Research has increasingly examined means of improving workers’ health and the working environment, in accordance with the increasing popularity of social change for improved welfare. National-level efforts to manage workers’ health and establish pleasant working environments have emerged; for example, the enactment of the Industrial safety and health act by the Ministry of Employment and labor [1], in addition to personal health management.

Psychiatric problems such as sleep disorder, neurological headaches, and depression are often considered personal problems rather than job-associated problems, and therefore receive less attention than relatively serious occupational health problems such as suicide, physical injury, and accidents [2]. Nonetheless, the International Classification of sleep Disorders (ICSD; the most widely used sleep disorder classification) nominates stress as the primary cause of sleep disorders: relationship stress or decreased psychosocial abilities weaken physical resilience and psychological integrity [3]. Further, occupational risk factors (e.g., shift work, job-related stress) are related to insomnia [4].

The ICSD-3, a standard for ICSD diagnosis, classifies general sleep disorders including insomnia into seven categories: 1) insomnia, 2) sleep-related breathing disorders, 3) central disorders of hypersomnolence, 4) circadian rhythm sleep-wake disorders, 5) parasomnias, 6) sleep-related movement disorders and 7) other sleep disorders [5]. Of these, insomnia (sleep abnormalities that persist despite suitable opportunities and environment) is the most common of the sleep disorders. Its prevalence is high worldwide (10–30 % in western countries; 15–21 % in Asian countries [6–11]), suggesting that the total prevalence of all sleep disorders is considerably higher. Further, regarding South Korea, 17 % of the population reports experiencing insomnia ≥3 times a week [12] and 31 % of the general adult population reports sleep problems such as difficulty falling asleep, frequent waking, and early morning awakening [13]. Further, long-lasting sleep disorders predict reduced productivity and work efficiency among workers, ultimately leading to decreased company productivity. Sleep disorder increases the risk of reduced workers’ well-being and quality of life; sleep disorder therefore presents a serious problem that requires attention.

Jang et al. developed the Occupational Stress Scale for Korean Workers; Kim et al. used that scale to examine the correlation between occupational stress and sleep disorders. They found that individuals experiencing higher job stress experience insomnia more frequently [14]. Heo et al. found that physical factors and psychosocial factors increase individuals’ risk of sleep disturbance; they examined data from the 2nd Korean Working Conditions Survey (KWCS) [15].

This study analyzed job-related stress’ effects on sleep disorder, as addressed in previous studies, as well as factors newly identified to affect sleep disorder (e.g., discrimination experience and emotional labor), using data from the third KWCS. It is possible that workers, especially in the rapidly increasing service industries like telemarketing, the insurance business, etc., experience stress in their relationships with customers. Furthermore, labor experiences and efficiency may be affected by organizational culture and structure, as well as relationships between individuals because of the generalization of production systems per unit, module, and team. For this reason, we focused on relational and organizational job stress factors.

Our analysis extends the understanding of sleep disorders to include correlations with novel factors. Additionally, in contrast to earlier research we analyzed the interactions between gender and each job stress factor, and we presented the results of the analyses separately by gender. Therefore, this article clearly demonstrates the gender differences in exposure and risk of stress and disease. Furthermore, this analysis addressed the associations between work and sleep disorders, thereby examining job-related factors’ effects on the development of occupational diseases.

This study found that job-related stress factors generally increased the risk of sleep disorders. In particular, discrimination experience, and variables related to emotional labor and overall satisfaction showed OR values of over 2.00, even after correcting for all related variables. This, then, is a strong correlation. Further, the results of the Hosmer Lemeshow goodness-of-fit test measurements for each logistic regression were 0.217 for women and 0.308 for men, both of which were > 0.05 and were therefore adequate. To rule out the possibility of multicollinearity, we assessed the variance inflation factor (VIF), which is an index measuring how much the variance of an estimated regression coefficient increases because of collinearity. A VIF value greater than 10 is considered to indicate high multicollinearity [18]. All the variables in our analyses showed adequate VIF values, since they were smaller than 1.5.

Because there are reportedly gender differences in stress exposure and risk [19, 20], we analyzed the interactions between gender and each job stress factor. Discrimination experience was significantly correlated with gender (and the effect of discrimination on sleep disorders was larger among women than men, as seen in Fig. 1); however, no other significant interactions were found. Therefore, we performed the regression analysis separately by gender, and found that the OR and statistical significance trends were similar between the two genders, but women’s OR values for discrimination experience and direct customer confrontation were much higher than those of men.

Previous studies have examined the relationship between each category and sleep disorder; however, they have only examined factors associated with the presence or absence of general sleep disorders and insomnia. We added “sleep disorders due to one’s job,” which indicates “work-relatedness;” therefore, we analyzed work-related sleep disorder’s correlation with five variables common in previous research to better situate this study’s results within the broader research context. The present results corresponded well to the previous correlation analyses (these results are shown in the Additional file 1). Slight differences (i.e., significance at larger values of p) in this study’s co-analysis of additional categories may reflect sample size differences, independent variable selection, categorization differences, and differing definitions of dependent variables.

