Gender differences in psychosocial factors related to severe stress in a construction company in Japan

Abstract

BACKGROUND:

In Japan, women’s participation in the workforce has been increasing. However, there are few studies about gender differences in stress factors in the workplace.

OBJECTIVES:

We aimed to clarify gender differences in psychosocial factors in the workplace related to severe stress.

METHODS:

In total, 930 male and 339 female employees working in a construction company in Japan participated in this study. We used individual responses to the Brief Job Stress Questionnaire (BJSQ) conducted by their company for the most recent year between 2016 and 2019. Data were categorized by gender, and we extracted the factors contributing to severe stress using logistic regression.

RESULTS:

We identified 142 men and 32 women who had severe stress. Logistic regression analysis showed that six factors (“Qualitative job overload”, “Job control”, “Interpersonal conflict”, “Poor physical environment”, “Supervisor’s support”, and “Job satisfaction and life satisfaction”) were significantly associated with the presence of severe stress in men. In women, five factors (“Qualitative job overload”, “Job control”, “Interpersonal conflict”, “Suitable work”, and “Job satisfaction and life satisfaction”) showed a significant association with severe stress.

CONCLUSION:

Factors associated with severe stress were “Poor physical environment” and “Supervisor’s support” in men, and “Suitable work” in women.

Keywords

Workplace occupations gender role mental health

1 Introduction

Psychosocial stress at the workplace has been related to physical symptoms [1], accidents [2], and mental disorders including depression [3]. The relationship between psychosocial stress factors in the workplace and mental disorders has been revealed based on the Job Demand – Control model (JDC model) [4]. Several questionnaires such as the National Institute for Occupational Safety and Health (NIOSH) [5] and the Job Content Questionnaire (JCQ) [6] were developed based on the JDC model, and have become commonly used. Additionally, some studies have shown that differences in psychosocial stress factors can depend on nationality or occupation, and many have commonly shown that high job demands, low job control, low social support, and interpersonal conflict were factors that cause high stress situations [7, 8]. Also, it was reported that men and women experience and cope with stress differently [9].

In order to prevent employee health problems, the importance of educating employees in work-related stress management was suggested [10], and intervention programs for stress in the workplace have been developed. In the Cochrane review [11] and other reviews of intervention programs [12], cognitive behavioral therapy and relaxation programs were reported to be moderately effective in preventing stress. Additionally, some studies have reported direct organizational interventions to manage certain stress factors in the workplace, such as the job schedule [13, 14]. However, there have been fewer reports about interventions for specific stress factors in the workplace than about other interventions in the workplace such as psychoeducation using cognitive behavioral therapy, and evidence for those interventions conducted in the workplace remains poor [11]. In addition, some studies have reported mental health promotion programs for coping with workplace stress that target men. However, there are few reports about the similarities between men and women with respect to workplace stress, and there has been little consideration of gender differences [15]. Based on the background information provided in previous studies, we suspected that although there are many previous studies about psychosocial stress in workplace, studies focusing on differences in gender and practical interventions based on those differences are scarce.

In Japan, the 2018 Industrial Safety and Health survey conducted by the Ministry of Health, Labor and Welfare revealed that 58% of employees had strong anxiety and stress about their work or their working life [16]. Moreover, the number of mental health problems related to occupational accidents had increased since 1999 [17]. The Ministry of Health, Labor and Welfare released the “Guideline for Mental Health Promotion of Workers” in 2006, and in 2016, the Japanese Industrial Safety and Health Law was extended to include a stress check program that makes regular yearly screenings for high psychosocial stress in the workplace mandatory for enterprises with 50 or more employees. The purpose of this program is primary prevention of mental disorder. Namely, the program requires employers to provide a survey of employees’ psychosocial stress, inform each employee of their results, arrange an interview with a physician for employees experiencing high stress at the employee’s request, and follow the physician’s recommendations for improving adverse working situations. Additionally, a manual for practice was prepared, and recommends using the Brief Job Stress Questionnaire (BJSQ) to screen for psychosocial stress among employees in the workplace. Previous studies using the BJSQ have revealed prediction of long-term sickness absence with 1 year follow-up using the BJSQ [18], a relationship between BJSQ score and the risk of turnover [19], and a relationship between occupational position and job satisfaction [20]. However, reports about interventions in actual workplaces using the BJSQ are scarce. Although Kachi et al. mentioned gender differences in their report about the relationship between turnover and psychosocial stress in the workplace [19]. There is a shortage of literature about gender characteristics of stress factors causing severe stress in the workplace.

