Changes in work/sleep patterns due to the COVID-19 pandemic are associated with psychological distress among Japanese workers

Objectives The coronavirus disease 2019 (COVID-19) pandemic has significantly impacted working life quality. This study investigated whether changes in work/sleep patterns due to the pandemic are related to poor psychological status among Japanese workers during the third wave of the COVID-19 pandemic (December 2020). We hypothesized that workers who experienced drastic changes in working hours and sleep duration would be at an increased risk of psychological distress. Methods A cross-sectional self-administered Internet survey was conducted that included questions regarding socio-demographics, lifestyle, health, and occupational background and conditions. Multivariable logistic regression models were utilized to examine the association between psychological distress and a combination of changes in working hours and sleep duration. Results Among 25,762 workers, decreased work hours and decreased sleep duration presented 2.59 times higher odds of psychological distress (95% confidence interval [CI] = 2.05–3.28) compared to those who had no changes in work hours combined with unchanged sleep duration (reference group). Increased work hours and decreased sleep duration were associated with 1.98 times higher odds of psychological distress (95% CI = 1.64–2.39). Conclusion Our observations confirmed that decreased sleep duration could be a key factor for psychological distress, irrespective of working hours. Interestingly, workers with a combination of decreased work hours and sleep duration posed the highest risk of psychological distress. Decreased work hours accompanied by financial difficulties in the early stage of the pandemic may have caused decreased sleep duration, resulting in a high prevalence of psychological distress. Our study underlined the importance of sleep management in maintaining workers’ mental health, moreover, the need to consider situations and conditions of other daily tasks, such as work hours, for better sleep management.


Introduction
The prolonged novel coronavirus disease 2019  pandemic has drastically changed our daily lives. While it has brought about some positive changes, such as less commuting and more time spent with family, it has negatively impacted mental health (Bueno-Notivol et al., 2021;Cénat et al., 2021;COVID-19 Mental Disorders Collaborators, 2021;Nochaiwong et al., 2021;OECD, 2021;Tanaka and Okamoto, 2021). Globally, due to the COVID-19 pandemic, the prevalence of mental health issues, including psychological distress, was higher than prior to the pandemic (Cénat et al., 2021;Nochaiwong et al., 2021). The estimated pooled prevalence of psychological distress among the general population was up to 50.0% in the first year of the pandemic (Cénat et al., 2021;Nochaiwong et al., 2021). Due to such psychological distress, total suicide rates have increased in Japan by 16% between July and October 2020 (Tanaka and Okamoto, 2021).
During the first year of the pandemic, many workers experienced changes in working conditions, such as forced teleworking, changes in job demands, and income reduction (financial difficulties). Moreover, workers experienced changes in their time use, especially the time used for working and sleeping.
In terms of work hours, experience of changes in work hours varies by workers. Some workers, such as health care and research laboratory workers, experienced increases in work hours during the first year of pandemic (Ischia et al., 2022;Loibner et al., 2022). According to the white paper from the Ministry of Health, Labour and Welfare, Japan (2021), workers in the information and communication industry also experienced an increase in work hours in December 2020 compared to December 2019.
In contrast, other industries, such as aviation, had a dramatic drop in demand (OECD, 2020), and the workers experienced decreases in working hours. Remote work also changed working hours; a reduction of work hours by 26 min was observed during remote work (Hallman et al., 2021), while a private survey in Japan reported that more than half of remote workers experienced longer working hours than working in the office (Japanese Trade Union Confederation, 2020). Another study also concluded that the time spent at the workstation increased by 1.5 h during a typical day for remote workers (Awada et al., 2021).
Regarding changes in sleep duration, the total sleep time increased by 1.5 h in duration at the end of the quarantine/stay-at-home condition compared to the pre-pandemic period in many countries (AMHSI Research Team et al., 2020;Raman and Coogan, 2022). Several studies on the pandemic have also indicated that remote work was associated with longer sleep duration (Hallman et al., 2021;Raman and Coogan, 2022); 34 min longer sleep duration for workdays (Hallman et al., 2021).
The Whitehall II prospective cohort study concluded that long working hours were significantly associated with shortened sleep (Virtanen et al., 2009). Furthermore, working more than 55 h per week (vs. normal work hours of 35-40 h) was related to an adjusted odds ratio (OR) of 2.25 (95% confidence interval (CI) 1.62-3.12) for short sleep of <7 h. In the same study, repeated exposure to long working hours was also associated with shortened sleep duration (OR = 2.80, 95% CI 1.20-6.53) (Virtanen et al., 2009). Given that work hours are inversely related to sleep duration, a study investigated the combined effects of working hours and sleep duration on mental health (Nakata, 2011); employees who worked more than 10 h per day with short sleep duration (<6 h/day) were at the highest risk for experiencing depressive symptoms (OR = 2.69, 95% CI 1.67-4.33) compared to those with 6-8 h of work hours and more than 6 h of sleep per day. However, the above findings are from a pre-pandemic situation; drastic changes in our work and life conditions, including work/sleep patterns, may have been caused by COVID-19, leading to poor mental health. These past studies considered ordinal work hours and sleep duration, but not societal changes posed by the COVID-19 pandemic.
Therefore, we sought to explore how changes in work/sleep patterns due to the pandemic have impacted workers' psychological status. This study aimed to investigate the combined association of work hours and sleep duration on psychological distress among Japanese workers during the third wave of the COVID-19 pandemic in Japan. We hypothesized that workers with increased work hours and decreased sleep duration due to the pandemic would have an increased risk of psychological distress.

