This study investigated the prevalence and risk factors of insufficient sleep syndrome (ISS), and factors associated with daytime dysfunction in the disorder in Japanese young adults.
In this cross-sectional study, a web-based questionnaire survey was used to assess demographic variables, sleep habits and quality, depressive symptoms, and health-related quality of life (HRQOL) in 2,276 participants aged 20–25.
Eleven percent of participants were classified as having ISS. Multiple logistic regression analysis revealed that the presence of ISS was significantly associated with social status (student or full-time employee). The participants with ISS had significantly higher depression scores and lower mental component summary scores than healthy sleepers. In the participants with ISS, a delayed sleep-wake schedule was extracted as a factor associated with worse mental component summary.
Results indicate a relatively high proportion of Japanese young adults suffer from ISS, and that the condition is associated with a social status of student or full-time employee. Moreover, a delayed sleep-wake schedule may lead to further deterioration of mental HRQOL in ISS-affected persons.
Morita Y, Sasai-Sakuma T, Asaoka S, Inoue Y. Prevalence and correlates of insufficient sleep syndrome in Japanese young adults: a web-based cross-sectional study. J Clin Sleep Med 2015;11(10):1163–1169.
Previous epidemiological studies have indicated that 7–8 hours of sleep duration is associated with lower risk of physical health problems.1,2 It has also been reported that young adults' cognitive functioning is maintained when sleeping for 8 hours per day compared to 4 or 6 hours per day.3 Thus, a nocturnal sleep duration of 7–8 h is thought to be necessary to maintain good health of young adults. However, many young adults, including university students and incipient workers, have shorter than recommended sleep duration, especially on weekdays; thus, a significant proportion of this population are partially but chronically sleep deprived. In particular, young adults in Asian countries reportedly sleep for fewer hours per night than young in the United States or European countries.4 According to a previous study, around 70% of Japanese young adults in their twenties had a nocturnal sleep duration less than 7 hours, and as many as 36% of them slept less than 6 hours per night.5 Given this, it is thought that many Japanese young adults are chronically sleep deprived.
Adequate sleep is important for the maintenance of physical and mental health in young people. Much evidence suggests that insufficient sleep results in a number of negative effects on health and health-related behaviors in young adults6, including obesity,7–9 elevated daytime sleepiness,5,10,11 depressed mood,4,12 and suicidal ideation.13,14 However, the most pronounced consequence of chronic sleep deprivation is excessive daytime sleepiness. This symptom is likely to be associated with other daytime dysfunctions including depression, anxiety, or fatigue in adolescents.15 Moreover, daytime sleepiness is known to be associated with decline in school performance.16 Considering that many young people have a shorter than ideal sleep duration, they may be at an increased risk for these daytime dysfunctions.
Current Knowledge/Study Rationale: Insufficient sleep syndrome (ISS) in young adults has become a problematic issue in modern Japanese society; however, neither the prevalence nor factors associated with the disorder have been completely examined. Sleep-wake pattern as well as sleep duration is associated with depression, hence it was hypothesized that habitual sleep pattern would have an impact on daytime function in ISS-affected persons.
Study Impact: We demonstrated that 11% of Japanese young adults suffered from ISS and that delayed sleep-wake schedule of them was associated with worse mental health-related quality of life. Sleep hygiene education to prevent not only insufficient sleep but also sleep schedule delay is necessary to maintain daytime functions in young persons.
