The present study explored the work-related and personal factors among a special working population, namely, TFT-LCD factory workers. The multiple regression model showed that older age, repetition, and posture were associated with shoulder symptoms; however, psychological stress and work-related physical fatigue were not associated with shoulder symptoms. The results were similar to the population excluding participants with acute musculoskeletal disorders. Workers with more repetition, posture, or force risks reported a higher proportion of shoulder symptoms (Fig. 2.), and these findings support the combined effect of similar ergonomic risks for shoulder disorders. This combined effect approach was similar to the ‘Key Indicator Method for Manual Handling Operations’ and combines the main risk factors for force, repetition, posture, and others into a single risk score . Furthermore, we identified sex differences in the effect of exposure to the risk factors for shoulder symptoms. For men, older age, repetition, posture, and force were associated with shoulder symptoms. For women, posture was associated with shoulder symptoms. This finding implies that force was more important for shoulder disorders in men. Furthermore, permanent psychological stress was associated with shoulder symptoms in univariable regression, but the finding in the regression model was not found after considering other factors, such as ergonomic risks. To the best of our knowledge, this is the first study to investigate shoulder symptoms and physical examination of subacromial impingement syndrome accompanied by potential personal factors, ergonomic risks, and psychological stress among workers in an electronics factory.
The ergonomic risks or musculoskeletal disorders among workers in the TFT-LCD industry are rarely addressed. Only two studies indicated that high ergonomic risks for the shoulder area were associated with poor arm support and the discrepancy between the workstation and the workers’ anthropometry [9, 32]. No comprehensive analysis has considered different ergonomic risks (e.g., repetition, posture, force, and vibration) as well as personal factors and psychological stress for shoulder disorders. Nevertheless, studies in the semiconductor industry, which shares similar work procedures with those of the TFT-LCD industry, have indicated that shoulder symptoms are among the most prevalent musculoskeletal disorders [33,34,35]. In one of the earliest studies in 1986, Kilbom et al. indicated that flexion and abduction of the upper arm were associated with shoulder symptoms . Chandrasakaran et al. showed that prolonged sitting and trunk bending were associated with shoulder symptoms . Chee et al. indicated that prolonged sitting in awkward postures with the characteristics of a forward bent neck and tables that are too high may result in shoulder pain . Furthermore, due to the high accuracy requirements for tasks, such as inspection or manual assembly, workers tend to bend their necks forward to give optimal visual conditions and could cause shoulder symptoms . Aghilinejad et al. found that the use of magnification loupes may improve the visibility of electronic parts as well as improve the postures of assembly workers and may reduce musculoskeletal discomfort . The aforementioned findings were similar to those of the present study, which demonstrated that posture was an independent ergonomic risk for shoulder symptoms in the two regression models, indicating that awkward or extreme joint positions, joints held in fixed positions, stretching to reach items or controls, and working overhead were significant risk factors. The possible reasons accounting for the association between posture and shoulder symptoms are that arm elevation or prolonged sitting with awkward posture (e.g., bent neck) may place additional load on the musculoskeletal system of the shoulder. One of the main pathophysiological mechanisms of shoulder disorder (e.g., subacromial impingement syndrome) is compression of the tendons between the humeral head and the coracoacromial arch and ischaemia by impingement or increased intramuscular pressure as a result of arm elevation .
Regarding the association between vibration and shoulder symptoms, the meta-analysis indicated low to very low evidence for an association between shoulder disorders and hand-arm vibration (OR = 1.3) . In positive association studies, the study populations were special working populations, such as forestry workers, rock drill workers, construction workers, and railroad engineers [39,40,41,42], and the hospital served as the recruitment location . Hagberg et al. indicated that the exposure factors associated with rotator cuff tendinitis in different occupational groups were not the same . In the present study, an association was not found in the regression model, which was similar to earlier studies for electronics workers [8, 36]. It is possible that the difference in study populations or locations recruited may explain these inconsistent findings. A high prevalence of shoulder pain was found in the electronics industry, which could be related to repetitive lifting tasks, repetitive operating machines, and monotonous short cycles of tasks . Chee et al. found that repetitive tasks could increase the risk of shoulder pain . A longitudinal study in France indicated that repetitive work under time constraints contributed to the development of chronic neck and shoulder disorders after adjustment for age . Furthermore, Jonsson et al. showed that reorganizing monotonous and repetitive work into a more diverse pattern may improve work-related upper limb musculoskeletal disorders after a 2-year follow-up study . These findings among electronics workers are consistent with the present study, indicating that work that involves repeating the same motion every few seconds, a sequence of movements over twice per minute, and over half of the cycle time in the same sequence of motion were significant risk factors.
Regarding the association between force and shoulder symptoms, one systematic review revealed that shoulder load (OR = 2.0) and hand force exertion (OR = 1.5) were associated with shoulder disorders . Another systematic review indicated that the occurrence of subacromial impingement syndrome was associated with high maximal voluntary contraction, lifting, and high hand force (OR = 2.8–4.2) . Repetitive tasks using mechanical force that put stress on small areas increased the prevalence of neck or shoulder pain in the department of manual assembly in 18 electronics factories . One possible mechanism accounting for the association between force and shoulder symptoms is that the direction of the force performed increases muscular activity levels, especially in overhead work . Similar to the above studies, an association between force and shoulder symptoms was found (crude OR = 1.76–2.05) in the univariate analysis (Table 4), but an association was not found after the multiple regression. One possible reason accounting for the lack of association is that the production process is typically automated and the process changes to light objects in the electronics industry.
