key: cord-0952048-9suab0nh
authors: Tomono, Misa; Yamauchi, Takashi; Suka, Machi; Yanagisawa, Hiroyuki
title: Impact of overtime working and social interaction on the deterioration of mental well‐being among full‐time workers during the COVID‐19 pandemic in Japan: Focusing on social isolation by household composition
date: 2021-07-23
journal: J Occup Health
DOI: 10.1002/1348-9585.12254
sha: 2841e7b79401c4ca1f039f450e0d6b63f9c75591
doc_id: 952048
cord_uid: 9suab0nh

OBJECTIVES: The coronavirus disease (COVID‐19) pandemic has forced many employees to alter both their work style and lifestyle. This study aimed to examine how the combination of changes in overtime working hours and social interaction affects the full‐time employees’ mental well‐being, focusing on the difference in household composition. METHODS: In November 2020, we conducted a cross‐sectional Internet survey that included 4388 Japanese men and women aged 25–64 years, who continued the same full‐time job during the pandemic. We performed a logistic regression analysis using a combination of the changes in overtime working hours and social interaction as an independent variable, and the presence/absence of deterioration of mental well‐being as the dependent variable. RESULTS: Overall, 44% of participants reported the deterioration of mental well‐being compared to before the outbreak. The multivariate analysis revealed that the participants coded as “increased overtime/decreased interaction” were significantly associated with the deterioration of mental well‐being compared to those with “unchanged overtime/unchanged interaction” (odds ratio [OR] 2.14, 95% confidence interval [CI] 1.59–2.89). Moreover, this association was relatively stronger among single‐person households (OR 2.66, 95% CI 1.50–4.69). CONCLUSIONS: The negative combination of increasing overtime working hours and decreasing social interaction may have an impact on the deterioration of mental well‐being during the COVID‐19 pandemic, and this association was comparably strong among single‐person households. In the pandemic, it is necessary to pay close attention to both overtime working hours and the presence of social interaction to address the mental well‐being among employees.

Mental well-being of employees is one of the major public and occupational health issues worldwide, especially in developed countries. Several studies have shown that there was a small to moderate association between long working hours and the onset of depressive symptoms. 1, 2 In addition, regarding employees' mental health, it has been suggested that the transition of the amount of overtime work plays a crucial role in terms of the deterioration of mental well-being. 3 Focusing on the coronavirus disease (COVID-19) pandemic, owing to the scientific and medical uncertainty of SARS-COV-2, public health policy has relied on nonpharmaceutical interventions to prevent spreading of the virus. Like many countries, the Japanese government also declared the state of emergency due to COVID-19 in April 2020 requiring residents to avoid going out ( Figure 1 ). Although this political strategy may be effective in terms of infectiology, there was a drawback regarding mental well-being because of the lost connection with others. Recent studies reported that the loss of in-person social interaction in the pandemic crisis has led to psychological loss among adults. 4, 5 Recent studies have shown that not only work tasks/duties, but also several life-related factors, including a personal relationship with others, might have an impact on mental health among workers. 6, 7 According to these findings, it is crucial to analyze how the combination of job-related and life-related factors, namely overtime working hours and social interaction, respectively, affects the mental well-being of employees during the COVID-19 pandemic when both work style and lifestyle have been changing. 4, 5, 8 Moreover, among single-households, there is a greater chance to lose the connection with others under the "stay-home" requirement because of the absence of cohabitants, compared to those who live in a multi-person household. Thus, we hypothesized household composition (ie, single-person households vs multi-person households) would play an important role as a moderator which influences the level of strength of the relation between the combination of overtime working hours and social interaction and the deterioration of mental well-being.

