Abstract
Daily rest period (DRP) refers to the interval between the end of one workday and the start of the next. This study examined the joint association of DRP and sleep duration with subsequent sick leave among Japanese daytime employees. A total of 5,593 participants were assessed for DRP and sleep duration at baseline and for sick leave at a one-year follow-up. They were categorized into ten groups based on their DRP and daily sleep duration. Logistic regression analyses for individuals experiencing sick leave for longer than a month showed that the combination of short DRP (<11 h) and sleep duration (<6 h) had a higher odds ratio (4.981, 95% confidence interval [CI]=1.126–22.046) than the reference group. Furthermore, the combination of short DRP and normal sleep duration had a higher odds ratio (8.152, 95% CI=1.801–36.902) than the reference group. Short DRP was associated with subsequent long sick leave.
Keywords: Quick return, Sleep quantity, Absenteeism, Sickness absence, Long working hours
Daily rest period (DRP) refers to the interval between the end of one workday and the start of the following workday. Short DRP indicates long working hours and inhibits recovery from fatigue due to work. In this regard, short DRP, especially <11 h (quick return), is associated with health problems1) and sick leave2, 3). Thus, the European Union’s (EU) working time directive states that EU workers have the right to take “a minimum DRP of 11 consecutive hours every 24 h”. In Japan, the work interval system requires employers to ensure that their employees receive a certain number of rest hours between the end of one workday and the beginning of next. Ensuring a minimum DRP through these directives and systems may prevent health problems and sick leave.
DRP includes activities and/or behaviors that are normally performed outside working hours, i.e., daily sleep, leisure activities, commuting, and other nonwork activities. Among these, sleep duration is also associated with health problems4) and sick leave5). Meanwhile, a short DRP may result in shorter sleep opportunity, since it is difficult to reduce commuting times or times for other daily activities (e.g., meals and bathing) required to maintain a minimum standard of living. Although, DRP has been associated with sleep duration, i.e., a shorter DRP results in a shorter sleep duration6), our previous cross-sectional survey of Japanese daytime employees (n=13,306) found that some employees had short sleep duration (<6 h) within a sufficient DRP (15 h: standard 8 h work shift with a 1 h lunch break; i.e., no overtime work) (0.7%), whereas others had normal sleep duration (≥6 h) during quick return (6.8%)7). In addition, the combinations of quick return and short sleep duration, of the sufficient DRP and short sleep duration, and of quick return and normal sleep duration were significantly associated with health problems (sleep difficulty and poor mental health). This suggests that ensuring sufficient DRP and normal sleep duration is important for worker health7). However, quick return with or without short sleep duration was negatively associated with sick leave (≥1 d per the previous year)8). This finding is contradictory to the results of previous studies2, 3, 5). This may be due to several reasons; one of which is the unique Japanese work/rest environment (culture). In Japan, sick leave is not ruled by law. Therefore, some workers cannot take sick leave (even if they want to), and others do not want to take sick leave because of minor illnesses. For example, 700 individuals in Japan aged ≥15 yr were asked whether they would take sick leave from school or work due to cold symptoms and a temperature of 37°C. Although this was during the COVID-19 pandemic period, 23.0% of respondents responded that they could not take sick leave from work or school (even if they wanted to), and 17.0% individuals indicated that they would not take sick leave9). Therefore, the sickness-absence rate is lower among Japanese than that in other countries10). Among workers with short DRP (i.e., long working hours), they are unable to take sick leave because they are simply too busy. Another possible reason is the healthy worker effect, i.e., unhealthy workers are generally excluded from those with quick return and/or short sleep duration. Meanwhile, little evidence exists on the temporal relationship between DRP and sleep duration with sick leave, due to the limitations of cross-sectional surveys, which simultaneously assessed the exposure and outcomes. Therefore, prospective cohort data are required to clarify this causal relationship.
In this regard, the present study examined the joint association of DRP and sleep duration with subsequent sick leave among a sample of Japanese daytime employees. It hypothesized that the combination of a short DRP and short sleep duration, of a short DRP and normal sleep duration, and of a sufficient DRP and short sleep duration are associated with increased risk of sick leave relative to the combination of a sufficient DRP and normal sleep duration.
