characteristics, such as comfort at the workstation, workstation alteration, wearing glasses, and physical activity. Occupational characteristics consisted of daily working time, daily working time with a VDT, consecutive work with a VDT, company experience, Hi-Tech experience, and additional tasks required beyond VDT work; all derived from the questionnaire. Occupational characteristics also included job description (programming, management, etc.) which was collected from the Human Resource Manager of the company for accuracy. Job stress was determined via a questionnaire aimed at assessing perceived stress in Hi-Tech settings (Toviana, 1999). Subjects were asked to determine whether each of the 11 statements affected their perceived on-the-job stress based on a 15 Lickert scale (see Table 3). Each subject received a stress score comprised of the average score of 11 variables. This unication was possible after the internal reliability was found to be high (Cronbach a 0:85). The mean job stress score was entered as a continuous variable into the logistic regression models. Ergonomic risk factors were assessed through direct observation of employees postures at their

workstations using the rapid upper limb assessment (RULA) tool (McAtamney and Corlett, 1993). RULA is a validated tool that assesses biomechanical and postural loading on the upper limbs. The RULA scoring system generates an action list indicating the level of intervention required to reduce the risks of injury. For instance, a nal score of 1 or 2 indicates an acceptable humancomputer interface, while a score of 7 represents a completely unacceptable interface, emphasizing the need to make immediate modications. The RULA assessment was performed by one of the authors (KS) in order to avoid interobserver variation. The assessment represents a moment in the work cycle; therefore, an a priori decision was made to examine the longest adopted posture during two 30-min work cycles (with a 1-h interval in between) for enhanced validity. The correlation (Pearson) between the two observations was medium to high (r 0:420:7); hence, the mean score of the two observations was used for analysis. For data analysis, three RULA scores were used: the nal arm/wrist score, nal neck/ trunk/leg score, and the nal score. The RULA score was entered as a continuous variable into the logistic regression models. 2.3. Statistical methods Descriptive and analytical statistics were carried out using chi-square test, ANOVA, and logistic regression analysis while calculating OR of the different categories compared to the reference category. Condence intervals were in the 95% range. The logistic regression analysis was performed using the backward method to detect as many as possible covariates independently associated with UEMSS. Variables that remained in the nal model were analyzed again in a regression model using the Enter method. Job stress was entered into the models in order to address an important concern in the literature (Bongers et al., 2002). Because one model comprising the different musculoskeletal symptoms is not possible in logistic regression, a number of models were used. Thus, the type one error might be greater than stated.

ARTICLE IN PRESS
572 K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 Table 1 General characteristics of study population General characteristics Age 2329 3039 4067 Gender Male Female Marital status Single/divorced Married Number of children 0 1 X2 Glasses No Yes VDT use on weekends (h) No Yes Physical activity No Yes Workstation altered No Yes No. (%)

Chi-square test was used to compare categorical variables, while ANOVA was used to compare means. Correlation between continuous variables was determined using Pearson correlation and the signicance was two tailed, with a 0:01: Data were analyzed using SPSS 11

41 (48.8) 30 (35.7) 13 (15.5) 57 (67.9) 27 (32.1) 27 (32.1) 57 (67.9) 47 (56.0) 17 (20.2) 20 (23.8) 33 (39.3) 51 (60.7) 20 (23.8) 64 (76.2) 45 (53.6) 39 (46.4) 57 (67.9) 27 (32.1)

3. Results 3.1. Population description The study population consisted of a majority of male subjects (67.9%) with an average age of 32. More than half (61.9%) of the subjects worked as computer programmers or in a related eld (quality assurance, technical support, system administrators). The rest were management, marketing, and secretarial (Table 1). Most (87.8%) subjects between the ages of 2329 worked as programmers, compared to 15.4% of those between the ages of 40 and 67 (Po0:001). More than half (60.7%) of the study population wore glasses, most of them (76.1%) used a VDT on weekends. Less than half (46.4%) the workers were physically active (e.g., aerobic activity at least three times a week and/or resistance training at least twice a week). A third (32.1%) of the employees altered their workstation in order to improve subjective comfort. 3.2. Occupational characteristics Table 2 shows that 65.5% of employees worked 810 h daily and the rest more than 10 h a day. The employees worked an average of 7.8 h on the VDT a day (S.D. 2.07). Employees worked an average of 20.9 months in the HiTech company, but 54% had worked at a different Hi-Tech company beforehand. Almost all (95.2%) employees worked 5 days a week and 48.8% reported having additional tasks to their computer work, consisting of ling, reading and writing hardcopy documents, and using hardware.