To date, several previous studies have investigated “sleep deficiency” or “insufficient sleep” among various generations.4,5,17,18 In the 3rd edition of the International Classification of Sleep Disorders (ICSD-3),19 insufficient sleep syndrome (ISS) was classified as one of the central hyposomnolence disorders characterized by longer sleep durations on weekends than on weekdays, as well as shorter sleep times during week-days and daytime sleepiness. To identify the presence/absence of ISS, recent studies have thoroughly investigated sleep schedules on both weekdays and weekends in order to evaluate how differences in sleep duration during the week could be a marker of sleep debt.20–23 ISS is defined by hypersomnia symptoms due to chronic sleep debt resulting from self-induced sleep restriction or wake extension.24 Although people of all ages can develop ISS, this condition is thought to occur frequently in young people because of their increased need for sleep, later bedtime resulting from their lifestyle (e.g., living alone, watching TV, or using internet at midnight),25 and cultural factors. As insufficient sleep is reportedly associated with the development of depression,4 it is important to clarify the prevalence of ISS and factors associated with the condition in young adults. Thus far, only a few studies have investigated ISS prevalence based on diagnostic criteria from the ICSD-3.21,23,26 Furthermore, virtually no study has addressed this issue among post-adolescent samples. Considering that bedtimes and rise times during weekdays are later among university as compared to high-school students,11 ISS prevalence during post-adolescence should be examined.
Moreover, factors associated with daytime dysfunction in ISS-affected persons have not yet been investigated. Recent reports have described that sleep-wake pattern as well as sleep duration is associated with daytime function. For example, the late midpoint of nocturnal sleep, which is defined as an intermediate time between sleep onset time and rise time,27 on school days was associated with depression among adolescents.28 Given this, sleep-wake schedule as well as sleep duration should be considered as factors associated with daytime dysfunction in ISS-affected persons.
The present study was conducted to assess the prevalence of ISS in Japanese young adults to analyze factors associated with the presence of ISS and explore factors associated with daytime dysfunction in ISS-affected persons.
Participants and Procedures
This cross-sectional web-based questionnaire survey was part of the Young Generations' Sleep Health Survey, of which the goal was to investigate the impact of sleep schedule on daytime function in new university graduates.19 Therefore, the target population of the survey was persons aged 19 to 25 years. This survey was administered during a week in February 2012. The ethics committee of the Neuropsychiatric Research Institute reviewed and approved this study. Informed consent was obtained from all participants prior to enrollment. The questionnaire consisted of a series of items designed to assess demographic variables, current and former social status (e.g., worker, student, or other), sleep habits, subjective sleep quality, depressive symptoms, and health-related quality of life (HRQOL).
Of the 3,904 participants who accessed the survey web page, 3,613 participants responded to the questionnaire (92.5%). Exclusion criteria were as follows: (1) excessive body mass index (BMI ≥ 50 [n = 5]); (2) absent (n = 175) or invalid answers to the questionnaire items (e.g., selecting the same number on every question; answers that were inconsistent with those of other questions [n = 48]); (3) use of hypnotics at the time of the investigation (n = 165); (4) shift work (n = 723); and (5) age < 19 years (n = 102). After excluding data from 1,218 participants according to these criteria, data from 2,395 were subjected to subsequent analyses. Data from an additional 119 participants were excluded from analyses because of outlier bedtime, rise time, and sleep duration (≥ mean ± 3 standard deviation [SD]). Thus, data from 2,276 participants were analyzed.
In the questionnaire, the following demographic information was assessed: age, sex, body weight, height, smoking (Do you currently smoke? Yes or No), drinking (Do you currently drink alcohol? Yes or No), family configuration (Please indicate your family configuration – living with families or living alone), residential area (participants selected from a list of 47 prefectures in Japan), social status (Please specify your affiliation in society – student [enrolled in technical college, junior college, university, graduate school, or other school], full-time employee, or other [part-time employee, housewife, or unemployed]).
Participants' habitual sleep pattern and sleep quality during the preceding month were assessed using the Japanese version of the Pittsburgh Sleep Quality Index (PSQI).20 Participants were also asked to answer questions about their usual bedtime and rise time on weekdays and weekends (hours: minutes). Sleep duration was calculated using their usual bedtime, rise time, and sleep latency (time in bed minus sleep latency). Total PSQI score was calculated using sleep durations during week-days, which more accurately reflects a habitual sleep-wake schedule as compared to durations on weekends.
