Abstract
Objectives
In the Brief Job Stress Questionnaire (BJSQ), items A12–A14 assess interpersonal conflict in the workplace. A12 (disagreement) and A14 (friendliness) are typically summed, which implicitly assumes that disagreement is adverse. However, employees who perceive high disagreement in a friendly work environment may be able to voice differing opinions without concern, a pattern consistent with psychological safety, and may form a low-risk group overlooked by summed-score approaches. This study investigated how combinations of perceived disagreement and friendliness are associated with long-term sickness absence (LTSA) due to mental disorders.
Methods
A baseline survey using the BJSQ was conducted among 14,687 employees of a financial services company without LTSA in the previous three years. LTSA and diagnoses during one-year follow-up were obtained from personnel records. Participants were classified into four groups based on positive/negative responses to A12 and A14: high-disagreement and high-friendliness (HD+HF), high-disagreement and low-friendliness (HD+LF), low-disagreement and high-friendliness (LD+HF), and low-disagreement and low-friendliness (LD+LF). Cox proportional hazards models estimated hazard ratios (HRs) and 95% confidence intervals (CIs) for LTSA due to mental disorders, using the HD+LF group as reference. Interaction between disagreement and friendliness was also tested.
Results
After adjusting for potential confounders, HRs were significantly lower for both the HD+HF group (HR=0.18, 95% CI: 0.07–0.44, p<.001) and the LD+HF group (HR=0.26, 95% CI: 0.14–0.50, p<.001), with overlapping CIs. The interaction was marginally significant but imprecise (HR=4.89, 95% CI: 0.90–26.6, p=.066).
Conclusions
Perceived friendliness in the workplace remained protective against LTSA due to mental disorders, even with disagreement.
Keywords: absenteeism, disagreement and friendliness, job stress, longitudinal studies, Stress Check System, survival analysis
Introduction
It has been a decade since the implementation of Japan’s mandatory Stress Check System for workplaces with 50 or more workers, which aims to provide primary prevention of mental health problems among workers1). Under this system, employers are required to administer a self-report questionnaire annually to workers to assess psychological burden (i.e., job stressors, stress reactions, and buffer factors), with each worker’s own results provided directly to them by designated Stress Check administrators. Workers identified as “high stress” may request a physician interview. Aggregated group-level results, derived from individual responses, are provided to employers for workplace environment improvement, with employers encouraged to utilize these findings. In the Stress Check survey, the Brief Job Stress Questionnaire (BJSQ)2) is recommended as the self-report questionnaire1) because it efficiently assesses various job stressors, stress reactions, and buffer factors using only 57 items. For group analysis, the Job Stress Assessment Diagram3) has been widely used to estimate workplace health risk based on selected BJSQ items reflecting the Demand-Control-Support model4). Recently, however, some workplaces have expanded group analyses to include a broader range of BJSQ items or subscales, reflecting increasing interest in more detailed evaluations of the workplace environment5).
Thus, as group analyses using the BJSQ are derived from individual employees’ self-reported responses and increasingly rely on detailed items or subscales, it becomes important to examine more closely how specific items function and are interpreted. In this study, we focus on items A12 (“There are differences of opinion within my department,” hereafter disagreement) and A14 (“The atmosphere in my workplace is friendly,” hereafter friendliness). Although items A12–A14 are conventionally summed (after reverse-coding A12) to assess interpersonal conflict in the workplace, one of the job stressors2), we did not consider A13 (“My department does not get along well with other departments”) because it primarily reflects interdepartmental relations rather than employees’ experiences within their own department, which were the focus of this study. The conventional summed-score approach implicitly interprets disagreement as uniformly adverse, positioning employees who perceive high disagreement and low friendliness as having higher-risk interpersonal profiles, whereas those who perceive low disagreement and high friendliness are positioned as having lower-risk profiles. Under this approach, employees who perceive high disagreement and high friendliness, as well as those who perceive low disagreement and low friendliness, tend to be treated as representing intermediate levels of risk.