Sick leave was divided into 3 periods in the previous year as follows: short (1–6 d), medium (1 wk to 1 month), and long (>1 month) sick leave. Since shiftwork and regular-night shifts can disturb circadian rhythms and influence workers’ health, this study only focused on daytime employees without such schedules.
This study was approved by the Research Ethics Committee of the National Institute of Occupational Safety and Health, Japan (2021N-1-19). It uses data from the web-based longitudinal study of the work environment and daily lifestyle (WELWEL), a web-based prospective cohort study of employees in Japan11). The baseline and follow-up surveys were conducted in February 2022 and February 2023, respectively. Data was collected online by a third-party research company, which sent participation invitations to 421,825 registrants, of which 115,094 accessed the survey website. The first 20,000 self-selecting registrants who met the distribution ratios (number of respondents in each of 20 industry types, age groups [20–29, 30–39, 40–49, and 50–59 yr old], and gender) for Japanese employees based on the population census12) and did not meet the exclusion criteria (e.g., working <20 h per week) participated in the baseline survey11). In the follow-up survey, the participants were invited to participate in the survey; 7,970 of 20,000 participants completed it. A comparison between participants and dropouts was conducted in another study13) (e.g., participants who dropped out tended to be more likely to be female and younger).
Regarding the measurements, exposures were DRP and sleep duration at the baseline. The average starting and ending work hours of the previous month were obtained, with answers provided in 5 min intervals. DRP was calculated as the interval between the end of one workday and the start of the following workday. The sleep duration question gathered information on how many hours the participants normally slept, with answers given in 5 min intervals.
Outcomes were sick leave at the 1 yr follow-up. Sick leave was assessed using a single question, “How long have you been absent from work because of sickness or poor health during the past year”? for which there were four response options: not at all, 1–6 d, 1 wk to 1 month, and >1 month. These cutoff points were based on previous Japanese cohort studies that used cutoff points of 1 wk14) and 1 month15) of sick leave.
Potential confounders included individual characteristics, sick leave, and sleep difficulty at the baseline. Individual characteristics included gender, age, height, weight, smoking status (dichotomized into nonsmoker and current smoker), alcohol consumption (dichotomized into nonconsumption and ≥1 time/month), marital status (dichotomized into unmarried and married), preschool child care (yes or no), family caregiving (yes or no), academic history (dichotomized into senior high school or less [i.e., ≤12 yr] and some college or higher [i.e., >12 yr]), annual household income (grouped into categories of <4 million, 4–8 million, and ≥8 million yen)11), industry type (dichotomized into secondary and tertiary industries), occupation (dichotomized into managerial and nonmanagerial workers), frequency per week of remote working from home during the previous month (dichotomized into <5 d per week and ≥5 d per week [fully remote]), COVID-19 involvement (dichotomized into no involvement and involvement [COVID-19 infected participants, family, and/or co-workers and/or participants was a close contact of someone with COVID-19]), and perceived stress in the workplace (dichotomized into yes or no). Sick leave at the baseline was also collected in the same manner as described above the 1 yr follow-up. The Japanese version of the Athens Insomnia Scale (AIS), a self-administered questionnaire that has been proven to exhibit sufficient reliability and validity, was used to assess sleep difficulties16) at the baseline. The Japanese version of the AIS16) includes eight question items collecting information on sleep disturbances and daytime dysfunction in the previous a month, where each item is rated on a 4-point Likert scale. The final AIS score is the sum of the item scores (total range: 0–24), where higher scores indicate greater difficulty sleeping. Scores of 6 points or more are considered indicate pathological insomnia16).
Drawing on our previous cross-sectional survey7), we categorized the participants into 10 groups according to their DRP (<11, 11–12, 13–14, 15, or ≥16 h) and sleep duration (<6 or ≥6 h). In this case, DRP of <11 h (quick return) was set as short DRP. The National Sleep Foundation recommends 7–9 h of sleep per day and does not recommend <6 h sleep for adults aged 26–64 yr17). However, in Japan, few workers sleep for more than 7 h per day (15.3%)18). Therefore, sleep durations were divided into <6 h and ≥6 h, and set as short and normal sleep durations, respectively.
