Effects of Occupational Hazards on Job Stress and Mental Health of Factory Workers and Miners: A Propensity Score Analysis

This study is to evaluate the effects of different occupational hazards on job stress and mental health of factory workers and miners. A total of 6120 workers from factories and mining enterprises in seven districts and one district of Urumqi were determined using the stratified cluster random sampling method. The Effort-Reward Imbalance (ERI) questionnaire and the Symptom Checklist-90 (SCL-90) were used to evaluate the effects of occupational hazard factors on job stress and mental health of workers. The propensity score analysis was used to control the confounding factors. The occupational hazards affecting job stress of workers were asbestos dust (OR = 1.3, 95% CI: 1.09-1.55), benzene (OR = 1.25, 95% CI: 1.10-1.41), and noise (OR = 1.39, 95% CI: 1.22-1.59). The occupational hazards affecting the mental health of workers were coal dust (OR = 1.19, 95% CI: 1.02-1.38), asbestos dust (OR = 1.58, 95% CI: 1.32-1.92), benzene (OR = 1.28, 95% CI: 1.13-1.47), and noise (OR = 1.23, 95% CI: 1.07-1.42). Different occupational hazards have certain influence on job stress and mental health of factory workers and miners. The enhancements in occupational hazard and risk assessment, occupational health examination, and occupational protection should be taken to relieve job stress and enhance the mental health of factory workers and miners.


Introduction
Occupational hazards are personal threats and injuries that occur in the process of production and labor [1] and usually classified into six categories: dust, chemical factors, physical factors, radiological factors, biological factors, and other factors [2]. The specific hazards, potential risk factors, harmful factors, and unsafe behavior of occupational groups may cause occupational accidents and even occupational diseases [3][4][5]. ILO reports that 2.34 million people die per year in workrelated accidents or diseases around the world, and approximately one-quarter of all deaths are related to an unhealthy environment [6]. In addition, 160 million people suffer from nonfatal work-related diseases [7]. According to incomplete statistics, there are 200 million workers exposed to various occupational hazards and more than 16 million people are working in toxic and harmful enterprises, involving more than 30 different job types [8,9]. However, the importance of safety at the workplace was often overlooked in understanding and practicing occupational health and safety [10].
Current studies mainly focus on the relationship between occupational hazards and occupational diseases [11][12][13]. However, psychosocial changes in the process of occupational exposure usually happen before physiological disorders, such as job stress and psychological health (anxiety disorder and depression) [14][15][16]. Nevertheless, these psychosocial hazards are often ignored and lack intervention [17,18]. In addition, excess psychological stress could decline psychological function and cause negative physiological responses, leading to emotional fluctuations and psychological health problems [19].
Job stress refers to people's response under physiological and psychological pressure caused by the imbalance between occupational needs and individual resources, needs, abilities, and knowledge [20]. With the development of society and the increase in life pressure, people bear more and more pressure from society, work, and life. Job stress has been regarded as the crisis and illness in modern society and life. Factory workers and miners were the occupational group with high demands, low reward, and low job control [21], and their working environment was full of dust, chemical factors, physical factors, and biological factors. Therefore, this paper is to evaluate the effects of occupational hazards on the job stress and mental health of factory workers and miners in Urumqi, China, using the propensity score analysis, in order to provide a theoretical basis to develop prevention and control policies.

Materials and Methods
2.1. Participants. This survey was carried out from January to May 2019. Workers with the occupational exposures to coal dust, silica dust, asbestos dust, benzene, lead, noise, and Brucella in factories and mines in Urumqi, China, were investigated. A total of 6500 factory workers and miners were initially selected using a cluster sampling method. Participants working at a production department, assembly line operation, quality assessment, and quality control department were included, and administrative personnel and management personnel were excluded. Those with work experience less than one year or taking psychoactive drugs were also excluded. According to the inclusion and exclusion criteria, 6130 participants were included in this survey. The cross-sectional study was conducted by an online questionnaire using a mobile phone. The respondents volunteered to participate in the survey, and the written informed consent was provided. Finally, 6120 questionnaires were collected, and 10 copies of continuous answer questionnaires were excluded.