However, disagreement may not always be adverse: employees who perceive high disagreement in a friendly work environment (i.e., high friendliness) may experience psychological safety rather than stress. According to Edmondson’s conceptualization, psychological safety involves the ability to express dissenting views while maintaining supportive and non-threatening interpersonal relationships6). In this sense, employees who perceive high disagreement and high friendliness may feel able to voice differing opinions without fear of embarrassment, rejection, or interpersonal conflict. Consistent with this interpretation, existing psychological safety scales include conceptually similar items, such as “Nobody in my unit will pick on me even if I have different opinions”7). Taken together, employees who perceive high disagreement and high friendliness may constitute another lower-risk group which differs in characteristics from those who perceive low disagreement and high friendliness, a condition which has been conventionally considered the most ideal under the summed-score approach. Given that psychological safety has been reported to be associated with reduced burnout, a well-known predictor of long-term sickness absence (LTSA)8,9), this group of employees may plausibly be at lower risk of LTSA. Examining disagreement in conjunction with friendliness as two distinct but interacting aspects of employees’ perceptions may better capture the different risk profiles associated with workplace interpersonal climate.
In this study, we used the same database of Japanese employees as that used in previous studies examining the predictive effects of “high stress” based on the Stress Check survey, job dissatisfaction, and their combinations on LTSA (i.e., sickness absence of one month or longer)10,11,12). Using this database, we investigated how combinations of perceived disagreement and friendliness (items A12 and A14 of the BJSQ) are associated with the risk of LTSA due to mental disorders. We hypothesized that employees who perceive high disagreement and high friendliness would have a lower risk of LTSA due to mental disorders.
Methods
Participants
A one-year prospective cohort study was conducted between July 2015 and July 2016 among employees of a financial services company in Japan. In July–August 2015, all 15,615 employees, excluding executives, secondees, overseas assignees, temporary agency workers, and employees on leave, were invited to participate in a baseline survey using the BJSQ, and 14,711 employees responded (response rate: 94.2%). Subsequently, based on personnel records, 14,687 employees (7,343 men and 7,344 women) without a history of LTSA in the three years preceding the baseline were included in the cohort and followed until July 31, 2016. Informed consent for the secondary analysis of previously collected anonymized data was obtained using an opt-out approach. The study procedures were reviewed and approved by the Kitasato University Medical Ethics Organization (No. B15-113) and the Ethics Committee of Medical Research, University of Occupational and Environmental Health, Japan (No. ER23-005).
Exposure: combination of disagreement and friendliness
The levels of perceived disagreement and friendliness in the workplace were assessed using items A12 and A14 of the BJSQ, respectively (see Introduction for item wording). The response options for both items are “1=Very much so,” “2=Moderately so,” “3=Somewhat,” and “4=Not at all.” Although the response option “3=Somewhat” may be interpreted as neutral or slightly positive in English, it represents a mildly negative evaluation in the original Japanese language in which the survey was administered. Therefore, responses 1 and 2 were classified as positive (high) and responses 3 and 4 as negative (low), and participants were categorized into four groups based on the combinations of these binary classifications. For brevity, high and low levels were denoted as H and L, and disagreement and friendliness as D and F, respectively, yielding the following shorthand group labels: the HD+HF, HD+LF, LD+HF, and LD+LF groups.
Outcome: long-term sickness absence due to mental disorders
Information on the occurrence of LTSA during the follow-up period was obtained from the personnel records of the company under study. In this company, employees are required to submit a medical certificate from their attending physician to the human resources department when applying for sickness benefits. Accordingly, the personnel records included the start date of LTSA and the type of illness leading to LTSA (i.e., mental disorders, musculoskeletal disorders, cardiovascular diseases, or cerebrovascular diseases). The records also included information on separation from employment. Therefore, employees who experienced LTSA due to non-mental disorders and those who separated from employment during the follow-up period were treated as censored cases. The follow-up period started on the date on which each participant completed the BJSQ and ended on the earliest of the following: the start date of LTSA, the date of separation from employment, or July 31, 2016.
Potential confounders
Information on potential confounders, namely age, sex, years of service, occupation, and job position, was also obtained from the personnel records of the company under study. Age and years of service were treated as continuous variables; sex had two categories (men and women); occupation had four categories (sales, claims service, administrative, and others); and job position had five categories (manager, staff, senior employee, temporary employee, and others). These variables were used in the analyses without further recoding.