The population of working women in Japan has increased to an estimated 26.93 million, based on a survey conducted by the Ministry of Internal Affairs and Communications in 2020. Also, the “M-shaped curve” (i.e., the curve showing that the employment rate of women aged around 30 decreases temporarily and then increases after their child-rearing years) has changed from the lowest bottom of 51.1% (30–34 years old) in 1989 to 73.4% (35–39 years old) in 2018. Those data suggest an increase in the number of women working with shorter breaks. Women’s participation in the workforce influences the gender ratio in the workplace, and means that men’s roles and careers, not only those of women, are changing.

For example, men’s acquisition rate of child care-leave in Japan is rising slightly [21], and the number of hours men spend on housework is also increasing [22]. According to those changes, new health management systems for both men and women will be needed to adapt to new health problems in new situations. A previous study also suggested the importance of taking a female perspective, as a minority perspective, in typically male occupations like bus driving [23]. Therefore, it is important to identify psychosocial stress factors specific to each gender in the workplace, and for professionals and employers to develop programs for health maintenance.

This study aimed to reveal gender differences in psychosocial stress factors causing severe stress using data of the BJSQ collected during a stress check program conducted by one company in Japan. The results of this study are expected to help develop interventions in actual workplaces.

2 Methods

2.1 Setting

This study was conducted in one Construction Company in Japan. In this company, a stress check program using the BJSQ has been conducted to protect employee health since 2016. Questionnaires including the BJSQ were sent by mail, and employees submitted completed responses by mail.

2.2 Participants

Participants were employees who had completed the stress check program using the BJSQ conducted by their company between 2016 and 2019. Individual data for 1 year were used, and we extracted the most recent data when participants had results for more than 1 year because we judged that a longer tenure at the company would show more accurate stress factors. The average of reply rate of the BJSQ each year was 65.7±5.98%. We found no significant difference between men and women who completed the BJSQ and those who did not in each year. The company provided data through the staff of the health management department, who were co-researchers in this study.

The design and conduct of this study have been approved by the Research Ethics Committee of the Graduate School of Medicine, Nagoya University, Japan (authorization number: 2019–0090). Informed consent was obtained from participants using the opt-out method. All procedures were in accordance with the ethical standards of the Research Ethics.

2.3 Variables and data source/measurement

The BJSQ is a self-administered questionnaire recommended for stress check programs as set out in new occupational health policies launched by the Japanese government on December 1, 2015 in 2016 [24], and is based on the NIOSH Generic Job Stress Questionnaire (NIOSH GJSQ) and includes the NIOSH stress model [5]. The BJSQ has three basic components based on the NIOSH stress model (job stressors, stress responses, buffering factors) and one comprehensive perspective component (satisfaction with working life). Namely, job stressors are stressors causing stress responses in workplace, stress responses are concrete physical and mental symptoms as the responses to such situations, and buffering factors constituted “Daily life social support by family, co-workers and supervisors”, which buffers the stress reaction. Based on those four components, the BJSQ has the following 20 factors and 57 items. Job stressors include “Quantitative job overload” (three items), “Qualitative job overload” (three items), “Physical demands” (one item), “Job control” (three items), “Skill utilization” (one item), “Interpersonal conflict” (three items), “Poor physical environment” (one item), “Suitable work” (one item), and “Intrinsic rewards” (one item). Stress responses include “Lassitude” (three items), “Irritation” (three items), “Fatigue” (three items), “Anxiety” (three items), “Depression” (six items), and “Physical stress response” (eleven items). Buffering factors include “Supervisor’s support” (three items), “Coworkers’ support” (three items), and “Support from family and friends” (three items). Satisfaction with working life includes “Job satisfaction” (one item) and “Life satisfaction” (one item). Item responses are measured on a four-point Likert scale. All BJSQ scales have demonstrated acceptable or high levels of internal consistency, reliability, and factor-based validity [25].

In the stress check program manual, which includes the BJSQ by the Ministry of Health, Labor and Welfare, two calculation methods for each factor are recommended, the total score method and the raw conversion method. In this study, we used the raw conversion method, as recommended for use in the research field but not for practical application such as for company use by the Tokyo Medical University Department of Preventive Medicine and Public Health, which was one of the organizations which developed the BJSQ [26]. In the raw conversion method, scores are calculated based on a raw conversion algorithm from the five-point Likert evaluation score and are calculated according to a conversion table. A low score means severe stress situations and strong stress in all factors. Moreover, an individual’s stress severity can be judged by the score of each calculation method. In the raw conversion method we used, severe stress is shown in a two-score pattern: 1) the sum of the scores of the factors related to stress response is under 12 points, and 2) the sum of the scores of factors related to job stressors and buffering factors (excluding “Job satisfaction” and “Life satisfaction”) is under 26 points and the score of stress response factors is under 17 points.

As demographic variables, we used gender, age, job type, position, and tenure at the company. We were also provided those data by the staff of the health management department.

2.4 Outcome

We used severe stress as calculated by the BJSQ as the main outcome. The score of each factor was calculated by the raw conversion method. In this method, a low score means more severe stress status or heavy stress situations in all factors.