Depressive symptoms and HRQOL were assessed using a 12-item version of the Center for Epidemiologic Studies Depression Scale (CES-D)21 and the Short Form Health Survey (SF-8). The SF-8 is a simpler version of the SF-36, which is useful for evaluating quality of life.22,23 The SF-8 consists of the following domains: vitality, social functioning, mental health, role-emotional, general health, physical functioning, role-physical, and bodily pain. Based on the scores of these items, a mental component summary and physical component summary were calculated, and general population averages for these scores were set at 50 points. Therefore, participants with a score < 50 points were inferred to have worse HRQOL.23
Diagnostic criteria for ISS in the ICSD-3 are as follows: (A) having excessive daytime sleepiness or daytime lapse, (B) having shorter sleep duration than expected for age, (C) shortened sleep time lasting ≥ 3 months, (D) sleeping longer on weekends or vacation, (E) sleepiness resolved by an extension of sleep time, and (F) symptoms not explained by another sleep disorder (e.g., mental or neurologic disorders) or the effects of medication.
Based on the ISS definition according to the ICSD-3 and previous reports on behavioral ISS using the ICSD-2 diagnostic criteria,24–26 participants were defined as having ISS when they met all of the following criteria: (1) usual experience of excessive daytime sleepiness or daytime lapse (Q8 scores on PSQI ≥ 1), which corresponds to criterion A of ICSD-3; (2) sleep duration (defined as time in bed minus habitual sleep latency) on weekdays < 7 h, which corresponds to criterion B; and (3) 2 h longer sleep duration per day on weekends than on weekdays, which corresponds to criterion D.
Participants who slept > 7 h on either weekdays or weekends and did not experience daytime sleepiness, and those who slept < 7 h on weekdays but slept 2 h longer on weekends and did not have daytime sleepiness were classified as “healthy sleepers.” In addition, participants who slept > 7 h on either weekdays or weekends and had daytime sleepiness were classified as “possible hypersomnia.” Participants who slept < 7 h on weekdays, and did not sleep 2 h longer on weekends than on weekdays and had daytime sleepiness were classified as “possible insufficient sleep syndrome (ISS).” Participants who slept < 7 h on weekdays, did not sleep 2 h longer on weekends than on weekdays, and did not have daytime sleepiness were classified as “possible short sleepers.”
We first compared demographics, subjective sleep parameters, CES-D scores, SF-8 scores, and PSQI scores among the 5 subject groups using one-way ANOVA followed by a post hoc Bonferroni test.
Secondly, to examine the factors associated with ISS, logistic regression analyses were conducted with the following independent variables: gender, age, BMI, social status, currently smoking, currently drinking, living alone, and area of residence. Thirdly, logistic regression analyses were conducted to explore factors associated with daytime dysfunction in the subjects with ISS setting “physical component summary scores < 50,” “mental component summary score < 50,” and “CES-D ≥ 12” as dependent variables; and gender, age, BMI, social status, sleep duration on weekdays, weekends over-sleep time and midpoint of nocturnal sleep on weekdays as independent variables. All the dependent variables were initially examined in univariate models, and multivariate logistic regression analysis was carried out for all variables that showed significant associations in univariate models to control for confounding factors and to determine the main correlates. Wald statistics were used to test the significance of the odds ratios generated by the regression analysis. ORs and 95% confidence intervals (CIs) are shown in Table 1.
Factors associated with ISS (n = 2,276).
Factors associated with ISS (n = 2,276).
All analyses were conducted using software (SPSS, V11.5; SPSS Inc., Chicago, IL, USA). Statistical significance was indicated by a p value < 0.05.