To sum up, the psychological well-being of employees has been associated with the complexity of both work-related and life-related issues, especially during the COVID-19 pandemic. While, to our best knowledge, no previous studies have examined the effects of the combination between overtime working hours and social interaction on the deterioration of mental well-being, this study aimed to investigate the association between overtime working hours, social interaction, and deterioration of mental well-being by household composition F I G U R E 1 Trend of the number of confirmed cases of COVID-19 in Japan among full-time employees during the COVID-19 pandemic in Japan, using data from a web-based questionnaire survey conducted in Japan in November 2020 (i.e., approximately 9 months following the beginning of the COVID-19 outbreak in Japan in February). In our study, we developed a conceptual model among overtime working hours, social interaction, deterioration of mental well-being, and household composition (Figure 2 ). By shedding light on the link between overtime work, social interaction, and deterioration of psychological well-being, this study contributes to raising awareness regarding this unique vulnerability of employees under a catastrophic situation, including a pandemic.

In November 2020, a web-based survey was conducted among Japanese men and women aged 25-64 years, who lived in the metropolis of Tokyo and the three surrounding prefectures (Kanagawa, Saitama, and Chiba). Participants were recruited from a web-based research panel maintained by a leading research company in Tokyo, Japan (Rakuten Insight Inc.). Recruitment emails were sent to 61 129 randomly selected registrants of Japanese nationals. Applicants for participation in the survey were accepted in the order of receipt until the number of participants reached the quotas for gender and age groups (25-34, 35-44, 45-54, and 55-64 years old), and a total of 9247 responses (15.1%) were obtained over the 2-day recruitment. After excluding 121 respondents those who gave incomplete/inconsistent answers (i.e., those who answered the same option throughout the questionnaire and those who completed answering it too soon) from 9247 potential participants, 8000 subjects on the quotas for sex and age group were randomly selected. All participants volunteered to participate in the survey after reading a description of the purpose and procedure of the survey. Consent to participate was implied by the completion and submission of the survey.

To examine the association between overtime working hours, social interaction, and deterioration of psychological well-being during the pandemic among full-time employees, we set the exclusion criteria for eligibility as follows: (i) those who were unemployed (n = 1530), (ii) those who were self-employed (n = 478), (iii) those who were doing a part-time job (n = 1168), (iv) those who had changed the job following the first wave (i.e., March 2020) (n = 390), and (v) those who had been infected prior to participating in the survey (n = 46). To evaluate the changes of the amount of overtime working hours before and after the outbreak of the COVID-19 pandemic, we removed the participants who had switched jobs after March 2020, the time point when the first wave of cases started. Similarly, to investigate the changes of mental well-being among COVID-19-free participants, we excluded those who had been infected before the survey. Of 8000 participants, 4388 (1515 females and 2873 males; mean age [standard deviation], 43.1 [10.75] years) were eligible according to these criteria.

To assess the deterioration of mental well-being between January 2020 (ie, time point before the COVID-19 outbreak in Japan) and November 2020 (ie, time point right after the beginning of the third wave of the novel coronavirus), for each item of the WHO-Five Well-being Index (WHO-5) (ie, "I have felt cheerful and in good spirits," "I have felt calm 

Moderator and relaxed," "I have felt active and vigorous," "I woke up feeling fresh and rested," and "My daily life has been filled with things that interest me."), 9, 10 we asked the participants to answer whether they experienced any changes in their psychological state ("Have your current psychological wellbeing changed from that at the time point before starting the COVID-19 outbreak [ie, around January 2020]?"). The WHO-5 is among the most widely used questionnaires assessing subjective mental well-being. For each statement, participants responded to the question among the following three options: "increasing" = 1, "unchanged" = 0, "decreasing" = −1. Since Cronbach's alpha coefficient was 0.88, we calculated the total score for the five items for each participant. As relatively few reported an improvement in their psychological well-being (ie, total score: 1≤), we coded the participants into the following two categories: "the presence of deterioration of mental well-being" (ie, total score: −5 -−1) and "the absence of deterioration of mental well-being" (ie, total score: 0 -5).

To investigate the change of overtime working hours through the pandemic, we asked the participants whether the current overtime work was increased or decreased compared to the period prior to the COVID-19 pandemic (ie, around January 2020). The participants were offered three options: "increase," "not change," and "decrease."