Multiple logistic regression analyses were conducted to estimate odds ratios (ORs) for the 10 groups with short (1–6 d), medium (1 wk to 1 month), and long (>1 month) sick leave, with a 95% confidence interval (CI). Crude ORs were estimated (model 1). Meanwhile, model 2 was adjusted for the sick leave of the independent baseline variable (i.e., short, medium, or long sick leave). Model 3 was adjusted for the variable of model 2 and AIS (<6 or ≥6). Finally, model 4 was adjusted for the variables of model 3 and demographic factors (gender, age, body mass index, smoking status, alcohol consumption, marital status, preschool childcare, family caregiving, academic history, annual household income, industry type, occupation, full working from home, COVID-19 involvement, and perceived stress in the workplace). The combination of 15 h DRP (sufficient DRP) and ≥6 h sleep group (normal sleep duration) was set as the reference group. The ORs and 95% CIs for the <11 h DRP and <6 h sleep group (quick return and short sleep duration) group, <11 h DRP and ≥6 h sleep (quick return and normal sleep duration) group, and 15 h DRP and <6 h sleep (sufficient DRP and short sleep duration) group were used to examine the hypotheses. The level for statistical significance was set at α=0.05. All statistical analyses were conducted using IBM SPSS Statistics version 23.0 for Microsoft Windows.
The data for 2,377 (of 7,970) participants were excluded because they were shift or night workers (n=979), worked less than five days a week (n=596), were being treated for a sleep disorder (n=176), had used sleeping pills within the previous year (n=577), were outliers (± 3 standard deviations [SDs] from the mean of the eligible participants for sleep duration, start and end of working hours, and/or DRP: n=671), and/or had inconsistent data (sleep duration longer than the DRP, n=240). The final sample included 5,593 Japanese daytime employees.
Table 1 presents the demographics of the final sample (n=5,593) at the baseline. Of the participants, 27%, 3%, and 1%, experienced short, medium, and long sick leave, respectively.
Table 1. Participant characteristics at the baseline (n=5,593).
| n or mean (% or SD) | ||
|---|---|---|
| Gender, female, n (%) | 2,243 (40%) | |
| Age, years, mean (SD) | 45.1 (9.0) | |
| Body mass index, mean (SD) | 22.3 (3.6) | |
| Alcohol consumption, ≥1 time/week, n (%) | 3,234 (58%) | |
| Smoking status, current smoker, n (%) | 1,273 (23%) | |
| Marital status, married, n (%) | 3,141 (56%) | |
| Childcare, yes, n (%) | 799 (14%) | |
| Family caregiving, yes, n (%) | 287 (5%) | |
| Education, ≥12 yr, n (%) | 4,319 (77%) | |
| Annual household income, n (%) | ||
| <4 million yen | 1,292 (23%) | |
| 4–8 million yen | 2,568 (46%) | |
| ≥8 million yen | 1,733 (31%) | |
| Occupation, managerial workers, n (%) | 905 (16%) | |
| Industry type, secondary industry, n (%) | 1,803 (32%) | |
| Remote working, ≥5 d per week, n (%) | 371 (7%) | |
| COVID-19 involvement, involvement, n (%) | 922 (17%) | |
| Perceived stress in the workplace, yes, n (%) | 3,525 (63%) | |
| Athens Insomnia Scale, n (%) | ||
| <6 | 3,598 (64%) | |
| ≥6 (insomnia) | 1,995 (36%) | |
| Sickness leave, n (%) | ||
| not at all | 3,882 (69%) | |
| more than 1 d and less than 1 wk | 1,481 (27%) | |
| more than 1 wk and less than 1 month | 174 (3%) | |
| more than 1 month | 56 (1%) | |
| Daily rest period, n (%) | ||
| <11 h | 62 (1%) | |
| 11–12 h | 460 (8%) | |
| 13–14 h | 1,974 (35%) | |
| 15 h | 2,333 (42%) | |
| ≥16 h | 764 (14%) | |
| Sleep duration, n (%) | ||
| <6 h | 1,176 (21%) | |
| ≥6 h | 4,417 (79%) | |
Table 2 presents the incidence rates and results of logistic regression for short, medium, and long sick leave. Logistic regression analyses for long sick leave revealed that the combination of quick return and <6 h sleep (OR=4.981, 95% CI=1.126–22.046: model 1) and quick return and ≥6 h sleep (OR=8.152, 95% CI=1.801–36.902: model 1) had higher ORs than the reference group. These significant associations remained after adjustment for covariates (models 2 to 4).
