HUMEDITロゴ

The Psychological Impact of Emergencies on COVID-19 on Part-Time Workers: A Nationwide Follow-up Study in Japan.

The Psychological Impact of Emergencies on COVID-19 on Part-Time Workers: A Nationwide Follow-up Study in Japan.

Summary

Background

The COVID-19 epidemic caused an increase in mental health problems and unemployment due to the economic recession. This study aimed to assess the psychological impact of the emergency. We estimated changes in mental health, quality of life, and unemployment experiences of the general workforce during the first COVID-19 outbreak in Japan.

Method

The authors conducted a nationwide follow-up survey. Japanese general workers aged 15-59 were surveyed via the Internet between March 26~April 6, 2020 and June 26~July 2, 2020. Questionnaire items covered employment status and socioeconomic factors, and epidemiological research centers were used. Depression Scale (CES-D) and EQ-5D-5L to assess depression and health-related quality of life (HR-QOL). Propensity score analysis was used to analyze differences in outcomes between full-time and non-full-time employees . Multiple linear regression analysis was used to examine the relationship between unemployment and CES-D scores.

Result

Two hundred thirty-one subjects were included in the analysis. Changes in both CES-D scores and usefulness were not significantly different between the two groups. However, there was a significant difference with respect to the unemployment rate associated with higher CES-D scores.

Conclusion.

This study demonstrates that the mental health of informal workers is not negatively affected after a COVID-19 emergency in Japan. Unemployment is an important factor affecting the mental health of the general workforce.

Keywords: covid-19, general workforce, mental health, propensity score analysis, quality of life, job security, unemployment, web survey

Background

Coronavirus disease has a major impact on global public health and is spreading worldwide. As of August 3, 2020, 17,918,582 cases of COVID-19 have been confirmed, resulting in 686,703 deaths worldwide [1]. Many countries have closed their external borders and implemented nationwide lockdowns in an effort to temporarily, but successfully, contain the spread of COVID-19. However, this measure has resulted in a reduction in the workforce across economic sectors and many job losses [2].

In Japan, a state of emergency was declared in seven prefectures, including Tokyo and Osaka, on April 7, 2020. In response to a sharp increase in "untraceable" new infections, the state of emergency expanded nationwide, and on April 16, 13 prefectures were designated as "Special Alert Zones. With the state of emergency in effect, the government urged citizens to stay away from home, maintain social distance, stay at home, and follow travel restrictions, and closed non-essential businesses for about a month [3,4].

The COVID-19 epidemic has resulted in increased mental health problems and unemployment due to the economic recession. A large cross-sectional study conducted in China showed a decreased risk of depression, anxiety, and insomnia once people were in work [5]. Several studies reported rates of depression and anxiety of 18.7% and 21.6%, respectively, among Spaniards early in the COVID-19 outbreak; Spanish adults restrained by COVID-19 exercise restrictions showed an inverse association between current physical activity and current perceived anxiety and mood [6, 7]. The Organization for Economic Cooperation and Development (OECD) noted a 2.9% increase in unemployment, reflecting the impact of the COVID-19 containment measures [2]. Recently, the Japanese government reported that the mental health status of unemployed and informal workers may be particularly vulnerable during the ongoing COVID-19 crisis [8].

The number of informal workers in our country is increasing. Early employment includes part-time, temporary, and fixed-term employment, with men and women accounting for 22% and 53% of paid employment in 2020, respectively [9]. There has been a negative impact on the mental health of informal workers due to job insecurity, and unemployment status has been associated with psychological conditions such as depression, anxiety, and poor health outcomes due to unemployment. A previous study suggested that precarious employment was associated with twice the risk of serious psychological distress among middle-aged Japanese men [10]. In addition, the transition from full-time

In East Asia, employment in a different employment situation was also associated with the development of severe depressive symptoms [11, 12]. It is well known that poor mental health, including depression, is an independent risk factor for suicide and is associated with a lower quality of life compared to healthy subjects [13, 14].

