Computer use at work is associated with self-reported depressive and anxiety disorder

In order to investigate the association between computer use and DAD in participants from different occupational groups, education, and job status, stratified analyses were applied as seen in Additional file 1: Tables S1. In sales and service workers, computer use for 1/2 ~ 3/4 of the work time (OR 1.66, 95 % CI 1.11–2.48) and for >3/4 of the work time (OR 2.07, 95 % CI 1.35–3.17) had higher OR for DAD compared to computer use for <1/2 of the work time. In other occupational group, computer use time was not significantly related with DAD. The combined effect of computer use and occupational group on DAD, shown in Additional file 2: Table S2, produced no statistically significant relationship. In regard to education, high school and college education had a significantly high OR in workers who used computers for >3/4 of the work time (Additional file 1: Table S1). The university graduate or above group had a marginal association with DAD for computer use for >3/4 of work time, and the combined effect of computer use and education on DAD is displayed in Additional file 2: Table S2. When examining job status, responses of self-employed, employer, and the others had a statistically significant association with DAD only for those using computers for >3/4 of their work time (OR 2.22, 95 % CI 1.46–3.38; Additional file 1: Table S1). In full-time employees, computer use for >3/4 of the work time was related to DAD (OR 1.46, 95 % CI 1.04–2.05). The combined effect of computer use and education on DAD is exhibited in Additional file 2: Table S2.

The aim of this study was to investigate the association between computer use at work and self-reported depressive and anxiety disorder in a nationally representative sample of Korean workers. The results showed that participants using computers for more than 3/4 of the work time had a higher odds ratio of DAD after adjusting for socio-demographic factors. From combined association of computer use proportion and the total working hours, this study revealed that computer use for >3/4 of the work time was associated with DAD for both short and long working hours. Sales and service workers as well as those with high school and college education, with a high proportion of computer use had a significantly higher odds ratio of DAD with an increasing trend. For job status, the third category of self-employed, employers, and the others and the category of full-time employees were significantly related to a higher odds ratio of DAD with a dose-response relationship.

This result is consistent with other previous studies. Ye et al. has shown that over 5 h of computer work is associated with high scores on the 12-item General Health Questionnaire, a screening instrument for common mental disorders and psychiatric well-being. Nakazawa et al. demonstrated that adverse mental symptom scores in computer workers were significantly related with using VDT for a long duration compared to using VDT for a short duration [10]. Moreover, several studies about the relationship between daily computer use and mental health among students and young adults have indicated that high use of computers or internet addiction increased pathologic symptoms such as depression, sleep disturbances, and insomnia [20–22].

Several hypotheses have been proposed. Firstly, computer work could be a source of job stress which is highly mentally demanding with requiring more attention and cognitive concentration [7, 23]. Although computers made the tasks efficiently processed, work intensification was also simultaneously developed. Work intensification is defined as an increase in the demands, pace, or qualitative requirements of the tasks and may be related to a high level of job strain and distress. In the highly skilled workers, creative and complexed tasks are demanded by using computers. In the low skilled workers, almost computers works are highly controlled and repetitive. These tasks are monotonous but still requiring constant attention such as accounting in sales department and handling customer service with information (e.g., customers’ call center). In both area, computer works are mentally demanding and therefore, job stress can be high and cognitive problems such as concentration, memory, and decision-making could be affected by increased computer job demands [23]. Further, these job stress can be aggravated with work intensification. With the technology developed, computer use may intensify the pace of work with computer monitoring system and multitasking [7, 24, 25]. Secondly, computer works are usually highly sedentary. Recently, systematic reviews investigated that sedentary works including computer work were related with depression and anxiety [26–28]. Decreased physical activity could be one of the mechanism of sedentary behaviors related depression and anxiety. Thirdly, social withdrawal hypothesis suggested that computer work could remove individuals from social interactions and increase risk for depression [7, 24]. Fourthly, technology-related stress which is caused by frustrating experiences with hardware or software problems can increase stress and anxiety in computer workers with high job demands. Lastly, sleep problems and musculoskeletal disorders related with computer work could affect the pathophysiology of the development of depression and anxiety [29, 30].

We found that computer use for more than 3/4 of total work time is associated with a more than 2-fold odds ratio of DAD among the workers with both short and long working hours. There was a dose-response relationship between computer use and DAD after stratification by working hours. Many previous studies have revealed that long working hours are related to depression and anxiety disorders [31, 32]. In addition to long working hours, high computer use may have additional psychologically adverse health effects through increased job stress, work intensification and the pervasive use of computers in the home.. With work intensification, the pervasive use of the computer and work extension can blur the boundaries between work and private life, which can further increase demands and responsibilities and shorten time for private life and leisure.