Descriptive Information about Study Participants
Participants (n = 2,276) consisted of 1,030 males (45.3%) and 1,246 females (54.7%) aged 23.1 ± 1.6 (mean ± SD) years, with a BMI of 21.1 ± 3.3 kg/m2. Of these, 845 participants (37.1%) were students, and 852 participants (37.4%) were full-time employees. In addition, 218 (9.6%) were currently smoking, 760 (33.4%) participants h were currently drinking, and 755 participants (33.2%) lived alone. The mean sleep duration on weekdays was 6 h 35 min ± 1 h 20 min, whereas on weekends, the duration was 7 h 46 min ± 1 h 28 min. The mean bedtime on weekdays was 0:40 ± 1 h 24 min and 1:17 ± 1 h 32 min on weekdays. The mean rise time on weekdays was 7:40 ± 1 h 34 min and 9:29 ± 1 h 50 min on weekends. The difference in the sleep duration between weekends and weekdays was 1 h 11 min ± 1 h 31 min.
Among all analyzed participants, 1,417 (62.3%) reported a sleep duration < 7 h on weekdays (ISS criterion A), 712 (31.3%) indicated a weekend nocturnal sleep duration ≥ 2 h longer than that of weekdays (ISS criterion B), and 680 (29.9%) experienced daytime sleepiness (ISS criterion C). Consequently, 249 participants (10.9%) met all 3 criteria and were classified as having ISS (Figure 1). Meanwhile, 1,049 participants (46.1%) were classified as healthy sleepers; 547 participants (24.0%), 171 participants (7.5%), and 260 participants (11.4%) were classified as possible short sleepers, possible hypersomnia, and possible ISS, respectively.
ISS, insufficient sleep syndrome.
Participant flowchart.ISS, insufficient sleep syndrome.
Comparisons of Daytime Function and Severity of Sleep Disturbance among the Subject Groups
Table 2 shows the total scores of CES-D, SF-8, and PSQI for the 5 categorized groups. There was significant difference in the total scores of CES-D among the groups (F(4,2271) = 20.0, partial η2 = 0.03, p < 0.001). CES-D score was significantly higher in the groups with ISS, possible hypersomnia, and possible ISS, than in the healthy sleepers (p < 0.001, p = 0.009, p < 0.001) or possible short sleepers (p < 0.0001, p = 0.009, p < 0.001).
Comparison of daytime function and severity of insomnia among the subject groups.
Comparison of daytime function and severity of insomnia among the subject groups.
For the SF-8, a significant difference was found in physical component summary scores and mental component summary scores among the groups (F(4,2271) = 4.54, partial η2 = 0.008, p = 0.008; F(4,2271) = 7.66, partial η2 = 0.013, p < 0.001). Post hoc tests revealed that physical component summary score was significantly lower in the possible hypersomnia group than in the healthy sleepers (p = 0.009). As for mental component summary, the score of the ISS group was significantly lower than that of the healthy sleepers and possible short sleepers (p = 0.002, p < 0.001).
PSQI score differed significantly among groups (F(4,2271) = 109.2, partial η2 = 0.16, p < 0.0001). A post hoc test revealed that groups with ISS, possible hypersomnia, possible short sleepers, and possible ISS had significantly higher PSQI scores than healthy sleepers (p < 0.001, p < 0.001, p = 0.001, p < 0.001). The scores of the groups with ISS, possible hypersomnia, and possible ISS were significantly higher than scores of the short sleeper group (p < 0.001, p < 0.001, p < 0.001). In addition, the groups with ISS and possible ISS showed significantly higher PSQI scores than the possible hypersomnia group (p < 0.001, p = 0.001). No significant difference was found in PSQI score between the groups with ISS and possible ISS (p = 0.129).
Factors Associated with the Presence of ISS
To examine factors associated with ISS, we conducted a series of logistic regression analyses. The following 7 variables were tested in the model: gender, BMI, social status, currently smoking, currently drinking, and living with families. As shown in Table 1, social status (students: OR = 2.13, 95% CI = 1.43–3.18; full-time employees: OR = 2.35, 95% CI = 1.58–3.49) was significantly associated with the presence of ISS in univariate logistic regression analyses. A multivariate model of logistic regression analysis was not carried out because no other independent variables showed a significant association with the presence of ISS.