Statistical analysis
First, to describe the distribution of participants’ baseline characteristics, simple tabulations of the potential confounders (i.e., age, sex, years of service, occupation, and job position) were conducted for each of the four groups defined by the combinations of high and low levels of disagreement and friendliness. Additionally, cumulative hazard curves for LTSA due to mental disorders were plotted for these four groups using the Kaplan-Meier method, providing a visual overview of the incidence over the follow-up period. Incidence rates were also calculated for each group.
Next, as the main analysis, Cox proportional hazards models were fitted. To examine the main effects of disagreement and friendliness, hazard ratios (HRs) and 95% confidence intervals (CIs) for LTSA due to mental disorders were estimated for each variable separately, using the high disagreement group and the low friendliness group as the respective references. Subsequently, to examine the risk associated with each combination of disagreement and friendliness, HRs and 95% CIs were estimated for the LD+HF, HD+HF, and LD+LF groups, using the HD+LF group as the reference. Separately, the interaction between disagreement and friendliness was tested by including a product term (disagreement × friendliness) in the model, using the same reference group. Three models were examined: Model 1 (crude), Model 2 (adjusted for age and sex), and Model 3 (further adjusted for years of service, occupation, and job position). The proportional hazards assumption was evaluated using Schoenfeld residuals for the fully adjusted model (Model 3), and the global test did not indicate any statistically significant violation for the main effects of disagreement (p=.129) and friendliness (p=.329), or the combination of disagreement and friendliness (p=.111).
Finally, since LTSA due to non-mental disorders (i.e., musculoskeletal disorders, cardiovascular diseases, or cerebrovascular diseases) and separation from employment can be considered competing events for LTSA due to mental disorders, these events were treated not as censored but as competing events in a Fine-Gray subdistribution hazard model, which was performed as a sensitivity analysis. As in the main analysis, subdistribution HRs (SHRs) and 95% CIs were estimated for the main effects of disagreement and friendliness, the three combination groups (LD+HF, HD+HF, and LD+LF), and the interaction term between disagreement and friendliness.
The significance level was set at p<.05 (two-sided), and all statistical analyses were conducted using Stata/MP 17.0 (StataCorp., College Station, TX, USA).
Results
The distribution of baseline characteristics across the four groups defined by the combinations of high and low levels of disagreement and friendliness is shown in the upper panel of Table 1. Mean age ranged from 40.0 to 43.4 years, and the mean years of service ranged from 12.0 to 13.9 years, indicating no substantial differences across the groups. The proportions of men and women were also similar, although the HD+LF group included slightly fewer men (39.2%) than the other groups (43.9% to 53.0%). Occupation showed modest variation, with the proportion of employees in sales ranging from 45.4% to 56.6% and that in claims service from 27.8% to 39.7%. A similar pattern was observed for job position categories, with staff members comprising the majority in all groups (58.5% to 68.9%). Overall, baseline characteristics were broadly comparable across groups.
Table 1. Baseline characteristics and long-term sickness absence (LTSA) risk due to mental disorders by combinations of high and low levels of disagreement and friendliness (N=14,687).