The current focus of the COVID-19 outbreak is not only on the medical outcomes of infected patients, but also on the mental health of affected individuals and the general population. Therefore, it is worthwhile to investigate changes in mental health and quality of life during the COVID-19 crisis. The relationship between employment security and mental health problems after the lifting of the state of emergency in Japan has not yet been examined. The purpose of this study is to assess the psychological impact of the state of emergency. We estimated changes in mental health, quality of life, and unemployment experiences of the general workforce during the first COVID-19 outbreak in Japan.

Method

Study design and data collection

This nationwide survey was conducted online through a platform of over 2 million candidates, targeting general workers aged 15-59 in Japan. It was managed by Cross Marketing Corporation (Tokyo, Japan), a company specializing in survey research.
The first survey was conducted from March 26, before the COVID 19 state of emergency was declared, through April 6, 2020. A follow-up survey was administered to the same cohort of respondents from June 26 to July 2, 2020, after the emergency was lifted. The initial survey was administered until data were collected from 3,000 respondents, and the response rate for the follow-up survey was based on the general response rate for the web-based survey, covering 70% of the initial cohort. The initial survey ensured a representative sample of the Japanese population with respect to age, gender, and residential area. The residential areas of Japan were divided into 10 regions (Figure 1). No incentives or rewards were offered to participants in this study.

Questionnaire Operational status

The authors defined the following four types of work status; subjects selected their status during self-report. Full-time employees are defined as company employees who are guaranteed lifetime employment until retirement, hired directly by the employer, and employed as full-time employees.

Non-permanent employees:

Company employees with a labor contract with a fixed term, such as part-timers, dispatched workers, and contract and commissioned workers.

Civil Servant:

A public official of a state or local government incorporated into a non-profit organization.

Self-employed:

Self-employed workers, including sole proprietors and freelancers. The target population for this study was workers employed by general companies. Government employees were excluded because in Japan employment is guaranteed until retirement. Self-employed workers were also excluded because they are not employees.

Figure 1: Japan's 10 regions

日本の10地域

Socioeconomic and medical status

The questionnaire covered age, gender, region, marital status, children, family living together, education, industry, firm size, personal income, family income, average monthly overtime hours, union membership, head of household, exercise, smoking history, drinking history, commuting time, and average sleep duration.

Information was also collected on the history of cardiac disease, cerebrovascular disease, cancer, Alzheimer's disease, physical disease with chronic pain, epilepsy, and depression, according to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) [15]. For this analysis, the number of disorders was defined as the primary factor. The Japanese version of the Sense of Coherence (SOC) scale was used to measure stress coping skills, with a final score ranging from 13 to 91. The scale included 13 items, with higher scores indicating better stress coping ability [16].

Return to origin

The degree of depressive symptoms and HR-QOL were assessed in both the first and second surveys. The Japanese version of the CES-D was used to measure depressive symptoms. This scale consists of 20 items, and participants were asked to rate the frequency with which they experienced depression-related symptoms over the past week. CES-D scores ranged from 0~60, with scores above 16 usually indicating depressive symptoms. The CES-D has high sensitivity, specificity, and internal consistency for identifying risk for depression [17].

A five-dimensional EQ-5D-5L instrument was used to assess respondents' HR-QOL. The EQ-5D-5L consists of five items across five levels, i.e., mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. The resulting general preference-based scale reflects subjective values assigned to specific health-related outcomes ranging from -0.025~1, with 0 indicating death and 1 indicating perfect health in the Japanese value set. We refer to these as utility weights [18].

The difference in CES-D scores and utility values between the two time frames was defined as the outcome in this analysis. We also investigated the experience of emergency unemployment in the second survey.

