Pathways to Adolescent Health: Sleep Regulation and Behavior PDF

JOURNAL OF ADOLESCENT HEALTH 2002;31:175–184 SUPPLEMENT ARTICLE Pathways to Adolescent Health: Sleep Regulation and Behavior RONALD E. DAHL, M.D. AND DANIEL S. LEWIN, Ph.D. Abstract: There are several converging reasons to focus among some high schools © Society for Adolescent...

JOURNAL OF ADOLESCENT HEALTH 2002;31:175–184 SUPPLEMENT ARTICLE Pathways to Adolescent Health: Sleep Regulation and Behavior RONALD E. DAHL, M.D. AND DANIEL S. LEWIN, Ph.D. Abstract: There are several converging reasons to focus among some high schools © Society for Adolescent Med- on sleep regulation in relation to healthy adolescent icine, 2002 development: (a) Sleep appears to be particularly impor- tant during periods of brain maturation; (b) there are KEY WORDS: substantial biological and psychosocial changes in sleep and circadian regulation exist across pubertal develop- Adolescence ment; (c) interactions between physical and psychosocial Affect regulation Sleep domains can lead to dramatic alterations in sleep patterns and habits during adolescence; (d) increasing evidence that many adolescents frequently obtain insufficient sleep exists; (e) there is mounting evidence that sleep This paper primarily address aspects of sleep regu- deprivation has its greatest negative effects on the con- lation in adolescent development at the interface trol of behavior, emotion, and attention, a regulatory between biological and psychosocial domains. How- interface that is critical in the development of social and ever, the main points require some understanding of academic competence, and psychiatric disorders; (f) the sleep physiology and some familiarity with terminol- most obvious direct health consequences of insufficient ogy regarding sleep stages. Therefore, the first sec- sleep are high-risk behaviors associated with substance tion provides a general overview of sleep, its physi- abuse and automobile accidents; (g) substantial evidence ology and development, as background and context for bidirectional effects between sleep and behavioral/ for specific discussions of sleep changes in adoles- emotional regulation exists. Although the past decade has seen research progress in these areas, there continue cence in later sections. to be major gaps in existing knowledge and a paucity of well-controlled studies to guide specific health policy decisions and recommendations regarding sleep in ado- Sleep: An Overview lescence. In particular, there is need for improved under- Sleep is not simply rest. Sleep is an active process standing of the acute and chronic effects of inadequate wherein some brain regions show the same (or sleep in adolescents, guidelines for defining adequate increased) activity as during wakefulness. Moreover, sleep in adolescents, and a better delineation of the links there are several aspects of sleep including the con- among sleep, behavior, and affect regulation. Finally, tinuity, timing, and patterning of different stages of this paper briefly examines one specific application of sleep that are necessary for the restorative process to this knowledge area regarding early starting times occur. For example, if subjects are permitted to sleep ad libitum for 8 to10 hours but are awakened every 15 minutes for brief periods, the following day they will From the Western Psychiatric Institute and Clinic, University of Pittsburgh Medical School, Pittsburgh, Pennsylvania. report tiredness, fatigue, and emotional changes sim- Address correspondence to: Ronald E. Dahl, M.D., Director, Child ilar to having obtained insufficient amounts of sleep and Adolescent Neurobehavioral Laboratory and Professor of Psychiatry. Similarly, if subjects are permitted ad libitum and Pediatrics, Western Psychiatric Institute and Clinic, 3811 O’Hara Street, Pittsburgh, PA 15213. E-mail: [email protected] amounts of sleep but are selectively deprived of one Manuscript accepted August 22, 2002. sleep stage, such as rapid eye movement (REM) or © Society for Adolescent Medicine, 2002 1054-139X/02/$–see front matter Published by Elsevier Science Inc., 360 Park Avenue South, New York, NY 10010 PII S1054-139X(02)00506-2 176 DAHL AND LEWIN JOURNAL OF ADOLESCENT HEALTH Vol. 31, No. 6S delta sleep, they also report daytime consequences modern human brain continue to reflect these links, [3,4]. such that social stresses evoke powerful feelings of This restorative function of sleep is essential not threat and sleep disruption, but feelings of love, only for optimal physical and mental function but caring, and social connection create a sense of safety also for survival. Sleep is observed in every species of and promote sleep. living creature that has ever been studied. Animals Finally, it is important to consider how safety/ deprived of sleep perish; experiments with rats show threat perception and its capability to disrupt sleep survival without sleep is comparable to survival change during adolescent development. The devel- without food. Yet the specific function of sleep opment of the vigilance or threat perception and remains a scientific mystery despite extensive stud- response system shows a significant increase across ies aimed at elucidating the purpose of this state. pubertal development; this was likely to have been There are, however, a few themes that have adaptive as adolescents were making the physical emerged from this area of investigation, relevant to