3.3. Perceived job stress Out of 11 stress factors, having to nish a project within a strict deadline was perceived as the most stressful (mean 3.1 on a ve-point scale, S.D. 1.2) followed by conict between work and family obligations (mean 2.9, S.D. 1.39; see Table 3). In addition, each subject received a job stress score, a mean score of 11 components (see Fig. 1). The mean job stress score of the study population was 2.54 (S.D. 0.81), while the minimum score was 1 and the maximum 4.18. Perceived job stress of men and women was almost identical (2.52 and 2.57, respectively). Computer programmers received a higher job stress score than management; however, the

ARTICLE IN PRESS
K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 Table 2 Occupational characteristics Occupational characteristics Job description Computer programmers Management, marketing, etc. Daily working time (h) 810 10.112 Daily working time with VDT (h) 15 68 8.112 Consecutive work with a VDT (h) p1 41 Company experience (months) 124 2560 Hi-Tech experience (years) p2 2.14 4.130 Additional tasks None Meetings, management, etc. Reading, writing, ling, etc. No. (%) 573

14 12 10
52 (61.9) 32 (38.1) 55 (65.5) 29 (34.5) 10 (11.9) 37 (44.0) 37 (44.0) 37 (44.0) 47 (56.0) 63 (75.0) 21 (25.0) 29 (34.5) 25 (29.8) 30 (35.7) 25 (29.8) 18 (21.4) 41 (48.8)

Frequency

8 6 4 2 0 1.00 1.50 2.00 2.50 3.00 3.50 4.00 1.25 1.75 2.25 2.75 3.25 3.75 4.25 Mean job stress scores

Fig. 1. Distribution of the mean job stress scores of the study population.

difference was not statistically signicant (2.67 and 2.32, respectively; P 0:061). 3.4. Prevalence rates The highest prevalence rate of musculoskeletal pain or discomfort was in the neck/shoulder region: 47.6% reported of having a problem in the last year and 15.5% in the last week. The prevalence of UEMSD in hand/wrist/nger region was 32.1% in the last year compared to 4.8% in the last week. 3.5. Ergonomic factors RULA: The mean wrist score of 2.3 indicates that the wrists of employees were in extension (sagittal plane) of up to 151 in both observations. The mean neck score of 1.9 means that the necks of the population were in either exion, rotation, or side bending. The nal mean RULA score of 4.5 indicates that the employees postures at their workstations need to be investigated further and altered in some cases. There were no employees in the Hi-Tech company who received RULA scores of 12, which indicates an acceptable posture. The nal RULA score ranged from a minimum of 3 to a maximum of 6.5 (see Fig. 2). The mean arm/wrist RULA score was identical in men and women

Table 3 Means and standard deviations (S.D.) of job stress factors Job stress factor 1. Working many hours 2. Multiple tasks 3. Having to nish a project within a strict deadline 4. Lack of appreciation from superiors 5. Communication problems within team/company 6. Lack of professional challenges 7. Situations of uncertainty 8. Conict between work and family obligations 9. Loneliness/lack of support 10. Support from supervisors in resolving work-related issues 11. Contradictory demands from superiors Mean 2.7 2.8 3.1 2.5 2.1 2.4 2.9 2.9 2.1 2.1 2.1 S.D. 1.2 1.2 1.1 1.4 1.2 1.3 1.4 1.4 1.3 1.3 1.3

ARTICLE IN PRESS
574 K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581

40

30 Frequency

20

10

0 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 RULA: final score
Fig. 2. Distribution of the nal RULA score of the study population. Fig. 3. A Hi-Tech employee at his workstation.