2.5 Statistics

We extracted the psychosocial factors contributing to severe stress using logistic regression for each gender. We set the objective variables based on a comparison between participants with severe stress (severe group) and those without severe stress (non-severe group) in each gender using the Welch t-test. Values of p < 0.05 were considered statistically significant. We calculated the odds ratio for high stress with accompanying 95% confidence intervals (CIs). This risk is the proportion of all cases of severe stress attributable to high stress, assuming a causal association.

3 Results

Participants: We collected BJSQ data from 2016 to 2014. The participants included in the analysis were 911 men and 332 women. The mean age of participants was not significantly different based on the Welch t-test: men, 46.3±14.8 years old; and women, 54.1±13.6 years old. We also extracted 142 men (15.5%) and 32 women (9.7%) as cases of severe stress. All severe cases were defined as a stress response score under 12 points. The mean ages of men in the severe and non-severe stress groups were 43.5±13.9 years old and 46.8±15.0 years old, respectively, and the participants in the non-severe stress group were significantly older than the severe stress group based on the Welch t-test. On the other hand, in women, the mean age was 52.1±14.0 years old in the severe group and 54.4±13.6 years old in the non-severe group, and there was no significant difference between the two groups. In terms of employment status, participants were full-time workers, contract workers, and part-time workers. Also, job types of full-time workers included general workers, specialized workers, industrial engineers, and managers. Detailed information about the study participants is shown in Table 1.

Table 1
Participants’ basic information

Participants (n = 1243)

Men (n = 911, 73.3%) Women (n = 332, 26.7%)

Severe group (n = 142) Non-severe group (n = 769) Severe group (n = 32) Non-severe group (n = 300)

Mean age, y 46.3±14.8 54.1±13.6

43.5±13.9 46.8±15.0 52.1±14.0 54.4±13.6

Employment status and job type (n, %)

Full time worker

General worker 247 (27.1%) 63 (19.0%)

49 (34.5%) 198 (26.0%) 7 (21.9%) 56 (18.6%)

Specialized worker 280 (30.7%) 14 (4.22%)

42 (29.5%) 238 (31.2%) 2 (3.25%) 12 (4%)

Industrial engineer 148 (16.2%) 43 (13.0%)

29 (20.4%) 119 (15.6%) 4 (12.5%) 39 (13%)

Manager 15 (1.64%) 0 (0%)

3 (2.11%) 12 (1.64%) 0 (0%) 0 (0%)

Contracted worker 135 (14.8%) 21 (6.33%)

10 (7.04%) 125 (1.57%) 1 (3.12%) 20 (6.67%)

Part-time worker 86 (9.55%) 191 (57.5%)

9 (6.33%) 77(10.1%) 18 (56.2%) 173 (57.7%)

	Participants (n = 1243)
Mean age, y	46.3±14.8	54.1±13.6
	43.5±13.9	46.8±15.0	52.1±14.0	54.4±13.6
Employment status and job type (n, %)
Full time worker
General worker	247 (27.1%)	63 (19.0%)
	49 (34.5%)	198 (26.0%)	7 (21.9%)	56 (18.6%)
Specialized worker	280 (30.7%)	14 (4.22%)
	42 (29.5%)	238 (31.2%)	2 (3.25%)	12 (4%)
Industrial engineer	148 (16.2%)	43 (13.0%)
	29 (20.4%)	119 (15.6%)	4 (12.5%)	39 (13%)
Manager	15 (1.64%)	0 (0%)
	3 (2.11%)	12 (1.64%)	0 (0%)	0 (0%)
Contracted worker	135 (14.8%)	21 (6.33%)
	10 (7.04%)	125 (1.57%)	1 (3.12%)	20 (6.67%)
Part-time worker	86 (9.55%)	191 (57.5%)
	9 (6.33%)	77(10.1%)	18 (56.2%)	173 (57.7%)

BJSQ: Table 2 shows the average scores of the participants for each factor. The average score of each factor was around 3 points, indicating a normal level. Therefore, the study participants did not seem to have specific stress caused by any of the factors investigated. Also, each gender was divided into two groups, a severe group and a non-severe group, based on the presence/absence of severe stress, respectively, and scores of each factor were compared using the Welch t-test. In men, all factors showed a significant difference between the severe group and non-severe group. In women, however, all factors showed a significant difference except for “Physical demands” and “Skill utilization”. The average scores of each factor for both groups are shown in Table 3.

