A nationwide web-based survey of factors associated with depressive symptoms among Japanese workers

Abstract

Background:

Precarious employment has affected mental health, and limited data are available on the association of low stress tolerance with depressive symptoms among Japanese workers.

Aims:

This study aimed to examine the relationship between stress tolerance and depressive symptoms among Japanese workers, including company employees, civil servants and self-employed persons in various industries.

Methods:

We conducted a nationwide cross-sectional study. From March 26 to April 6, 2020, we performed a web-based survey of Japanese workers. The questionnaire included questions on socioeconomic factors, the SOC scale that assesses stress tolerance, the CES-D, and the EQ-5D-5L. Multivariate regression analyses were performed to determine the factors associated with depressive symptoms.

Results:

We included 3,001 participants in the analysis. A high SOC score, adequate sleeping time and frequency of exercise were associated with higher depressive symptoms. Employment status and long working hours were not associated with depressive symptoms. Younger workers had lower SOC scores than older workers. An inverse correlation between the SOC score and CES-D score was found among Japanese workers.

Conclusions:

Improving stress tolerance among younger workers is needed to prevent worsening mental health regardless of employment status for Japanese workers.

Keywords

Depression employment status mental health quality of life sense of coherence scale web-based survey

Introduction

Mental health illnesses among workers negatively affects work and productivity. A poor mental health status, including depression, is an independent risk factor associated with a lower quality of life (QOL) compared with a healthy mental status, and mental health illness is a health condition that most interferes with work (Ina et al., 2011; Yoshimoto et al., 2020) . The lifetime and 12-month prevalence of major depressive disorder in Japan were estimated as 6.6% and 1.2%, respectively (Ishikawa et al., 2016).

Recently, a web-based longitudinal study including 10,000 Japanese workers reported that 37.1% of the baseline cohort had depressive symptoms (Nishimura et al., 2020). Additionally, depression is considered a primary risk factor for suicide, particularly in older people (Conwell et al., 2002; Dhungel et al., 2019). Therefore, mental health improvement for workers with depression and the early detection of depressive symptoms for healthy workers are required to raise their labor productivity and prevent suicide in the future.

In Japan, there are several social problems that lead to decreased mental health in the labor environment. The Organization for Economic Co-operation and Development (OECD) indicated that among workers in other developed countries, Japanese workers had significantly longer working times and the lowest labor productivity in 2018 (Organization for Economic Co-operation and Development (OECD), 2020a, 2020b). Recently, several studies have indicated a significant association between long working hours and psychological distress in Japan (Bannai et al., 2015; Ogawa et al., 2018).

The number of non-permanent workers has been increasing in Japan (Statistics Bureau, 2020). Non-permanent workers have experienced adverse effects on their mental health due to employment instability, and unemployment status has been shown to be related to psychological statuses such as depression and anxiety and worsened health outcomes due to job loss. Therefore, these aspects of the social environment have been considered important factors contributing to negative impacts on the mental health of workers.

Several studies have demonstrated the predictive factors associated with increased depressive and anxiety symptoms. In Japan, a cohort study conducted by Sairenchi et al. (2011) revealed that stress tolerance may predict the onset of depression in Japanese workers, and Urakawa et al. (2012) reported that increasing stress tolerance may reduce negative job stress responses and subjective symptoms among general workers. Recently, a large-scale cross-sectional study conducted by Kikuchi H clarified that Japanese workers who worked more hours of overtime showed significantly higher anxiety and depression than those who worked fewer hours of overtime among both males and females (Kikuchi et al., 2020).

Moreover, a Korean study suggested that head of household status, sex, and precarious employment were associated with the development of severe depressive symptoms, and a previous study indicated that precarious employment was associated with a doubled risk of serious psychological distress among middle-aged Japanese men (Jang et al., 2015; Kachi et al., 2014).

The results of these previous studies lead to the hypothesis that these factors may be associated with depressive symptoms among workers in Japan. However, to our best knowledge, limited data are available on the relationship between stress tolerance and depressive symptoms for each employment status in Japan. Therefore, this study aimed to clarify the relationship between stress tolerance and depressive symptoms adjusted for the employment status and other socioeconomic factors among Japanese workers using nationally representative data.

Materials and methods

Study design and data collection

This cross-sectional nationwide survey was conducted online in Japan from March 26 to April 6, 2020, with general workers aged between 15 and 59 years through a platform with more than two million candidates. The survey was administered by the Cross Marketing Corporation, Tokyo, Japan, which specializes in questionnaire research. We ensured that the sample was representative of the Japanese population in terms of age, sex, and residential area during the collection phase. Residential area was based on the division of Japan into 10 regions. The target sample size of this study was 3,000 respondents.

The study was approved by the medical ethics committee of Niigata University of Health and Welfare (No. 18385-200318; date: April 18, 2020), and it was conducted in accordance with the principles of the 1964 Declaration of Helsinki and its later amendments. This article is based on a web-based survey, and the study did not involve any interventions conducted on human subjects by any of the authors. The respondents provided explicit consent, and the data were completely anonymous in the survey.

Questionnaire

The questions covered socioeconomic factors including employment status age, sex, residential area, marital status, children, number of family members living in the household, education level, industry, company size, personal income, family income, average overtime per month, status of labor union, householder, frequency of exercise, smoking history, drinking history, commute time, and average sleeping time.

