Insomnia, insufficient sleep and delayed sleep-wake phase in university students
The present study revealed the current statuses of university students with regard to insomnia, insufficient sleep, and delayed sleep-wake phase. In a previous systematic review on insomnia in university students, the prevalence of insomnia ranged from 9.4 to 38.2% [35,36,37]. The prevalence of insomnia among the university students included in the present study fell within this range. In the present study, 24.9, 2.7, 1.9, and 4.4% of students were categorized as having insomnia, BIISS (without DSWP), DSWP (without BIISS), and BIISS + DSWP, respectively. In contrast to the results of previous studies in which short sleep was judged only by the cut-off value of sleep duration (e.g., < 6 h) [1, 4, 8], we defined BIISS by including the following criteria; presence of daytime sleepiness and oversleeping on weekends. Likewise, DSWP was defined by assessing bedtime and MEQ score. Sleep duration and delayed sleep phase are correlated [38]; however, a number of previous studies that evaluated the impact of late chronotypes did not consider sleep duration [3, 39, 40]. Therefore, we defined BIISS, DSWP, and BIISS + DSWP more strictly by following the aforementioned criteria in order to investigate the actual impact of insufficient sleep and delayed sleep-wake phase. As a result, the prevalence of BIISS was less than 10% in the present study; however, about half of the total participants were suspected to have insufficient sleep because 38% of them reported that their sleep duration was < 6 h on weekdays and half of them extended their sleep duration by 2 h on weekends and/or showed ESS score ≥ 11. Similarly, the actual prevalence of delayed sleep-wake phase could also be higher than that reported in the present study (6%) because 16% of the students were categorized as evening-type individuals. The sleep duration of students in our study was short; this is in line with the results of previous studies, which reported that the sleep duration of Japanese university students was shorter than those of university students in other countries [1]. In the present study, the prevalence of evening-type individuals (16%) was lower than that of previous studies, which reported results that ranged from 20 to 35% [41,42,43].
Characteristics of each sleep problems
The present study provided the characteristics of the aforementioned sleep problems considering not only participants’ sociodemographic information or lifestyle but also their sleep-related symptoms. Although their sleep-related symptoms were not confirmed by face to face interview or validated screening scale, this is the first study to investigate these symptoms and their associations with sleep problems.
The prevalence of REM related symptoms such as hypnagogic hallucination and sleep paralysis were 19 and 11%, respectively. In the general population, the prevalence of hypnagogic hallucination was 31% in the 15–44 age group and the symptom was associated with mental health disorders including anxiety disorders, depression disorders, and short sleep duration [44]. In contrast, the prevalence of sleep paralysis was 2% in the 15–24 age group [45], and 13% in university students [46] and the symptom was also associated with mental/physical disease, non-restorative sleep, nocturnal leg cramps, and nightmares [45, 47]. In the present study, the prevalence of NREM/REM parasomnia was 17 and 50%, respectively. NREM/REM parasomnia are likely to be associated with insufficient sleep and a delayed sleep phase. In the case of REM parasomnia, nightmares are induced by stressful or traumatic events and are linked to insomnia [48]. Taking these results into account, these sleep-related symptoms may cause sleep problems or mental/physical dysfunction in the young adult population.
Hypnagogic hallucination (27%), sleep paralysis (15%), and REM/NREM parasomnia (57%/24%) other than insomnia symptoms (i.e., difficulty initiating sleep, difficulty maintaining sleep, early morning awakening) were more prevalent in the insomnia group than in the students without insomnia. This is consistent with the results of previous studies, which reported that hypnagogic hallucination was most frequent in individuals who had difficulty initiating sleep [49]. In the insomnia group, the percentage of the students who took a nap for over an hour was significantly higher than that in the students without insomnia. This may be as a result of unsatisfactory nocturnal sleep due to insomnia. Contrarily, taking a long nap during the day may disturb the sleep-wake rhythm and evoke hypnagogic hallucination. In a previous study, forced two-hour nap in the early evening shortened REM latency and worsened sleep efficiency following nocturnal sleep in young adults [50]. Generally, a short nap has a positive effect on daytime performance and reduces fatigue; therefore, appropriate timing and length of naps should be considered [51].
In the BIISS group, the prevalence of sleep-related symptoms did not differ from that recorded in the students without BIISS, despite the longer commute and earlier class start times. Interestingly, all students with BIISS were female who did not live alone. Thus, for these university students, it was suggested that insufficient sleep might be affected by external environmental factors.
Difficulty awakening (80%) and NREM parasomnia (35%) were more prevalent in the DSWP group than in the students without DSWP. The high prevalence of difficulty awakening in this group was reasonable because the students with DSWP needed to wake up earlier than their internal clock time. In contrast to the BIISS group, commute time was shorter and school start time later in the DSWP group than in the group without DSWP. Students with short commutes and late school start times were generally on time for school even though they wake up relatively later due to their age-dependent eveningness. Furthermore, it is possible that students with DSWP take late start classes as a coping strategy; that is, these two school lifestyles can be either cause DSWP or be an effect of it.
In the BIISS + DSWP group, difficulty initiating sleep (39%), difficulty awakening (74%), and hypnagogic hallucination (35%) were more prevalent than in the students without BIISS + DSWP. In contrast to the students with only DSWP who could have a sufficient length of nocturnal sleep, the students with BIISS + DSWP would be forced to get up early. Moreover, their bedtime was quite late (2:20 ± 1:01) and one fourth of them had difficulty initiating sleep. Given this, the BIISS + DSWP group included the students who have an evening-type chronotype due to their delayed internal clock but could wake up early in order to adjust to their social schedule. Therefore, students with BIISS + DSWP, most of whom are thought to be unaware of their delayed intrinsic circadian rhythm, may have profoundly worsened health in the long run.