Table 2

BJSQ scores in men and women

		Men (n = 911)		Women (n = 332)
		Mean	SD	Mean	SD
Job stressors	Quantitative job overload	3.25	1.15	3.63	1.00
	Qualitative job overload	3.37	0.98	3.92	0.92
	Physical demands	2.49	1.06	2.24	1.15
	Job control	2.97	1.09	3.40	1.05
	Skill utilization	2.69	0.85	2.59	0.82
	Interpersonal conflict	2.95	0.99	3.16	1.01
	Poor physical environment	2.73	0.99	2.71	0.99
	Suitable job	2.80	1.09	3.24	1.13
	Intrinsic rewards	2.66	1.10	2.93	1.15
Psychological response	Lassitude	2.81	1.19	3.11	1.27
	Irritation	2.96	1.12	3.19	1.19
	Fatigue	2.91	1.11	2.91	1.02
	Anxiety	3.11	1.17	3.52	1.18
	Depression	3.17	1.22	3.45	1.14
	Physical stress response	2.94	1.06	2.94	2.05
Buffering factors	Supervisor’s support	2.68	1.11	2.49	1.12
	Coworkers’ support	2.47	1.01	2.46	1.02.
	Support from family and friends	3.21	1.40	3.51	1.36
	Job satisfaction and life satisfaction	3.06	0.91	3.23	0.94

SD: standard deviation.

Table 3

Comparison of BJSQ scores in the severe group and non-severe group in men and women

		Men (n = 911)						Women (n = 332)
		Severe (n = 142)		Non-severe (n = 769)		T^†	p-value^†	Severe (n = 32)		Non-severe (n = 300)		T^†	p-value^†
		Mean	SD	Mean	SD			Mean	SD	Mean	SD
Job stressors	Quantitative job overload	2.32	1.19	3.43	1.05	–11.4	<0.01**	2.73	1.08	3.73	0.942	–5.75	<0.01**
	Qualitative job over load	2.83	1.08	3.47	0.933	–7.56	<0.01**	3.38	0.888	3.98	0.904	–3.68	<0.01**
	Physical demands	2.19	1.08	2.55	1.04	–3.76	<0.01**	2.421	1.28	2.23	1.13	0.876
	Job control	2.08	0.975	3.14	1.03	–11.5	<0.01**	2.52	1.13	3.50	1.00	–5.30	<0.01**
	Skill utilization	2.51	0.946	2.73	0.827	–2.90	<0.01**	2.44	0.704	2.616	0.835	–1.17
	Interpersonal conflict	2.12	0.988	3.11	0.914	–11.7	<0.01**	2.26	0.994	3.26	0.968	–5.68	<0.01**
	Poor physical environment	2.22	1.03	2.83	0.955	–6.62	<0.01**	2.05	1.04	2.79	0.964	–4.16	<0.01**
	Suitable work	2.08	1.08	2.93	1.04	–9.03	<0.01**	2.38	0.888	3.34	1.11	–4.83	<0.01**
	Intrinsic rewards	1.98	1.04	2.79	1.07	–8.45	<0.01**	2.02	0.797	3.03	1.14	0.233	<0.01**
Psychological response	Lassitude	1.63	0.875	3.03	1.11	–14.6	<0.01**	1.67	1.03	3.34	1.18	–7.68	<0.01**
	Irritation	1.59	0.756	3.21	1.10	–21.9	<0.01**	1.61	0.652	3.37	1.11	–13.6	<0.01**
	Fatigue	1.50	0.623	3.18	0.982	–26.9	<0.01**	1.58	0.743	3.06	0.945	–8.78	<0.01**
	Anxiety	1.54	0.714	3.41	0.994	–27.1	<0.01**	1.64	0.773	3.73	1.03	–14.3	<0.01**
	Depression	1.46	0.633	3.49	1.03	–31.6	<0.01**	1.52	0.563	3.96	0.978	–19.3	<0.01**
	Physical stress response	1.72	0.840	3.17	0.936	–17.3	<0.01**	1.50	0.662	3.10	0.962	–9.49	<0.01**
Buffering factors	Supervisor’s support	1.91	0.917	2.82	1.09	–10.6	<0.01**	1.94	0.885	2.55	1.12	–3.69	<0.01**
	Coworkers’ support	1.98	0.986	2.56	0.999	–6.52	<0.01**	1.76	0.698	2.54	1.02	–5.84	<0.01**
	Support from family and friends	2.74	1.45	3.29	1.38	–4.37	<0.01**	2.76	1.39	3.60	1.33	–3.44	<0.01**
	Job satisfaction and life satisfaction	2.21	0.864	3.22	0.835	–13.0	<0.01**	2.29	0.905	3.33	0.888	–5.82	<0.01**

**p < 0.01. ^†Scores between the two groups were compared using the t-test (Welch t-test). SD, standard deviation.

Logistic regression: In men, significant differences were observed between the severe group and non-severe group in all factors. Therefore, we used all factors as independent variables in the logistic regression and set the presence of severe stress as the objective variable. We conducted logistic regression using forward selection (likelihood ratio). As a result, we used the Hosmer Lemeshow goodness of fit to assess the model fit, and it was significantly satisfactory (χ2 = 8.54 p = 0.382). The discrimination rate of presence of severe stress using this model was 89.9%. We therefore considered this model to be satisfactory. In this model, the following six factors were selected with logistic regression analyses, and all six contributed significantly to the presence of severe stress: “Qualitative job overload”, “Job control”, “Interpersonal conflict”, “Poor physical environment”, “Supervisor’s support”, and “Job satisfaction and life satisfaction”.

