NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) and Institute of Medicine (US) Forum on Adolescence; Graham MG, editor. Sleep Needs, Patterns, and Difficulties of Adolescents: Summary of a Workshop. Washington (DC): National Academies Press (US); 2000.

Cover of Sleep Needs, Patterns, and Difficulties of Adolescents

Sleep Needs, Patterns, and Difficulties of Adolescents: Summary of a Workshop.

Show details


Workshop participants heard from a panel of researchers who reviewed findings from the United States and abroad on sleep patterns and problems in adolescents. They discussed findings indicating that the factors contributing to teenagers' sleep loss lie in both the biological and the social realms.

Mary A. Carskadon, director of the E. P. Bradley Hospital Sleep Research Laboratory and professor in the Department of Psychiatry and Human Behavior, Brown University School of Medicine, noted several major trends in adolescent sleep patterns. Data from cross-sectional surveys of students show that, from ages 10 to 17, students' self-reported bedtimes become later and later, on both weekdays and weekends (Carskadon, 1990; Wolfson and Carskadon, 1998). In middle adolescence, rising times become earlier during the week, due largely to school starting times. High school starting times, which typically are earlier than those of middle and elementary schools, have moved to even earlier hours in recent years. Many begin at or before 7:30 a.m., largely due to the timing and availability of school buses. Thus, while sleep needs remain unchanged, Carskadon said, adolescents are spending less time sleeping, and alterations in sleep schedules during the week compared with those on the weekend are becoming more pronounced. This is in sharp contrast to the stable pattern of sleep found in younger children, who get the same amount of sleep during the week as on weekends—an average of 10 hours a night, Carskadon noted.

The effects of restricted sleep on sleep structure, mood, and performance in children and young people have been evaluated under different conditions (Carskadon and Dement, 1981). In a longitudinal study of sleep and sleepiness in young people, researchers assessed children in a summer “sleep camp” laboratory at Stanford University (Carskadon, 1982). Researchers began studying the children when they were 10 to 12 years of age and followed them every summer for 4 to 6 years. Researchers measured their sleep according to the Multiple Sleep Latency Test (MSLT), a standard measure of sleepiness; the test is administered at designated periods throughout the day to determine the time it takes subjects to fall asleep (Carskadon et al., 1986).

In the laboratory the young study subjects wore electrodes that gauged their physiological reactions in sleep and while they were awake. Each night they had the same 10-hour window of time available for sleep, with sleep latency—the time it takes to fall asleep—tested throughout the day at 2-hour intervals. Starting with the hypothesis that the amount of sleep needed would decrease with age to a typical adult 7.5 hours a night, the study assessed the youngsters at various stages of pubertal development to shed light on the issue of sleep needs.

The results showed that the younger children slept 9 hours and 20 minutes on average and awoke spontaneously. As they progressed through adolescence, they continued to get the same amount of sleep, but they no longer woke spontaneously before the end of the sleep window. At midpuberty, adolescents also became sleepier during the day. According to the MSLT, prepubertal and early adolescents were unable to fall asleep in the daytime, but at midpuberty, even with 9 hours and 20 minutes of sleep, daytime drowsiness appeared and worsened. These older adolescents struggled to stay awake throughout the day, whereas the younger adolescents had no problem at all.

A sleep habits survey administered to more than 3,000 Rhode Island 9th to 12th graders revealed that the median amount of reported sleep in this group was 7.5 hours (Wolfson and Carskadon, 1998). A quarter of these students reported sleeping 6.5 hours or less. For two-thirds of the students, bedtime was after 11 p.m. on school nights; 91 percent rose at 6:30 a.m. or earlier. Seventy percent of the teenagers delayed both bedtime and wake-up time by an hour or more on weekends to try to catch up on their sleep. Sleeping late on Saturday and Sunday, however, usually fosters a later sleep onset on Sunday night. Despite this, sleeping in on weekends allows adolescents to pay back some of their weeknight sleep debt, some workshop participants observed.