preparations to take on adult roles, with increased clinical issues in Adolescent Health. First, whatever demands on threat appraisal. its purpose, sleep seems to be particularly important In summary, sleep is an active process in the brain during periods of brain maturation. Across species, that is necessary for restorative functions, particu- greater sleep requirements are observed in maturing larly during development. Because sleep requires a vs. fully mature individuals. period of relative unresponsiveness, it is naturally Another central principle about sleep, which may restricted to safe places and times. Threat-related have particular relevance to aspects of adolescent arousal interferes with sleep (for adaptive purposes). development, is the close link between sleep and These effects on sleep undergo significant matura- perceptions of threat/safety. Sleep, at a behavioral tional changes including developmental changes as- level, involves a fundamental loss of awareness and sociated with the pubertal transition. responsiveness to the external environment. During sleep, most sensory information stops at the level of the thalamus, preventing perception of (and behav- Sleep Physiology and Terminology ioral response to) potential threats in the environ- Sleep stages are defined by the patterns of three ment. As a result, most species have evolved mech- electrophysiologic measures: the electroencephalo- anisms to ensure that sleep behavior is limited to safe gram (EEG), electromyogram, and electro-occulo- places (burrows, nests, and temporal niches rela- gram. These measures are used to divide sleep into tively safe from predators). Unless one feels safe, it is broad categories of REM sleep and non-REM sleep prudent to avoid turning off vigilance and respon-. siveness. Thus, it also makes sense that any percep- REM sleep also is called paradoxical sleep because tion of threat and the accompanying increased it has aspects of deep sleep and light sleep at the arousal is antithetical to going to sleep. same time. On one hand, REM appears deep because This link between sleep and safety has great the changes in the body (loss of muscle tone) and clinical relevance to sleep problems including an subcortical brain systems, such as temperature reg- important source of sleep disruption among adoles- ulation and control of respiration and heart rate, are cents. The point here is that social and emotional more profound than in any other stage of sleep. On cues are critical to creating the feelings of safety that the other hand, higher cortical brain functions are promote sleep. This is because for most of human quite active and resemble wakefulness. Dreaming is history (and as is true for most social primates today) associated closely with REM sleep. Further, it is safety from predators was primarily through a pro- relatively easy to awaken a person from REM sleep tective social group. Early humans were sur- (alertness returns relatively briskly, compared with rounded by large nocturnal-hunting carnivores with- deep non-REM sleep). REM periods occur in cycles out physically safe sleep sites. (Humans can not sleep of approximately 90 minutes throughout the night, in trees or on cliff edges because we lose all muscle with the longest and most intense REM periods tone during REM sleep.) In the human ancestral occurring just after the body temperature reaches a environment, a close-knit social group provided pro- minimum, around 5 AM.. tection against predators. The human brain evolved Non-REM sleep is subdivided further into Stages under conditions in which this sense of social belong- 1, 2, 3, and 4. Stages 3 and 4, also called delta or ing and social connectedness formed the underpin- slow-wave sleep, represent the deepest sleep in hu- ning for feelings of safety. Natural tendencies in the man beings. The length of this deep delta sleep December 2002 SLEEP REGULATION AND BEHAVIOR 177 Figure 1. Sleep Pattern in an Early School-aged Child increases in proportion to how long one has been out the night (with longer REM sleep periods toward awake (i.e., there is more delta sleep and deeper delta early morning). sleep following sleep loss or chronic sleep distur- Age-related (developmental) processes also exert bances). Children experience extremely large profound influences on sleep regulation. Total sleep amounts of deep slow-wave sleep, which gradually decreases from 16 hours/day in the newborn period decreases as they get older. During this deep sleep to approximately 8 hours/night by age 18 years [8,9]. (usually 1 to 3 hours after going to sleep), it is At age 1 year, the average child sleeps approximately extremely difficult to arouse a child, and if aroused, 11 hours/night with another 2.5 hours of sleep they often appear disoriented, confused, and cogni- obtained in two separate daytime naps. By age 3 tively slow. Confused partial arousals, including years, the average child gets 10.5 hours of sleep each sleep walking, talking, and night terrors, usually night with one 1.5-hour nap. In the United States, a emerge from this state. typical child ceases daytime naps at about 4 to 5 Knowledge of the patterning of these sleep stages years of age. It is also important to emphasize that is critical to understanding the organization of sleep there is considerable individual variation in sleep stages and their regulation. Figure 1 shows a typical requirements as well as cultural influences on sleep