Study design and population
This Internet-based survey targeted Japanese workers, including sectional employees, managers, and executive managers as well as temporary/contract employees, self-employed, small office/home office workers, but excluding part-time workers, homemakers, students, and the unemployed. We distributed a nationwide health survey to workers registered in the research company's panel from December 20 to 26, 2020, during the third wave of the COVID-19 pandemic in Japan (the detailed procedure is presented in our protocol paper; Fujino et al., 2021). After excluding invalid responses containing inconsistent answers, short response time (≤360 s), extremely low body weight (<30 kg), short height (<140 cm) (n = 6,051), and those who were already diagnosed with depression (n = 1,274), a total of 25,762 respondents were submitted for the final analysis.

Exposure measures
Participants were asked whether they had experienced changes in work hours and sleep duration due to the COVID-19 pandemic with the answering choice of increased, unchanged, or decreased. To investigate the combined association, we categorized work hours and sleep duration into nine work/sleep patterns: (i) increased work hours and increased sleep duration, (ii) increased work hours and unchanged sleep duration, (iii) increased work hours and decreased sleep duration, (iv) unchanged working hours and increased sleep duration, (v) unchanged working hours and unchanged sleep duration (reference group), (vi) unchanged work hours and decreased sleep duration, (vii) decreased work hours and increased sleep duration, (viii) decreased work hours and unchanged sleep duration, and (ix) decreased work hours and decreased sleep duration.

Outcome measure: Psychological distress
Psychological distress was assessed using the Japanese version of the Kessler 6 (K6) (Kessler et al., 2002), which is a short screening scale for psychological distress (Furukawa et al., 2008). K6 includes six questions regarding psychological states during the past 30 days, with five answering choices (0 = none of the time to 4 = all the time) (Furukawa et al., 2008). The scores for the six questions were then summed, and the cutoff score for mild psychological distress was set at 5 or higher, as it was suggested to be the optimal cutoff score for the Japanese population (Kessler et al., 2002;Furukawa et al., 2008;Sakurai et al., 2011).

Control variables
Control variables included socio-demographics (sex, age groups, marital status, educational level, living with a child or children under 12 years old [yes/no], residential area where the state of emergency was declared [yes/no], and subjective financial situation [very difficult/ slightly difficult/average/comfortable/very comfortable]), lifestyle (smoking, drinking frequency, exercise frequency, time for house chores during the pandemic [increased/unchanged/decreased], and time spent with family during the pandemic [increased/unchanged/ decreased]), health (body mass index, diseases currently being treated [yes/no]), and occupational background and conditions (type of industry, number of employees, one-way commuting time, teleworking preference, telework frequency, changes in job stress due to the COVID-19 pandemic [increased/unchanged/decreased], job control, and social support at work). Job control and social support at work were assessed using the job content questionnaire (Karasek et al., 1998) and divided into low/high groups by the median score of each: job control (64.0) and social support (22.0).