In women, we set 11 objective variables, excluding the two factors that were not significant in the Welch t-test. As a result of logistic regression, the Hosmer Lemeshow goodness of fit was significantly satisfactory (χ2 = 3.82, p = 0.872), and the discrimination rate was 92.6%. Therefore, we found this model to be satisfactory. We found that the following five factors significantly contributed to the presence of severe stress in women: “Qualitative job overload”, “Job control”, “Interpersonal conflict”, “Suitable work”, and “Job satisfaction and life satisfaction”. Detailed information including odds ratios and 95% confidence intervals for the logistic regression results is shown in Table 4.

Table 4

Logistic regression analysis results of the presence of severe stress evaluated by the BJSQ in men and women

	OR	95% CI	p-value
Men
Quantitative job overload	0.508	0.411–0.627	<0.01**
Job control	0.627	0.495–0.795	<0.01**
Interpersonal conflict	0.628	0.483–0.815	<0.01**
Poor physical environment	0.733	0.585–0.918	<0.01**
Supervisor’s support	0.683	0.524–0.890	<0.01**
Job satisfaction and life satisfaction	0.380	0.280–0.516	<0.01**
Women
Quantitative job overload	0.408	0.251–0.663	<0.01**
Job control	0.617	0.390–0.978	<0.05*
Interpersonal conflict	0.597	0.357–0.997	<0.05*
Suitable work	0.554	0.329–0.935	<0.05*
Job satisfaction and life satisfaction	0.414	0.229–0.748	<0.01**

*p < 0.05, **p < 0.01. OR, odds ratio; CI, confidence interval.

4 Discussion

4.1 Factors common to both men and women

Factors related to severe stress common to both men and women were as follows: “Qualitative job overload”, “Job control”, “Interpersonal conflict”, and “Job satisfaction and life satisfaction”. “Job demands” and “Job control” are core factors in the JDC model [4], and many reports have indicated that these are key factors related to severe stress [27 –29]. Additionally, a previous study showed a correlation among “Job demands”, “Job control”, and “Job satisfaction” in both men and women [7]. These correlations have also been observed in previous studies in the construction industry [30, 31]. Therefore, the results of this study were consistent with previous findings. In addition, in both men and women in this study, “Qualitative job overload” was a factor related to severe stress. In the construction industry, Rodriguez et al. reported that high cognitive demands are required for technological innovation, which results in increased quality of job demands required [32]. The quantity of job demands has been the focus of International Organization for Standardization in the workplace (ISO 44001) or “work system reform” promoted by the Japanese government since 2018 [33]. However, the results of this study showed the necessity of focusing on the quality of job demands. We need to pay attention to this point in developing support for employees based on this result.

Regarding “Interpersonal conflict”, some previous studies reported that this was a factor significantly related to severe stress in women [2, 34]. However, results of this study showed this factor was significantly related to severe stress in both men and women. In order to clarify this result, more information, such as who employees have stressful work relationships with (e.g., co-workers or supervisors), is required. We plan to investigate this further in a future study.

4.2 Factors specific to men

“Supervisor’s support” (odds ratio: 0.683) was found to be an important factor related to severe stress. The relationship between supervisor’s support and stress states prospected from gender has been previously reported. Sakagami [35] pointed out the importance of supervisor’s support in males. Boschman et al. [30] conducted a study in the construction industry and found that when male supervisors in management roles had less support from their boss, the risk of depression was 7.5 times higher than the general population.

On the other hand, the influence of support from supervisors has been found to differ between men and women [7, 9]. Some previous reports suggested a gender difference in the influence of support from supervisors from the viewpoint of strategies to deal with problems in the workplace [7, 34]. They indicated that men tended to deal with problems directly, while women tended to use emotional strategies like demonstrating understanding or sympathy. Therefore, men often needed support from their supervisors in order to resolve problems, and access to this support often made a big impact on their mental health. The present study was conducted in a construction company in which the majority of employees were men. Moreover, the main roles were fulfilled by men and men had more responsibility than women because of the nature of the occupation. Therefore, it seemed natural that the factor “Supervisor’s support” was specific to men in this study.