To assess the changes in social interaction with a close person(s) whom participants could consult about their problems or stress during the COVID-19 pandemic, first, participants were asked whether they had such a person (or persons) other than family members. Subsequently, if they answered yes, they were asked whether they contacted such a person (or persons) more than once per month in January 2020 and November 2020. Based on the answers to these questions, each participant was coded into the following three categories: "no close contact from January 2020 to November 2020," "decreased frequency of interaction from January 2020 to November 2020," and "no change in the frequency of interaction from January 2020 to November 2020." The frequency distribution of each category was unbalanced (ie, "decreased the frequency of interaction from January 2020 to November 2020": 7.2%, "no close contact from January 2020 to November 2020": 22.5%, "no change in the frequency of interaction from January 2020 to November 2020": 70.3%), and cross-tabulation analysis indicated that the trend in the results regarding "no close contact from January 2020 to November 2020" and "decreased frequency of interaction from January 2020 to November 2020" was similar. Thus, in this study, we coded each participant into either of the following two categories: "no close contact, or decreased the frequency of interaction from January 2020 to November 2020" and "no change in the frequency of interaction from January 2020 to November 2020."

To further examine how the work-life mutual relation in terms of overtime working hours and social interaction, respectively, associated with the deterioration of mental wellbeing, we constructed a variable combination patterns of overtime working hours and social interaction. Regarding an interaction term by categorical variables, creating the values (ie, subtracting the mean value from the original value for each variable) makes the interpretation of the original variables difficult and, therefore, is not recommended in general. Therefore, in this study, we created the combination variable between these two categorical variables with six categories (i.e., (i) decreased overtime/decreased interaction, (ii) decreased overtime/unchanged interaction, (iii) unchanged overtime/decreased interaction, (iv) unchanged overtime/ unchanged interaction, (v) increased overtime/decreased interaction, and (vi) increased overtime/unchanged interaction) for statistical analysis. To identify groups that may be at particular risk for deterioration of psychological well-being, the "unchanged overtime/unchanged interaction" combination category was set as the reference category and constructed with other combinations because that is the combination where least influence on mental well-being is expected to be observed.

To measure the household composition of participants, we asked them the number of cohabitants. According to the answer, we coded the participants into two categories: "singleperson household" and "multi-person household."

Following the previous literatures, 3, 11 we measured current mental well-being ("poor" or "not poor"), employment status (regular employee or non-regular employee), occupation (clerical work, sales, service, professional work, and others [eg, transport-related occupation]), presence/absence of the increase in frequency of working from home, presence/

TOMONO eT al.

absence of the increase in amount of housework, household income ("≦5 million yen", "5 million< and ≦8 million yen" yen, and "8 million< yen"), sleep hours ( ) as potential confounders. In terms of current mental well-being, we asked the participants the mental status of the past 2 weeks, using the Japanese version of the WHO-5, 9, 12 which has high psychometric properties. Each option is self-rated on a 6-point Likert scale ("at no time" = 0 to "all the time" = 5). Total scores range from 0 to 25, with scores <13 suggesting "poor" psychological well-being. We coded the participants into either of the following two categories: "poor" and "not poor."

We conducted cross-tabulation and chi-square tests to examine the association between each variable and the presence/ absence of deterioration of psychological well-being, reporting effect size to supplement P-value.

To examine the association between the combination of overtime working hours and social interaction to the deterioration of psychological well-being, we carried out multivariate logistic regression analysis with the presence/absence of deterioration of psychological well-being as the dependent variable. Models were adjusted for demographic variables and other potential confounders (ie, employment status, occupation, presence/absence of the increase in working from home, amount of housework, household income, household composition, current mental well-being, sleep hours, frequency of drinking, presence/absence of close people with COVID-19, and presence/absence of constant information gathering regarding COVID- 19) .