We defined employment status as the following four types: permanent workers—company employees who are guaranteed lifetime employment until retirement, are hired directly by their employers, and have full-time employment; nonpermanent workers—company employees with fixed-term labor contracts, such as part-time employees, dispatched employees from temporary labor agencies, and contract or entrusted employees; civil servants—public service workers in national or local governments who are incorporated into nonprofit organizations; self-employed persons—self-employed individuals such as sole proprietors and freelancers.

The medical histories included cerebrovascular disease, cancer, Alzheimer’s disease, physical disorders with chronic pain, epilepsy, and depression, which are associated with the incidence of mental illnesses according to the Diagnostic & Statistical Manual of Mental Disorders (DSM-5) (Kuroki et al., 2016). The number of diseases was defined as a primary factor in this analysis. We used the Japanese version of the Sense of Coherence (SOC) scale to measure stress-coping ability, with a final score ranging from 13 to 91. This scale includes 13 items, and higher scores indicate better stress-coping ability (Ohta et al., 2015). SOC is a factor that estimates stress tolerance and a protective factor for mental health. The concept of SOC constitutes three core components termed comprehensibility, manageability, and meaningfulness.

We measured the degree of depressive symptoms using the Japanese version of the Center for Epidemiologic Studies Depression Scale (CES-D). This scale consists of 20 items that ask participants to rate how often they experienced symptoms associated with depression over the past week; the CES-D score ranges from 0 to 60, with a score above 16 usually indicating depression symptoms. The CES-D has good sensitivity, specificity, and high internal consistency for identifying the risk of depression (Lewinsohn et al., 1997).

We used the five-dimensional EQ-5D-5L instrument to assess the respondents’ QOL. The EQ-5D-5L consists of five items on mobility, self-care, usual activities, pain/discomfort, and anxiety/depression, which are rated according to five levels. The resulting generic preference-based measure reflects subjective values for specific health-related outcomes ranging from −0.025 to 1, with 0 indicating death and 1 indicating perfect health in the Japanese value set; we call this score the utility value (Shiroiwa, Ikeda, et al., 2016).

Statistical analysis

First, we performed a multiple linear regression analysis and a multiple logistic regression analysis to identify the determinants of the CES-D score and depressive symptoms, respectively. The logistic regression coefficients were transformed to odds ratios (ORs) with 95% confidence intervals (CIs). The independent variables used in the analysis were socioeconomic factors, comorbidity, and degree of SOC. There was no criterion to define the degree of SOC; therefore, we defined SOC categories according to the quartile of the SOC score. The participants were divided into four categories based on their SOC scores (poor, low, moderate, and high). Independent variables with variance inflation factors (VIFs) less than three were included in the multiple linear regression analysis to avoid multicollinearity. In the multiple logistic regression analysis, the area under the curve (AUC) and Hosmer-Lemeshow test were used to assess the discrimination and goodness-of-fit of the model, respectively.

Second, we described the trend in SOC scores by age group and in CES-D scores by SOC category. Trends between an ordinal independent variable and a continuous or ordinal dependent variable were examined using the Jonckheere-Terpstra test. Last, we also estimated the health-related utility values for each age group and for participants with depressive symptoms. The differences in the utility values of males and females in each subgroup were tested using the Wilcoxon rank-sum test. All statistical tests were two-sided, and p-values less than 0.05 were considered significant. All analyses were conducted using STATA 16.1 (College Station, Texas, USA: StataCorp LP).

Results

Characteristics of the participants

We collected data from 3,001 subjects, excluding housewives, students, and unemployed people in the web-based survey. The characteristics of the participants are shown in Table 1. The standardized differences indicates the difference in the independent variables between males and females. The numbers of males and females were 1,514 and 1,487, respectively. The median [interquartile range] of the SOC score was estimated to be 52 points [47 to 57 points]. In total, 838, 852, 574, and 737 participants had poor, low, moderate, and high SOC, respectively.

Table 1.

Characteristics of the participants.