Table 2. Daily rest period (DRP), sleep duration, and sick leave.
| DRP, sleep duration | Incidence ratio (%) | Odds ratio (95% confidence interval) | ||||
|---|---|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |||
| Short sick leave: 1–6 d | ||||||
| <11 h, <6 h (n=38) | 26.3% | 0.875 (0.422–1.813) | 1.025 (0.480–2.189) | 0.937 (0.438–2.005) | 0.952 (0.444–2.041) | |
| <11 h, ≥6 h (n=24) | 25.0% | 0.816 (0.322–2.068) | 0.789 (0.298–2.093) | 0.764 (0.288–2.023) | 0.705 (0.263–1.887) | |
| 11–12 h, <6 h (n=150) | 29.3% | 1.017 (0.706–1.465) | 1.026 (0.698–1.506) | 0.966 (0.657–1.420) | 0.943 (0.638–1.395) | |
| 11–12 h, ≥6 h (n=310) | 23.2% | 0.741 (0.559–0.982) | 0.795 (0.593–1.067) | 0.775 (0.577–1.041) | 0.750 (0.556–1.012) | |
| 13–14 h, <6 h (n=438) | 24.9% | 0.811 (0.639–1.030) | 0.830 (0.647–1.066) | 0.801 (0.623–1.029) | 0.795 (0.616–1.026) | |
| 13–14 h, ≥6 h (n=1,536) | 28.4% | 0.971 (0.837–1.126) | 1.032 (0.883–1.207) | 1.027 (0.878–1.201) | 1.000 (0.852–1.173) | |
| 15 h, <6 h (n=429) | 28.4% | 0.973 (0.772–1.227) | 0.937 (0.734–1.196) | 0.887 (0.694–1.135) | 0.912 (0.711–1.169) | |
| 15 h, ≥6 h (n=1,904) | 29.0% | 1.000 | 1.000 | 1.000 | 1.000 | |
| ≥16 h, <6 h (n=121) | 25.6% | 0.844 (0.554–1.284) | 0.810 (0.521–1.260) | 0.781 (0.501–1.217) | 0.806 (0.515–1.262) | |
| ≥16 h, ≥6 h (n=643) | 29.7% | 1.035 (0.851–1.259) | 1.114 (0.907–1.369) | 1.116 (0.908–1.372) | 1.104 (0.891–1.367) | |
| Medium sick leave: 1 wk to 1 month | ||||||
| <11 h, <6 h (n=38) | 5.3% | 0.631 (0.151–2.647) | 0.599 (0.142–2.536) | 0.540 (0.127–2.293) | 0.533 (0.124–2.293) | |
| <11 h, ≥6 h (n=24) | 8.3% | 1.033 (0.241–4.434) | 1.111 (0.259–4.769) | 1.063 (0.247–4.570) | 1.104 (0.253–4.813) | |
| 11–12 h, <6 h (n=150) | 8.0% | 0.988 (0.536–1.823) | 0.985 (0.532–1.824) | 0.917 (0.494–1.703) | 0.983 (0.523–1.848) | |
| 11–12 h, ≥6 h (n=310) | 11.0% | 1.400 (0.945–2.073) | 1.444 (0.973–2.142) | 1.401 (0.943–2.080) | 1.436 (0.955–2.159) | |
| 13–14 h, <6 h (n=438) | 5.9% | 0.717 (0.467–1.102) | 0.726 (0.472–1.117) | 0.696 (0.452–1.073) | 0.697 (0.449–1.081) | |
| 13–14 h, ≥6 h (n=1,536) | 7.2% | 0.885 (0.687–1.141) | 0.893 (0.692–1.153) | 0.889 (0.689–1.147) | 0.926 (0.713–1.202) | |
| 15 h, <6 h (n=429) | 10.7% | 1.365 (0.964–1.932) | 1.335 (0.941–1.895) | 1.257 (0.883–1.789) | 1.224 (0.855–1.752) | |
| 15 h, ≥6 h (n=1,904) | 8.1% | 1.000 | 1.000 | 1.000 | 1.000 | |
| ≥16 h, <6 h (n=121) | 8.3% | 1.024 (0.525–1.996) | 1.003 (0.512–1.966) | 0.971 (0.496–1.903) | 0.844 (0.427–1.671) | |