Statistical analysis

Propensity scores were matched, adjusted for confounding factors in the permanent and non-permanent groups, and outcomes were evaluated [19]. Propensity scores were estimated using multiple logistic regression models with clinical measures such as socioeconomic factors and comorbidities, baseline CES-D scores, and utility. According to the propensity scores, nearest neighbor matching was performed using the estimated propensity scores without replacement and with a caliper of 0.2 standard deviations of the propensity scores in a 1:1 ratio. In matched subjects, absolute standardized differences in means and proportions of these variables were used to confirm propensity scoring balance between the two groups. In the matched cohort, we compared CES-D score change, utility change, and the rate of unemployment experienced by full-time and non-full-time employees. Categorical variables were compared using Pearson's chi-square test and continuous variables were compared using Student's t-test. Finally, we assessed the psychological impact of unemployment with respect to changes in CES-D scores. Multiple linear regression analysis was performed to identify the determinants of change. The independent variables used in the analysis were working status, unemployment experience, gender, age, number of comorbidities, region, marital status, personal income, family income, average hours worked per day, trade union membership, head of household, exercise, smoking, alcohol consumption, average sleep duration, SOC score, and baseline CES-D score The Categorical and ordinal variables were transformed into dummy variables. The authors considered the interaction between work status and unemployment experience to assess the extent of psychological effects in the two groups.

All statistical tests were two-tailed, and a p-value less than 0.05 was considered significant. All analyses were performed using STATA 16.1 (College Station, Texas, USA: StataCorp LP).

Result

Data Collection

In the first study, data were collected from 3001 subjects (excluding housewives, students, and the unemployed); 2351 subjects responded to the follow-up survey. Subsequently, 161 self-employed and 132 civil servants were excluded from the analysis. Finally, 1373 full-time and 685 non-full-time employees were included in the propensity score matching for full-time and non-full-time employees, respectively (Figure 2).

Trend Score Analysis

Propensity score matching identified 497 subjects from both the non-permanent and permanent groups. Thus, a total of 994 subjects were included in subsequent analyses. Table 1 shows the differences in baseline characteristics between the non-permanent and permanent groups before and after matching. All baseline variables included in the model were well balanced within standardized differences or close to 0.1 after matching. The c statistic for the propensity score was estimated to be 0.862, ranging from 0.846 to 0.878, indicating good discrimination between the two groups. The web survey allowed us to collect data without missing values. Figure 3 compares changes in CES-D scores, changes in utility, and unemployment rates. The change in CES-D score was estimated to be -0.706 for full-time employees and -0.575 for non-full-time employees (p = 0.807). The change in utility was also not significantly different between the two groups (permanent 0.014 vs. non-permanent 0.009, p = 0.533). However, there was a significant difference with respect to unemployment rates, and the data were matched (permanent 7.20% vs. non-permanent 11.47%, p = 0.022); the risk ratio for unemployment was estimated at 1.583 (95% confidence interval = 1.063 to 2.358).

Figure 2: Flowchart of study participants

試験参加者のフローチャート

Multiple Regression Analysis

The results of the multiple regression analysis presented in Table 2 showed that experience of unemployment was a factor associated with increased CES-D scores (p = 0.003). There was no significant difference in the interaction between work status and unemployment (p = 0.340). Two or more comorbidities (p = 0.044) and an average working day of 10~12 hours (p = 0.027) were associated with higher CES-D scores. In particular, higher SOC scores (p < 0.001), marital status (p = 0.032), and baseline CES-D scores (p < 0.001) were associated with lower CES-D scores.

Consideration

The physiological consequences of COVID-19 outbreaks are a global concern. This study examined changes in CES-D scores, health-related utility, and unemployment in the general workforce during a COVID-19 status emergency, based on data from a nationwide web-based questionnaire in Japan. After matching subjects' backgrounds, there were no statistically significant differences in these scores. However, unemployment in the non-permanent group was statistically higher than in the permanent group, even after adjusting for baseline factors. The authors' findings suggest that there was a deterioration in employment conditions after the emergency, especially for non-permanent workers. Unemployment was found to worsen the psychological condition of the general Japanese workforce.

Several studies have demonstrated predictors associated with increased depression and anxiety. In Japan, a cohort study conducted by Sairenchi et al. [20] found that SOC can predict the onset of depression in Japanese workers. [21] reported that increased SOC may decrease negative work stress reactions and subjective symptoms in general workers. A recent large cross-sectional study conducted by Kikuchi et al. [22] found that Japanese workers with longer overtime hours exhibited significantly higher anxiety and depression than those with less overtime, both among men and women. Furthermore, a Korean study suggested that head of household status, gender, and precarious employment were associated with the development of severe depressive symptoms [11].