| Low disagreement and high friendliness (LD+HF) group (n=9,222) |
High disagreement and high friendliness (HD+HF) group (n=3,450) |
Low disagreement and low friendliness (LD+LF) group (n=620) |
High disagreement and low friendliness (HD+LF) group (n=1,395) |
|||||
|---|---|---|---|---|---|---|---|---|
| Mean (SD) | n (%) | Mean (SD) | n (%) | Mean (SD) | n (%) | Mean (SD) | n (%) | |
| Baseline characteristics | ||||||||
| Age (years) | 41.9 (13.0) | 40.0 (11.2) | 43.4 (12.0) | 41.5 (11.1) | ||||
| Sex | ||||||||
| Men | 4,887 (53.0) | 1,637 (47.4) | 272 (43.9) | 547 (39.2) | ||||
| Women | 4,335 (47.0) | 1,813 (52.6) | 348 (56.1) | 848 (60.8) | ||||
| Years of service | 12.0 (10.5) | 13.3 (9.48) | 13.4 (10.1) | 13.9 (9.63) | ||||
| Occupation | ||||||||
| Sales | 4,186 (45.4) | 1,766 (51.2) | 295 (47.6) | 789 (56.6) | ||||
| Claims service | 3,663 (39.7) | 960 (27.8) | 242 (39.0) | 390 (28.0) | ||||
| Administrative | 1,370 (14.9) | 717 (20.8) | 81 (13.1) | 215 (15.4) | ||||
| Others | 3 (0.03) | 7 (0.20) | 2 (0.32) | 1 (0.07) | ||||
| Job position | ||||||||
| Manager | 1,483 (16.1) | 631 (18.3) | 66 (10.6) | 163 (11.7) | ||||
| Staff | 5,399 (58.5) | 2,313 (67.0) | 374 (60.3) | 961 (68.9) | ||||
| Senior employee | 404 (4.38) | 72 (2.09) | 37 (5.97) | 44 (3.15) | ||||
| Temporary employee | 1,933 (21.0) | 427 (12.4) | 141 (22.7) | 226 (16.2) | ||||
| Others | 3 (0.03) | 7 (0.20) | 2 (0.32) | 1 (0.07) | ||||
| LTSA risk due to mental disorders | ||||||||
| Follow-up and event summary | ||||||||
| Person-days | 3,300,920 | 1,239,879 | 220,514 | 497,597 | ||||
| Number of events | 27 | 7 | 2 | 15 | ||||
| Incidence rate (/100,000 person-days) | 0.82 | 0.56 | 0.91 | 3.01 | ||||
| Cox proportional hazards model (HR [95% CI], p-value) | ||||||||
| Model 1 a | 0.27 (0.14–0.51), p<.001 | 0.19 (0.08–0.46), p<.001 | 0.30 (0.07–1.31), p=.110 | 1.00 (reference) | ||||
| Model 2 b | 0.26 (0.14–0.50), p<.001 | 0.18 (0.07–0.43), p<.001 | 0.31 (0.07–1.36), p=.121 | |||||
| Model 3 c | 0.26 (0.14–0.50), p<.001 | 0.18 (0.07–0.44), p<.001 | 0.30 (0.07–1.32), p=.111 | |||||
| Fine-Gray subdistribution hazard model (SHR [95% CI], p-value) d | ||||||||
| Model 1 a | 0.27 (0.14–0.51), p<.001 | 0.19 (0.08–0.46), p<.001 | 0.30 (0.07–1.31), p=.109 | 1.00 (reference) | ||||
| Model 2 b | 0.26 (0.14–0.50), p<.001 | 0.18 (0.07–0.43), p<.001 | 0.31 (0.07–1.36), p=.121 | |||||
| Model 3 c | 0.26 (0.14–0.49), p<.001 | 0.18 (0.07–0.43), p<.001 | 0.30 (0.07–1.34), p=.115 | |||||
SD, standard deviation; HR, hazard ratio; CI, confidence interval; SHR, subdistribution hazard ratio.
Crude (no adjustment).
Adjusted for age and sex.
Further adjusted for years of service, occupation, and job position.
LTSA due to non-mental disorders (i.e., musculoskeletal disorders, cardiovascular diseases, or cerebrovascular diseases) and separation from employment were treated as competing events.
The cumulative hazard curves for LTSA due to mental disorders across the four groups are shown in Figure 1. Throughout the follow-up period, the HD+LF group exhibited the highest cumulative hazard, with the curve rising more steeply and earlier than in the other groups. The LD+LF and LD+HF groups showed relatively low cumulative hazard levels, with broadly similar trajectories, although the LD+LF curve rose slightly earlier in the follow-up. In contrast, the HD+HF group appeared to have the lowest cumulative hazard, with only a modest increase over time. The corresponding incidence rates were 0.82, 0.56, 0.91, and 3.01 per 100,000 person-days for the LD+HF, HD+HF, LD+LF, and HD+LF groups, respectively, as presented in the lower panel of Table 1.
Fig. 1.