(4.0); however, womens mean neck/trunk/leg score was higher than men (4.7 and 4.2, respectively, P 0:057). Women also received higher nal RULA scores than men (mean 4.7 and 4.4, respectively, P 0:132). VDT workstations: The nal RULA score (mean 4.5) emphasizes a poor VDT workstation design. The primary work surface was too high, 75 cm, in all workstations without the option of adapting the height of the surface (Salvendy, 1997). In addition, conventional keyboards were used with a positive slope of 201 and the tilt feet were up creating an additional slope (Simoneau et al., 1999). The computer screen was placed at the same primary work surface. Additionally, too much hardware (faxes, additional PCs) was placed on the work surface, which might lead to lack of alignment between keyboard and screen (see Fig. 3). Subjective comfort at workstations: 33.3% of the employees reported feeling uncomfortable in their workstations. Employees feeling uncomfortable scored a higher RULA wrist score than those who felt comfortable. 3.6. Relationship of risk factors to UEMSS univariate analysis Neck/shoulder symptoms: Table 4 presents the association between symptoms and risk factors.

Subjects who wear glasses reported 3.5 more neck/ shoulder symptoms in the last week than those who do not wear glasses. Furthermore, employees working over 2 years in the Hi-Tech company had 1.6 more symptoms in the last year than employees who worked o2 years in the company. Subjects who worked 410 h a day had three times more symptoms in the last week than those who work p10 h a day. Hand/wrist/ngers symptoms: Subjects employed 44 years in Hi-Tech companies reported symptoms 2.6 times more than subjects employed p2 years in the Hi-Tech industry. Moreover, subjects working between 2 and 4 years reported 3.5 more symptoms than the reference group (Table 5).

3.7. Relationship between risk factors and UEMSSlogistic regression analysis Neck/shoulder symptoms in the last year: Three statistically signicant risk factors were found (Table 6): being married (5.2 times higher risk than non-married), company experience, and less Hi-Tech experience. Physical inactivity, found as an independent risk factor, was not statistically signicant (OR 2.8; P 0:054). Neck/shoulder symptoms in the last week: The main predictor of symptoms in the last week was subjective discomfort at the VDT workstation

ARTICLE IN PRESS
K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 Table 4 Prevalence of neck/shoulder symptoms by independent variablesunivariate analysis Symptoms Independent variables Gender Male Female Marital status Not married Married Job description Computer programmers Management, marketing, etc. VDT use on weekends (h) No Yes Glasses No Yes Company experience (months) 124 425 Daily working time (h) 810 10.112 N Last year (%) P-value Last week (%) P-value 575

0.142 57 27 27 57 52 32 20 64 33 51 63 21 55 29 42.1 59.3 0.071 33.3 54.0 0.091 40.4 59.4 0.075 65.0 42.2 0.443 42.4 51.0 0.044 41.3 66.7 0.583 45.5 51.7 9.1 27.6 15.9 14.3 6.1 21.6 20.0 14.1 17.3 12.5 22.2 12.3 14.0 18.5

0.596

0.333

0.758

0.498

0.055

1.000

0.053

Table 5 Prevalence of hand/wrist/ngers symptoms by independent variablesunivariate analysis Symptoms Independent variables Gender Male Female Marital status Not married Married Hi-Tech experience (years) p2 2.14 4.130 N Last year (%) P-value Last week (%) P-value

0.509 57 27 27 57 29 25 30 42.1 59.3 0.401 25.9 35.1 0.022 13.8 48.0 36.7 3.4 8.0 3.3 11.1 1.8 14.0 18.5

0.591

0.096

0.663

(Table 7). Daily working time (410 h a day) was also a risk factor. Hand/wrist/nger symptoms in the last year: An increase of 1 point in the RULA score more than

triples (3.2 times) the odds for hand/wrist/nger symptoms (Table 8). In addition, daily working time and Hi-Tech experience were also identied as independent risk factors.