A propensity score analysis adjusted for these factors between the permanent and non-permanent groups. There was no statistically significant difference in CES-D scores between the two groups, but there was a slight improvement in scores for both groups. Although informal workers reported higher rates of unemployment compared to workers with regular contracts, no significant effect on mental health was observed in our data. The authors' results did not confirm our hypothesis regarding HR-QOL. However, because the EQ5D-5L includes dimensions of anxiety/depression, worsening psychological conditions affect HR-QOL. We hypothesized that an increase in adverse mental health effects should decrease HR-QOL among non-regular workers. Job security is an important factor in maintaining the mental health of non-regular workers. Increased unemployment may increase suicide rates during and after an outbreak of COVID-19. We suggest that unemployment is a factor that negatively impacts mental health. Previous studies have shown that working long hours is associated with an increased risk of depression in Japan. Multiple regression analysis showed similar results [23].

Table 1: Baseline characteristics before and after propensity score matching

傾向スコアマッチング前後のベースライン特性
傾向スコアマッチング前後のベースライン特性
傾向スコアマッチング前後のベースライン特性

CES-D=Center for Epidemiological Studies Depression Scale, SOC=sense of consistency, SD=standard deviation

[A] High infection area includes Tokyo, Saitama, Chiba, Kanagawa, and Hokkaido.
[B] Low (less than 4 million yen), medium (4 million yen to 8 million yen), and high (8 million yen or more)
[C] Exercises are defined as Moderate exercise for approximately one hour with light breathing
[D] Industry is defined as: primary industry, including agriculture, forestry, and fisheries; secondary industry, including mining, gravel quarrying, construction, and manufacturing; electricity, gas, heat, water, and information and communications; transportation and postal services; wholesale and retail trade; finance and insurance; real estate and goods rental; scientific research; professional and technical services; accommodation; food services; living-related and personal services; amusement services; educational and learning support; medical care and welfare; multiple services; other services; government; and unclassifiable industries.

Figure 3: Box-and-whisker plot of changes in CES-D score and health-related utility, and background plot of unemployment experience for permanent and nonpermanent groups. (N.S: no significant difference, *: p < 0.05, **: p < 0.01, ***: p < 0.001)

図3:CES-Dスコアと健康関連の効用の変化の箱ひげ図、および永続的グループと非永続的グループの失業

Before Matching=Before Matching, After Matching=After Matching, Change in CES-D=Change in CES-D score, Change in utility=Change in utility, Unemployment experience=Unemployment experience, Permanent=Permanent, Non-employment=Part-time, working status=Working status

Twenge et al [24] reported a slight increase in the prevalence of depression in the United States from April 2020 to May 2020. The prevalence of diagnosed depression could not be estimated. However, half of the subjects in this data confirmed depressive symptoms, as indicated by a CES-D score of 16 points or higher. Therefore, we considered that this was not an optimal situation and that increased unemployment could lead to an increased incidence of depression in Japan in the near future.

Web-based surveys are a reliable method for epidemiological studies [25, 26]. However, the study had several limitations. First, approximately 25% of participants were excluded from follow-up because a second response to the web-based survey was not obtained. Younger participants were more likely to not respond to the follow-up survey. Thus, some selection bias remained with respect to follow-up data compared to initial survey data. However, such selection bias may have had only a minimal effect on our results because the initial survey ensured adequate representation of the Japanese population. Second, because this study included workers employed by commercial firms, self-employed workers and civil servants were excluded from the analysis. Civil servants were considered to be engaged in public service, while self-employed workers generally worked independently. Because of the limited number of subjects in the follow-up data, further data collection is needed to investigate the mental health status of workers in future studies.

Finally, in the follow-up questionnaire, unemployment was defined as job loss or termination during an emergency. It was not possible to collect more information on the reasons for unemployment. Moreover, the anonymous self-reporting nature of the survey precluded the use of additional approaches to validate respondents' socioeconomic status or clinical history. Despite these limitations and the short time frame of the prospective survey, the statistical analysis presented in this study may serve as important information for future health and economic policies related to the COVID-19 crisis in Japan.