Kaplan-Meier cumulative hazard estimates of long-term sickness absence (LTSA) due to mental disorders across four groups defined by high (H) and low (L) levels of disagreement (D) and friendliness (F) (N=14,687)
In Model 1 (crude) of the main analysis using the Cox proportional hazards models, the main effects of disagreement and friendliness indicated that both low disagreement and high friendliness were associated with a lower risk of LTSA due to mental disorders, although statistical significance was reached only for friendliness (HR=0.65, 95% CI: 0.37–1.13, p=.127 and HR=0.32, 95% CI: 0.18–0.57, p<.001, respectively). For the group-combination analysis, using the HD+LF group as the reference, HRs were significantly lower in both the HD+HF group (HR=0.19, 95% CI: 0.08–0.46, p<.001) and the LD+HF group (HR=0.27, 95% CI: 0.14–0.51, p<.001), with overlapping CIs. Although the LD+LF group also showed an HR below 1.0, the number of events was small, and the estimate did not reach statistical significance (HR=0.30, 95% CI: 0.07–1.31, p=.110). The interaction between disagreement and friendliness was marginally significant, with a wide CI (HR=4.82, 95% CI: 0.89–26.2, p=.069). These findings were largely unchanged in Model 2 (adjusted for age and sex) and Model 3 (further adjusted for years of service, occupation, and job position). Similarly, the SHRs and 95% CIs estimated from the sensitivity analysis using the Fine-Gray subdistribution hazard model, which treated LTSA due to non-mental disorders and separation from employment as competing events, were consistent with those obtained from the Cox proportional hazards models. The results of the group-combination analysis are presented in the lower panel of Table 1, whereas the results of the main effects of disagreement and friendliness and their interaction are shown in Supplementary Table 1.
Discussion
This study aimed to investigate how combinations of perceived disagreement and friendliness in the workplace (items A12 and A14 of the BJSQ) are associated with the risk of LTSA due to mental disorders. Consistent with our hypothesis, employees who perceived high disagreement and high friendliness exhibited a lower risk. Those who perceived low disagreement and high friendliness similarly exhibited a lower risk. The risk appeared to be lower among those who perceived low disagreement and low friendliness, but the estimate was unstable due to the small number of events. The interaction between disagreement and friendliness was marginally significant but imprecise. These findings remained largely unchanged after adjustment for potential confounders.
Employees who perceived high disagreement and high friendliness exhibited a lower risk of LTSA due to mental disorders. Among such employees, differing opinions may be more likely to be expressed and received in a constructive manner, rather than being perceived as a source of interpersonal conflict. This perceived interpersonal climate may share certain conceptual features with psychological safety, in which individuals can voice dissenting views without fear of negative interpersonal consequences6), and may also reflect fulfillment of basic psychological needs, such as autonomy (expressing differing opinions) and relatedness (supportive interpersonal climate), as suggested by Self-Determination Theory13). This pattern is also in line with the idea that psychological safety may help reduce burnout, which in turn could lower the risk of LTSA, as introduced earlier8,9). In addition to this interpretation, an alternative explanation may also be considered. Although “department” (A12) and “workplace” (A14) are defined at the organizational level in the BJSQ, employees may perceive these items as referring to different levels of their work environment. In large companies, such as the one examined in this study, a single department may encompass multiple subunits, and employees may interpret disagreement (A12) as reflecting differences across a broader organizational context, while interpreting friendliness (A14) as reflecting the interpersonal climate within their more immediate work setting. From this perspective, the combination of high disagreement and high friendliness may indicate strong intra-unit cohesion alongside inter-unit differences, consistent with the notion of in-group cohesion described in Social Identity Theory14). In addition to psychological safety, this pattern may also reflect supportive relationships within employees’ immediate work setting, which may contribute to better mental health outcomes. Taken together, these findings suggest that the meaning of perceived disagreement may depend on the broader interpersonal context in which it occurs.
A lower risk of LTSA due to mental disorders was also observed among those who perceived low disagreement and high friendliness. In the context of Japan’s collectivist cultural tendencies, which emphasize harmony and smooth interpersonal relationships15), employees who perceive such an interpersonal climate may experience their work environment as comfortable, low-stress, and generally favorable, making this pattern plausible. At the same time, the marginally significant interaction between disagreement and friendliness warrants attention, suggesting that the difference in risk between employees who perceive high disagreement and high friendliness and those who perceive low disagreement and high friendliness may be noteworthy, even if not statistically definitive. In particular, the HR for the interaction term being substantially greater than 1.0 can be interpreted as indicating that at least one group—here, the HD+HF group—had a lower risk than the LD+HF group, which lies diagonally opposite the reference group (HD+LF). Although employees who perceive low disagreement and high friendliness may view their workplace as conflict-free and comfortable, it is possible that in some cases they refrain from expressing differing opinions to maintain workplace harmony, which could partly contribute to the slightly higher risk observed relative to employees who perceive high disagreement and high friendliness. However, as noted above, the wide 95% CI for the interaction term indicates that this finding (i.e., the HR between the HD+HF and LD+HF groups) should be interpreted as tentative rather than robust.