ARTICLE IN PRESS
576 K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 Table 8 Determinants of hand/wrist/nger symptoms in the last year logistic regression Variable N OR 95% CI

Table 6 Determinants of neck/shoulder symptoms in the last year logistic regression Variable Gender Male Female Marital status Not married Married Physical activity No Yes Company experience (months) 124 425 Hi-Tech experience (years) p2 2.14 4.130 Job stress Continuous (per 1 score) N OR 95% CI

57 27 27 57 45 39 63 21 29 25 30 84

1 2.53 1 5.24 2.85 1 1 5.41 5.16 1.95 1 1.33

0.837.68

Daily working time with VDT (h) 27 31 7.19 33 9.112 20 Hi-Tech experience (years) p2 29 2.14 25 4.130 30 RULA hand/wrist score Continuous (per 1 score) 84 84

1 4.39 1.73 1 6.28 4.85 3.18 0.99

1.2715.17 0.397.56

1.3819.84 0.988.31

1.5225.96 1.2019.50 1.148.85 0.511.94

1.4619.95 1.1523.07 0.556.91

Job stress Continuous (per 1 score)

0.712.48

Table 7 Determinants of neck/shoulder symptoms in the last week logistic regression Variable Daily working time (h) 810 10.112 Comfort at workstation Uncomfortable Comfortable Job stress Continuous (per 1 score) N OR 95% CI

55 29 28 56 84

1 3.91 6.38 1 0.96

1.0414.59 1.5426.47

0.412.23

4. Discussion This prevalence study examines a Hi-Tech communication software company in northern Israel consisting of 91 employees. It was chosen after 10 medium- and large-scale companies refused, claiming the study would lead to unwanted expenses on ergonomic equipment. The

company examined is relatively small, without the hierarchy, rules, and regulations associated with large-scale companies. The study population is unique in some ways: 62% are computer programmers, who are young (mean age of 32), mostly men (68%), and all work long hours a day (average 10.1 h). The prevalence of neck/shoulder symptoms in the last year (47.6%) in the company was higher than found in Blatter and Bongers (2002) study (10.3%), and similar to Jensen et al. (2002) study (neck 43%; shoulder 35%). A higher prevalence of neck/shoulder symptoms was found by Bergqvist et al. (1995a), 61.5%, and by Cook et al. (2000), 59.9%. The higher prevalence in these studies could be explained by the higher percentage of women. Other studies by Sauter et al. (1991), Bernard et al. (1994), Yu and Wong (1996), and Polanyi et al. (1997) found lower prevalence rates than this study (27%, 22%, 31%, and 16%, respectively). This discrepancy may stem from the use of a more severe case denition based on the NIOSH questionnaire and not on the Nordic. The NIOSH questionnaire uses a body part diagram, similar to the Nordic diagram, but adds questions to determine frequency, duration, and intensity of symptoms (Baron et al., 1996). The prevalence of hand/wrist/nger symptoms in the last year was 32.1%, and 4.8% in the last

ARTICLE IN PRESS
K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 577

week. Bergqvist et al. (1995a), Marcus and Gerr (1996), Cook et al. (2000), and Jensen et al. (2002) found similar prevalence rates in the last year in their studies (29.9%, 34.2%, 36.8%, and 26%, respectively). A number of studies found lower prevalence rates between 13% and 21.7% (Hales et al., 1994; Polanyi et al., 1997; Sauter et al., 1991; Yun et al., 2001), but case denitions were different (some included physical examination). This study, like many others, found a higher prevalence of UEMSS in women compared to men (Bernard et al., 1994; Bergqvist et al., 1995b; Burt et al., 1990; Gerr et al., 2002; Grieco et al., 1989; Jensen et al., 2002; Karlqvist et al., 2002). However, Cook et al. (2000) and Hales et al. (1994) did not nd such an association. The higher prevalence of neck/shoulder symptoms in women could be explained by a differential in exposure. The results of the study indicate that women received higher neck/trunk/leg RULA scores than men (mean 4.7 and 4.2, respectively, P 0:057). This could be explained by the fact that women have different anthropometry than man, i.e., their average body size is smaller and thus more susceptible to the effects of a high unadjustable primary working surface. A number of studies employing physical examination to determine UEMSD recorded a higher prevalence of disorders in women, thereby supporting the ndings of the present study (Bergqvist et al., 1995b; Gerr et al., 2002; Karlqvist et al., 1999). No association was found between job description and UEMSS. The prevalence of UEMSS among computer programmers was not higher than the rest of the company. However, six times more programmers than non-programmers reported feeling uncomfortable at their workstation. No interaction between job description and gender was found. Unlike other studies, no association was found between age and UEMSS (Cook et al., 2000; Gerr et al., 2002; Punnett and Bergqvist, 1997). The lack of association might be attributed to the narrow age distribution in the study population. In addition, no association between UEMSS and job stress was found, both in univariate and multivariate analysis. This nding is not consistent with previous reports focusing on ofce work (Bergqvist et al., 1995a; Bongers et al.,