Following the declaration of the state of emergency, the Japanese government encouraged its citizens to refrain from unnecessary travel and to avoid going out unless necessary. The closure of many public and commercial facilities, except for essential businesses, was strongly urged. In other countries, most citizens restrained themselves until the state of emergency was lifted, although no lockdowns or other mandatory measures were taken. Japan succeeded in taking countermeasures against the virus at the end of June. However, according to Japan's Ministry of Health, Labor and Welfare [27,28], the number of unemployed people due to COVID-19 was estimated at 48,206 on August 25. Japan is facing the second wave of COVID-19 and the situation may remain severe in the medium to long term. Unemployment is expected to increase in certain industries such as manufacturing, food services, and tourism. Therefore, changes in mental health and suicide rates should be closely monitored.

Table 2: Multiple regression models of socioeconomic indicators and changes in CES-D scores (N = 2351)

Variable

Coefficient

95%Decline in confidence interval

95%Upper confidence interval

Pvalue

Part-time Work

0.135 0.720 0.989 0.757

Experience of Unemployment

2.358 0.793 3.923 0.003

Irregular Employment*Experience of Unemployment

1.148 3.507 1.210 0.340

Gender

0.301 1.055 0.452 0.433

Age

0.015 0.050 0.020 0.397

Number of comorbidities
0 (base)
1
2 or more
Area of High Infection
Married


1.000
0.939
3.284
0.566
0.821


0.238
0.092
0.125
1.570


2.116
6.476
1.256
0.072


0.118
0.044
0.108
0.032

Personal Income
Low (base) Medium

Amount of Money


1.000
0.078
0.551


1.122
2.099


0.967
0.998


0.884
0.486

Family Income
Low (base)
During (a certain time when one did or is doing something)
Amount of Money


1.000
0.465
1.003



1.417
2.189


0.486
0.184


0.338
0.098

Average daily working hours
Less than 8~10 hours
10~12 hours More than 12 hours


1.000
0.091
1.574
0.467


0.888
0.179
2.210



0.706
2.969
3.145


0.822
0.027
0.732
Union Member 0.194 0.928 0.540 0.604
Head of the Household 0.235 0.617 1.087 0.589

Motion
None (base)

Once every two weeks
Once a week
At least twice a week


1.000
0.531
0.537
0.203


0.639
0.472
1.126


1.701
1.545
0.719


0.373
0.297
0.665

Smoking
None (base)
Yes, sir.

A past (i.e. a personal history)


1.000
0.604
0.011



1.151
1.441


0.611
0.766


0.548
0.548

Beverage
None (base)
Yes, sir.
A past (i.e. a personal history)


1.000
0.604
0.011


1.306
1.527


0.097
1.549


0.091
0.989

Hours of Sleep
Less than 4 hours
4~6 hours
6~8 hours (standard)
More than 8 hours


0.750
0.052
1.000
0.999

1.381
0.838

2.880
0.735

0.490
0.898
SOC Score 0.134 0.172 0.097 <0.001

BaselineCES-DScore

0.383 0.420 0.347 <0.001

Constant

14.721 11.913 17.528 <0.001

CES-D Epidemiology Research Center Depression Scale, SOC Consistency*: P < 0.05, **: P < 0.01, ***: P < 0.001

Conclusion.

In conclusion, this study found that the mental health of informal workers was not adversely affected by the COVID-19 status emergency in Japan. The authors suggested that a history of unemployment was a factor associated with reduced mental health, with about 10% of informal workers experiencing unemployment between two periods. The COVID-19 crisis is still in its early stages, and systematic policies, including infection control and economic measures, are needed to prevent the mental health of the general workforce from deteriorating. Further studies are needed to assess long-term mental health outcomes and rates of depression during the COVID-19 crisis in Japan.