example of a young prepubertal child’s sleep pattern. and napping behavior; for example, daytime naps Stages 1, 2, 3, and 4 are shown as progressively lower continue through adulthood in some cultures. steps on the vertical axis with REM sleep indicated Changes in sleep needs during school age and into by striped boxes at an intermediate level, indicating adolescence show a complex pattern, with different the paradoxical relationship to sleep depth. This findings on school nights vs. weekends and holidays, child, illustrated in Figure 1, fell briefly into Stage 1, as will be discussed in greater detail in subsequent descended to Stage 2, and then to Stages 3 and 4 in chapters. which she remained for approximately 1 hour. Just before midnight, sleep returned to Stage 2 with a 30-second arousal, followed by another hour of deep Biological Changes in Sleep/Circadian Stage 3 and 4. The first REM period occurred about Regulation During Pubertal Development 1:00 AM with additional REM sleep periods occur- There are four areas of sleep regulation showing ring approximately every 60 to 90 minutes through- changes during adolescent maturation: (a) There is a 178 DAHL AND LEWIN JOURNAL OF ADOLESCENT HEALTH Vol. 31, No. 6S decrease in the duration and depth of non-REM optimal levels. In studies of adolescents in natural (Stages 3 and 4) and REM sleep, (b) a more adultlike environments (when adolescents obtain less sleep pattern of REM sleep develops, (c) there are increases than in the laboratory situations), dramatically ele- in daytime sleepiness, and (d) there is a shift in the vated levels of daytime sleepiness are often observed circadian pattern toward a more owl-like tendency. In many cases, the levels of sleepiness in high for later bedtimes and wake-up times. Each of school students are near the threshold seen in sleep these will be described briefly. disorders such as narcolepsy or sleep apnea. Although the amount of REM and Stage 3/4 sleep declines somewhat during school-age years, these levels show a marked decrease across adolescence. Although most data have come from cross-sectional Circadian Changes at Puberty studies, our research group (Dahl, Ryan, Birmaher, et During adolescence there are various alterations in al., unpublished data) is just completing a large addition to the duration and organization of sleep longitudinal study of normal adolescents, restudied stages. These include important maturational each year for 4 to 5 years during pubertal develop- changes in the timing of sleep and the influence of ment, showing the same patterns of sleep changes. In the biological clock or circadian regulation system. essence, there is a shift toward lighter sleep (Stages 1 This circadian system is involved not only in the and 2) and less of the deep Stages 3 and 4 and REM timing of sleep but also the timing of brain hormone sleep. There is also a decrease in the threshold of release, body temperature regulation, and the archi- arousal from infancy through adolescence. tecture within sleep. The system is largely influenced Adolescent maturation is also associated with a by the suprachiasmatic nucleus located in the hypo- relative shortening of the interval from sleep onset to thalamus, which is sensitive to light cues and social the first REM period (shorter REM latency) and a schedules and plays a modulatory role in the timing decrease in the REM density (rate of eye movements of melatonin release. occurring within REM). One important general principle about this circa- In contrast to these maturational changes in sleep dian system, relevant to adolescent sleep patterns, is patterns, developmental changes regarding total that it adapts slowly to changes in sleep/wake need for sleep during adolescence reveals a much schedules. Thus, adolescents rapidly shift sleep/ more complex picture. On one hand, adolescents wake schedules between school nights and week- typically get by on less sleep (at least on school ends/vacations, which can be particularly problem- nights). However, there are several reasons to believe atic. that this decrease occurs independent of biology. A second important general principle about the There is no evidence that adolescents need less sleep. circadian timing system is that it adapts more easily More specifically, a series of studies by Carskadon to delays in the sleep/wake schedule rather than to and colleagues [12–14] show that daytime sleepiness advances. This is why it is naturally easier to stay up increases near midpuberty, even when the total later and sleep in later on weekends and why it is amount of sleep is held constant across puberty. easier to travel two to three time zones to the west In a classic set of studies at Stanford University, rather than to the east. Carskadon and Dement had children return to The relevance of these principles to adolescent sleep camp each summer with the total amount of sleep patterns is straightforward. Many adolescents time in bed held constant year after year. They have abrupt changes in the timing of their sleep showed that even with the same amount of night- between regular school schedules, requiring early time sleep as in previous summers, objective mea- morning awakening, and the late bedtimes, with sures of sleepiness on EEG nap studies showed quick shifts back to late bedtimes and sleeping in on increased sleepiness as maturation proceeded to weekends and vacations. For example a typical ado- Tanner Stages 