Research Methods.
A self-administered electronic questionnaire was used to investigate job stress and mental health.  [15], which was on the basis of the imbalance between effort and reward causing a series of changes in physiological, psychological, and cognitive functions and body health. The ERI questionnaire consisted of three sections: effort (E, 6 items), reward (R, 11 items), and overcommitment (6 items), totaling 23 items. Each item was assigned a score ranging from 1 (highly disagree) to 5 (highly agree) [20] and was attributed the same weight. The index of an effort-reward ratio was ERI = E/ðR * CÞ, where C was the ratio of the number of items of E to the number of items of R. In this paper, C = 6/11. The ERI index of >1, 1, and <l indicated a high effort-low reward return, a balanced effort-reward return, and a low effort-high reward return, respectively. The higher the ratio of the ERI, the worse the job stress [22].

Mental Health
Investigation. The Symptom Checklist-90 (SCL-90) was widely used in psychiatric outpatient examination of its high authenticity in evaluating various mental health surveys [23,24]. SCL-90 involved emotions, consciousness, thinking, feeling, interpersonal relationships, habits, diet, sleep, etc. It can independently assess the state of one's body, cognition, emotion, social function, and spirit. The 90 items were divided into 9 dimensions: somatization, obsessive-compulsive symptom, sensitivity of interpersonal relationship, depression, anxiety, hostility, horror, bigotry, and mental diseases. Each of 90 items ranged from 1 to 5. The higher score showed a worse psychological symptom. According to the Chinese norm, the further examination was needed at the total score of 160 higher, any item score of 2 higher, or the number of positive items more than 43 [25].

Quality Control.
The online electronic questionnaires were built using the Questionnaire Star Platform. The preexperiment was conducted in order to ensure the humanistic design and logicality in the answering process. With the assistance of the department managers, the informed consent explaining the questionnaire survey's purpose and the voluntary participation principle was set to all volunteers. A twodimensional code was used to conduct on-site online investigation. Our staff explained the purpose, significance, content, and requirements to ensure the quality and effectiveness of the online investigation. The data of each online questionnaire was valid data based on the technical design. The continuous answer questionnaires were excluded.

Statistical
Methods. Statistical analysis was performed using R (version 3.4.4). The balance of confounding factors was assessed by the propensity score analysis. The multiple logistic regression was employed for multivariate analysis. The Mann-Whitney U test was used to compare two groups, and the Kruskal-Wallis H test was used to compare more groups of nonparametric (nonnormally distributed) data. A chi-squared test was used for the counting data, and multiple linear regression was employed for multivariate analysis. The significance level (α) was set at 0.05.

General Demographic Characteristics of Factory Workers and Miners.
Among the 6120 factory workers and miners, 4017 were men (65.64%) and 2103 were women (34.36%). The occupational hazard factors in which workers are involved were coal dust (1446, 23.63%), silica dust (622, 10.16%), asbestos dust (35,15.28%), benzene (1947, 31.81%), lead (353, 5.77%), noise (4545, 74.26%), and brucellosis (108, 1.76%) ( Table 1). The survey results showed that 2898 (47.35%) of factory workers and miners suffered from job stress and others (3222, 52.65%) were normal. In order to control confounding factors and eliminate the bias of general demographic characteristics, the propensity score matching was used by selecting a total of 14 factors as the matching covariates, including sex, ethnicity, education level, labor contracts, professional title, work schedule, marital status, monthly income, height, weight, age, working years, working hours per day, and working days per week. The caliper value was set at 0.02. Finally, 2498 pairs of positive and control groups were screened out by a 1 : 1 proximity matching method as shown in Table 2.

Comparison of Propensity Score Matching in SCL.
The survey results showed that among 6120 of the factory workers and miners, 2342 (38.27%) suffered from mental health problems and 3778 (61.73%) did not. Using PSM as described in Section 3.2, 2152 positive and control groups were screened out. The differences of sex, ethnicity, education level, labor contracts, work schedule, monthly income, hypertension, working hours per day, and working hours per week were eliminated after matching (Table 5).