As a subgroup analysis, we performed multivariate logistic regression analysis by household composition to examine the association of overtime working hours, social interaction, and deterioration of mental well-being. P < .05 was considered statistically significant. All analyses were performed using SPSS version 25 (IBM, Chicago, IL, USA). Table 1 shows participants' characteristics. In total, a large volume of participants (n = 1583, 36.1%) was in the group of "unchanged overtime/unchanged interaction" in terms of the combination of overtime working hours and social interaction. Roughly one-fourth of participants (n = 1110, 25.3%) reported living as a single-person household. Onethird of them (n = 1458, 33.2%) were doing clerical work, about quarter (n = 1066, 24.3%) were doing professional work, 15.7% (n = 690) were doing sales work, 6.4% (n = 279) were working in sales, and around one-fifth were performing other types of work (n = 895, 20.4%). Table 2 shows the characteristics of participants by the presence/absence of deterioration of psychological wellbeing. Almost half the participants (n = 1931, 44%) answered that their psychological well-being deteriorated compared to before starting the COVID-19 outbreak. Moreover, of the combination of overtime working hours and social interaction, the frequency distribution of each category differed significantly by the presence/absence of deterioration of mental well-being. Table 3 summarizes the results of multivariate logistic regression analysis using the presence/absence of deterioration of mental well-being as the dependent variable. Regarding overtime working hours and social interaction, participants coded as "increased overtime/decreased interaction" reported greater deterioration of psychological well-being (odds ratio [OR]: 2.14, 95% confidence interval [CI]: 1.59-2.89) compared to those in the category of "unchanged overtime/unchanged interaction" as a reference, while those who coded as "decreased overtime/decreased interaction" also reported more deterioration in psychological well-being (OR:1.93, 95% CI: 1.48-2.53). Focusing on single-person households in the subgroup analysis, those who were in the category of "increased overtime/decreased interaction" (OR: 2.66, 95% CI: 1.50-4.69) and "decreased overtime/decreased interaction" (OR: 2.08, 95% CI: 1.24-3.47) reported significantly more severe deterioration of mental well-being.

This study examined the association between overtime working hours, social interaction, and deterioration of psychological well-being during the COVID-19 pandemic, focusing on full-time employees in Japan. Approximately 45% of participants reported that they experienced deterioration of mental well-being between January 2020 and November 2020. The multivariate analysis revealed that the combination of the increase in the amount of overtime working hours and decrease in social interaction was significantly associated with deterioration of mental well-being. Moreover, this association was relatively stronger among single-person households. Participants coded as "decreased overtime/decreased interaction" also tended to report the deterioration of psychological well-being considerably.

Almost half of the study participants had reported deterioration of mental well-being since the COVID-19 outbreak started in Japan. According to previous studies, the deterioration of mental well-being associated with lockdown or the restrictions related to COVID-19 were reported globally. [13] [14] [15] In contrast, with regard to the COVID-19-related policy, Japanese citizens had undergone a unique circumstance. The Japanese government declared a state of emergency on April 7, 2020, which authorized prefectural governors to "request" residents to refrain from going out of their homes for nonessential reasons and restrict the use of stores and facilities. Therefore, enforceable measures are extremely limited in the declarations and are much less restrictive than the lockdowns introduced in some regions such as Europe. However, during this limited lockdown, the deterioration of mental well-being can be seen among half of the participants, and this study can suggest that even request-based restriction may be associated with the deterioration of mental well-being among workers. In contrast, Sasaki and colleagues found that the psychological well-being remained almost the same among non-healthcare workers while healthcare workers tended to experience deterioration during the COVID-19 outbreak. 11 In their report, the participants were asked about their mental condition right after the first wave (May 2020), earlier than we asked our participants in the present research. Due to this timing gap, it is possible to consider that the work style and lifestyle of employees had changed through the second wave, and this may have made the mental well-being worse. Recent meta-analyses have suggested that the association between long working hours and depressive symptoms was stronger in Asian countries, including Japan. 1 Another study has indicated the associations between personal problems such as relationships with others and psychological distress among employees. 6 Based on these indications, this study focused on how the combination of work-related and life-related factors, the amount of overtime working hours, and the presence of social interaction, respectively, were associated with employees' mental well-being, and it found out the negative consequence associated with the combination of increasing overtime working hours and decreasing social interaction. Particularly during the COVID-19 crisis, also as known for the period of isolation, this negative combination between overtime working hours and social interaction, therefore, must be paid attention to.

