Sufficient sleep is essential for the physical growth, emotional stability, and maintenance of cognitive function in adolescence1). Excessive daytime sleepiness attributable to sleep deprivation or a sleep disorder is known to reduce work efficiency and to cause traffic and industrial accidents2). Chronic sleep deprivation among adolescents inhibits pre-frontal lobe functions, such as working memory, judgment, and insight, resulting in impairment of learning and school performance. The increased emotional lability and depression associated with sleep deprivation also makes it difficult for sleep-deprived students to adjust to school life3-12).

Modern-day adolescents get much less sleep than was the case in the 20^th century. In particular, excessive use of the internet and other media has reduced sleep time13, 14). It was noted that the average sleep duration of adolescents was 7.6-8.6 hours, 0.4-1.4 hours less than what is needed5). Adolescents in higher grades usually get even less nighttime sleep and experience more marked differences when sleep duration on weekdays and weekends is compared10, 15-17). In Japan, where the educational environment is similar to that of Korea, sleep duration was 7.2-7.8 hours among 10-12^th graders, thus shorter than that of adolescents in the West18).

Korean parents place great emphasis on education, and adolescents often take extra classes or private lessons, contributing to chronic sleep deprivation. In particular, 11-12^th graders in Korea have an average nighttime sleep duration of 4.9-5.5 hours, much less than students in Japan, where parents also place great emphasis on education19-21). This severe chronic nighttime sleep deprivation can lead to excessive daytime sleepiness and lack of attention in class, an increase in emotional lability and depression, and may also be associated with increases in violence and suicide among adolescents22). A previous report indicated that about 70% of Korean adolescents were worried about nighttime sleep quality and over 20% complained of excessive daytime sleepiness. However, the daytime sleep patterns and sleepiness of Korean adolescents have never been studied using an age-specific measure of sleepiness, such as the well-known pediatric daytime sleepiness scale (PDSS)23).

In the present study, we examined the sleep patterns and daytime sleepiness of Korean adolescents using a Korean version of the PDSS and studied the relationship between PDSS data, and adjustment to school life and academic achievement.

We administered sleep questionnaires to 3,379 students in grades 5-12, in their classrooms, in schools situated in the southern region of Gyeonggi-do and Seoul (Korea), from May to November 2009. A total of 3,201 students successfully completed the survey (5-6^th grade: 532, 7-9^th grade: 302, 10-12^th grade: 2545). The male/female ratio was about equal for students in grades 5-9. However, more boys than girls were in grades 10-12 (79.6% boys; 20.4% girls) (Table 1).

The entire sleep questionnaire contained 27 items, and included four questions on sleep/wake patterns, two questions on time spent at school, two questions on sleep satisfaction, three questions on abnormal sleep behavior, two questions on school adjustment, one question on performance, three questions on after-school activities (including the taking of private lessons), two questions on emotional state, and eight questions from the Korean translation of the PDSS (Appendix 1). We surveyed sleep patterns on weekdays and weekends during the previous 4 weeks. The sum of total sleep time (sTST) was calculated by addition of average nighttime (nTST) and daytime total sleep time (dTST). The dTST were computed as median values: 15 minutes of dTST for nap of less than 30 minutes a day, 45 minutes for nap between 30 minutes and 1 hour, 1 hour 30 minutes for nap of less than 2 hours, and 2 hours 30 minutes for more than 2 hours of nap. Questions about emotional state, abnormal sleep behavior and school adjustment were assessed using a Likert-scale format. Two items about school adjustment were the questions about participants' happiness and loneliness during school time. School performance was measured by asking students "What's your grade point average in last semester?" The self-reported choices were: 90 or above, 80-89, 70-79, 60-69, and less than 60. Two questions about mood state were how often the responders were calm and got angry with themselves or others over the last 4 weeks. Responses to the questions were "all the time (6-7 days/week)," "frequently (3-5 days/week)," "sometimes (1-2 days/week)," "seldom," or "never."

We also assessed participants' daytime sleepiness using the translated version of PDSS. Items about PDSS were scored from 0 to 4 (never = 0; seldom = 1; sometimes = 2; frequently = 3; always = 4). Total PDSS score (ranging from 0-32) was calculated by adding scores on each of the eight questions; a higher score indicated more sleepiness. Means and standard deviations for the PDSS score across grade level were assessed. Correlation coefficient between nTST, dTST, sTST and PDSS were assessed. ANOVAs were performed to examine group differences such as mood, grade, school adjustment and performance for TST and PDSS. Computing statistical reliability of PDSS, Cronbach's alpha was calculated. All statistical analyses were performed using SPSS version 17.0 (SPSS Inc. Chicago, IL). The study protocol was approved by the institutional review board of the CHA Medical Center.

Our study indicates that the average school day nTST of Korean adolescents is 7.95±1.05 hours for 5-6^th graders, 7.57±1.05 hours for 7-9^th graders, and 5.78±1.13 hours for 10-12^th graders. The school day sTST are 8.15±1.12 hours for 5-6^th graders, 8.17±1.20 hours for 7-9^th graders, and 6.87±1.40 hours for 10-12^th graders, thus making sleep deprivation less severe. Nonetheless, 9 hours of sleep is considered necessary for adolescents, and all students were thus sleep-deprived10, 24). In particular, the nTST of 9-12^th graders was significantly less than that of younger students, but the sTST did not differ significantly between these two groups because of an increase in dTST. However, the sTST of 9-12^th graders remained almost 2 hours less than the recommended sleep time.