Regarding sex differences, the association between shoulder symptoms and the frequency of forceful exertions was higher for women than men in a sex-stratified analysis . Women are considered to be at a higher risk of shoulder disorders (e.g., rotator cuff syndrome) than men, possibly reflecting both biological predisposition and exposure to work-related repetitive biomechanical constraints [16, 17]. The biological distinctions between men and women, including anatomy, strength, hormones, neuromuscular control, and musculoskeletal flexibility , suggest a different vulnerability to these work-related risk factors for shoulder disorder. An association between force and shoulder symptoms was not found for women in the two regression models (Table 5). The sex difference may result from differences in the type of task assigned, which means different exposures to the constraints at work [16, 17]. Women and men in the same industry may have different tasks, interactions between equipment and tool dimensions, and work activities . In the present study, a majority of men workers (30.5%) were assigned to tasks that involved the handling of heavy objects, whereas fewer women workers (10.4%) were assigned to tasks that involved the handling of heavy objects. Therefore, it is possible that men workers had a higher opportunity of exposure to force risk than women. Furthermore, the results of the present study from the regression model found that repetition, posture and force (approximately significant) were risk factors for shoulder symptoms in men, whereas posture (approximately significant) was a risk factor in women (Table 5). Further investigation is needed to elucidate whether specific task assignments are associated with shoulder symptoms. Regarding posture and shoulder symptoms in women, earlier studies among women workers in electronics factories revealed that shoulder symptoms were the most common musculoskeletal disorders [33, 34]. Kilbom et al. indicated that flexion and abduction of the upper arm were associated with shoulder symptoms . Miranda et al. found that the risk of chronic shoulder disorders was associated with working in awkward postures in women (adjusted OR = 2.3) . The present findings in the regression model are consistent with earlier studies showing that posture (crude OR = 2.42–2.52; adjusted OR = 2.12, which was approximately significant) was associated with shoulder symptoms in women (Table 5).
The present study indicated that permanent psychological stress was a significant risk factor for shoulder symptoms in women based on univariable regression (Table 5). This finding is consistent with an earlier study that found that women may have jobs with higher psychosocial stress (e.g., high demands, low control), negatively impacting musculoskeletal health . Although the biological pathway for shoulder disorders is biomechanical, psychological factors (e.g., work stress) may function as intermediating factors affecting these ergonomic risk factors . Two possible reasons accounting for the association between psychological stress and shoulder symptoms for women are that the hardness of shoulder muscle for women is larger than that of men and that women are more sensitive to symptoms of their shoulder . Furthermore, women reported higher levels of work overload, stress, and conflict compared with men due to the combined stress from the workplace and family (e.g., taking care of children) . Women may accumulate risk factors related to work activities and activities of daily living; thus, high job demand contributes to the development of chronic neck and shoulder pain independently of age . The association between psychological stress and shoulder symptoms in women was not significant after the multiple regression, and only posture were associated. The present findings are inconsistent with earlier studies. It is possible that the use of different methods to assess psychological stress could explain the different findings given that the single-item question on the frequency of stress at work was applied in the present study. Further investigation is needed to explore the issue using the different psychosocial stress models (e.g., job demand-control model, effort–reward imbalance model) for high-risk groups in the industry instead of the single-item question.
Age is a predictor for shoulder symptoms in earlier studies [18, 19]. The reason may be related to the pathophysiological mechanisms of increasing degeneration of the shoulder tendons and the development of osteoarthrosis in shoulder joints [17, 48]. Although ageing may play a role in shoulder symptoms in the working population, contradictory findings have been indicated among some working populations exposed to high biomechanical risks . For example, Silverstein et al. indicated that age was marginally significant for shoulder disorders . The present study found that age is a risk factor for shoulder symptoms only for men in the regression model possibly because the modification of the age effect was different for different sexes . This present finding of examining subacromial impingement syndrome was similar to that of earlier studies that showed that subacromial impingement syndrome was a common cause of musculoskeletal pain in the general working population [12, 17, 52], and a correlation between shoulder symptoms and clinical signs of rotator cuff tendinitis by physical examination was identified .
Several limitations should be noted. First, this was a study in a single facility, and the generalizability of this study requires further assessment. Second, the cross-sectional design restricts the inference of causal relationships and can only determine the association between relevant risk factors and shoulder symptoms. Third, the multivariable logistic regression model did not consider other potential confounding variables for shoulder symptoms, such as lack of sufficient rest, work organization factors, physical activity in spare time, muscular endurance in the arms, and job satisfaction [19, 34, 54]. Fourth, the risk assessment tool used was mainly by self-administered questionnaire, and not by experts. The ergonomic risk factors analysed for repetition, posture, and force were only analysed for more than 2 h per shift; thus, one risk factor or combined risk factors with less this duration were not included. Moreover, the prolonged periods of exposure and working longer hours each day were not assessed. Thus, the results of the present study may be underestimated. Fifth, physical examination for subacromial impingement syndrome was only performed among workers with shoulder symptoms. The positive rate of subacromial impingement syndrome remains uncertain in workers without shoulder symptoms, and in the viewpoint of early prevention, further examination for non-shoulder symptom workers may explore subclinical shoulder cases. On the other hand, given that not all participants did receive a physical examination and shoulder symptoms served as outcomes, misclassification of health status may have occurred. Further investigation is needed to assess work-related shoulder disorders, such as integrating comprehensive exposure assessments into the intensity, duration, and frequency of ergonomic risk factors. This work may aid our understanding of the pathophysiological mechanisms of shoulder disorders as well as attributional fractions of relevant risk factors. Studies to improve the knowledge of sex and the physical and psychosocial aspects of job interactions could enhance workplace job design and policy on the prevention of work-related shoulder disorders.