Interestingly, those who coded as "decreased overtime/ decreased interaction" and "decreased overtime/unchanged interaction" also tended to reveal the deterioration of psychological well-being. Considering the data from 2020 by the Ministry of Health, Labour and Welfare in Japan, overtime pay included around 7% of the total monthly income. Moreover, the amount of overtime pay in a month was reduced by approximately 10%; there was a reduction of >30% in the service business specifically, compared to the previous year among full-time workers. 16 From this perspective, the COVID-19 pandemic made economic activity near-standstill because of stay-home requirements from the government, and some employees had undergone the reduction of overtime work, which might be related to the financial difficulty, accompanied by the feeling of isolation because of social distancing. Thus, based on this viewpoint, it may be possible that whereas decreasing overtime seems to be associated with a reduction of job stress, financial distress due to reduction of overtime work has a negative impact on mental well-being among employees during the pandemic. This type of combination would also have an impact on employees' mental well-being. In this study, the household composition itself was not associated with the deterioration of mental well-being, whereas it was also revealed that living in a single-person household was associated with worsening of a harmful impact of "increased overtime/decreased interaction" combination on the psychological well-being. In terms of the relation between household composition and mental well-being, several studies had been already conducted, yet there are inconsistent. A previous study found no or little association between the number of cohabitants itself and stress levels during the COVID-19 pandemic. 17 In contrast, other studies reported that singleperson households had experienced low mental well-being compared to multi-person households. [18] [19] [20] This study revealed that living alone itself may not be associated with the deterioration of mental well-being directly, which was consistent with a previous report, 17 whereas the presence of cohabitants would moderate the effect of the combination of the increase of overtime working hours and lack of social interaction under the COVID-19 situation. From this consequence, it is suggested that employees living in a single-person household would have more chances of experiencing difficulties because of the absence of cohabitants. In fact, the British government implemented the "social bubble" strategy to reduce the negative effect on the mental well-being of single-person households under the lockdown, 21 and our findings may imply the effectiveness of this strategy on the promotion/maintenance of psychological well-being among those in single-person households in public health crises, including a pandemic.

As far as we know, this is the first study that intended to shed some light on the relation between overtime working hours, social interaction, and long-term deterioration of mental well-being during the COVID-19 pandemic. However, the study also has potential limitations to note. First, to contact the participants during the pandemic, it was necessary to use a web-based survey. In addition, applicants for participation in the survey were accepted in the order of receipt until the number of participants reached the quotas. Due to this procedure, there is a possibility that the subjects who are potentially interested in this study may gather faster to apply. Furthermore, to remove the participants who did not devote an appropriate amount of attention when answering the question, the research company excluded the respondents who answered the same option throughout the questionnaire and those who completed answering it too soon, yet there has not been sufficient evidence to support the reliability and validity of these methods. For these reasons, there might be selection biases that could limit our findings. Moreover, because of the Internet-based questionnaire, all participants answered the questions in the form of a self-report. Second, the cross-sectional nature of this study made it difficult to determine the causality between overtime working hours, social interaction, and deterioration of mental well-being. In addition, although the questionnaire item was developed to assess the changes in mental well-being, the amount of overtime working hours, the opportunity of social interaction, and other variables, it is almost impossible to exclude recall biases from the survey. Third, in the adopted questionnaire, we could not obtain the information about, for example, the accessibility of preventive measures (eg, face mask and medical supply) against the COVID-19, the presence of financial vulnerability, or fear of job loss. Fourth, the questionnaire was written in Japanese and distributed to citizens in the metropolis of Tokyo and the three surrounding prefectures. Furthermore, the participants answered the questionnaire after the second wave of COVID-19, as shown in Figure 1 . The state of mental well-being in the general public might differ by the time point during the period. 22 Therefore, it is uncertain whether our findings are applicable to the population living in other countries or regions or different time periods. Further investigation is required to generalize the present findings to different populations. Finally, the eligibility of this study made it difficult to analyze the difference in deterioration of mental well-being between genders given the fact that the gender composition of the study sample was unbalanced (ie, singlehousehold men: 14.5%, multi-household men: 51.0%, singlehousehold women: 10.8%, multi-household women: 23.7%). However, female workers, especially those who are in multiperson households, may have unique vulnerability to deterioration of mental well-being due to an increase in the burden of household chores in the "stay home" recommendation during the COVID-19 pandemic. Thus, in future studies, it is worth focusing on female-specific vulnerability to deterioration of mental well-being during the pandemic.