Factors Associated with Daytime Dysfunction in Subjects with ISS
To examine factors associated with deterioration of HRQOL (mental component summary and physical component summary) and depression, we conducted a series of logistic regression analyses. The following 7 variables were examined in the model: gender, age, BMI, social status, sleep duration on weekdays, difference in sleep duration between weekdays and weekends, and the midpoint of nocturnal sleep on weekdays. As shown in Table 3, all the variables were associated neither with worse physical component summary nor with depression, whereas the midpoint of nocturnal sleep was significantly associated with worse mental component summary in univariate logistic regression analyses (OR = 1.4 4, 95% CI = 1.09–1.91). A multivariate model of logistic regression analysis was not conducted because no other independent variables showed a significant association with worse mental component summary.
Factors associated with daytime dysfunction in subjects with ISS (n = 246).
Factors associated with daytime dysfunction in subjects with ISS (n = 246).
This study revealed the prevalence of ISS, associated factors for the condition, and factors associated with daytime dysfunction in ISS-affected young adults. As mentioned above, the prevalence of ISS has not been extensively investigated.21,23,26 To our knowledge, there have been two studies of BIISS prevalence in adolescents: the first showed a 10.4% prevalence rate in Norwegian adolescents,21 and the second indicated an 18.8% rate in Korean adolescents.23 The present findings indicate that the ISS prevalence rate in Japanese young adults is almost identical to that of Norwegian adolescents. Moreover, the rate seemed to be quite high in this population, although a direct comparison with that of other generations has not yet been made.
Results of the logistic regression analyses revealed that ISS was significantly associated with social status (i.e., ISS was more prevalent in students or full-time employees than part-time employees, housewives, and unemployed persons). Previous findings indicate an inverse relationship between hours spent working and sleep duration.34,35 In the Japanese working population, longer working hours were associated with shorter sleep duration (< 6 h/day) in both men and women.36 Thus, full-time employees may be more susceptible to ISS because they work a greater number of hours and consequently sleep for a shorter period during the night. In addition, students might be at greater risk for ISS because their sleep-wake patterns are likely to be delayed during school days, thus elevating nocturnal preference and ultimately decreasing total nocturnal sleep time.37
In the present study, participants with ISS had a greater incidence of depressive mood and worse HRQOL relative to healthy sleepers. This finding is consistent with several previous reports.23,38–40 For instance, one study reported that adolescents with ISS scored higher on the Beck Depression Inventory than did those without ISS, particularly those who regularly obtained sufficient sleep (≥ 7 h of nocturnal sleep).23 Similarly, a short duration of nocturnal sleep is reportedly associated with decreased physical and mental health.41,42 Furthermore, daytime sleepiness resulting from insufficient sleep has been reported to be associated with the presence of depressive mood in adolescents.43 Thus, the increased rate of depressive mood and poor HRQOL observed in the ISS participants of the present study were possibly due to short sleep duration or its resultant daytime sleepiness. In this regard, depressive mood and mental HRQOL in the subjects with ISS were worse than those of possible short sleepers, although sleep duration on weekdays was shorter in possible short sleepers than in subjects with ISS. Furthermore, the degree of depressive mood and HRQOL in ISS subjects was similar to those of subjects with possible ISS or possible hypersomnia. These results might indicate that the presence of daytime sleepiness plays a major role in the formation of daytime dysfunction in individuals with ISS.
In the present study, we examined factors associated with the deterioration of HRQOL and depressive mood. As a result, the midpoint of sleep on weekdays appeared as a factor associated with the deterioration of the mental component of HRQOL, despite a lack of a significant association between the midpoint of sleep and depressive mood. Considering that a later midpoint of sleep is associated with decreased well-being,44 a delayed sleep-wake schedule might be associated with worse of mental HRQOL or well-being. Furthermore, long-term exposure to a delayed sleep-wake schedule will likely aggravate one's mental health, which could lead to depression.28 Thus, young adults keeping an optimal sleep-wake schedule should be desirable for maintaining adequate mental health. One previous study reported that computer use and general social activities after school among older adolescents were strongly associated with delayed bedtimes.45 Therefore, for purposes of sleep education, activities that likely enhance alertness before bedtime should be avoided.