Influences on Adolescent Sleep Patterns

The various factors that influence how much adolescents sleep cluster into two major areas. One is intrinsic—the biological processes going on internally in adolescents; the other is the external factors—social, academic, and environmental—that play a significant role in their sleep habits.

Intrinsic Factors

Internal processes themselves fall into two types. One is the biological timing system—the circadian rhythms of approximately 24-hour intervals that influence when and how much we sleep. The second is the internal system that tallies the balance of sleeping and waking—the sleep/wake homeostasis system: when sleep is deprived, more sleep is needed. Thus, as discussed at the workshop, “sufficient” sleep can be defined as the amount that satisfies the homeostatic process and is not associated with daytime sleepiness. This is analogous to the daily caloric requirement to maintain a stable weight.

Research findings suggest that changes occur in the “biological clock” during adolescence. As a result, teenagers have a natural tendency to fall asleep later and to wake up later. This is referred to as sleep phase delay. Carskadon described research on college students that restricted their sleep to 5 hours a night for several nights. This study found that daytime sleepiness increased with each night of restricted sleep, indicating the cumulative effect of sleep loss. The research also showed that even with restricted sleep students felt more alert in the evening, encouraging the tendency to stay up late again (Carskadon and Dement, 1981). If additional tests of sleep latency are carried out at 8 and 10 p.m., a student who struggled and dozed through the early afternoon becomes energetic and internally stimulated in the evening, often past midnight.

Another study that looked at the effects of the biological clock did so by examining melatonin secretion. As night falls, melatonin is “turned on,” preparing the body for sleep. Toward dawn, it shuts off, as cortisol secretion increases. Carskadon discussed research on 10 adolescents (five boys and five girls; mean age of 13.7) who were put on a fixed sleeping schedule for 10 days at home. Their schedules were checked by sleep logs, telephone calls, and wrist actigraphy (a device worn to measure daily activity levels). They then were assessed in a laboratory setting on a 28-hour schedule that controlled for all environmental and psychosocial influences on sleep (e.g., lights, television, radio). A correlation was found between subjects' melatonin secretion and their stage of development. The results indicated that melatonin onset occurs later in adolescents, making it difficult for them to go to sleep earlier at night. At the same time, the hormone “turns off” later in the morning, making it harder for them to wake up early (Carskadon et al., 1998, 1999).

At the workshop, Carskadon said that more research is needed to determine whether the apparent changes in melatonin secretion found in this study are a primary intrinsic phenomenon. Comparisons with adults and adolescents under the same conditions are required for more definitive examination. “While it may not be an immutable biological process,” Carskadon told workshop participants, “it sets the stage for other psychosocial and environmental conditions that make it easier for these adolescents to stay awake.” Adolescent development in general, she added, is “a handshake” between biology and behavior, not just one or the other.

The circadian system is governed by the 24-hour alteration of light and darkness. These findings in a laboratory setting in which light and darkness are controlled suggest that the circadian system can be reset with controlled light exposure. At the same time, Carskadon noted that it may take less light to affect this system. An interesting question is whether teenagers' sensitivity to the ambient light they are exposed to in the evening—including from television, computers, and video games—might contribute to this evening arousal (Minors et al., 1991).

Other participants noted that it is not just sleep loss that is troublesome in adolescents but also the enormous variation in their weekday/weekend sleep patterns. While some argued strongly that allowing students to sleep in on weekends was essential for reducing their sleep debt, others pointed out problems: a youngster who gets up at 6 a.m. on weekdays and then sleeps until noon or later on weekends is experiencing “a Washington to Hawaii time zone change twice a week,” said Richard Ferber, a workshop discussant. In effect, the body is in a physiologically wrong time zone.