| ≥16 h, ≥6 h (n=643) | 9.2% | 1.148 (0.838–1.572) | 1.141 (0.831–1.565) | 1.143 (0.833–1.569) | 0.957 (0.688–1.332) | |
| Long sick leave: >1 month | ||||||
| <11 h, <6 h (n=38) | 5.3% | 4.981 (1.126–22.046) | 6.072 (1.364–27.034) | 5.191 (1.153–23.368) | 5.835 (1.222–27.868) | |
| <11 h, ≥6 h (n=24) | 8.3% | 8.152 (1.801–36.902) | 9.935 (2.182–45.246) | 9.355 (2.045–42.787) | 8.853 (1.860–42.142) | |
| 11–12 h, <6 h (n=150) | 2.0% | 1.830 (0.540–6.206) | 2.230 (0.653–7.620) | 1.999 (0.581–6.876) | 2.497 (0.700–8.912) | |
| 11–12 h, ≥6 h (n=310) | 1.0% | 0.876 (0.260–2.955) | 0.826 (0.239–2.853) | 0.793 (0.229–2.745) | 1.027 (0.293–3.594) | |
| 13–14 h, <6 h (n=438) | 0.9% | 0.826 (0.282–2.420) | 0.873 (0.294–2.592) | 0.805 (0.269–2.409) | 0.846 (0.276–2.591) | |
| 13–14 h, ≥6 h (n=1,536) | 1.0% | 0.944 (0.491–1.815) | 1.020 (0.525–1.982) | 1.018 (0.523–1.980) | 1.085 (0.550–2.141) | |
| 15 h, <6 h (n=429) | 2.1% | 1.921 (0.874–4.225) | 1.605 (0.706–3.647) | 1.455 (0.635–3.335) | 1.399 (0.592–3.304) | |
| 15 h, ≥6 h (n=1,904) | 1.1% | 1.000 | 1.000 | 1.000 | 1.000 | |
| ≥16 h, <6 h (n=121) | 1.7% | 1.507 (0.349–6.503) | 1.580 (0.356–7.017) | 1.534 (0.348–6.768) | 1.315 (0.286–6.048) | |
| ≥16 h, ≥6 h (n=643) | 1.7% | 1.561 (0.748–3.255) | 1.447 (0.680–3.081) | 1.491 (0.701–3.173) | 1.206 (0.548–2.658) | |
Reference (1.00)=daily rest period of 15 h with sleep duration ≥6 h. Model 1=Crude model. Model 2=Adjusted for the presence or absence of sick leave of the independent variable (i.e., sick leave for 1–6 d, 1 wk to 1 month, or >1 month) in the baseline. Model 3=Model 2 + Athens Insomnia Scale (<6, ≥6). Model 4=Model 3 + gender, age, body mass index, smoking status (0: nonsmoker, 1: current smoker), alcohol consumption (0: nonconsumption, 1: more than once per week), education (0: <12 yr, 1: ≥12 yr), marital status (0: unmarried, 1: married), childcare (0: no, 1: yes), family caregiving (0: no, 1: yes), annual household income (<4 million yen, 4–8 million yen, ≥8 million yen), industry type (0: tertiary industry, 1: secondary industry), occupation (0: nonmanagerial workers, 1: managerial workers), remote working from home (0: <5 d/week, 1: ≥5 d/week), perceived stress at the workplace (0: no, 1: yes), and COVID-19 involvement (0: no involvement, 1: involvement). Significant odds ratios (p<0.05) and their 95% confidence intervals values are presented in boldface.