Variable	Category	Total (N = 3,001)		Male (N = 1,514)		Female (N = 1,487)		Standardized difference
Variable	Category	N	%	N	%	N	%	Standardized difference
Employment status	Permanent worker	1,709	57.0	994	65.7	715	48.1	0.4910
	Nonpermanent worker	939	31.3	310	20.5	629	42.3
	Civil servant	193	6.4	105	6.9	88	5.9
	Self-employed	160	5.3	105	6.9	55	3.7
Age	15–19	105	3.5	22	1.5	83	5.6	0.2419
	20–29	710	23.7	393	26.0	317	21.3
	30–39	688	22.9	347	22.9	341	22.9
	40–49	813	27.1	410	27.1	403	27.1
	50–59	685	22.8	342	22.6	343	23.1
Living area	Hokkaido	121	4.0	58	3.8	63	4.2	0.0438
	Touhoku	202	6.7	102	6.7	100	6.7
	South Kanto	912	30.4	468	30.9	444	29.9
	North Kanto and Koushin	227	7.6	118	7.8	109	7.3
	Hokuriku	116	3.9	59	3.9	57	3.8
	Toukai	357	11.9	183	12.1	174	11.7
	Kinki	490	16.3	242	16.0	248	16.7
	Chugoku	167	5.6	83	5.5	84	5.6
	Shikoku	83	2.8	41	2.7	42	2.8
	Kyusyu	326	10.9	160	10.6	166	11.2
Marital status	Single	1,562	52.1	815	53.8	747	50.2	0.0720
Marital status	Married	1,439	48.0	699	46.2	740	49.8	0.0720
Children	No	1,924	64.1	976	−64.5	948	63.8	0.0149
Children	Yes	1,077	35.9	538	−35.5	539	36.2	0.0149
Number of family members living in the household	0	349	11.6	201	13.3	148	10.0	0.1238
	1	731	24.4	357	23.6	374	25.2
	2	705	23.5	332	21.9	373	25.1
	3	657	21.9	338	22.3	319	21.5
	4⩽	559	18.6	286	18.9	273	18.4
Number of comorbidities	0	2,683	89.4	1,344	88.8	1,339	90.0	0.0572
	1	286	9.5	150	9.9	136	9.1
	2⩽	32	1.1	20	1.3	12	0.8
Education level	Junior high school	84	2.8	42	2.8	42	2.8	0.5140
	High school	870	29.0	416	27.5	454	30.5
	Associate degree or diploma	707	23.6	222	14.7	485	32.6
	Bachelor’s degree	1,163	38.8	708	46.8	455	30.6
	Master’s degree or doctorate	177	5.9	126	8.3	51	3.4
Industry^†	Primary sector	29	1.0	16	1.1	13	0.9	0.3136
	Secondary sector	678	22.6	438	28.9	240	16.1
	Tertiary sector	2,160	72.0	1,003	66.2	1,157	77.8
	Industries unable to be classified	134	4.5	57	3.8	77	5.2
Company size (number of employees)	<50	1,018	33.9	412	27.2	606	40.8	0.3280
	50–99	392	13.1	190	12.5	202	13.6
	100–299	396	13.2	223	14.7	173	11.6
	300–999	410	13.7	219	14.5	191	12.8
	1,000⩽	705	23.5	429	28.3	276	18.6
	Not applicable	80	2.7	41	2.7	39	2.6
Personal income (JPY million)^‡	Low	1,900	63.3	722	47.7	1,178	79.2	0.7006
	Middle	819	27.3	571	37.7	248	16.7
	High	282	9.4	221	14.6	61	4.1
Family income (JPY million)^‡	Low	954	31.8	421	27.8	533	35.8	0.1837
	Middle	1,181	39.4	612	40.4	569	38.3
	High	866	28.9	481	31.8	385	25.9
Average hours of overtime per month	<10 hours	1,884	62.8	755	49.9	1,129	75.9	0.5604
	10–44 hours	847	28.2	573	37.8	274	18.4
	45–79 hours	209	7.0	143	9.4	66	4.4
	80 hours⩽	61	2.0	43	2.8	18	1.2
Labor union	No	2,141	71.3	1,058	69.9	1,083	72.8	0.0653
Labor union	Yes	860	28.7	456	30.1	404	27.2	0.0653
Householder	No	1,467	48.9	444	29.3	1,023	68.8	0.8593
Householder	Yes	1,534	51.1	1,070	70.7	464	31.2	0.8593
Exercise^§	None	1,804	60.1	786	51.9	1,018	68.5	0.3521
	1 time every 2 weeks	328	10.9	206	13.6	122	8.2
	1 time in a week	376	12.5	240	15.9	136	9.1
	2 times in a week	493	16.4	282	18.6	211	14.2
Smoking history	No	2,422	80.7	1,131	74.7	1,291	86.8	0.3111
Smoking history	Yes	579	19.3	383	25.3	196	13.2	0.3111
Drinking history	No	1,557	51.9	680	44.9	877	59.0	0.2843
Drinking history	Yes	1,444	48.1	834	55.1	610	41.0	0.2843
Commute time (one-way)	<30 minutes	1,669	55.6	753	49.7	916	61.6	0.2524
	30–59 minutes	976	32.5	541	35.7	435	29.3
	60 minutes⩽	356	11.9	220	14.5	136	9.1
Average sleeping time	<4.0 hours	89	3.0	55	3.6	34	2.3	0.0838
	4.0–5.9 hours	673	22.4	342	22.6	331	22.3
	6.0–7.9 hours	1,876	62.5	941	62.2	935	62.9
	8.0 hours⩽	363	12.1	176	11.6	187	12.6
SOC^¶	Poor	838	27.9	407	26.9	431	29.0	0.0887
	Low	852	28.4	451	29.8	401	27.0
	Moderate	574	19.1	273	18.0	301	20.2
	High	737	24.6	383	25.3	354	23.8

Note. SOC = sense of coherence.

†

Industry was categorized into four types: the primary sector, including agriculture and forestry and fisheries; the secondary sector, including the mining and quarrying of stone and gravel, construction, and manufacturing; the tertiary sector, including electricity gas heat supply and water, information and communications, transport and postal activities, wholesale trade and retail trade, finance and insurance, real estate and goods rental and leasing, scientific research, professional and technical services, accommodations, eating and drinking services, living-related and personal services and amusement services, education learning support, medical health care and welfare, compound services, other services, and government; and industries unable to be classified.

‡

Low (less than JPY 4 million), middle (between JPY 4 million to JPY 8 million), and high (more than JPY 8 million).

Exercise is defined as moderate exercise with light breathing for approximately 1 hour.

Categories were defined based on the interquartile range of the overall SOC score.

Regression analyses

The results of the multiple linear regression analysis and multiple logistic regression analysis are shown in Table 2. The variable number of family members living in the household was excluded in the analysis because the VIF was over 3. The mean VIF of the included independent variables was estimated to be 1.43. The AUC was estimated to be 0.764 and to range from 0.747 to 0.780, and the Hosmer-Lemeshow test showed no statistically significant difference (p = .101). Therefore, the multiple logistic regression analysis indicated the good fit of the model.