3 and 4, compared with similar mea- lescent on Eastern Time going to bed at 2:00 AM on sure when children were prepubertal (Tanner Stages Friday and Saturday nights and sleeping in until 1 and 2). Thus these data are consistent with a noon will shift their circadian system to Pacific Time possible increase in sleep needs during pubertal within a few days. However, the shift to an earlier maturation. time compatible with going to school on Eastern Moderate increases in daytime sleepiness during Time will require several days of a stable schedule to pubertal development are evident in these experi- shift the temperature and hormone rhythms com- mental situations, even when nighttime sleep is at pletely [19,20]. Therefore, many adolescents, partic- December 2002 SLEEP REGULATION AND BEHAVIOR 179 ularly those that oversleep or miss an occasional day caffeine and nicotine, is also common and increasing of school during the middle of the week, may expe- among adolescents. rience jet lag–like symptoms of fatigue, difficulty Finally, social stresses including fears, anxieties, falling asleep at night, and difficulty awakening in and emotional arousal can greatly interfere with the morning. sleep in adolescents. Extensive physical and emo- In the most severe version of this problem, labeled tional change, social turmoil, and novel experiences delayed sleep phase syndrome, adolescents and their and challenges occur during adolescence and tran- families often battle for months about late night siently interfere with going to sleep. It is also impor- bedtimes and great difficulties awakening on school tant to emphasize that the cognitive components of days. In essence, the adolescents are trying to the ability to fall asleep undergoes substantial awaken during their body temperature minimum, changes at some point in adolescence (i.e., younger when their body is not prepared to be awake and children’s rumination, worry, and distressing active. The important principle here is that slow, thoughts are much less likely to interfere with going steady, and consistent changes in the sleep/wake to sleep than those of adolescents ). timing will permit the circadian system to realign to a more appropriate pattern. In reality many adoles- cents use extreme efforts over short weekly intervals of time and are operating in opposition to the phys- Interactions Between Physical Changes in iology. On a clinical level, one of the difficulties Sleep Regulation and Psychosocial Influences adolescents are often reluctant to change their week- During adolescence, the pendulum of sleep and end schedules to realign their systems, even if they arousal regulation tends to encounter multiple forces are motivated to get up for school. pushing in one direction: away from sleep toward Finally, a critical aspect of this pattern that has high arousal. This occurs for several reasons: The recently come to light from work by Carskadon et al. biological depth of sleep decreases, and there are is that the tendency toward these delayed circa- multiple circadian changes at puberty; the elements dian patterns appears to appear during puberty. of modern cultures such as artificial light and easy That is, there appears to be a biological change in the access to a multitude of stimulating activities in- circadian system from a more larklike pattern to a creases arousal; and new social stresses, emotional more owl-like pattern at about midpuberty. Current variability, and increases in the cognitive basis of work by Carskadon and colleagues (personal com- generating worries and perceived threats have de- munication, June 1998), using dim light melatonin veloped. onset studies, has provided further evidence of a Many adolescents also experience chronic tension biological basis for self-selection of later bedtimes between sleep needs and circadian limits. A typical during adolescence. and preferred adolescent summer sleep schedule is 3:00 AM to noon. The shift to the school schedule is incomplete because the circadian system requires a more gradual and consistent transition over at least a Psychosocial Influences Contributing to Sleep couple of weeks. Thus, adolescents often revert to Problems in Adolescents their old familiar schedule of late bed- and wake-up In addition to these biological changes leading to- times. The consequences of this delayed sleep phase ward late night schedules, there are extensive psy- is that adolescents have difficulty falling asleep (dur- chosocial influences that interfere with adolescent ing a period of increased body temperature) and sleep. These include: a desire to stay up late at night waking up (during the body temperature minimum) to engage in adultlike social activities, self-deter- until their core body temperature rhythm is ad- mined bed times, and easy access to a vast range of vanced. This creates a vicious cycle of sleep depriva- stimulating and arousing activities. Historically there tions/jet lag–like symptoms. were very few options for late night activities after Not only does this lead to sleep related problems, dark for most adolescents. However, in the current but there are also complex interactions with emo- era with access to hundreds of cable television chan- tions, attention, and behavior. This can perhaps be nels, telephones, the Internet, video games, and best illustrated by a clinical example. One 14-year- late-night social activities, adolescents have innu- old student in midpubertal