Discussion
Job stress has been proven as a risk factor for the adverse physiological function (such as declined strength, stiffened body, and disorders in sense and memory) and psychological reaction (such as distracted attention and reduced working will and desire) [26,27]. Ahn et al.'s study has also shown that exposure to occupational hazards adversely affects the individuals' health and exacerbated job insecurity [28]. Factory workers and miners belong to a special professional group, who work in a special environment of high temperature, high pressure, darkness, or dust but have low income and social status. Some studies have already proven that people living in harsh environments have a higher risk of developing mental illnesses and the special environments affect the degree of job stress [29][30][31]. Our research showed that 47.35% and 38.27% of factory workers and miners suffered from job stress and mental health problems, respectively, suggesting that the status of occupational mental health of factory workers and miners should be paid more attention.
Propensity score matching (PSM) is a statistical matching technique that attempts to estimate the effect of a treatment and intervention by accounting for the covariates that predict receiving the treatment. It attempts to reduce the bias due to confounding variables that could be found in an estimate of the treatment effect obtained from simply comparing outcomes among units that received the treatment versus those that did not [32]. This method can adjust covariates more objectively and scientifically and thereby get more intuitive conclusions. In recent research, the main risk factors for job stress involved sex, professional title, working age, working hours, chronic diseases, and so on. However, the relationship  between the risk factors and occupational stress or mental health was less reported. Therefore, this study is the first one to use PSM to explore the relationship between occupational hazards and job stress and mental health. Thus, 14 demographic characteristics of factory workers and miners and the control group, such as sex, ethnicity, and educational level, were adjusted to reduce the confounding bias and thereby obtain more objective results.  The results showed that the influences of occupational hazards on the job stress of factory workers and miners were statistically significant, and the risks of job stress increased 1.30 times, 1.25 times, and 1.39 times with exposure to asbestos dust, benzene, and noise, respectively. Coal dust, asbestos dust, benzene, and noise were the risk factors for  Occupational dust was an important health risk in modern society, and the adverse health impacts of coal mine particulate matter had been well known [33,34]. Changes in the body's functioning due to occupational dust exposure, such as developing respiratory symptoms and decreased pulmonary functions, could make workers feel more tired at work and thereby increase the possibility of work errors, leading to psychological burden. Some studies have confirmed the negative effects of chemical substance exposure on tests of psychomotor function and short-term memory [35]. Hence, the long-term occupational exposure to benzene would affect body metabolism, making the workers prone to job stress and psychological problems. Noise was one of the commonest occupational hazards of the modern world. Lai and Huang found that exposure to a high-noise environment would cause endocrine disorders, headaches, blood pressure, and fatigue [36]. In addition, hearing and cognitive ability shared the same neurophysiological conduction, and workers needed to pay more attention of message reception and understanding in a noisy working environment. It reminded us that setting a barrier between workers and occupational hazards, standardizing operation procedures, and providing protection education training were effective measures to reduce the potential harm and adverse health to workers. Moreover, the use of personal protective devices, such as gloves, glasses, aprons, safety footwear, and dust masks, should be paid more attention by both employees and employers. Improvement of occupational health and safety management has a wide range of benefits for companies, such as the reduction of absenteeism and labor accident and the increase in workers' motivation and productivity [37].
The present study has some limitations to be improved upon in future studies. Firstly, there were still continuously repeated answers in the online questionnaire. Secondly, the cross-sectional investigation cannot prove causality between variables, and the relationship between the factors and diseases needs further investigation. Thirdly, the inclusion and exclusion criteria were not sufficient, such as people's family history in terms of mental and behavioral problems. Then, the effect of PSM was not significant in the case of respondents having similar background. Finally, the influence of single occupational hazard on the mental health of workers and miners was preliminarily described qualitatively, while the influence of pollution level and the interaction of two or more occupational hazards on mental health should be verified in the future studies.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.