Long working hours and depressive symptoms: systematic review and meta-analysis of published studies and unpublished individual participant data

Working hours and the onset of depressive disorder: a systematic review and metaanalysis

The effect of changes in overtime work hours on depressive symptoms among Japanese white-collar workers: a 2-year follow-up study

Public perceptions and experiences of social distancing and social isolation during the COVID-19 pandemic: a UK-based focus group study

Psychological impact of COVID-19 lockdown: an online survey from India

Associations between psychological distress and the most concerning present personal problems among working-age men in Japan

Job stress, burnout, work-life balance, well-being, and job satisfaction among pathology residents and fellows

Changes in health status, workload, and lifestyle after starting the COVID-19 pandemic: a web-based survey of Japanese men and women. Environ Health Prev Med

Nihongo-ban "WHO-5 seishinteki kenkoujyoutai hyou" no shinraisei narabini datousei

Psychiatric Research Unit. Who-five well-being index

The deterioration of mental health among healthcare workers during the COVID-19 outbreak: A population-based cohort study of workers in Japan

Reliability and validity of the Japanese version of the World Health Organization-Five Well-Being Index in the context of detecting depression in diabetic patients

Mental health before and during the COVID-19 pandemic: a longitudinal probability sample survey of the UK population

Mental health burden of the COVID-19 outbreak in Germany: predictors of mental health impairment

Using internet-based psychological measurement to capture the deteriorating community mental health profile during COVID-19: observational study

Monthly labour survey

Who is the most stressed during the COVID-19 pandemic? data from 26 countries and areas

Office for National Statistics. The cost of living alone

Relationship between living alone and common mental disorders in the 1993, 2000 and 2007 National Psychiatric Morbidity Surveys

Chronic stress and psychological well-being: evidence from Thailand on household crowding

Making a support bubble with another household

Increase in suicide following an initial decline during the COVID-19 pandemic in Japan

Impact of overtime working and social interaction on the deterioration of mental well-being among full-time workers during the COVID-19 pandemic in Japan: Focusing on social isolation by household composition

The data cannot be shared publicly as they contain information that could compromise the privacy of research participants, based on the "Ethical Guidelines for Medical and Health Research involving Human Subjects" set forth by the Japanese government. Under these guidelines, in principle, institutional review boards in Japan do not allow researchers to publicly share data obtained from human subjects with a third party. Accordingly, in the application requesting ethical approval for this study to the institutional review board of the Jikei University School of Medicine, Tokyo, Japan, we declared that this study would be conducted under the condition that only researchers approved by the review board at the time of the ethical application can access the dataset. Upon reasonable request from a third party, the corresponding author, MS (E-mail: suka@jikei.ac.jp), will consult with the review board (E-mail: rinri@jikei.ac.jp) regarding permission for data sharing.

This work was supported by the JMWH Bayer Grant from the Japan Society for Menopause and Women's Health.

Ethics approval: The study protocol was approved by the institutional review board of the Jikei University School of Medicine (No. 32-304 (10386)). This study has been conducted in accordance with the Ethical Guidelines for Medical