Table 2.

Results of the multivariate regression models.

Variable	Category	Linear regression model of factors associated with CES-D scores				Logistic regression model of factors associated with depressive symptoms
Variable	Category	Coefficient	Lower 95% CI	Upper 95% CI	p Value	Odds Ratio	Lower 95% CI	Upper 95% CI	p Value
Employment status	Permanent worker	1 (Reference)				1 (Reference)
	Nonpermanent worker	−0.010	−0.894	0.874	.982	0.931	0.751	1.156	.519
	Civil servant	1.442	−0.440	3.324	.133	1.308	0.815	2.099	.266
	Self-employed	0.369	−1.205	1.943	.646	1.001	0.675	1.484	.997
Gender	Male	1 (Reference)				1 (Reference)
Gender	Female	0.647	−0.144	1.438	.109	1.348	1.111	1.635	.002
Age	15–19	−0.374	−2.391	1.643	.716	0.692	0.424	1.129	.141
	20–9	1(Reference)				1 (Reference)
	30–39	0.289	−0.745	1.322	.584	0.878	0.684	1.128	.309
	40–49	−0.317	−1.355	0.722	.550	0.837	0.651	1.077	.166
	50–59	−1.430	−2.574	−0.285	.014	0.687	0.520	0.908	.008
Residential area	Hokkaido	−0.674	−2.471	1.124	.462	1.072	0.689	1.666	.758
	Touhoku	0.231	−1.241	1.703	.758	0.975	0.681	1.397	.890
	South Kanto	1 (Reference)				1 (Reference)
	North Kanto and Koushin	−0.367	−1.755	1.020	.604	1.098	0.782	1.542	.591
	Hokuriku	1.488	−0.352	3.329	.113	0.996	0.631	1.573	.986
	Toukai	−0.475	−1.640	0.690	.424	0.906	0.681	1.204	.495
	Kinki	0.277	−0.760	1.313	.601	0.987	0.767	1.271	.921
	Chugoku	0.408	−1.168	1.985	.611	1.284	0.875	1.884	.201
	Shikoku	1.666	−0.461	3.793	.125	1.232	0.734	2.068	.431
	Kyusyu	−0.962	−2.188	0.263	.124	0.953	0.707	1.286	.754
Marital status	Single	1 (Reference)				1 (Reference)
Marital status	Married	−1.490	−2.520	−0.460	.005	0.703	0.547	0.904	.006
Children	No	1 (Reference)				1 (Reference)
Children	Yes	0.106	−0.923	1.135	.840	1.102	0.855	1.420	.454
Number of comorbidities	0	1 (Reference)				1 (Reference)
	1	5.958	4.798	7.117	.000	2.874	2.106	3.921	.000
	2⩽	8.044	4.736	11.353	.000	12.457	3.816	40.668	.000
Education level	Junior high school	2.097	−0.022	4.216	.052	1.872	1.072	3.268	.027
	High school	1 (Reference)				1 (Reference)
	Associate degree or diploma	−0.218	−1.182	0.746	.658	1.020	0.807	1.290	.868
	Bachelor’s degree	−0.365	−1.279	0.550	.434	0.928	0.744	1.159	.511
	Master’s degree or doctorate	−0.537	−2.153	1.080	.515	1.116	0.750	1.660	.588
Industry	Primary sector	−6.050	−9.573	−2.527	.001	0.301	0.114	0.791	.015
	Secondary sector	−0.124	−0.968	0.719	.773	0.941	0.767	1.154	.558
	Tertiary sector	1 (Reference)				1 (Reference)
	Industries unable to be classified	0.292	−1.352	1.937	.727	0.867	0.584	1.287	.479
Company size (number of employees)	<50	1 (Reference)				1 (Reference)
	50–99	0.927	−0.187	2.040	.103	1.026	0.783	1.344	.853
	100–299	0.795	−0.331	1.921	.166	1.107	0.842	1.456	.467
	300–999	1.211	0.087	2.334	.035	1.176	0.893	1.549	.248
	1,000⩽	0.543	−0.465	1.551	.291	1.080	0.846	1.380	.537
	Not applicable	−0.768	−3.465	1.929	.577	1.046	0.539	2.028	.895
Personal income (JPY million)	Low	1 (Reference)				1 (Reference)
	Middle	−0.255	−1.353	0.842	.648	0.959	0.735	1.252	.761
	High	−0.986	−2.624	0.651	.238	0.731	0.487	1.096	.129
Family income (JPY million)	Low	1 (Reference)				1 (Reference)
	Middle	−0.816	−1.778	0.145	.096	0.894	0.708	1.130	.349
	High	−1.634	−2.828	−0.440	.007	0.776	0.579	1.039	.088
Average hours of overtime per month	<10 hours	1 (Reference)				1 (Reference)
	10–44 hours	0.039	−0.789	0.866	.927	0.953	0.779	1.166	.639
	45–79 hours	1.066	−0.332	2.464	.135	1.172	0.833	1.647	.362
	80 hours⩽	1.985	−0.449	4.418	.110	1.472	0.816	2.658	.199
Labor union	No	1 (Reference)				1 (Reference)
Labor union	Yes	0.601	−0.163	1.366	.123	1.084	0.900	1.305	.397
Householder	No	1 (Reference)				1 (Reference)
Householder	Yes	−0.157	−1.025	0.711	.722	0.913	0.740	1.126	.395
Exercise	No custom	1 (Reference)				1 (Reference)
	1 time every 2 weeks	0.306	−0.822	1.435	.595	1.324	1.002	1.750	.048
	1 time in a week	−0.677	−1.733	0.379	.209	0.820	0.634	1.059	.129
	2 times in a week	−1.748	−2.694	−0.802	.000	0.665	0.527	0.838	.001
Smoking history	No	1 (Reference)				1 (Reference)
Smoking history	Yes	−0.089	−0.977	0.799	.844	0.941	0.756	1.170	.583
Drinking history	No	1 (Reference)				1 (Reference)
Drinking history	Yes	−0.093	−0.801	0.614	.797	1.062	0.894	1.262	.493
Commute time (one-way)	<30 minutes	1 (Reference)				1 (Reference)
	30–59 minutes	0.185	−0.605	0.974	.646	1.148	0.947	1.393	.160
	60 minutes⩽	0.489	−0.659	1.637	.404	1.090	0.823	1.442	.548
Average sleeping time	<4.0 hours	3.492	1.472	5.513	.001	1.886	1.110	3.205	.019
	4.0–5.9 hours	1.200	0.360	2.041	.005	1.255	1.021	1.542	.031
	6.0–7.9 hours	1 (Reference)				1 (Reference)
	8.0 hours⩽	−0.591	−1.656	0.474	.276	0.679	0.526	0.877	.003
SOC	Poor	10.959	10.011	11.907	.000	9.761	7.582	12.566	.000
	Low	8.232	7.301	9.163	.000	7.360	5.755	9.413	.000
	Moderate	6.939	5.917	7.961	.000	5.470	4.207	7.112	.000
	High	1 (Reference)				1 (Reference)
Constant		12.274	10.480	14.069	.000	0.253	0.162	0.396	.000