development was given merable options for stimulating and arousing activ- freedom to choose his own summer. His typical sleep ity. Access to pharmacological stimulation, including schedule, as were most of his friends, was from 2:30 180 DAHL AND LEWIN JOURNAL OF ADOLESCENT HEALTH Vol. 31, No. 6S AM until noon. When school started in the fall, he Specific factors associated with less sleep among tried to go to bed at 10 PM. but was unable to fall adolescents include: early school starting times, aca- asleep until 3 AM; not only was he trying to sleep demic pressures, extracurricular activities, anxiety, during the wrong phase of his circadian cycle, but he stress, and work schedules. In this study, more was also nervous and excited about going back to than 50% of high school students hold part-time jobs school. It was extremely difficult to get up at 6 AM to and 28% reported working more than 20 hours/ perform his adolescent grooming ritual in time to week during school schedules. There were much make the 6:45 AM school bus higher rates of sleep complaints in students who For a few days, he was able to get by in an worked more than 20 hours/week. extremely sleep-deprived state on sheer motivation In summary, the biological tendency for sleep until the first weekend rolled around. Then he not delay in adolescence, social influences toward sleep only reverted to his more usual sleep pattern but also delay, access to a multiple stimulating activities, was able to catch up by sleeping until as late as 3 PM greater freedom to self-select bed times, and the on a Saturday afternoon. increased stress and anxiety leading to difficulty This pattern persisted for several weeks, and he falling asleep can all interact to produce significant never successfully shifted his circadian cycle to an disruptions of sleep in many teenagers. earlier bedtime but instead overrode the system for a few days at a time. At least once a week, he would oversleep and be late for school. He also began falling asleep in class. Conflict with his parents and Consequences of Insufficient Sleep teachers increased. Every morning was an extreme in Adolescents battle. He began to miss entire days of school, and A surprising lack of empirical data examining effects when he did attend, he was increasingly negative of sleep deprivation and insufficient sleep among and tired and had difficulty paying attention. Social adolescents exists. There are, however, extensive conflict, emotional stress, and school failure began to circumstantial evidence, clinical evidence, and adult contribute to increased sleep difficulties at night data relevant to these questions. In brief, the most because he would try to go to sleep at midnight and direct effect of inadequate sleep is sleepiness during lie in bed awake and worried about his ability to get low-stimulation activities such as paying attention in up on time the following morning. He also became a classroom, driving, or doing repetitive activities. increasingly anxious about the looming possibility of Many schools surveyed suggest that falling asleep in failing at school and the seriousness of his parents’ school has become a very common behavior in many threats of punishment. This vicious cycle led to a high schools. pattern of chronic sleep deprivation, extreme emo- A separate but related factor is tiredness, in con- tional distress, and great difficulties with attention trast to frank sleepiness. Tiredness can be defined as and self-control. He eventually met diagnostic crite- the feeling of fatigue that makes it difficult to moti- ria for depression, and attention deficit-hyperactivity vate or initiate certain types of behavior, particularly disorder (ADHD), although the relative contribution those behaviors associated with long-term goals or of chronic sleep deprivation were very difficult to negative consequences. Tiredness and symptoms of disentangle from the psychiatric disorders. fatigue can be prominent in adolescents, even if they Studies that have examined school starting time remain awake; these symptoms may also contribute show no relationship with bedtime. For example, to longer-term sequelae. high school students in a district with an early A particularly important health consequence is starting time (7:15 AM) had a mean rise time of 5:58 the negative synergy between insufficient sleep and AM and mean school night bedtime of 11:10 PM. A alcohol use (i.e., there is an additive impact of comparable school district with a late starting time of alcohol and sleep deprivation on cognitive abilities 8:30 AM had students reporting a mean rise time of and motor coordination). This is a particularly dan- 7:06 AM and a mean school night bedtime of 11:20 gerous combination for young, recently licensed PM. There were no significant differences in bed- drivers. time, only in wake-up time and amount of sleep on Another important domain of sleep deprivation school nights. Other investigators have reported effects is in relation to mood. The relationship be- similar results [24,25]. Data also suggest that at least tween sleep and depression will be discussed further transient or moderate sleep deprivation occur in 10% in the next section. However, there are also more to 40% of high school students [26 –28]. general effects on mood lability (i.e., following sleep December 2002 SLEEP REGULATION AND BEHAVIOR 181 loss, children and adolescents often report less con- greater risk for