In the JDC-S model, which is an advanced version of the JDS model, support from others is an important axis, in addition to “Job demands” and “Job control”. High stress situations in the model are indicated by high job demand, low job control, and low support from others. However, Fila et al. [7] indicated that men’s support situations fit this model, but those of women did not. Although they showed that support was an important factor influencing workers’ well-being, they suggested that support factors be treated independently from the other two factors of “Job demands” and “Job control”. Namely, this suggestion showed that support factors could have different tendencies by gender. Therefore, we thought that our finding that “Supervisor’s support” was significant only in men was in agreement with that suggestion. Additionally, our results suggested the necessity of future study focusing on gender differences in social support because social support is one of the most general intervention methods in the workplace.

“Poor physical environment” (odds ratio: 0.773) was another factor specific to men. This factor evaluates the degree of comfort in the workplace environment with respect to aspects like noise, temperature, and brightness. The target company was a construction company, and most employees work outside. Therefore, this result seemed to reflect those situations. However, those were only speculative, because we did not investigate the workplaces of the participants in this study. In this study, we only could suggest that it is necessary for management to show some ingenuity, even if those environments are unavoidable because of the nature of the job.

4.3 Factors specific to women

Our results showed “Suitable work” was specific to women. This factor consisted of the question “My work is suitable for me”. This meant that women experienced severe stress because they did not feel their work was a good match for them.

Padkapayeva et al. [9] revealed that differences between men and women had two background aspects, sex and gender, and women’s problems in the workplace included gender problems as a background factor. A previous interview study with women working in a company with a majority of male employees indicated that they felt they needed to recognize the culture of their male-dominated company and adjust accordingly [36]. The authors of that study developed a women-workplace culture questionnaire (WWQ) and found that perceived burdens on women, sexual harassment and social support, in addition to the individual’s burden, were related to stress and mental disorder [36]. Additionally, Reinholdt et al. [37] also mentioned stress factors related to minority status and being a woman in extremely gender-segregated occupations. Hence, in male-dominated companies, it seemed that the level of fit with the company culture and occupational demands could be factors related to severe stress in female employees. The results of this study are suspected to be related to that finding. However, we did not investigate the level of fit with the company culture or the gender roles in the target company, as in the WWQ. Therefore, we could not determine whether the level of fit to the company culture was a factor related to severe stress in our female participants. The BJSQ did not include these perspectives, but our results suggested the importance of screening that includes level of fit to the company culture and company gender roles, especially in companies that are male dominated.

4.4 Study limitations

Our study had several limitations. Firstly, this study was conducted in one company and used cross-sectional data for 4 years. Our data were influenced by the culture of the company, especially the gender culture. Therefore, we need further study to confirm our results and conclusion. Moreover, we extracted the most recent data when participants had complete results for more than 1 year because data for the first year of employment were not thought to reflect the correct situation. However, it is possible that interventions for mental disorder, for example sickness absence, occurred during the study period. We did not consider such interventions in this study. Secondly, some odds ratios were significant; however, those effects were a small or moderate contribution to severe stress. In this study, we aimed to extract factors contributing to severe stress by using the BJSQ, which is generally used in companies to screen employee health. The strength of using the BJSQ was that it is an appropriate tool for developing interventions in actual workplaces with staff engaging in occupational safety and health because it has already been used in many companies. From that perspective, our results are expected to be helpful in suggesting new viewpoints, even if the effect of this study was not large. Based on this limitation, in order to confirm our results, we need further research across several companies and with more data. Moreover, in order to reveal the influence of those factors on mental health, we need to investigate changes in the factors over a long period of time. It is also important to consider the existence and contents of interventions based on screening using the BJSQ, and we need holistic data to do that. Thirdly, our analyses did not include basic information such as occupation, age, and working hours. As mentioned above, previous studies showed a difference in stress factors depending on the kind of work or occupational position. Basic background information might have influenced the significance of stress factors revealed in this study. We saw this study as first step in revealing the psychosocial factors related to severe stress in the workplace and developing actual intervention methods, and we only investigated the perspective of gender differences, which is one of the most important topics in Japanese companies nowadays. In a future study, we hope to use more holistic data to further develop the base for interventions in the workplace.

5 Conclusion

In conclusion, we extracted the psychosocial factors that contribute to severe stress in the workplace by analyzing data from the BJSQ. Factors common to both men and women were “Qualitative job overload”, “Job control”, “Interpersonal conflict”, and “Job satisfaction and life satisfaction”. Factors specific to men were “Poor physical environment” and “Supervisor’s support”, and the one factor specific to women was “Suitable work”. Some factors contributing to severe stress were different between men and women, and our results suggested the importance of paying attention those in their managements.

Ethical approval

The study was approved by the Ethics Committee of the Graduate School of Medicine, Nagoya University, Japan (authorization number: 2019–0090). The study was carried out according to the principles of the 1964 Declaration of Helsinki and its later amendments.

Informed consent

Informed consent was obtained from participants using the opt-out method.

Conflict of interest

None to report.

Footnotes

Acknowledgments

The authors are extremely grateful to all the participants as well as the company staff.