Note. CI = confidence interval; SOC = sense of coherence; CES-D = Center for Epidemiologic Studies Depression Scale.

The two models showed that participants who were aged between 50 and 59; were married; had one or more than two comorbidities; worked in the primary sector; exercised two times a week; had an average sleeping time of 4.0 to 5.9 hours or less than 4.0 hours; and had poor, low, or moderate SOC had significantly different CES-D scores and depressive symptoms than the rest of the sample. Being aged between 50 and 59, being married, working in the primary sector, and exercising two times per week were associated with lower CES-D scores and depressive symptoms. Having one and more than two comorbidities, sleeping an average of 4.0 to 5.9 hours or less than 4.0 hours per night, and having a poor, low or moderate SOC were associated with higher CES-D scores and depressive symptoms.

Trend analyses

We described the proportion of participants in each SOC category in each age group and the distribution of CES-D scores among the SOC categories in Figures 1 and 2, respectively. Figure 1 shows the trend of older subjects having high SOC scores and younger subjects having low SOC scores. Figure 2 shows the inverse correlation between the CES-D score and SOC score. There was a trend of subjects with high SOC scores having low CES-D scores (p < .001) and subjects with low SOC scores having high CES-D scores (p < .001).There were significant differences in both trends.

Figure 1.

Mosaic plot of the proportion of participants in each SOC category by age group.

Figure 2.

Box plot of CES-D scores by SOC category.

Health-related utility value estimation

Table 3 shows the descriptive statistics of the estimated health-related utility values. The mean values were 0.868, 0.868, 0.884, 0.889, and 0.893 for participants aged 15 to 19, 20 to 29, 30 to 39, 40 to 49, and 50 to 59, respectively. The mean values for subjects with or without depressive symptoms were estimated to be 0.825 and 0.944, respectively. The proportion of subjects with depressive symptoms was 51.1% (1,532/3,001) in our data. There were no statistically significant differences in the utility values of males and females in any subgroup.

Table 3.

Descriptive statistics of the estimated health-related utility values.

Variable	Category	Total		Male		Female		p Value
Variable	Category	Mean	SD	Mean	SD	Mean	SD	p Value
Age	15–19	0.868	0.180	0.890	0.179	0.862	0.181	.394
	20–29	0.868	0.176	0.866	0.176	0.869	0.175	.852
	30–39	0.884	0.157	0.879	0.175	0.889	0.137	.463
	40–49	0.889	0.144	0.890	0.153	0.889	0.136	.359
	50–59	0.893	0.137	0.898	0.142	0.889	0.133	.092
CES-D score	No depressive symptoms (<16 point)	0.944	0.119	0.943	0.129	0.944	0.105	.076
CES-D score	Depressive symptoms (16 point⩽)	0.825	0.164	0.817	0.169	0.833	0.159	.064

Note. SD = standard deviation; CES-D = Center for Epidemiologic Studies Depression Scale.

Discussion

Diversification of employment types refers to various types of employment other than permanent employment in the last two decades in Japan. Increasing depression among nonpermanent workers has been identified nationally for a long time. In the present study, we examined the factors associated with depressive symptoms based on data from a Japanese nationwide web-based questionnaire. Our findings show that employment status is not a factor affecting the mental health of Japanese workers and indicate that other socioeconomic factors are related to their mental health condition.