This study has several limitations. First is that it was a web-based cross-sectional survey in which the majority of participants were frequent Internet users46; thus, a sampling bias might have occurred. A recent study indicated that increased Internet use leads to shorter sleep duration.47 It has also been suggested that shorter sleep duration and Internet use are independently, but additively, associated with multiple adverse health consequences.48 Thus, our sample might not be representative of the general young adult population in Japan, and the 11.2% prevalence rate of ISS observed in the present study might be higher than the actual rate for this population. A second limitation was that this study used the score of Q8 on PSQI as an indicator of daytime sleepiness, which is the core symptom of ISS. In order to more accurately evaluate daytime sleepiness, objective measurements, such as a multiple sleep latency or psychomotor vigilance test, would be useful in future studies. Although sleep duration was estimated based on information obtained from the PSQI, sleep duration was not entirely correct due to a lack of information about wake after sleep onset. Similarly, a clinical interview is typically desirable for making the proper diagnosis of ISS, but we could not conduct clinical interviews in this study. In addition, due to our use of the PSQI, we evaluated subjects' sleep-wake pattern only during the preceding one month, which was shorter than the period required for making a diagnosis of ISS according to the ICSD-3 criteria (three months).
Third, there might have been insufficient sleepers among the subjects who were not included in the ISS group. In fact, 15.8% (361) of subjects classified as healthy sleepers had sleep schedules similar to those of the ISS group, although they did not report subjective daytime sleepiness; thus, it is possible that these persons underestimated their daytime sleepiness. Furthermore, we excluded 723 shift workers, which might include a certain number of individuals with ISS, from the analysis. This was done because shift work schedules in Japan rotate rapidly, resulting in various bedtime patterns; thus, no information regarding shift work schedules within this population could be obtained. Similarly, we excluded people using hypnotics from the analysis since its use may have an influence on sleep-wake patterns or daytime sleepiness. Taking these limitations into consideration, the actual prevalence of ISS among the Japanese young adult population might actually be higher than what was observed in the current study.
Finally, the midpoint could possibly reflect not only subjects' sleep-wake schedules but also their age.49 However, in the present study, all subjects were within a relatively narrow age range. In addition, age was not significantly associated with daytime dysfunction in the logistic regression analyses. Thus, the midpoint of nocturnal sleep during weekdays was inferred to reflect subjects' sleep characteristics, possibly resulting from a social rhythm rather than age.
In conclusion, the prevalence of ISS in Japanese young adults was estimated to be high, and status as a student of full-time employee was associated with the presence of ISS. Persons with ISS might be more susceptible to depressive mood and worse HRQOL. Moreover, a midpoint delay of nocturnal sleep in ISS-affected subjects might lead to dysfunctional of mental health-related quality of life. The present results suggest that young people should obtain an adequate amount of sleep with an appropriate sleep-wake phase.
This was not an industry supported study. Dr. Yuichi Inoue has received research support from the Hisamitsu Pharmaceutical Co., Inc, Nippon Boehringer Ingelheim Co., Ltd., Philips Respironics GK, Alfresa Pharma Corporation, Takeda Pharmaceutical Company Limited, MSD K.K., Pacific Medico Co., Ltd., Otsuka Pharmaceutical Co., Ltd., Eisai Co., Ltd., Mitsubishi Tanabe Pharma Corporation, GlaxoSmithKline K.K., Astellas Pharma Inc., Sanofi-aventis K.K., and Yoshitomiyakuhin Corporation. The other authors have indicated no financial conflicts of interest.