Statistical analysis
Multivariable logistic regression models were utilized to investigate the association between psychological distress and combined changes in work hours and sleep duration. We first adjusted for sex and age groups (Model 1). In the next model (Model 2), we adjusted for marital status, educational level, living with a child or children under 12 years old, residential area where the state of emergency was declared, smoking, drinking frequency, exercise frequency, time for house chores, time with family, body mass index, and diseases currently being treated, in addition to Model 1. In Model 3, we additionally adjusted for occupational background and conditions, such as type of industry, number of employees, one-way commuting time, teleworking preference, telework frequency, job control, social support, and job stress during the pandemic. Finally, in Model 4, we further adjusted for the financial situation because work hours, sleep duration, and psychological states were more likely related to the financial situation during the pandemic among workers.
In addition to the above analyses, we recorded the financial situation into two groups (difficult/not difficult) and performed a multivariate logistic regression analysis to investigate the impact of psychological distress on changes in work hours and sleep duration, in tandem with the financial situation, resulting in the creation of 18 groups.
All analyses were performed using SPSS Version 25.0 (IBM Corp., Armonk, NY, United States). The level of significance was set at p < 0.05. Table 1 presents the characteristics of the study participants according to their psychological state. In this study, 9,827 participants (38.1%) had mild psychological distress. A higher proportion of younger (age 20-39 years) and unmarried participants fell into the groups with psychological distress in comparison to the group without psychological distress. Nearly half of the participants with psychological distress experienced difficulty or slight difficulty in relation to their financial situation, while nearly 30% in those without psychological distress. In the group with psychological distress, 46.3% of participants experienced increased job stress and 60.0% of them experienced low social support at work. The other characteristics were similar in both groups. Table 2 highlights the characteristics of the study participants by nine groups of changes in work/sleep patterns due to the pandemic. In the group with decreased working hours and decreased sleep duration, more than 30% of participants had experienced extreme difficulty in relation to their financial situation. In the group with increased working hours and decreased sleep duration, 81.6% experienced increases in job stress due to the pandemic.

Results
Independent associations between psychological distress and work hours and sleep duration, respectively, as estimated by stepwise logistic regression models, are denoted in Table 3. In Model 4, participants with increased working hours had significantly higher odds of experiencing psychological distress (odds ratio [OR] = 1.15, 95% confidence interval [CI] = 1.03-1.28) compared to those with unchanged work hours, while decreased work hours showed an insignificant association. Regarding sleep duration, those who experienced decreased sleep duration showed OR of 1.97 (95% CI = 1.79-2.18) compared to unchanged sleep duration. Table 4 presents associations between psychological distress and changes in work/sleep combined patterns by stepwise logistic regression models. Notably, participants had more risk of experiencing Frontiers in Psychology 04 frontiersin.org