SOC is a factor that can effectively predict the onset of depression. Our results demonstrated that the OR of depressive symptoms for workers with poor SOC was 10, which was higher than that of workers with high SOC; therefore, improving SOC could be an effective measure to prevent the onset of depression and to decrease the suicide rate due to depression among Japanese workers. We also report a new finding that younger workers had lower SOC than older workers in Japan. This finding likely suggests the low stress tolerance of younger people in Japan. Additionally, we were able to confirm an inverse correlation between the CES-D score and SOC score; workers with low SOC scores had a high prevalence of depressive symptoms in our study.

Similar results have been reported in other countries. A cross-sectional study showed that SOC scores were inversely related to scores on the Beck Depression Inventory and Beck Anxiety Inventory in adolescent females aged 15.9 to 17.7 years in Stockholm, Sweden (Blom et al., 2010). A nationwide study of 2003 Swedish men and women with a mean age of 44.2 years showed an inverse correlation between SOC and depressive symptoms (Larsson & Kallenberg, 1996).

According to a previous study, precarious employment and overtime work were factors that decreased the mental health of workers (Bannai et al., 2015; Jang et al., 2015; Ogawa et al., 2018). However, there were no statistically significant differences in the factors of employment status and average overtime hours worked per month in our data. These results indicate that the employment environment in Japan has recently improved. In fact, the average annual working time in Japan has decreased by approximately 100 hours in this decade (Organization for Economic Co-operation and Development (OECD)). We consider these factors to still be significant concerns in Japan. Further research focused on specific industries is needed to clarify the relationship between these factors and mental health.

Yamauchi et al. (2019) also suggested that both long working hours and sleep-related problems were positively associated with safety outcomes, including near misses and injuries among workers in Japan. Therefore, we expect that mental health problems due to long working hours will decrease for Japanese workers in the future.

We also identified factors associated with improved mental health. Adequate sleep and exercise contribute to improved mental health. A cross-sectional study conducted by Ohtsu et al. (2013) suggested that Japanese workers with long weekday working hours tended to have a short sleep duration on weekdays and holidays. It is essential for workers with higher CES-D scores to avoid working long hours to prevent short sleep durations and to try to improve their lifestyles to improve their mental health conditions.

We found a trend of younger workers having lower SOC and health-related utility than older workers. Older people have lower health-related utility than younger people because older people have a high prevalence of comorbidities and a high unemployment rate (Shiroiwa, Fukuda, et al., 2016). The reason for this trend in our study was that older persons who were unemployed and were unable to work due to comorbidities were excluded from our survey.

Our data suggested a high prevalence of depressive symptoms. Half of the subjects had depressive symptoms as indicated by a CES-D score of 16 points or above. However, we believe that COVID-19 pandemic did not influence our results because our survey was conducted before nationwide infection spread of the virus and the declaration of a state of emergency in Japan. Although Table S1 shows the number of participants in each industry, we could not identify the number of “frontliners” or “essential workers” and analyze any other psychiatric illness in our data.

Several studies have reported on the negative impact on mental health due to the global COVID-19 outbreak (Ma et al., 2020; Verma & Mishra, 2020; Tanaka et al., 2021). In Japan, a cohort study revealed that unemployment is a factor associated with worsening mental health. They suggested that employment security did not affect changes in the depressive symptom score at the beginning of the COVID-19 pandemic in Japan (Saito et al., 2021). Unemployment due to COVID-19 has been increasing rapidly since the COVID-19 outbreak in Japan; the unemployment rate was estimated at 2.9% in 2020 (The Japan Institute for Labor Policy and Training, 2021). Although our results showed that differences in employment status did not affect the psychological condition of Japanese workers, we should continue observing changes in the mental health of nonpermanent workers with precarious employment.

Additionally, we considered that workers with higher stress tolerance can sustain their mental health positively even under challenging situations. By contrast, workers with lower stress tolerance could be more negatively affected by temporary job loss than workers with higher stress tolerance. Our findings highlight the importance of designing societal support based on employment status and age, particularly for younger workers with lower stress tolerance with precarious employment in the pandemic.

We also estimated health-related utility values for several subgroups based on a nationally representative sample of workers in Japan. Utility is an important factor in health technology assessment (HTA). A cost-effectiveness study in HTA requires utility values to be calculated as effectiveness parameters and can be used to make decisions regarding health policy and drugs (Shiroiwa et al., 2017). We believe that our data can support future analyses of cost-effectiveness in the psychological and occupational health fields. Additionally, the estimated utility values based on Japanese representative data can be indicators to assess societal policies or services for occupational health considering changes in QOL.

Web-based surveys are a reliable method for epidemiological research (Ekman et al., 2006; Smith et al., 2007). However, this study had several limitations. First, younger participants aged between 15 and 19 represented only 3.5% of our sample. Therefore, some selection bias remained in our cohort compared with the Japanese population. However, we believe that this selection bias had minimal impact on our results because we succeeded in ensuring the representation of the Japanese population in the survey and obtaining a large cohort in the survey.

Second, we asked about the participants’ employment status and industry. However, our questionnaire was unable to elicit further information such as position and occupation. Hence, we must collect more detailed information to examine participants’ mental health conditions in future research. Finally, we were unable to use an additional approach to verify the respondents’ socioeconomic statuses or clinical histories because of the anonymous, self-reported nature of the survey. Despite these limitations, the statistical analysis presented in this study can provide important information for future health policies related to the mental health of workers in Japan.