2002; Cook et al., 2000; Marcus and Gerr, 1996). This inconsistency might be explained by the unique characteristics of the present study population, comprised of young, strongly motivated HiTech employees. The current study suggests an association between physical activity and neck/shoulder symptoms. The odds for physically inactive employees to report neck/shoulder symptoms are approximately three times higher than for their physically active counterparts (OR 2.8, 95% CI 0.988.31). Although this association is not statistically signicant chances are that a larger sample size would detect a statistically signicant association (based on the condence interval). This association should be examined in a larger-scale study. A number of studies found no association between physical inactivity and higher odds for musculoskeletal complaints, perhaps due to a different case denition (Karlqvist et al., 2002; Marcus and Gerr, 1996; Yun et al., 2001). This study had a more extreme case denition of physical activity: three times a week of aerobic activity and/or two times a week of muscular strength training. Engaging in physical activity less than the abovementioned frequency does not promote positive physiological and musculoskeletal changes (US Department of Health and Human Services, 1996). Katz (2000) similarly found that physically active students (on a sports team) had less of a chance for UEMSD than students who were not physically active. Subjects in the Hi-Tech company working 410 h a days had greater odds for reporting neck/shoulder symptoms in the last week (OR 3.9) than subject working less than 10 h daily. These ndings concur with other studies that found that working 69 h a day is a risk factor for UEMSD (Bernard et al., 1993, 1994; Yun et al., 2001). Additionally, subjects working 7.19 h had greater odds for hand/wrist/nger symptoms (OR 4.4), compared to subjects working p7 h. Bernard et al. (1993, 1994) and Yun et al. (2001) found similar results, that working 46 h is a risk factor for UEMSD. Faucett and Rempel (1994), Karlqvist et al. (2002), and Oxenburgh et al. (1985) found that even 2 h of VDT work is a risk for UEMSD.

ARTICLE IN PRESS
578 K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581

A higher prevalence of neck/shoulder symptoms was found in employees working 42 years in the company compared to those working o2 years. Yun et al. (2001) found that employees working 48 years in the bank experienced more shoulder symptoms (RR 1.5) and more wrist symptoms (RR 1.3) than subjects employed o2 years. In addition, an association was traced between Hi-Tech experience and UEMSS. Subjects working p2 years and between 2.1 and 4 years in the HiTech industry had higher odds for neck/shoulder symptoms in the last year than subjects working 44 years in Hi-Tech (OR 5.1 and 1.9, respectively). An opposite trend was discerned with hand/wrist/nger symptoms. Subjects employed 2.14 and 44 years in the Hi-Tech industry had higher odds for hand/wrist/nger symptoms than those employed p2 years (OR 6.3 and 4.8, respectively). These results might indicate a different association between the development of musculoskeletal symptoms in the neck/shoulder region and hand/wrist region to cumulative HiTech experience. Hence, neck/shoulder symptoms might appear at the initial stage of employment in the Hi-Tech industry, whereas it would take a few years for hand/wrist/nger symptoms to appear. This explanation is possible, but cannot be substantiated due to the cross-sectional nature of the study. In a prospective study, Gerr et al. (2002) established a temporal relationship between past computer use and the incidence of musculoskeletal symptoms, thus supporting the above-mentioned explanation. They found an increased risk of developing hand/arm symptoms among employees with X2 years of computer experience, whereas no such association arose between previous computer work and neck/shoulder symptoms. Marcus and Gerr (1996), in an earlier cross-sectional study, found higher odds for hand/arm symptoms only in the 46 years of VDT use category; whereas higher odds for neck/shoulder symptoms were present in all categories of VDT experience (o3, 46, 46). An extensive review by Punnett and Bergqvist (1997) concluded that VDT work was a direct causative agent for hand and wrist disorders mediated through repetitive nger motion and sustained muscle loading across the forearm and wrist. However, neck and shoulder disorders are