Discrimination experience significantly increased the risk of sleep disorders, including after correcting for other variables. Discrimination experience itself appears to affect general stress; gender discrimination experience leads to reduced job satisfaction, compensation satisfaction, and organizational commitment; age discrimination at work leads to disadvantages in work allocation and promotion opportunities; stress resulting from this dissatisfaction ultimately increases workers’ risk of sleep disorder [21–23].

High-risk groups in the direct customer confrontation and emotional stress categories reported sleep disorders significantly more frequently. Survey items in these categories examined if the worker frequently directly manages individuals who are not co-workers (i.e., customers, patients) and the worker’s frequency of emotional labor (which modern society is strongly focused on). Emotional labor predicts negative outcomes such as exhaustion and job dissatisfaction [24–26]. Additionally, stress level and prevalence of sleep disturbance were found to be high among female temporary workers in department stores; further, physicians and nurses (who are in close contact with patients) show high job-related stress and prevalence of sleep disturbance, corroborating this study’s results [27–29]. Requiring workers to endure negative emotions (personal implicit feelings) in order to follow company rules, and thereby distorting or exaggerating workers’ intrinsic nature, may thus lead to psychological and physical exhaustion, further provoking pathological phenomena including suicide and sleep disorder [30].

The rate of sleep disorder increased in groups with high job dissatisfaction, as found in the work dissatisfaction (detailed) and work dissatisfaction (overall) categories. This result corresponds with those of previous studies, which have additionally found that workers with low job satisfaction tended to be more stressed and more likely to experience whole-body fatigue, depression, anxiety and depression, and that low job satisfaction increases workers’ risk of insomnia [31, 32]. In contrast, the univariable analysis of organizational system dissatisfaction detected a greater OR of sleep disorder in the high-risk group; however, this difference decreased below significance after correcting for other variables. Variables that other studies defined as “organizational unfairness”(unreasonableness or unfairness of organizational policy, operational systems, or formal structure of the company) were found to be related to an increased risk of poor sleep quality with high organizational unfairness [33]. Expression of opinion difficulty was negatively correlated with workers’ risk of sleep disorder; this may reflect habitually passive workers who wish to do no more work than is required of them and are unmotivated to express their opinion, but instead prefer to passively accept information without understanding it; however, further research of this relationship is required.

Regarding high job demand, work intensity was positively correlated with workers’ risk of sleep disorder, corroborating previous results [14, 15, 34–36]. Job demand indicates work-related burden; high job demand may be a classical occupational stress factor and related to difficulty maintaining sleep and low sleep quality [37, 38].

Regarding insufficient job control, we found that difficulty controlling one’s job was negatively correlated with workers’ likelihood of reporting sleep disorder. Some previous studies have reported that workers facing greater work control difficulty reported insomnia more frequently; however, some of these studies’ hypothetical models did not achieve statistical significance [14, 39]. In contrast, some studies have found an independent effect between work control and insomnia, and have suggested that low job control affects workers’ stress more weakly and indirectly than job strain [15, 40]. This disparity may reflect differences in the measures used; however, future research should settle this dispute.

Regarding inadequate social support, workers in the high-risk group reported sleep disorder more frequently before correcting for the newly added variables; however, this difference became insignificant following correction for the newly added variables. Previous studies have also found that low social support in workplace is correlated with workers’ insomnia, citing conflict with colleagues or superiors as an important workplace stressor. In particular, individuals who work in the same workplace throughout their lifetime were found to prioritize interpersonal working relationships [14, 15].

Regarding job insecurity, the high-risk group reported sleep disorder more frequently, corroborating previous study results [14, 41]. Job insecurity indicates one’s stability in one’s position at work, capturing employment opportunities as well as instability. In South Korea, for example, most employees indicated exposure to serious threats to their psychological health from external factors such as downsizing and employment instability due to the rapidly changing socioeconomic situation since the financial crisis of 1997 [17].

Workers reporting a lack of reward were more likely to report sleep disorder. Inadequate compensation is a concept taken from the Effort-Reward-Imbalance (ERI) model: high effort and low compensation leads to high stress. In the present context, high ERI increased workers’ insomnia, corroborating previous studies [14, 15, 36].

Workers who reported working over 45 h per week and shift workers were significantly more likely to report sleep disorder. Numerous studies have also found this association, and have suggested desynchronization of circadian rhythms due to irregular sleep-wake schedule as the main cause of such work-induced sleep disorders [4, 32, 42, 43]. Other studies examining working hours per week have suggested a plausible causal pathway between overtime working hours and sleep disorder, mediated by poor recovery after work due to inadequate free time, and have therefore proposed the necessity of sufficient time for recovery from fatigue [34, 38, 44].