Discussion
This study investigated the association between changes in work/ sleep patterns and psychological distress among Japanese workers during the COVID-19 pandemic. To our knowledge, this is the first study that specifically examined the combined effect of changes in work hours and sleep duration from pre-pandemic to in-pandemic periods on psychological distress. This study primarily questioned whether changes in work hours and sleep duration would pose a higher risk of psychological distress compared to when the work hours and sleep duration remained unchanged. We found an increased risk of psychological distress among those who had decreased sleep duration, irrespective of changes in work hours. Notably, workers with decreased work hours and decreased sleep duration had the highest odds of psychological distress (OR = 2.59, 95% CI = 2.05-3.28), followed by workers with increased work hours and decreased sleep duration (OR = 1.98, 95% CI = 1.64-2.39). This finding outlines that reduced sleep duration is a major factor related to increased psychological distress.
Moreover, we found that the risk of psychological distress was the highest among those with a combination of decreased work hours and sleep duration in the early stages of the pandemic, which was contrary to our expectations. A possible key factor that led to this result was financial impact of the pandemic on this decreased work hours and decreased sleep duration group. We found that more than 30% of participants in this group reported experiencing extreme difficulty in relation to their financial situation (Table 2). This group also experienced a drop of OR from 2.93 to 2.59 when adjusted for the financial situation in the final model in Table 4. Additionally, in our sub-analysis, we found a higher risk of psychological distress  Table S1). Previous studies have also supported our hypothesis; people experiencing financial difficulties had higher rates of psychological distress during the COVID-19 crisis (Griffiths et al., 2021;Nagasu et al., 2021;OECD, 2021;Ettman et al., 2022;Sekścińska et al., 2022). The increased risk of psychological distress, with increased work hours and decreased sleep duration, was in line with a past crosssectional study in an ordinal pre-pandemic work situation (Nakata, 2011), which presented longer work hours of 10+ h per day with short sleep duration (<6 h/day) to have 2.69 times higher odds of having depressive symptoms compared to normal working hours (6-8 h/day with more than 6 h/day of sleep).
In the early stages of the pandemic, many workers experienced unstable working conditions, such as decreased quantitative job demands, decreased work hours, and reduced income. Despite this situation, the provision of job-related services and support (e.g., in-person job consultations and career seminars) was limited to reduced human contact, resulting in workers having poorer access to job-related services and support. According to statistics from the Ministry of Health, Labour and Welfare, Japan, the number of new job applications in public employment service centers in Japan dropped by approximately 10,000 from 2019 to 2020 (Ministry of Health Labour and Welfare, Japan, 2022). Such unstable working conditions and the limited provision of job-related services and support during the pandemic are more likely to lead to financial difficulties for several workers.
A previous occupational cohort study revealed strong associations between financial difficulties and sleep problems such as shortened sleep duration (Lallukka et al., 2012). In addition to financial difficulties and decreased sleep duration, uncertainty over the prolonged pandemic may have increased psychological distress among workers. Moreover, fewer socializing opportunities, limited access to mental health services and support, and fewer opportunities to seek help during the pandemic (Campion et al., 2020;Moreno et al., 2020) might have heightened the risk of psychological distress.
The strengths of this study are as follows. First, we assessed changes in work/sleep patterns among Japanese workers from pre-pandemic to in-pandemic periods, unlike previous studies that cross-sectionally assessed work/sleep patterns on typical workdays. Second, we employed self-reports to evaluate sudden changes in work/sleep patterns due to the pandemic. Most previous studies objectively assessed changes in working hours and sleep duration by the number of hours. However, evaluating the sudden changes caused by COVID-19 with numbers by self-report would lead to potential biases, that is, a recall bias in the number of hours before the pandemic. Additionally, sufficient working hours and sleep duration vary among individuals. Third, we included both changes in work hours and sleep duration due to the COVID-19 pandemic to examine the acute impact on psychological distress. Combining changes in work hours and sleep duration would reflect the time use of workers because they spend a large amount of time working and sleeping during the day.
However, this study had several limitations. First, although many workers experienced increased sleep duration in the early stages of the pandemic, our study focused on workers with decreased sleep duration. Second, this study utilizes an online closed survey, meaning that participants were recruited from the survey company's panel, which may have led to selection bias. Third, we were unable to confirm the participants' pre-pandemic psychological state. They may have already developed psychological distress prior to the pandemic. Fourth, we were unable to identify whether participants experienced decreased sleep duration because of the stressful situation from the pandemic, or whether they simply could not set aside time for sleep because they had other tasks to do. Finally, there might be other co-factors, such as napping, side jobs, and time-use on free days, which may have impacted psychological distress among Japanese workers.

Conclusion
This study confirmed that decreased sleep duration could be a key factor in psychological distress, irrespective of working hours. Moreover, contrary to our expectations, workers with a combination of decreased work hours and sleep duration posed the highest risk of psychological distress. Decreased work hours accompanied by financial difficulties in the early stage of the pandemic may have caused decreased sleep duration, resulting in a high prevalence of psychological distress. Our study underlined the importance of sleep management in maintaining workers' mental health as well as the need to consider situations and conditions of other daily tasks, such as work hours, for better sleep management.

Data availability statement
The datasets presented in this article are not readily available because it is not open to the public. Requests to access the datasets should be directed to YF, zenq@med.uoeh-u.ac.jp.

Ethics statement
The studies involving human participants were reviewed and approved by the Ethics Committee of the University of Occupational and Environmental Health, Japan and the Ethics Committee of the International University of Health and Welfare. The patients/ participants provided their written informed consent to participate in this study.

Author contributions
TT contributed to data analysis and writing. YF, ST, KM, MT, HA, KO, and RM contributed to the research design and data collection and provided useful suggestions during the analysis. AN supervised the analysis and drafting of the manuscript and reviewed the manuscript for intellectual content. All authors have contributed to the manuscript and approved the submitted version.

Funding
This study was supported and partly funded by a research grant from the University of Occupational and Environmental Health, Japan; the Japanese Ministry of Health, Labour and Welfare (H30-josei-ippan-002,