not caused by sustained loading alone, but rather by a combination of exposures. These ndings might further support the explanation of differences between the two regions. The primary association examined in this study is the association between ergonomic risk factors (via RULA scores) and UEMSS. In univariate analysis subjects with hand/wrist/nger symptoms in the last year had a higher RULA arm/wrist score, indicating poor posture of the region. The same association was found in logistic regression. This association could be explained by the fact that the wrist posture (sagittal plane) of employees with symptoms was in a greater extension than employees without symptoms (4151 compared to p151). The more the wrist is extended the greater the inter-carpal tunnel pressure becomes (Rempel et al., 1992, 1997a, b). Inter-carpal tunnel pressure over a period of time decreases the blood ow to the median nerve, which may lead to discomfort, pain, or even to CTS in extreme cases. The obstruction in the blood ow to the region may lead to a loss of feeling (sensory dysfunction). This leads the VDT worker to use excessive force when keying, which, in turn, will lead to additional damage to the median nerve. All employees at the Hi-Tech company are prone to wrist discomfort or pain because of the non-neutral position of their wrists and extensive keying. Employees whose wrists were in an extreme position (4151 or radial/ulnar deviation) experienced higher prevalence rates. McAtamney and Corlett (1993) found similar ndings. Subjects with a non-optimal arm/wrist position reported pain signicantly greater than subjects in an optimal position. Hedge et al. (1999) found that improving wrist posture lowered the prevalence of disorders. Gerr et al. (2002) found increased odds for wrist disorders when wrists are in extension (OR 1.96). Bergqvist et al. (1995b) and Sauter et al. (1991) found an association between nonoptimal keyboard placements and increased prevalence of wrist disorders. This study did not nd an association between neck posture (RULA score) and neck/shoulder symptoms. Sauter et al. (1991) found similar ndings to the present study; however, many studies found an association between poor neck

ARTICLE IN PRESS
K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 579

posture and a higher prevalence of neck and shoulder disorders (Bergqvist et al., 1995a; Cook et al., 2000; Heyer et al., 1990; Hochanadel, 1995; Hoekstra et al., 1994). The outcome in the present study could be explained by the fact that essentially all employees necks were not in an optimal position (either in exion or in extension). Therefore, employees with symptoms might have a predisposition and not necessarily overexposure. The fact that an association was found between arm/wrist posture and musculoskeletal symptoms and not in the neck/shoulder region could be explained by physiological differences between these regions. Gerr et al. (2002) found that after working 1 year with a VDT the incidence of arm/ wrist disorders increased 10-fold compared to a ve-fold increase in neck/shoulder disorders. This could indicate a different pathophysiology between the regions, as already mentioned. Subjective discomfort at the workstation was another variable examined. Subjects with extreme wrist positions (4151 extension or radial/ulnar deviation) reported discomfort at their workstations. These subjects also had a higher prevalence of musculoskeletal symptoms than subjects with a less extreme posture. Additionally, subjects who were uncomfortable reported more neck/shoulder symptoms than employees who were comfortable at their workstations. Hoekstra et al. (1994) found similar results: more employees who were uncomfortable at their workstation reported neck/ shoulder symptoms than those who were comfortable. The importance of these ndings is that the simple question: Do you feel comfortable at your workstation? was associated with the prevalence of symptoms. This question could be used as an initial screening tool. Subjects who indicate discomfort should be physically examined, accompanied by an ergonomic evaluation.

the anthropometric needs of the employees. These conditions, in turn, lead to poor posture as well as discomfort and pain. In addition, a number of subgroups are at risk: women, employees working 410 h a day, working 7.19 h a day with a VDT, and subjects employed 42 years in a Hi-Tech company. An essential conclusion drawn from this study is the importance of asking employees about their subjective feeling. The association between workstation discomfort and UEMSS emphasizes this need. These conclusions underline the need for implementing an intervention program focusing on arm/wrist posture, and for taking into account the special requirements of subgroups.