Other studies have examined teens under the conditions in which they normally function. A field study of adolescents with an early school starting time (7:20 a.m.) showed that many had an elevated rate of REM sleep 1 and fell asleep within 5 minutes during morning MSLT tests (Carskadon et al., 1998). “For these kids,” Carskadon said, “biological night is 8:30 a.m., when they are in second-period class.” The challenges of engaging such youngsters intellectually when they are in the trough of their circadian rhythms was vividly explored by Catherine Colglazier, a teacher in Virginia (see Box 1). In addition to the obstacles to learning, for those who drive to school the increased risk of accidents because of drowsiness is a serious concern.

Box Icon


A Teacher's Perspective. Fairfax County teacher Catherine Colglazier described a typical day at her McLean, Virginia, high school, where the doors open at 7:15 a.m. Students come in either fired up on caffeine or straggly and sleepy eyed. Because McLean (more...)

External Influences

While an inherent phase delay may make it difficult for teenagers to go to bed early at night, other factors clearly play a significant role in the amount of sleep they get. During adolescence, social obligations and opportunities increase, academic requirements become more demanding, and opportunities for work expand. Young people themselves often point to homework as a contributing factor; however, many adolescents actually spend little time on academic pursuits, Carskadon said. Those on the academic fast track, however, do devote more time to school work. Many adolescents are involved in extracurricular activities for many hours a week. For some young people in team sports, this may involve 20 hours or more a week. Many coaches help students plan their sleep, but some youngsters report that practice requirements are a significant factor in their not getting enough sleep, according to Carskadon.

All these things take place in an environment in which television, computers, telephones, video games, and socializing with friends are widely available to most young people, often without parental monitoring and regulation of time spent on these activities. Also, the majority of adolescents have part-time jobs, and many work more than 20 hours a week. Beyond 20 hours is considered to be the point at which working becomes problematic for kids going to school (National Research Council and Institute of Medicine, 1998). A 1994 survey of Rhode Island high school students found that 40 percent of 9th through 12th graders worked on average 20 hours a week (Wolfson, in press).

Working is understood to be an important adult role, Jeylan Mortimer, professor, Department of Sociology, University of Minnesota, observed at the workshop. Studies show that all adolescents, both boys and girls, expect to work during a good part of their adulthood (Mortimer et al., 1999). As noted earlier, adolescence is a time when young people are projecting themselves into the future, and some believe that working encourages “planful competence,” a capacity to think about opportunities, potentials, and interests and to plan for desirable outcomes in the future. Working also fosters work readiness—the importance of getting to work on time, proper attire and behavior in the work environment, and so on. Parents are very positive about their children's jobs, reporting that their children become more capable in managing their time and money and in developing social skills and other benefits (Mortimer et al., 1999). They recall the place of work in their own lives in encouraging these same positive traits.

There is, however, a case against adolescent work. It involves concerns that working too much draws young people away from school, reducing the amount of time available for homework and families. There is also growing evidence that working more than 20 hours a week during the school year is associated with a range of problems, including poor academic performance, use of alcohol and other drugs, and risk of involvement in sexual activity and delinquent behaviors (National Research Council and Institute of Medicine, 1998).

Working too much also impinges on the amount of sleep young people get. In the Rhode Island study noted above, analysis of sleep to work time revealed that, for every 10 hours worked, students lost 14 minutes of sleep per night. A student who works 20 hours a week loses approximately 3 hours of sleep per week. In the 5 percent of the sample who worked full time, students lost an hour of sleep per night or 7 hours weekly (Wolfson, in press).

International Comparisons

How do young people in the United States compare with those in other countries as far as sleep is concerned? Amy R. Wolfson, associate professor of psychology at Holy Cross College, presented findings from a number of studies that reveal both similarities and differences. It is important to note, she said, that a comprehensive dataset using similar measurement tools across countries does not exist.