The association between sleep problems and daytime dysfunction in school life
Delayed sleep-wake phase (DSWP and BIISS + DSWP) were associated with all instances of daytime school life dysfunction, i.e., absence, tardiness, falling asleep during class, and interference with academic achievement, whereas BIISS was associated with only tardiness and interference with academic achievement. Compared to DSWP, BIISS + DSWP showed higher aOR for absence and falling asleep during class. In students with BIISS + DSWP, most are thought to suffer from serious insufficient sleep as a result of being forced to get up early, despite their delayed internal clock. Thus, BIISS + DSWP is considered a relatively more serious condition compared to DSWP. Absence and tardiness may be caused by difficulty awakening in the morning. Absence is considered to be a more serious condition than tardiness, which was negatively associated with being female (aOR: 0.6). This finding is consistent with the previous report stating that men are more likely to show biological eveningness [2]. In this study, a quarter of the students internal clocks were considered delayed due to their difficulty in initiating sleep, many of whom may be unaware of their delay. Therefore, in order to prevent the acceleration of this phase delay, it is important that university students learn about age-dependent sleep problems i.e. delayed sleep-wake phase and resulting insufficient sleep, and to avoid staying up late and/or extending their sleep duration on weekends. In this present study, results showed that over 80% of the students had experienced falling asleep during class. Insomnia (aOR: 1.6), DSWP (aOR: 2.6) and BIISS + DSWP (aOR: 7.6) were all associated with falling asleep during class. The BIISS + DSWP group reported severe insufficient sleep i.e. 4.3 h on weekdays, and only 4.4 h on weekends, with a delayed sleep-wake phase i.e. 2:20 am weekday bedtime. Key aspects to improve university student daytime function are thus maintaining a regular sleep-wake cycle and extending sleep duration. Students who cannot improve daytime sleepiness and difficulty awakening even after these attempts should be examined for sleep disorders, such as circadian rhythm disorders, narcolepsy, sleep apnea syndrome, or restless legs syndrome. Association was confirmed between interference with academic achievement and BIISS (aOR: 1.9), DSWP (aOR: 2.6), and BIISS + DSWP (aOR: 2.1). This result coincided with the findings of previous studies that reported a similar association between sleep problems and worsened academic performance e.g. grade point or ranking [8, 9, 12].
Some previous reports emphasized that delay of class start time improved academic outcomes in adolescents because of the consequent adjustment in their internal clocks [52, 53]. In contrast, other studies reported that late class start times possibly lead to alcohol consumption and may have a negative influence on academic performance in university students [54]. Unlike adolescents (junior/senior high school students), university students can engage in activities such as alcohol consumption and working a part time job in the night, freely and without supervision. In this study, students who displayed alcohol consumption and smoking habits were more prevalent in the BIISS + DSWP and DSWP groups respectively. Although alcohol consumption and smoking are prohibited under the age of 20 by law in Japan, these students were not excluded from this study. Thus, the lower percentage of students in this study whom consume alcohol and smoke compared to other countries may thus be due to this law. Therefore, although we cannot conclude the association and causal relationship from this cross-sectional study without adjusting for confounding factors, a delayed sleep-wake phase might be influenced not only by students’ internal clock, but also by their lifestyle habits. Further studies should clarify whether BIISS + DSWP is associated with alcohol consumption itself or with the drinking habits are necessary. Furthermore, having a part-time job was associated with absence, tardiness, falling asleep during class, and interference with academic achievement, independent of sleep problems. Further studies should investigate the reasons for which students engage in part-time work and the detailed schedules of students with part-time jobs.
In this study, sleep medication was associated with absence (aOR: 5.0) and interference with academic achievement (aOR: 2.0). It was unclear why participants took sleep medication. However, it can be assumed that students with worse physical/mental health status might be prescribed sleep medication at the hospital, as indicated by the associations between poor health and tardiness or interference with academic achievement, and between stress and interference with academic achievement. However, this cross-sectional study could not draw causal inferences due to the small number (n = 8) of students on sleep medication. A longitudinal study is thus needed in order to better understand the association between daytime dysfunction in school life and physical/mental health including sleep medication use.
Study limitations
This study had some limitations. First, differential diagnosis of sleep problems / sleep-related symptoms could not be done in this questionnaire-based study. Second, universities were not randomly selected, because survey data was collected from the universities where the research collaborators were able to directly contact the ethics committee. Therefore, our study sample may not be representative of Japanese university students and sampling bias might exist especially since the percentage of female students was high; this was because women’s university and faculties (e.g., nursing) participated in this study. Eveningness is more severe in males [2]; therefore, the impact of sleep problem on school life may be stronger in male students. Third, the causal relationship between sleep problems and daytime function in school life could not be ascertained in this cross-sectional study. Further studies on prospective sleep evaluation using objective sleep parameters are needed. In addition, although the questionnaire inquired about sleep-related symptoms and daytime function in school life, due to the lack of a time-frame measurement, misleading associations could have been found. Fourth, sleep-related symptoms could not be diagnosed using the validated questionnaires. Finally, the potential impact of covariates that were not included in our analyses should be considered, because various social or lifestyle factors may influence both sleep problems and daytime dysfunction in school life.