References
Baron, S., Hales, T., Hurrell, J., 1996. Evaluation of symptom surveys for occupational musculoskeletal disorders. American Journal of Industrial Medicine 29, 609617. Bergqvist, U., Wolgast, E., Nilsson, B., Voss, M., 1995a. Musculoskeletal disorders among video display terminal workers: individual, ergonomic, and work organizational factors. Ergonomics 38, 763776. Bergqvist, U., Wolgast, E., Nilsson, B., Voss, M., 1995b. The inuence of VDT work on musculoskeletal disorders. Ergonomics 38, 754762. Bernard, B., 1997. Musculoskeletal Disorders and Workplace Factors. National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication No. 97-141, Cincinnati, OH. Bernard, B., Sauter, S., Petersen, M., 1993. Health Hazard Report: Los Angeles Times, Los Angeles, CA. National Institute of Occupational Safety and Health. Bernard, B., Sauter, S., Fine, L., Petersen, M., Hales, T., 1994. Job task and psychosocial risk factors for work related musculoskeletal disorders among newspaper employees. Scandinavian Journal of Work, Environment & Health 20, 417426. Blatter, B., Bongers, P., 2002. Duration of computer use and mouse use in relation to musculoskeletal disorders of neck or upper limb. International Journal of Industrial Ergonomics 30, 295306. Bongers, P., Kremer, A., Lack, J., 2002. Are psychosocial factors, risk factors for symptoms and signs of the shoulder, elbow, or hand/wrist?: a review of the epidemiological literature. American Journal of Industrial Medicine 41, 315342. Burt, S., Hornung, R., Fine, L., 1990. Health Hazard Evaluation Report: Newsday Inc., Melville, NY. NIOSH (HETA Report 89-250-2046). Cook, C., Burgess-Limerick, R., Chang, S., 2000. The prevalence of neck and upper extremity musculoskeletal

5. Conclusion A high prevalence of musculoskeletal symptoms has been found in the neck/shoulder region and hand/wrist region. The ergonomic conditions in the company featured in this study point to poor workstation design and lack of compatibility with

ARTICLE IN PRESS
580 K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 VDU operators and associations with musculoskeletal symptoms: a cross sectional study focusing on gender differences. International Journal of Industrial Ergonomics 30, 277294. Katz, J., 2000. Prevalence of upper extremity musculoskeletal disorders in college students. The American Journal of Medicine 109, 586588. Kourinka, B., Jonsson, B., Kilbom, A., et al., 1987. Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms. Applied Ergonomics 18, 233237. Marcus, M., Gerr, F., 1996. Upper extremity musculoskeletal symptoms among female ofce workers: associations with video display terminal use and occupational psychosocial stressors. American Journal of Industrial Medicine 29, 161170. McAtamney, L., Corlett, N., 1993. RULA: a survey method for the investigation of work-related upper limb disorders. Applied Ergonomics 24, 9199. Oxenburgh, M., Rowe, S., Douglas, D., 1985. Repetition strain injury in keyboard operators: successful management over a 2 year period. Journal of Occupational Health and Safety (Australia, New Zealand) 1, 106112. Polanyi, M., Cole, D., Beaton, D., et al., 1997. Upper limb work related musculoskeletal disorders among newspaper employees: cross-sectional survey results. American Journal of Industrial Medicine 32, 620628. Punnett, L., Bergqvist, U., 1997. Visual display unit work and upper extremity musculoskeletal disorders: a review of epidemiological ndings. National Institute for Working LifeErgonomic Expert Committee Document No. 1. Arbetslivsininstitutet, Solna, Sweden. Rempel, D., Harrison, R., Barnhart, S., 1992. Work related cumulative trauma disorders of the upper extremity. JAMA 267, 838842. Rempel, D., Kier, P., Smutz, W., Hargen, A., 1997a. Effects of static ngertip loading on carpal tunnel pressure. Journal of Orthopedic Research 15, 422426. Rempel, D., Serina, E., Klinenberg, E., 1997b. The effect of keyboard keyswitch make force on applied force and nger exor muscle activity. Ergonomics 40, 800808. Salvendy, D., 1997. Handbook of Human Factors and Ergonomics, second ed. Wiley, New York, pp. 16381683. Sauter, S., Schleiffer, L., Knutson, S., 1991. Work posture, workstation design, and musculoskeletal discomfort in a VDT data entry task. Human Factors 33, 151167. Simoneau, G., Marklin, R., Monroe, J., 1999. Wrist and forearm posture of users of conventional computer keyboards. Human Factors 41, 413424. Toviana, Y., 1999. HiTech stress questionnaire. HiTech Life, SeptemberOctober, 2124, Hebrew. US Department of Health and Human Services, 1996. Physical activity and health: a report of the Surgeon General. US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Atlanta, GA, pp. 6173.