Abbreviation

OECD: Organization for Economic Cooperation and Development; DSM5: Diagnostic and Statistical Manual of Mental Disorders - 5th Edition; HRQOL: Health-Related Quality of Life; CES-D: Center for Epidemiological Studies-Depression Scale; SOC: Sense of Coherence

References

1. World Health Organization. Coronavirus disease 2019 (COVID-19) Situation; 2020. p. 196. https://www.who.int/emergencies/diseases/novel-corona virus-2019/situation-reports. Accessed on 7 August, 2020
2. Organization for Economic Co-Operation and Development (OECD). Record rise in OECD unemployment rate in April 2020. (2020) http://www.oecd.org/ sdd/labour-stats/unemployment-rates-oecd-update-june-2020.htm . Accessed on 27 August, 2020. Table 2 Multiple linear regression models of socioeconomic indicators and changes in the CES-D score (N = 2351) (Continued) Variables Coefficient Lower 95% CI Upper 95% CI P value SOC score −0.134 −0.172 − 0.097 < 0.001 *** Base-line CES-D score −0.383 −0.420 − 0.347 < 0.001 *** Constant 14.721 11.913 17.528 < 0.001 *** CES-D Center for Epidemiologic Studies Depression Scale, SOC Sense of coherence *: P < 0.05, **: P < 0.01, ***: P < 0.001
3. Looi MK. Covid-19: Japan prepares to extend state of emergency nationwide as “untraceable” cases soar. BMJ. 2020a;369:m1543. https://doi. org/10.1136/bmj.m1543.
4. Looi MK. Covid-19: Japan declares state of emergency as Tokyo cases soar. BMJ. 2020b;369:m1447. https://doi.org/10.1136/bmj.m1447.
5. Shi L, Lu ZA, Que JY, Huang XL, Liu L, Ran MS, et al. Prevalence of and risk factors associated with mental health symptoms among the general population in China during the coronavirus disease 2019 pandemic. JAMA Netw Open. 2020;3(7):e2014053. https://doi.org/10.1001/jamanetworkopen.2 020.14053.
6. González-Sanguino C, Ausín B, Castellanos MÁ, Saiz J, López-Gómez A, Ugidos C, et al. Mental health consequences during the initial stage of the 2020 coronavirus pandemic (COVID-19) in Spain. Brain Behav Immun. 2020; 87:172–6. https://doi.org/10.1016/j.bbi.2020.05.040. 7. López-Bueno R, Calatayud J, Ezzatvar Y, Casajús JA, Smith L, Andersen LL, et al. Association between current physical activity and current perceived anxiety and mood in the initial phase of COVID-19 confinement. Front Psychiatry. 2020;11:729. https://doi.org/10.3389/fpsyt.2020.00729.
8. Ueda M, Stickley A, Sueki H, Matsubayashi T. Mental health status of the general population in Japan during the COVID-19 pandemic. Psychiatry Clin Neurosci. 2020. https://doi.org/10.1111/pcn.13105.
9. Statistics Bureau. Ministry of Internal affairs and communications, Labour Force Survey monthly results. (2020) http://www.stat.go.jp/data/roudou/ sokuhou/tsuki/pdf/gaiyou.pdf. Accessed on June 2020 (Japanese).
10. Kachi Y, Otsuka T, Kawada T. Precarious employment and the risk of serious psychological distress: a population- based cohort study in Japan. Scand J Work Environ Health. 2014;40(5):465–72. https://doi.org/10.5271/sjweh.3442.
11. Jang SY, Jang SI, Bae HC, Shin J, Park EC. Precarious employment and newonset severe depressive symptoms: a population-based prospective study in South Korea. Scand J Work Environ Health. 2015;41(4):329–37. https://doi. org/10.5271/sjweh.3498.
12. Yoo KB, Park EC, Jang SY, Kwon JA, Kim SJ, Cho KH, et al. Association between employment status change and depression in Korean adults. BMJ Open. 2016;6(3):e008570. https://doi.org/10.1136/bmjopen-2015- 008570.
13. Turecki G, Brent DA. Suicide and suicidal behaviour. Lancet. 2016;387(10024): 1227–39. https://doi.org/10.1016/S0140-6736(15)00234-2.