Previous studies have found that short working hours were related to a higher prevalence of depression, possibly due to other chronic illnesses. The U-shaped relationship between working hours and health outcomes has been investigated in previous studies [33], especially in regard to cardiovascular diseases, and could be explained by the healthy worker effect [34]. Individuals who work less than 40 h per week may be precarious workers such as part-time workers or temporary workers. Recently, many studies have revealed that precarious work is related to deteriorated mental health and that it is bi-directionally associated with mental disorders [35, 36].

In the general working population, it is important to identify the groups vulnerable to DAD. In order to evaluate the modifying effect of job type on DAD, we stratified the occupational groups into 4 categories. Clerical and professional working groups were the 1 doing the most computer work, accounting for 64 % of the total computer workers, while substantial computer work was also performed by service, sales, and manual workers, accounting for 36 %. Sales and service workers with high computer use were linked to DAD, while clerical and professional workers were not. As previously mentioned, sales and service workers in computer work are faced with high job demands and low job control [7]. The greater amount of computer work may increase the monotony and tedium of their job and exacerbate job stress with poor psychosocial working condition [24, 37]. In addition, emotionally stressful tasks dealing with customers and complaints such as call centers can have adverse effects on mental health with low job control and high demands [38]. Low support from supervisor and co-workers can increase the risk for DAD with electronic operating and monitoring systems [39]. In manual workers, the relationship between computer use and DAD was not statistically significant probably because of the small number of computer users. However, high school and college education were related to an increased odds ratio of DAD. Previous studies revealed that low socioeconomic status, including low education level, was generally connected to common mental disorders and persistent morbidity [13–15]. Additionally, because more monotonous and repetitive tasks combined with high job demands and low job control tend to be allocated more in low-educated groups than in highly educated groups, they could be affected by high job strain, a known risk factor for depression and anxiety. Job insecurity and low opportunity of career advancement in low education level may also increase the risk for DAD [40].

Self-employed, employer, and others using computers for >3/4 of the work time had a significantly higher odds ratio of DAD with a dose-response relationship between computer use and DAD. The significantly elevated risk of DAD among the self-employed or employers may be partly due to work engagement or workaholism without social contact [41]. Besides, 80.7 % of these workers excluding employer with employee and entrepreneurs were self-employed, unpaid family workers and other workers which can be categorized into informal employment. They are away from job security, social support, and workplace regulations and are exposed to the risk factors of DAD [42]. Full-time employees using computers for >3/4 of the work time also had a higher odds ratio of DAD with a dose-response relationship. However, in part-time employees we could not observe the relationship between computer use and DAD due to small numbers of high computer users.

Several limitations of this study should be mentioned. First, DAD was detected by a single question on a binary scale about the experience of the disorders in the last 12 months. Therefore, reliability was not assured. A validated instrument or professional diagnosis could strengthen the reliability of the definition of the mental disorders. Second, although we used large representative worker data, there was a limitation in the relatively small number reporting DAD. The prevalence of DAD (1.46 %) was lower than in previous Korean mental health surveys (2 % and 6 %). Third, as KWCS had a cross-sectional design, the association between computer use and DAD could be bi-directional and the definite causality of computer work to DAD could not be investigated. Depressive workers might perform more slowly or work shorter hours than the euthymic workers. Fourth, KWCS used a categorical question about the proportion of computer use during work time. Therefore, the effect of the actual duration of computer use could not be investigated. Lastly, although KWCS was designed using stratified clustered sampling, we could not analyze KWCS with survey weight approaches due to limited data availability.

Based on our results, computer use at work was associated with DAD in South Korea. We need to recognize computer use as risk factor of depressive and anxiety disorders. Computer work more than 3/4 of the work time could be related with mental health problems and the association could be higher in worker with long working hours more than 50 h. Especially, workers with low education, low skilled, and self-employed workers are highly associated with DAD and could be susceptible group to protect from computer use and other poor working conditions. Management of psychosocial working conditions includes long working hours, high demands, low control, and low social support have to be considered in these susceptible groups. Organizational support including job security and social interaction could be a start point of improving working conditions and the effort to reduce job demands and enhance the job control are needed. Personal effort are also important in balancing work and private spheres which are blurred by the computer work systems. Since in real situations these suggestions for improvement of working conditions are hardly achieved by only the effort of enterprises and individuals, the organizational support of government are needed at this point which includes policies and guidelines of improvement of working condition and promotion of workplace health for computer work.

Longer hours of computer work was associated with an increased risk of DAD in South Korean workers. To our knowledge, this study is the first report demonstrating the association between computer use at work and DAD in a representative sample from a national population census. In the sales and service workers, those with a high school or college education, and the self-employed and employers, the association between computer use and DAD was prominent. This finding may be helpful in developing workplace health promotion guidelines for workers suffering from high computer use at work.