symptoms in computer mouse users. International Journal of Industrial Ergonomics 26, 347356. Faucett, J., Rempel, D., 1994. VDT related musculoskeletal symptoms: interactions between work posture and psychosocial work factors. American Journal of Industrial Medicine 26, 597612. Ferreira, J., Conceicao, G., Saldiva, P., 1997. Work organization is signicantly associated with upper extremities musculoskeletal disorders among employees engaged in interactive computertelephone tasks of an international bank subsidiary in Sao Paulo, Brazil. American Journal of Industrial Medicine 31, 468473. Gerr, F., Marcus, M., Ensor, C., et al., 2002. A prospective study of computer users: I. Study design and incidence of musculoskeletal symptoms and disorders. American Journal of Industrial Medicine 41, 221235. Grieco, A., Molteni, G., De Vito, G., Sias, N., 1989. Epidemiology of musculoskeletal disorders due to biomechanical overload. Ergonomics 41, 12531260. Hales, T., Sauter, S., Peterson, M., et al., 1994. Musculoskeletal disorders among visual display terminals users in a telecommunications company. Ergonomics 37, 16031621. Hedge, A., Morimoto, S., McCrobie, D., 1999. Effects of keyboard tray geometry on upper body posture and comfort. Ergonomics 10, 13331349. Heyer, N., Checkoway, H., Daniel, W., Horstman, S., Camp, J., 1990. Self reported musculoskeletal symptoms among video display terminal operators. In: Sakurai, H. (Ed.), Occupational Epidemiology. Elsevier Science Publication, B.V., Amsterdam, pp. 3034. Hochanadel, C., 1995. Computer workstation adjustment: a novel process and large sample study. Applied Ergonomics 26, 315326. Hoekstra, E., Hurrell, J., Swanso, N., 1994. Health Hazard Evaluation Report: Social Security Administration Teleservices Centers, Boston MA. NIOSH, Heta Report 920382-2450. Jensen, C., Finsen, L., Sogaard, K., Christensen, H., 2002. Musculoskeletal symptoms and duration of computer and mouse use. International Journal of Industrial Ergonomics 30, 265275. Karlqvist, L., Hagberg, M., Selin, K., 1994. Variation in upper limb posture and movement during word processing with and without mouse use. Ergonomics 37, 12611267. Karlqvist, L., Hagberg, M., Koster, M., Wenemark, M., Anell, R., 1996. Musculoskeletal symptoms among computer assisted design (CAD) operators and evaluation of a self assessment questionnaire. International Journal of Occupational Environment 2, 185192. Karlqvist, L., Bernmark, E., Ekenvall, L., Hagberg, M., Isaksson, A., Rosto, T., 1999. Computer mouse and trackball operation similarities and differences in posture, muscular load and perceived exertion. International Journal of Industrial Ergonomics 23, 157169. Karlqvist, L., Tornqvist, E., Hagberg, M., Hagman, M., Toomingas, A., 2002. Self-reported working conditions of

ARTICLE IN PRESS
K. Shuval, M. Donchin / International Journal of Industrial Ergonomics 35 (2005) 569581 WHO, 1989. Work with visual display terminals: psychosocial aspects and health. Journal of Occupational Medicine 31, 957968. Yu, I., Wong, T., 1996. Musculoskeletal problem among VDU workers in a Hong-Kong Bank. Occupational Medicine 46, 275280. 581

Yun, M., Lee, Y., Eoh, H., Lim, S., 2001. Results of a survey on the awareness and severity assessment of upper-limb workrelated musculoskeletal disorders among female bank tellers in Korea. International Journal of Industrial Ergonomics 27, 347357.