14. Ina K, Hayashi T, Nomura H, Ishitsuka A, Hirai H, Iguchi A. Depression, quality of life (QoL) and will to live of community-dwelling postmenopausal women in three Asian countries: Korea, China and Japan. Arch Gerontol Geriat. 2011;53(1):8–12. https://doi.org/10.1016/j.archger.2010.05.010.
15. Kuroki T, Ishitobi M, Kamio Y, Sugihara G, Murai T, Motomura K, et al. Current viewpoints on DSM-5 in Japan. Psychiatry Clin Neurosci. 2016;70(9): 371–93. https://doi.org/10.1111/pcn.12421.
16. Ohta M, Higuchi Y, Yamato H, Kumashiro M, Sugimura H. Sense of coherence modifies the effect of overtime work on mental health. J Occup Health. 2015;57(3):297–301. https://doi.org/10.1539/joh.14-0164-BR.
17. Lewinsohn PM, Seeley JR, Roberts RE, Allen NB. Center for Epidemiologic Studies Depression Scale (CES-D) as a screening instrument for depression among community-dwelling older adults. Psychol Aging. 1997;12(2):277–87. https://doi.org/10.1037/0882-7974.12.2.277. 18. Shiroiwa T, Ikeda S, Noto S, Igarashi A, Fukuda T, Saito S, et al. Comparison of value set based on DCE and/or TTO data: scoring for EQ-5D-5L health states in Japan. Value Health. 2016;19(5):648–54. https://doi.org/10.1016/j.jva l.2016.03.1834.
19. Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399–424. https://doi.org/10.1080/00273171.2011.568786.
20. Sairenchi T, Haruyama Y, Ishikawa Y, Wada K, Kimura K, Muto T. Sense of coherence as a predictor of onset of depression among Japanese workers: a cohort study. BMC Public Health. 2011;11(205):205. https://doi.org/10.11 86/1471-2458-11-205.
21. Urakawa K, Yokoyama K, Itoh H. Sense of coherence is associated with reduced psychological responses to job stressors among Japanese factory workers. BMC Res Notes. 2012;5(247):247. https://doi.org/10.1186/1756-0500- 5-247.
22. Kikuchi H, Odagiri Y, Ohya Y, Nakanishi Y, Shimomitsu T, Theorell T, et al. Association of overtime work hours with various stress responses in 59,021 Japanese workers: retrospective cross-sectional study. PLoS One. 2020;15(3): e0229506. https://doi.org/10.1371/journal.pone.0229506.
23. Nakata A. Long working hours, job satisfaction, and depressive symptoms: a community-based cross-sectional study among Japanese employees in small- and medium-scale businesses. Oncotarget. 2017;8(32):53041–52. https://doi.org/10.18632/oncotarget.18084.
24. Twenge JM, Joiner TE. U.S. Census Bureau-assessed prevalence of anxiety and depressive symptoms in 2019 and during the 2020 COVID-19 pandemic. Depress Anxiety. 2020. https://doi.org/10.1002/da.23077.
25. Smith B, Smith TC, Gray GC, Ryan MA. Millennium cohort study team. When epidemiology meets the internet: web-based surveys in the millennium cohort study. Am J Epidemiol. 2007;166(11):1345–54. https://doi.org/10.1 093/aje/kwm212.
26. Ekman A, Dickman PW, Klint A, Weiderpass E, Litton JE. Feasibility of using web-based questionnaires in large population-based epidemiological studies. Eur J Epidemiol. 2006;21(2):103–11. https://doi.org/10.1007/s10654- 005-6030-4.
27. Ministry of Health, Labour and Welfare. Employment Security. (2020) https:// www.mhlw.go.jp/english/policy/employ-labour/employment-security/index. html . Accessed on 27 August, 2020.
28. Japan Institute for Labour Policy and Training. Results of the “Survey on the Impact that Spreading Novel Coronavirus Infection has on Work and Daily Life.” (2020) https://www.jil.go.jp/english/index.html . Accessed on 27 August, 2020.