Challenges to International Data Comparisons

A number of issues make comparisons difficult across countries. Definitions of key terms, for example, vary greatly. In surveys done outside the United States, researchers discussing sleep latency—the time it takes to fall asleep—generally describe it as insomnia. It is not clear whether this corresponds to what in this country would result in a diagnosis of insomnia. Rather, it may refer to delaying bedtimes as opposed to genuine difficulty falling asleep.

It is also difficult to discern in studies conducted abroad whether total sleep time is based on an average across both weeknights and weekends. Obviously, what is considered a weekend varies. For example, in Israel the weekend is Friday afternoon through Saturday evening, with students returning to school on Sunday. In addition, minimal information exists regarding school schedules, so the research presented at the workshop reflected only what little could be determined about school start times in other countries.

Research conducted in Amsterdam looked at 1,500 12- to 18-year-olds with school starting times from 8 to 9 a.m. (Hofman and Steenhof, 1998). A second study, in Brazil, looked at a small sample of 12- to 16-year-olds with school starting times of 7:20 a.m. (Andrade et al., 1993). A third study, in Taiwan, assessed more than 900 13- to 15-year-olds (Gau and Soong, 1995); because Taiwan categorizes students academically very differently than does the United States, drawing comparisons with this research is particularly difficult. The total sleep time for those enrolled in this study who were in a more intensive academic track was an average of 7 hours per school night; those in the less intensive track slept about another half hour each night.

A self-report study surveyed students in Austria, Belgium, Hungary, Israel, Norway, Scotland, Spain, Sweden, Switzerland, and Wales (Tynjala et al., 1993). Sample sizes in the different countries ranged from 60 to more than 3,000 students ages 12 to 18. Students in Switzerland reported the most sleep, about 9.2 hours a night (summarized over the week) for 15-year-olds. Israeli and Finnish students reported the shortest total sleep time—between 8.2 and 8.5 hours in the 15 to 16 age bracket. Students also slept more on weekends.

Bedtimes varied generally by about an hour in most countries. Hungarian and Swiss teens reported the earliest bedtimes, on average before 10 p.m. In Spain even the younger students don't go to bed before 10:30 p.m. School starting times are between 7:00 and 7:30 a.m. in Brazil; 8 a.m. in Israel; and between 8 and 9 a.m. in Amsterdam (Holland), Finland, Norway, Great Britain, and Spain. Researchers who looked at European data for a World Health Organization study of health behavior noted that teenagers slept longer in countries in which parental control seemed to be more strict (Tynjala et al., 1993). This observation suggests the need for more research on parental involvement in and influence on teens' sleep schedules in this country and elsewhere.

Research in France studied more than 700 15- to 23-year-olds, focusing on the quality of sleep rather than sleep patterns: 41 percent had at least one sleep problem, and that tended to be the need for more sleep and difficulty waking up in the morning—very similar to U.S. findings (Vignau et al., 1997). In this study sample, researchers indicated that psychological distress was highly correlated with sleep problems.

In summary, available findings from other developed countries appear to parallel U.S. data on adolescent sleep, Wolfson said. Youngsters average 7.3 hours of sleep on school nights and 9.2 hours on weekends. The data also show that sleep time has decreased over the past 10 to 20 years.

Culture and Context

Wolfson emphasized that more detailed, controlled study of teenagers' sleep patterns in other industrialized countries could provide important insights into how students in the United States compare to their counterparts elsewhere. She stressed the need for such research to focus attention on the culture and context in which young people are growing and working. In the United States, for example, teenagers work more hours than their counterparts in other countries, as noted above. In other countries as well as the United States, cross-cultural analysis and documentation are clearly needed. Other factors that should be examined are parental roles in regulating adolescent sleep patterns both individually and as part of particular cultural contexts.



REM sleep is one of two basic sleep states. This type of sleep is indicated by rapid eye movements similar to those that occur in wakefulness and characteristic brain wave patterns. More REM sleep occurs toward the end of the night.

Copyright 2000 by the National Academy of Sciences. All rights reserved.
Bookshelf ID: NBK222804


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (335K)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...