Among employees who perceived low disagreement and low friendliness, the estimated risk of LTSA due to mental disorders was lower than that among those who perceived high disagreement and low friendliness; however, the difference was not statistically significant. Perceiving low disagreement may be associated with fewer experiences of interpersonal friction at work and, in this sense, may function protectively against the occurrence of LTSA due to mental disorders. However, when friendliness is also perceived as low, such perceptions may reflect limited communication and a closed or disengaged interpersonal climate, which may not be sufficient to effectively prevent LTSA due to mental disorders. As noted above, the number of cases in the LD+LF group was small in this study, and the analysis may therefore have lacked sufficient statistical power. Further studies using samples with a larger number of cases are needed to re-examine this association.
From a public health perspective, it is also informative to consider the absolute risk differences. The incidence rate of LTSA due to mental disorders (per 100,000 person-days) was 3.01 in the HD+LF group compared with 0.56 in the HD+HF group (see the lower panel of Table 1), corresponding to a difference of 2.45 (approximately 8.94 per 1,000 person-years). This suggests that a workplace climate characterized by high friendliness may have meaningful implications at the population level, despite the relatively low absolute risk. When compared with previous analyses in the same cohort12), the absolute risk difference associated with “high stress” (based on the Stress Check survey using the BJSQ) among employees with job dissatisfaction (3.91 for the high-stress group vs. 1.14 for the not high-stress group, a difference of 2.77) is of a comparable magnitude to that observed between the HD+HF and HD+LF groups. Although these contexts differ in baseline characteristics and cannot be directly compared, this observation suggests that the protective effect of high friendliness may be meaningful and potentially comparable in scale to the effect of high stress on the risk of LTSA due to mental disorders, supporting its public health relevance.
Several limitations of this study should be considered. First, employees who experienced high levels of stress related to their workplace environment and were at higher risk of LTSA due to mental disorders may have been less likely to respond to the baseline survey, potentially leading to underestimation of the associations. Given the high baseline response rate (approximately 95%), however, the impact of such selection bias is likely to have been minimal. Second, individual differences in personality traits and sensitivity to interpersonal disagreement were not assessed. Such characteristics may influence both the perception of workplace disagreement and the psychological impact of such environments, potentially affecting the observed associations16). Third, the outcome in this study was defined as LTSA due to mental disorders diagnosed by a physician, which presupposes help-seeking behavior. Individuals who developed mental health problems but did not seek medical care may have left their jobs before taking LTSA, resulting in incomplete case ascertainment. Nevertheless, sensitivity analyses treating separation from employment as a competing event yielded results largely consistent with the main analyses, suggesting limited impact on the findings. Finally, this study used data collected approximately 10 years ago from employees in a single financial services company, potentially limiting generalizability to current working conditions or other industries. However, disagreement and friendliness reflect fundamental aspects of workplace interpersonal climate that are likely to be broadly applicable, and the use of a single industry may have reduced industry-related confounding.
In conclusion, when perceptions of the workplace environment were evaluated by combining items A12 (disagreement) and A14 (friendliness) of the BJSQ, perceived friendliness remained protective against LTSA due to mental disorders, even in the presence of disagreement. This pattern may reflect aspects of the interpersonal climate that are conceptually related to psychological safety, in which differing opinions can be expressed and discussed without fear of negative consequences.
Author contributions
AI contributed to the conception or design of the work, the analysis and interpretation of the data, and drafting the work. KW, YO, AS, and NK contributed to the interpretation of the data and critically reviewed the work for important intellectual content. AT contributed to the acquisition and interpretation of the data, and critically reviewed the work for important intellectual content. All authors gave final approval of the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Supplementary material
This article contains supplementary material, which is available in the online version.
Supplementary Table 1
Sources of Funding
This study was supported by the Ministry of Health, Labour and Welfare (Health, Labour and Welfare Sciences Research Grants [Grant Number H27-Rodo-Ippan-004] and Industrial Disease Clinical Research Grants [Grant Numbers 230201-01 and 250301-01]) and the Japan Society for the Promotion of Science (JSPS KAKENHI [Grant Number JP24K13538]).