Logistic regression analyses for short sick leave revealed that the combination of 11–12 h DRP and ≥6 h sleep had lower OR than the reference (OR=0.741, 95% CI=0.559–0.982: model 1). However, this association was not significant (OR=0.795, 95% CI=0.593–1.067: model 2) after adjustment for the presence or absence of short sick leave in the baseline (i.e., models 2 to 4).
Logistic regression analyses for medium sick leave showed no significant effect of the combination of DRP and sleep duration (all p>0.05). These nonsignificant associations remained after adjustment for covariates (models 2 to 4).
This study examined the joint association of DRP and sleep duration with subsequent sick leave among Japanese daytime employees. It was found that the combination of quick return and short sleep duration was associated with long sick leave, and this supports the hypothesis. Previous studies have reported that quick return and short sleep duration were associated with increased risk of sick leave2, 3, 5), and this study found that overlapping quick return and short sleep duration lead to long sick leave. DRP was also closely related to the number of working hours, i.e., short DRP indicates long working hours, which lead to severe fatigue. Meanwhile, short sleep duration in short DRP may limit recovery from such fatigue, increasing the risk of health problems. Previous studies also reported that both quick return and short sleep duration were associated with various health problems4, 6, 8, 19). These health problems may increase the sick leave.
The combination of quick return and normal sleep duration was associated with long sick leave, and this supports the hypothesis. Thus, workers who suffer quick return have the risk of a long sick leave, even if they have ≥6 h sleep during quick return. Workers with normal sleep duration in quick return may have a short interval between the end of the workday and bedtime, leading to a high arousal level prior to sleep onset due to the remnants of high arousal from work and poor sleep quality. Previous studies reported that pre-sleep cognitive arousal can prolong sleep latency, increase high frequency electroencephalogram (EEG) activity20) and decrease EEG delta power density21), suggesting that negatively affects sleep quality. Poor sleep quality, not necessarily pathological insomnia, may inhibit recovery from severe fatigue, and this may increase the risk of long sick leave. In addition, a short interval between the end of the workday and bedtime may reduce leisure time and time for activities with family, friends, and co-workers. Loneliness, impaired social support, and short conversation time are problems associated with mental health22, 23). Impaired social interaction due to short leisure time may lead to increased sick leave.
A combination of sufficient DRP and short sleep duration was not associated with long sick leave, and this does not support the hypothesis. Lullukka et al.5) reported that the association between sleep duration and sick leave tends to be U-shaped, i.e., optimal sleep duration is 7.6–7.7 h, and the shorter the sleep duration, the greater the sick leave. The present study separated groups by the sleep duration at 6 h, following our previous cross-sectional study7). Setting the cutoff point to 5 h of sleep, which has a higher risk of sick leave than 6 h5), may lead to the same results as the study which reported that short sleep duration was associated with sick leave. In addition, although <6 h sleep in 15 h DRP was not associated with sick leave in the present study, short sleep durations are associated with several health problems4). Thus, it is recommended to ensure 7–9 h of sleep per day to maintain good health17).
The results of this prospective cohort survey differ from those of our previous cross-sectional study, which found that the combination of quick return with/without short sleep duration was associated with decreasing risk of sick leave (≥1 d)8). This study shows a temporal relationship between exposure and outcome, providing more support for a causal relationship. Therefore, quick return can result in long sick leave among Japanese employees. Other previous prospective studies have also reported that quick return increases the risk for sick leave2, 3). Thus, it is necessary to ensure a sufficient DRP to prevent sick leave.
This study showed that, the joint association of DRP and sleep duration was associated with long sick leave but not with short or medium sick leave. Previous studies reported that shorter sleep durations were associated with increased incidence of the common cold19), and quick return was associated with increased risk of short sick leave2, 3). Thus, it was expected that short DRP and sleep duration would lead to short and medium sick leave, but this was not shown by the results. One possible reason for this may be the work/rest environment (culture) in Japan. As sick leave is not ruled by law, some workers cannot take sick leave (even if they want to) and do not want to take sick leave when they have minor illnesses9, 10). Therefore, if Japanese workers suffering disease that they could recover from in the short to medium term, such as a cold, it is possible that they would go to work regardless, and this could lead to a lack of increase in short and medium sick leave. Conversely, long working hours (short DRP) and short sleep durations are also associated with several serious illnesses (e.g., cardiovascular disease)4). In this case, worker who suffer a serious illness may take long sick leave.