Funding

This work was supported by JSPS KAKENHI under grant number JP18K17716.

References

Salvagioni

DAJ

, Melanda

, Mesas

, González

, Gabani

, De Andrade

. Physical, psychological and occupational consequences of job burnout: A systematic review of prospective studies. PLoS One. 2017;12(10):1–29.

Osca

, López-Araújo

. Work stress, personality and occupational accidents: Should we expect differences between men and women? Saf Sci [Internet]. 2020;124(November 2019):104582. Available from: https://doi.org/10.1016/j.ssci.2019.104582

Madsen

IEH

, Nyberg

, Magnusson Hanson

, Ferrie

, Ahola

, Alfredsson

, et al. Job strain as a risk factor for clinical depression: Systematic review and meta-analysis with additional individual participant data. Psychol Med. 2017;47(8):1342–56.

Karasek

, Robert

, Theorell

. Healthy work: stress, productivity, and the reconstruction of working life. Basic Books; 1990. pp. 2–81.

CDC- Organization of Work: Measurement Tools: NIOSH Generic Job Stress Questionnaire - NIOSH [Internet]. [cited 2020 Jun 10]. Available from: https://www.cdc.gov/niosh/topics/workorg/detail088.html

Karasek

, Brisson

, Kawakami

, Houtman

, Bongers

, Amick

. The Job Content Questionnaire (JCQ): an instrument for internationally comparative assessments of psychosocial job characteristics. J Occup Health Psychol. 1998;3(4):322–55.

Fila

, Purl

, Griffeth

. Job demands, control and support: Meta-analyzing moderator effects of gender, nationality, and occupation. Hum Resour Manag Rev [Internet]. 2017;27(1):39-60. Available from: https://dx.doi.org/10.1016/j.hrmr.2016.09.004

Živković

, Milenović

, Krstić

, Veljković

. Correlation between psychosocial work factors and the degree of stress. Work. 2021;69(1):235–45.

Padkapayeva

, Gilbert-Ouimet

, Bielecky

, Ibrahim

, Mustard

, Brisson

, et al. Gender/sex differences in the relationship between psychosocial work exposures and work and life stress. Ann Work Expo Heal. 2018;62(4):416–25.

10.

Petković

, Nikolić

. Educational needs of employees in work-related stress management. Work. 2020;65(3):661–9.

11.

Ruotsalainen

, Verbeek

, Mariné

, Serra

. Preventing occupational stress in healthcare workers. Vol. 2015, Cochrane Database of Systematic Reviews. John Wiley and Sons Ltd; 2015.

12.

Joyce

, Modini

, Christensen

, Mykletun

, Bryant

, Mitchell

, et al. Workplace interventions for common mental disorders: a systematic meta-review. Psychol Med. 2016;46(4):683–97.

13.

Ali

, Wolf

, Hammersley

, Hoffmann

, O’Brien

, Phillips

, et al. Continuity of care in intensive care units: A cluster-randomized trial of intensivist staffing. Am J Respir Crit Care Med. 2011;184(7):803–8.

14.

Lucas

, Trick

, Evans

, Mba

, Smith

, Das

, et al. Effects of 2- vs 4-week attending physician inpatient rotations on unplanned patient revisits, evaluations by trainees, and attending physician burnout: A randomized trial. JAMA - J Am Med Assoc. 2012;308(21):2199–207.

15.

Seaton

, Bottorff

, Jones-Bricker

, Oliffe

, DeLeenheer

, Medhurst

. Men’s Mental Health Promotion Interventions: A Scoping Review. Am J Mens Health. 2017;11(6):1823–37.

16.

Ministry of Health L and WJ. Abstract of result of occupational health surveillance in 2018 [Internet]. [cited 2020 May 28]. Available from: https://www.mhlw.go.jp/toukei/list/h30-46-50b.html

17.

Ministry of Health L and WJ. Listen to voice your heart: Portal site for working people including Suicide Prevention [Internet]. [cited 2020 May 28]. Available from: https://kokoro.mhlw.go.jp/infographics/02.html

18.

Tsutsumi

, Shimazu

, Eguchi

, Inoue

, Kawakami

. A Japanese Stress Check Program screening tool predicts employee long-term sickness absence: A prospective study. J Occup Health. 2018;60(1):55–63.

19.

Kachi

, Inoue

, Eguchi

, Kawakami

, Shimazu

, Tsutsumi

. Occupational stress and the risk of turnover: A large prospective cohort study of employees in Japan. BMC Public Health. 2020;20(1):1–8.

20.

Kawada

, Otsuka

. Relationship between job stress, occupational position and job satisfaction using a brief job stress questionnaire (BJSQ). Work. 2011;40(4):393–9.

21.