In conclusion, this study identified factors associated with depressive symptoms for Japanese workers. Although employment status did not affect their mental health condition, lower stress tolerance indicated increasing depressive symptoms. Our results provide updated evidence on the occupational health field of Japanese workers. Younger workers have lower stress tolerance than older workers. Improving the stress tolerance in younger workers is likely one of the most effective measures to prevent depression in Japan. The COVID-19 pandemic remains a challenge in Japan that may continue in the long-term period. Thus, suicide due to depression is expected to increase in the near future. The Japanese government should provide policies not only for economic security for specific industries such as food service and tourism but also social support for younger workers with lower stress tolerance to improve their mental health and prevent depression under challenging situations.

Supplemental Material

sj-pdf-1-isp-10.1177_00207640211017586 – Supplemental material for A nationwide web-based survey of factors associated with depressive symptoms among Japanese workers

Supplemental material, sj-pdf-1-isp-10.1177_00207640211017586 for A nationwide web-based survey of factors associated with depressive symptoms among Japanese workers by Shota Saito, Ruan Qi, Huyen Thi Thanh Tran, Kenji Suzuki, Toru Takiguchi, Shinichi Noto, Sachiko Ohde and Osamu Takahashi in International Journal of Social Psychiatry

Footnotes

Acknowledgements

We thank American Journal Experts (AJE) for English language editing.

Conflict of interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported in part by the Japan Society for the Promotion of Science KAKENHI (Grant number 19K19367).

ORCID iD

Shota Saito

Supplemental material

Supplemental material for this article is available online.

References

Bannai

Ukawa

Tamakoshi

(2015). Long working hours and psychological distress among school teachers in Japan. Journal of Occupational Health, 57(1), 20–27. https://doi.org/10.1539/joh.14-0127-OA

Blom

E. C. H.

Serlachius

Larsson

J. O.

Theorell

Ingvar

(2010). Low Sense of Coherence (SOC) is a mirror of general anxiety and persistent depressive symptoms in adolescent girls - a cross-sectional study of a clinical and a non-clinical cohort. Health and Quality of Life Outcomes, 8, 58. https://doi.org/10.1186/1477-7525-8-58

Conwell

Duberstein

P. R.

Caine

E. D.

(2002). Risk factors for suicide in later life. Biological Psychiatry, 52(3), 193–204. https://doi.org/10.1016/s0006-3223(02)01347-1

Dhungel

Sugai

M. K.

Gilmour

(2019). Trends in suicide mortality by method from 1979 to 2016 in Japan. International Journal of Environmental Research and Public Health, 16(10), 1794. https://doi.org/10.3390/ijerph16101794

Ekman

Dickman

P. W.

Klint

Weiderpass

Litton

J. E.

(2006). Feasibility of using web-based questionnaires in large population-based epidemiological studies. European Journal of Epidemiology, 21(2), 103–111. https://doi.org/10.1007/s10654-005-6030-4

Ina

Hayashi

Nomura

Ishitsuka

Hirai

Iguchi

(2011). Depression, quality of life (QoL) and will to live of community-dwelling postmenopausal women in three Asian countries: Korea, China and Japan. Archives of Gerontology and Geriatrics, 53(1), 8–12. https://doi.org/10.1016/j.archger.2010.05.010

Ishikawa

Kawakami

Kessler

R. C.

(2016). Lifetime and 12-month prevalence, severity and unmet need for treatment of common mental disorders in Japan: Results from the final dataset of World Mental Health Japan Survey. Epidemiology and Psychiatric Sciences, 25(3), 217–229. https://doi.org/10.1017/s2045796015000566

Jang

S. Y.

Jang

S. I.

Bae

H. C.

Shin

Park

E. C.

(2015). Precarious employment and new-onset severe depressive symptoms: A population-based prospective study in South Korea. Scandinavian Journal of Work, Environment & Health, 41(4), 329–337. https://doi.org/10.5271/sjweh.3498

Kachi

Otsuka

Kawada

(2014). Precarious employment and the risk of serious psychological distress: A population-based cohort study in Japan. Scandinavian Journal of Work, Environment & Health, 40(5), 465–472. https://doi.org/10.5271/sjweh.3442

10.

Kikuchi

Odagiri

Ohya

Nakanishi

Shimomitsu

Theorell

Inoue

(2020). Association of overtime work hours with various stress responses in 59,021 Japanese workers: Retrospective cross-sectional study. PLoS One, 15(3), e0229506. https://doi.org/10.1371/journal.pone.0229506

11.

Kuroki

Ishitobi

Kamio

Sugihara

Murai

Motomura

Ogasawara

Kimura

Aleksic

Ozaki

Nakao

Yamada

Yoshiuchi

Kiriike

Ishikawa

Kubo

Matsunaga

Miyata

Asada

Kanba

(2016). Current viewpoints on DSM-5 in Japan. Psychiatry and Clinical Neurosciences, 70(9), 371–393. https://doi.org/10.1111/pcn.12421

12.

Larsson

Kallenberg

K. O.

(1996). Sense of coherence, socioeconomic conditions and health: Interrelationships in a nation-wide Swedish sample. European Journal of Public Health, 6(3), 175–180. https://doi.org/10.1093/eurpub/6.3.175

13.

Lewinsohn

P. M.

Seeley

J. R.

Roberts

R. E.

Allen

N. B.

(1997). Center for Epidemiologic Studies Depression Scale (CES-D) as a screening instrument for depression among community-residing older adults. Psychology and Aging, 12(2), 277–287. https://doi.org/10.1037//0882-7974.12.2.277

14.