This study had several limitations. First, DRP, sleep duration, and sick leave were determined drawing on self-reported outcomes. Second, the data were collected using a web survey, which may have introduced sampling bias. Third, participants were self-selected employees, which may result in self-selection bias. Fourth, this cohort survey was conducted during the COVID-19 pandemic. The pandemic altered workers’ daily lifestyles and work environments, which may also influence sick leave in this study. Fifth, this study targeted daytime workers; thus, it is not clear whether similar results would be obtained from shift workers and regular night shift workers. Sixth, this study had a large sample (n=5,593), but few employees had quick return (n=62), and these were divided into two groups, those with (n=38) and those without (n=24) a short sleep duration. The low numbers of the employees in these groups may be the cause of the wide CI. Further prospective cohort studies with larger sample size are needed to confirm the findings.
In spite of these limitations, this study established a joint association of DRP and sleep duration with subsequent sick leave in Japanese daytime employees. Short DRP, regardless of the presence or absence of short sleep duration, was associated with subsequent long sick leave, suggesting that it is important to prevent quick return. Therefore, based on the EU’s working time directive and Japan’s work interval systems, ensuring a minimum DRP (>11 h) may be useful for preventing long sick leave.
Funding
This study was supported by a research grant from the National Institute of Occupational Safety and Health of Japan (N-P03-02).
References
- 1.Vedaa Ø, Harris A, Bjorvatn B, Waage S, Sivertsen B, Tucker P, Pallesen S. (2016) Systematic review of the relationship between quick returns in rotating shift work and health-related outcomes. Ergonomics 59, 1–14. [DOI] [PubMed] [Google Scholar]
- 2.Vedaa Ø, Pallesen S, Waage S, Bjorvatn B, Sivertsen B, Erevik E, Svensen E, Harris A. (2017) Short rest between shift intervals increases the risk of sick leave: a prospective registry study. Occup Environ Med 74, 496–501. [DOI] [PubMed] [Google Scholar]
- 3.Ropponen A, Koskinen A, Puttonen S, Härmä M. (2019) Exposure to working-hour characteristics and short sickness absence in hospital workers: a case-crossover study using objective data. Int J Nurs Stud 91, 14–21. [DOI] [PubMed] [Google Scholar]
- 4.Buysse DJ. (2014) Sleep health: can we define it? Does it matter? Sleep 37, 9–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lallukka T, Kaikkonen R, Härkänen T, Kronholm E, Partonen T, Rahkonen O, Koskinen S. (2014) Sleep and sickness absence: a nationally representative register-based follow-up study. Sleep 37, 1413–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ikeda H, Kubo T, Sasaki T, Liu X, Matsuo T, So R, Matsumoto S, Yamauchi T, Takahashi M. (2018) Cross-sectional Internet-based survey of Japanese permanent daytime workers’ sleep and daily rest periods. J Occup Health 60, 229–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Ikeda H, Kubo T, Izawa S, Nakamura-Taira N, Yoshikawa T, Akamatsu R. (2022) The joint association of daily rest periods and sleep duration with worker health and productivity: a cross-sectional web survey of Japanese daytime workers. Int J Environ Res Public Health 19, 11143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ikeda H, Kubo T, Izawa S, Nakamura-Taira N, Yoshikawa T, Akamatsu R. (2023) The joint association of daily rest periods and sleep duration with occupational stress and sickness absence: a cross-sectional web survey of Japanese daytime workers. J Occup Saf Health 16, 201–8. [Google Scholar]
- 9.Action Medical Research Clinical Reference Center, National Center for Global Health and Medicine Hospital. Antibiotic awareness survey report 2022. https://amr.ncgm.go.jp/pdf/20220930_report_press.pdf. Accessed April 2, 2024 (in Japanese).