Gender Equality Bureau. gender equality program [Internet]. Japanese Gender Equality Bureau. 2018 [cited 2020 May 28]. Available from: https://www.gender.go.jp/public/kyodosankaku/2018/201806/201806_02.html

22.

Survey on time use and leisure activities [Internet]. Japanese Ministry of Internal Affairs and Communications Statistics Bureau. 2016 [cited 2020 May 28]. Available from: https://www.stat.go.jp/data/shakai/2016/index.html

23.

Cunha

, Nogueira

, Lacomblez

. Beyond a man’s world: contributions from considering gender in the study of bus drivers’ work activity. Work. 2014;47(4):431–40.

24.

Kawakami

, Tsutsumi

. The Stress Check Program: A new national policy for monitoring and screening psychosocial stress in the workplace in Japan. J Occup Health. 2016;58(1):1–6.

25.

Kawakami

, Kobayashi

, Araki

, Haratani

, Furui

. Assessment of job stress dimensions based on the job demands-control model of employees of telecommunication and electric power companies in japan: reliability and validity of the japanese version of the job content questionnaire. Int J Behav Med. 1995;2(4):358–75.

26.

the Brief Job Stress Questionnaire [Internet]. Tokyo Medical University. Department of Preventive Medicine and Public Health. [cited 2020 May 28]. Available from: https://www.tmu-ph.ac/topics/stress_table.php

27.

Inoue

, Otsui

, Yoshioka

, Suzuki

, Ozawa

, Iwata

, et al. A simultaneous evaluation of occupational stress and depression in patients with lifestyle-related diseases. Intern Med. 2016;55(9):1071–5.

28.

Niedhammer

, Coindre

, Memmi

, Bertrais

, Chastang

. Working conditions and depression in the French national working population: Results from the SUMER study. J Psychiatr Res [Internet]. 2020;123(January):178-86. Available from: https://doi.org/10.1016/j.jpsychires.2020.01.003

29.

Ilies

, Dimotakis

, De Pater

. Psychological and physiological reactions to high workloads: Implications for well-being. Pers Psychol. 2010;63(2):407–36.

30.

Boschman

, van der Molen

, Sluiter

, Frings-Dresen

MHW

. Psychosocial work environment and mental health among construction workers. Appl Ergon [Internet]. 2013;44(5):748-55. Available from: https://dx.doi.org/10.1016/j.apergo.2013.01.004

31.

Chan

APC

, Nwaogu

, Naslund

. Mental Ill-Health Risk Factors in the Construction Industry: Systematic Review. J Constr Eng Manag. 2020;146(3):1–13.

32.

Rodriguez

, Spilski

, Hekele

, Beese

, Lachmann

. Physical and cognitive demands of work in building construction. Eng Constr Archit Manag. 2019;27(3):745–64.

33.

Ministry of Health L and WJ. For “work system reform” [Internet]. [cited 2020 May 28]. Available from: https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/0000148322.html

34.

Mazzola

, Schonfeld

, Spector

. What qualitative research has taught us about occupational stress. Stress Heal. 2011;27(2):93–110.

35.

Sakagami

. Qualitative job stress and ego aptitude in male scientific researchers. Work. 2016;55(3):585–92.

36.

Bergman

, Hallberg

LRM

. Women in a male-dominated industry: Factor analysis of a women workplace culture questionnaire based on a grounded theory model. Sex Roles. 2002;46(9-10):311–22.

37.

Reinholdt

, Alexanderson

. A narrative insight into disability pensioners’ work experiences in highly gender-segregated occupations. Work. 2009;34(3):251–61.

	Participants (n = 1243)
	Men (n = 911, 73.3%)		Women (n = 332, 26.7%)
	Severe group (n = 142)	Non-severe group (n = 769)	Severe group (n = 32)	Non-severe group (n = 300)
Mean age, y	46.3±14.8		54.1±13.6
	43.5±13.9	46.8±15.0	52.1±14.0	54.4±13.6
Employment status and job type (n, %)
Full time worker
General worker	247 (27.1%)		63 (19.0%)
	49 (34.5%)	198 (26.0%)	7 (21.9%)	56 (18.6%)
Specialized worker	280 (30.7%)		14 (4.22%)
	42 (29.5%)	238 (31.2%)	2 (3.25%)	12 (4%)
Industrial engineer	148 (16.2%)		43 (13.0%)
	29 (20.4%)	119 (15.6%)	4 (12.5%)	39 (13%)
Manager	15 (1.64%)		0 (0%)
	3 (2.11%)	12 (1.64%)	0 (0%)	0 (0%)
Contracted worker	135 (14.8%)		21 (6.33%)
	10 (7.04%)	125 (1.57%)	1 (3.12%)	20 (6.67%)
Part-time worker	86 (9.55%)		191 (57.5%)
	9 (6.33%)	77(10.1%)	18 (56.2%)	173 (57.7%)