Conflicts of Interest
KW and NK are employed by the Department of Digital Mental Health, Graduate School of Medicine, The University of Tokyo, a social cooperation program supported by a grant from 15 enterprises (https://dmh.m.u-tokyo.ac.jp/c), outside the submitted work. NK is employed by the Junpukai Foundation, outside the submitted work. The other authors declare that there are no conflicts of interest.
References
- 1.Ministry of Health, Labour and Welfare. Manual for Implementation of the Stress Check System under the Industrial Safety and Health Act. Ministry of Health, Labour and Welfare; 2015. (in Japanese). [Google Scholar]
- 2.Shimomitsu T, Haratani T, Nakamura K, et al. Final development of the Brief Job Stress Questionnaire mainly used for assessment of the individuals. In: Kato M, ed. The Ministry of Labor Sponsored Grant for the Prevention of Work-Related Illness, FY 1999 Report. Tokyo Medical University; 2000: 126-164 (in Japanese). [Google Scholar]
- 3.Kawakami N, Miyazaki S, Tanaka M, et al. Final development of the Job Stress Assessment Diagram and investigation of its usefulness in practice. In: Kato M, ed. The Ministry of Labor Sponsored Grant for the Prevention of Work-Related Illness, FY 1999 Report. Tokyo Medical University; 2000: 12-26 (in Japanese). [Google Scholar]
- 4.Johnson JV, Hall EM. Job strain, work place social support, and cardiovascular disease: a cross-sectional study of a random sample of the Swedish working population. Am J Public Health. 1988; 78(10): 1336-1342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Inoue R, Sekine Y, Watanabe K, et al. [Contrivances for group analysis using the Brief Job Stress Questionnaire and its utilization: A systematic review]. Sangyo Eiseigaku Zasshi. 2026; 68(1): 1-15 (in Japanese). [DOI] [PubMed] [Google Scholar]
- 6.Edmondson A. Psychological safety and learning behavior in work teams. Adm Sci Q. 1999; 44(2): 350-383. [Google Scholar]
- 7.Liang J, Farh CIC, Farh JL. Psychological antecedents of promotive and prohibitive voice: A two-wave examination. Acad Manage J. 2012; 55(1): 71-92. [Google Scholar]
- 8.Hallsten L, Voss M, Stark S, Josephson M. Job burnout and job wornout as risk factors for long-term sickness absence. Work. 2011; 38(2): 181-192. [DOI] [PubMed] [Google Scholar]
- 9.de Lisser R, Dietrich MS, Spetz J, Ramanujam R, Lauderdale J, Stolldorf DP. Psychological safety is associated with better work environment and lower levels of clinician burnout. Health Aff Sch. 2024; 2(7):qxae091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tsutsumi A, Shimazu A, Eguchi H, Inoue A, Kawakami N.. A Japanese Stress Check Program screening tool predicts employee long-term sickness absence: A prospective study. J Occup Health. 2018; 60(1): 55-63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Inoue A, Tsutsumi A, Kachi Y, Eguchi H, Shimazu A, Kawakami N. Psychosocial work environment explains the association of job dissatisfaction with long-term sickness absence: A one-year prospect study of Japanese employees. J Epidemiol. 2020; 30(9): 390-395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Inoue A, Kachi Y, Eguchi H, Shimazu A, Kawakami N, Tsutsumi A. Combined effect of high stress and job dissatisfaction on long-term sickness absence: A 1-year prospective study of Japanese employees. Environ Occup Health Pract. 2020; 2(1): eohp.2020-0002-OA. [Google Scholar]
- 13.Deci EL, Ryan RM. The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychol Inq. 2000; 11(4): 227-268. [Google Scholar]
- 14.Tajfel H, Turner JC. An integrative theory of intergroup conflict. In: Austin WG, Worchel S, eds. The Social Psychology of Intergroup Relations. Brooks/Cole; 1979: 33-47. [Google Scholar]
- 15.Markus HR, Kitayama S. Culture and the self: Implications for cognition, emotion, and motivation. Psychol Rev. 1991; 98(2): 224-253. [Google Scholar]
- 16.Luo J, Zhang B, Cao M, Roberts BW. The stressful personality: A meta-analytical review of the relation between personality and stress. Pers Soc Psychol Rev. 2023; 27(2): 128-194. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Table 1
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.