- 10.Chimed-Ochir O, Nagata T, Nagata M, Kajiki S, Mori K, Fujino Y. (2019) Potential work time lost due to sickness absence and presence among Japanese workers. J Occup Environ Med 61, 682–8. [DOI] [PubMed] [Google Scholar]
- 11.Izawa S, Nakamura-Taira N, Moriishi C, Yoshikawa T, Akamatsu R, Ikeda H, Kubo T. (2024) Protocol for a web-based study on the work environment and daily lifestyle of Japanese employees. Ind Health 62, 102–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ministry of Internal Affairs and Communications Japan. Population census. https://www.stat.go.jp/data/kokusei/2015/kekka.html. Accessed May 18, 2022 (in Japanese).
- 13.Izawa S, Yoshikawa T, Nakamura-Taira N, Moriishi C, Akamatsu R, Ikeda H, Kubo T. (under submission) Workers’ experiences of improvement in work environment and mental health: a web-based 1-year prospective study for Japanese employee. J Occup Health. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ishizaki M, Kawakami N, Honda R, Nakagawa H, Morikawa Y, Yamada Y, Japan Work Stress and Health Cohort Study Group (2006) Psychosocial work characteristics and sickness absence in Japanese employees. Int Arch Occup Environ Health 79, 640–6. [DOI] [PubMed] [Google Scholar]
- 15.Tsutsumi A, Shimazu A, Eguchi H, Inoue A, Kawakami N. (2018) A Japanese Stress Check Program screening tool predicts employee long-term sickness absence: a prospective study. J Occup Health 60, 55–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Okajima I, Nakajima S, Kobayashi M, Inoue Y. (2013) Development and validation of the Japanese version of the Athens Insomnia Scale. Psychiatry Clin Neurosci 67, 420–5. [DOI] [PubMed] [Google Scholar]
- 17.Hirshkowitz M, Whiton K, Albert SM, Alessi C, Bruni O, DonCarlos L, Hazen N, Herman J, Katz ES, Kheirandish-Gozal L, Neubauer DN, O’Donnell AE, Ohayon M, Peever J, Rawding R, Sachdeva RC, Setters B, Vitiello MV, Ware JC, Adams Hillard PJA. (2015) National Sleep Foundation’s sleep time duration recommendations: methodology and results summary. Sleep Health 1, 40–3. [DOI] [PubMed] [Google Scholar]
- 18.Ministry of Health, Labour and Welfare. Survey on state of employees’ health. https://www.mhlw.go.jp/toukei/list/h24-46–50.html. Accessed April 14, 2022 (in Japanese).
- 19.Prather AA, Janicki-Deverts D, Hall MH, Cohen S. (2015) Behaviorally assessed sleep and susceptibility to the common cold. Sleep 38, 1353–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wuyts J, De Valck E, Vandekerckhove M, Pattyn N, Bulckaert A, Berckmans D, Haex B, Verbraecken J, Cluydts R. (2012) The influence of pre-sleep cognitive arousal on sleep onset processes. Int J Psychophysiol 83, 8–15. [DOI] [PubMed] [Google Scholar]
- 21.Takahashi M, Arito H. (1994) Suppression of electroencephalogram delta power density during non-rapid eye movement sleep as a result of a prolonged cognitive task prior to sleep onset. Eur J Appl Physiol Occup Physiol 68, 274–80. [DOI] [PubMed] [Google Scholar]
- 22.Izawa S, Nakamura-Taira N, Yoshikawa T, Akamatsu R, Ikeda H, Kubo T. (2022) Conversation time and mental health during the COVID-19 pandemic: a web-based cross-sectional survey of Japanese employees. J Occup Health 64, e12334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Pearce E, Birken M, Pais S, Tamworth M, Ng Y, Wang J, Chipp B, Crane E, Schlief M, Yang J, Stamos A, Cheng LK, Condon M, Lloyd-Evans B, Kirkbride JB, Osborn D, Pitman A, Johnson S. (2023) Associations between constructs related to social relationships and mental health conditions and symptoms: an umbrella review. BMC Psychiatry 23, 652. [DOI] [PMC free article] [PubMed] [Google Scholar]