In this study most of 7-12^th graders attributed their night time sleep deprivation to use of internet or early school start time, not excessive after-school classes. Unsurprisingly it is suggested that Korean adolescents seem to accept their excessive after-school classes in school or tutoring in institution involuntarily in highly competitive educational environment. For 10-12^th graders arriving time at home is around 9-10 p.m. whereas for 7-9^th graders, 3-4 p.m. is usual arriving time in our study. Moreover, 63.5% of high school students had their private tutoring for 1 to 4 hours after school, and 77.9% of middle-school students in this study received tutoring. Undoubtedly excessive classes or tutoring is a major cause of sleep deprivation, even though excessive use of internet at night is also contributing to sleep deprivation in Korean adolescents.

The average PDSS score of all students was 16.63±5.72, but increased significantly as students became older. As PDSS score rose, emotional lability and abnormal sleep behavior also increased. Moreover, a high PDSS score was negatively correlated with academic achievement with statistical significance. Thus, we conclude that the Korean version of the PDSS accurately assesses both daytime sleepiness and other factors related to academic achievement by Korean adolescents.

In 1981, Carskadon, at Stanford University, commenced detailed studies of sleep deprivation in 12 adolescents5, 24). Previous works had shown that sleep latency (the length of time that it takes to accomplish the transition from full wakefulness to sleep) decreased as sleep deprivation rose, and that the extent of sleepiness could be measured objectively5). Subsequent studies showed that sleep deprivation increased daytime sleepiness, and had negative effects on all of cognition, memory, fine motor control, and mood3, 10, 25, 26). Moreover, if sleep deprivation persisted for a long period, higher brain function was impaired, BMI rose, blood pressure elevated, glucose intolerance increased, and adverse effects on the immune and cardiac system were evident26-29). Adolescents and children have similar physiological sleep requirements, but both the time of entry into and awakening from sleep occur about 2 hours later in adolescents, as a result of the physiological changes associated with puberty. Moreover, the level of daytime sleepiness becomes higher physiologically in adolescents than in children30). In the present study, we found that total sleep time did not differ greatly when elementary- and middle-school students were compared, but that PDSS scores of middle-school students were significantly higher than those of students in elementary school. It is thought that reduced nighttime sleep and changes in the sleep cycle increase daytime sleepiness in adolescents.

Studies in the USA on the relationship between sleep time and academic achievement indicated that adolescents slept for an average of 7.6 hours. However, students in grades 10, 11, and 12 had average school day sleep times of 7.3 hours, 7.0 hours, and 6.9 hours, respectively13). Moreover, insufficient sleep time had an adverse affect on student grades25). Studies on middle-school students in New Zealand found that school day sleep time was about 8.7 hours, close to the recommended amount, but that students in higher grades had less school day sleep and more sleep on weekends, as has also been found in other countries. In India, the sleep time of middle- and high-school students was 7.6-8.0 hours, whereas in Okinawa (Japan), the figure was 7.2-7.9 hours10, 15-18). Previous studies in Korea have found that the average sleep time of high-school students in grades 10-12 was only 4.9-5.5 hours, indicating that many such students were severely sleep-deprived19-21). In the present study, we considered daytime sleep hours when calculating sleep time totals, and found that sleep deprivation was not quite as severe as previously indicated. However, our results do confirm that Korean students are sleep-deprived.

The Multiple Sleep Latency Test (MSLT) and the Maintenance of Wakefulness Test (MWT) both objectively measure daytime sleepiness in adolescents, but these tools can be difficult to use, and are not widely employed in epidemiologic studies5). Simple survey methods are often preferable. The Epworth Sleepiness Scale (ESS) can be used for long-term measurement of sleepiness and the Stanford Sleepiness Scale (SSS) is valuable to assess for short-term daytime sleepiness1, 31-33). The ESS has been employed in previous studies of daytime sleepiness in Korean adolescents. However, this test explores sleepiness during driving, and asks whether sleep-inducing medication was taken; the test is thus more appropriate for adults rather than adolescents21). The PDSS has eight questions, each with five possible answers, and studies in the USA, Argentina, and Taiwan have found that this questionnaire accurately assesses adolescent daytime sleepiness23, 34, 35). In the present study, we determined that the Korean version of the PDSS was appropriate for evaluation of sleepiness in Korean adolescents, and that higher scores were associated with emotional lability and poor school performance. Thus, the PDSS is a simple test that provides accurate results and that can easily be used to test large populations of adolescents or children.

Our study had some limitations. First, the self-reported data on sleep patterns were dependent on subjective memory, and school record data were also imparted by the students. Second, there was a predominance of males in the higher grades, and our results may thus not be applicable to other student populations. Previous studies have found that female students tend to rise earlier and to have longer nighttime sleep than do males6, 19). Our results would have been more accurate had we measured sleep deprivation with actigraphy (which assesses rest/activity cycles by sensor), had we evaluated official school records, and had we tested the students using a psychomotor vigilance task (PVT).

In the present study, using a Korean translation of the PDSS, we found that Korean adolescents were severely sleep-deprived and, as a result, daytime sleepiness was a significant problem associated with excessive napping during the day. Sleep-deprivation problems were more severe in upper-grade students. Moreover, although daytime sleep of older students increased sTST, significant daytime sleepiness was still apparent. Many Korean adolescents appear to be locked into inappropriate split sleep states. This compromises adaptation to everyday life, reduces emotional stability, and worsens school performance. Finally, we showed that the Korean translation of the PDSS accurately assessed sleepiness in Korean adolescents and can validly be used to explore emotional status and school performance.

Korea has an excessively competitive educational system, and the need to take extra classes leads to severe restriction of nighttime sleep during the school week, resulting in significant daytime sleepiness. It seems likely that sleep deprivation is related to inattentiveness, depression, and an elevated suicide rate22). Adolescents require appropriate sleep time to assure normal growth and development; to improve school performance; and to guarantee physical, psychological, and spiritual health. We suggest that the importance of appropriate levels of adolescent sleep be taught at homes, schools, and in society.