Z. F.

Zhang

Luo

Liu

Zhang

(2020). Increased stressful impact among general population in mainland China amid the COVID-19 pandemic: A nationwide cross-sectional study conducted after Wuhan city’s travel ban was lifted. International Journal of Social Psychiatry, 66(8), 770–779. https://doi.org/10.1177/0020764020935489

15.

Nishimura

Sasaki

Yoshikawa

Kubo

Matsuo

Liu

Takahashi

(2020). Effect of work-related events on depressive symptoms in Japanese employees: A web-based longitudinal study. Industrial Health, 58(6), 520–529. https://doi.org/10.2486/indhealth.2020-0058

16.

Ogawa

Seo

Maeno

Ito

Sanuki

Maeno

(2018). The relationship between long working hours and depression among first-year residents in Japan. BMC Medical Education, 18(1), 50. https://doi.org/10.1186/s12909-018-1171-9

17.

Ohta

Higuchi

Yamato

Kumashiro

Sugimura

(2015). Sense of coherence modifies the effect of overtime work on mental health. Journal of Occupational Health, 57(3), 297–301. https://doi.org/10.1539/joh.14-0164-BR

18.

Ohtsu

Kaneita

Aritake

Mishima

Uchiyama

Akashiba

Uchimura

Nakaji

Munezawa

Kokaze

Ohida

(2013). A cross-sectional study of the association between working hours and sleep duration among the Japanese working population. Journal of Occupational Health, 55(4), 307–311. https://doi.org/10.1539/joh.12-0257-br

19.

Organization for Economic Co-operation and Development (OECD). (2020a, August 20). Average annual hours actually worked per worker. https://stats.oecd.org/Index.aspx?DataSetCode=ANHRS#

20.

Organization for Economic Co-operation and Development (OECD). (2020b, August 20). GDP per hour worked. https://data.oecd.org/lprdty/gdp-per-hour-worked.htm

21.

Sairenchi

Haruyama

Ishikawa

Wada

Kimura

Muto

(2011). Sense of coherence as a predictor of onset of depression among Japanese workers: A cohort study. BMC Public Health, 11, 205. https://doi.org/10.1186/1471-2458-11-205

22.

Saito

Tran

H. T. T.

Suzuki

Takiguchi

Ishigami

Noto

Ohde

Takahashi

(2021). Psychological impact of the state of emergency over COVID-19 for non-permanent workers: A nationwide follow-up study in Japan. BMC Public Health, 21(1), 334. https://doi.org/10.1186/s12889-021-10401-y

23.

Shiroiwa

Fukuda

Ikeda

Igarashi

Noto

Saito

Shimozuma

(2016). Japanese population norms for preference-based measures: EQ-5D-3L, EQ-5D-5L, and SF-6D. Quality of Life Research, 25(3), 707–719. https://doi.org/10.1007/s11136-015-1108-2

24.

Shiroiwa

Fukuda

Ikeda

Takura

Moriwaki

(2017). Development of an official guideline for the economic evaluation of drugs/medical devices in Japan. Value in Health, 20(3), 372–378. https://doi.org/10.1016/j.jval.2016.08.726

25.

Shiroiwa

Ikeda

Noto

Igarashi

Fukuda

Saito

Shimozuma

(2016). Comparison of value set based on DCE and/or TTO data: Scoring for EQ-5D-5L health states in Japan. Value Health, 19(5), 648–654. https://doi.org/10.1016/j.jval.2016.03.1834

26.

Smith

T. C.

Gray

G. C.

Ryan

M. A.

(2007). When epidemiology meets the internet: Web-based surveys in the Millennium Cohort Study. American Journal of Epidemiology, 166(11), 1345–1354. https://doi.org/10.1093/aje/kwm212

27.

Statistics Bureau. (2020). Ministry of internal affairs and communications, labor force survey monthly results June 2020. http://www.stat.go.jp/data/roudou/sokuhou/tsuki/pdf/gaiyou.pdf

28.

Tanaka

Tahara

Mashizume

Takahashi

(2021) Effects of lifestyle changes on the mental health of healthcare workers with different sense of coherence levels in the era of COVID-19 pandemic. International Journal of Environmental Research and Public Health, 18(6), 2801. https://doi.org/10.3390/ijerph18062801

29.

The Japan Institute for Labor Policy and Training. (2021, April 20). Statistical indicators. https://www.jil.go.jp/english/jli/documents/2021/030-05.pdf

30.

Urakawa

Yokoyama

Itoh

(2012). Sense of coherence is associated with reduced psychological responses to job stressors among Japanese factory workers. BMC Research Notes, 5, 247. https://doi.org/10.1186/1756-0500-5-247

31.

Verma

Mishra

(2020). Depression, anxiety, and stress and socio-demographic correlates among general Indian public during COVID-19. International Journal of Social Psychiatry, 66(8), 756–762. https://doi.org/10.1177/0020764020934508

32.

Yamauchi

Sasaki

Takahashi

Umezaki

Takahashi

Yoshikawa

Suka

Yanagisawa

(2019). Long working hours, sleep-related problems, and near-misses/injuries in industrial settings using a nationally representative sample of workers in Japan. PLoS One, 14(7), e0219657. https://doi.org/10.1371/journal.pone.0219657

33.

Yoshimoto

Oka

Fujii

Nagata

Matsudaira

(2020). The economic burden of lost productivity due to presenteeism caused by health conditions among workers in Japan. J Occup Environ Med, 62(10), 883–888. https://doi.org/10.1097/JOM.0000000000002001

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.09 MB