Earlier research has not fully established a cause-effect mechanism between job-related stress and sleep disorder; however, some studies have reported these variables to be mediated by increased cortisol levels [36, 45, 46]. Increased cortisol secretion and hyperactivity of the sympathetic nervous system following activation of the hypothalamic-pituitary-adrenal axis is a relatively direct biological response to stress and may elicit various physiological responses [47, 48]. More indirectly, behaviors detrimental to health (i.e., smoking, consuming alcohol) may manifest in response to stress [23]. Stressful life events may powerfully affect the development of insomnia; similarly, minor stressors (e.g. Job strain) in daily life are also correlated with sleep disorder [49]. Further, psychosocial factors are more strongly correlated with consistent insomnia than physical health problems [50]. Research thus supports the correlation between stress and sleep disorder. Future research should promote acceptance of job-related stress as importantly affecting sleep disorder, illuminate the disorder’s causal mechanism more clearly, and intervene to prevent the disorder’s development.

This study possesses the following limitations. First, measurement involved the individuals taking the survey subjectively diagnosing themselves with sleep disorder. The KWCS was not specifically designed to measure stress and sleep disorder; responders therefore subjectively judged if they had sleep disorder. This does not meet the diagnostic criteria of sleep disorder set out by the World Health Organization (WHO), the International Classification of Diseases (ICD-10-CM), or the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) by American Psychiatric Association (APA); it also diverges from the International Classification of Sleep Disorders (ICSD-2, 2005) of the American Academy of Sleep Medicine (AASM). The validity of sleep disorder diagnosis in this study may therefore have been limited; the survey also did not specifically examine sleep quality and quantity. Second, this study used a cross-sectional research design; this prevents valid inferences regarding variables’ chronology and causal relationships. Third, we did not use a standardized tool to measure job-related stress factors. Jang et al. developed the Occupational Stress Scale for Korean Employees; however, this study combined this measure’s questions with others taken from the KWCS. Additionally, the author altered some survey items following comparison of the two surveys; the measure of job stress was therefore non-standardized. Nonetheless, reliability analysis of the survey yielded Cronbach’s alpha values exceeding 0.6 in all categories, exceeding adequacy cut-offs and indicating acceptable response consistency. Fourth, the participants were current workers; workers who had retired due to accident, disease or health deterioration were not included in the analysis. The results may therefore reflect a “healthy worker” effect, leading to a general underestimation of sleep disorders’ prevalence. In contrast, the results may also reflect recall bias: individuals personally experiencing sleep disorders may have overestimated stress level as the measure was self-administered.

Despite these limitations, the results of this study remain significant. This study identified and analyzed significant stress factors not analyzed in previous research, and found them correlated with job-related stress and sleep disorder; this deepens the understanding of the existing factors, and extends the range of known factors affecting job-related stress. In particular, the present results indicate that experience of discrimination, direct customer confrontation, and emotional labor significantly elevates workers’ sleep disorder risks; therefore, these factors merit additional attention. In addition, regarding discrimination experience, the OR of sleep disorder was approximately 1.7 times higher for women than for men; hence, some gender differences in exposure and risk of stress exist. Women are more vulnerable to stress from discrimination and disease. Questions regarding discrimination should be subdivided into discrimination by gender, age, education level, region, and employment type. Moreover, items pertaining to dealing with emotional labor should be added to the existing stress measurement tools. Further, previous studies have examined only insomnia or sleep disorder; in contrast, this study examined whether workers believed their symptoms to be work-related. This more specifically captures the correlation between workers’ evaluations of sleep disorder and job-related stress. Doctors should consider that workers’ sleep disorders may be associated with the stress they experience in their workplace, rather than with a personal problem. An objective diagnosis and treatment system should therefore be established to more accurately explore the role of work-related psychosocial factors. Finally, this study used data from a large-scale survey examining workers from all over Korea; this raises the likelihood that the present results are representative.

Sleep disorders including insomnia may accompany various psychiatric disorders, or occur as initial symptoms of a disorder or constitute part of a latent period of further disorder. Further, sleep disorder is closely correlated with various physiological diseases (e.g., cardiovascular diseases, chronic pain, gastrointestinal diseases) [51]. The present results confirm that job-related stress significantly predicts sleep disorder; interventions should therefore reduce workplace stress by controlling workers’ labor or relocating stressed workers. Further, social efforts should aim to create workplace cultures that allocate recreational or break time in order to alleviate accumulated worker stress.

In conclusion, this study found a strong correlation between sleep disorder and relational and organizational job stress factors. Sleep disorder may lead to large decreases in workers’ quality of life and work efficiency. Awareness and interventions are therefore required to reduce workplace stress; further research of this topic is also required.