the development of sleep problems trol or inhibition over their moods such that they and mood problems. may be sillier in response to humorous situations, more angry in a frustrating situation, or more likely to cry when upset or sad). Interactions Between Sleep and Behavioral/ Considerable data also indicate that the control of Emotional Domains attentional focus and impulses are diminished fol- It is essential to underscore the overlap between lowing sleep deprivation in children. Several sleep regulation and behavioral/emotional problems investigators have found evidence of ADHD-like in children and adolescents. There are clearly two- symptoms associated with sleep deprivation and way interactions between these systems. The devel- some improvement in ADHD symptoms in children opment, regulation, and timing of sleep can be following treatment of sleep problems [30 –33]. This altered by behavioral/emotional disorders, and cog- is a very complex area with great clinical relevance nitive, behavioral, and emotional control during day- and is currently the focus of several ongoing studies. time hours can be influenced by the way children As has been presented in other papers [34,35], and adolescents sleep. Furthermore, daytime activi- there is also increasing evidence that sleep depriva- ties, changes in the environment, and stressful events tion effects are particularly prevalent in measures of can have profound transient effects on children’s higher cortical functions. Furthermore, we have sleeping patterns in the absence of any clear-cut shown in pilot studies that tasks that simultaneously psychopathology. Additionally, medications used to challenge cognitive and emotional processing appear treat psychiatric disorders often affect sleep, and to be particularly sensitive to sleep deprivation ef- sleep loss can exacerbate mood and behavioral symptoms. fects. We recently extended this work to more quan- Subjective sleep complaints are very common in tifiable domains, cognitive processes, and a subcor- children and adolescents diagnosed with major de- tical regulatory system: the maintenance of postural pressive disorder. Symptoms include insomnia (75% balance. Preliminary evidence shows precisely what of cases) and hypersomnia (25% of cases). Hy- we predicted regarding sleep deprivation effects. In a persomnia difficulties are reported more frequently similar study, Schlesinger et al. showed that in after puberty. Insomnia symptoms usually include young healthy university students, one night of sleep difficulty falling asleep and a subjective sense of not deprivation caused no significant effect on perfor- having slept deeply all night. Early-morning awak- mance of an inhibitory cognitive task and no effect enings are less prevalent in children and adolescents on postural balance; however, there was a marked than in adults with depression. Recently clinicians effect on the ability to perform both the cognitive and and researchers have seen increasingly more adoles- balance tasks simultaneously after sleep deprivation. cents with overlapping phase-delay disorders In a similar way, sleep deprivation may impair the and/or other sleep/wake schedule disorders with ability to perform both a cognitive and emotional depression [29,38,39]. Depressed adolescents fre- task at the same time. quently have difficulty falling asleep, are unable to Although these may seem to be relatively subtle get up or refuse to go to school, sleep in late in the effects, the foundation of social competence, an area day, complain of extreme daytime fatigue, and over of major struggle for adolescents, requires fluency in time shift to increasingly more delayed sleep/wake performance in such tasks. In particular, the ability to schedules. Likewise, significant increases in com- concentrate on a long-term goal or consequence plaints of depressed mood have been found in sur- while regulating emotional reactions in social situa- veys of adolescents who obtain less than 6 hours of tions is precisely the territory that many adolescents sleep each school night and/or report more than are struggling to navigate in their day-to-day lives. If 2-hour differences between school night and week- sleep deprivation results in impairment in this do- end bedtimes. Clinicians experienced with these main, it may have very significant consequences. problems have pointed out that in many cases it is Finally, as in many areas of child health, habits difficult to differentiate decreased motivation, school established in adolescence will persist into adult- refusal/anxiety, delayed circadian phase, attention hood. Sleep habits are much the same. Therefore, difficulties, and depressive symptomatology. Clearly chronic sleep deprivation, difficulty falling asleep, both sleep patterns and behavioral symptoms must and difficulty waking up in the morning become be carefully assessed for prevention, accurate diag- established during adolescence and may confer nosis, and/or treatment planning. 182 DAHL AND LEWIN JOURNAL OF ADOLESCENT HEALTH Vol. 31, No. 6S Sleep Studies in Depression images, relaxing thoughts, and engaging imagery Polysomnographic studies in adult depression show associated with positive emotions can be an impor- a clear pattern of altered sleep regulation, and there tant aspect of treatment. Similarly, careful attention are subjective complaints of poor sleep quality. How- to sleep/wake schedules, including consistent bed ever, in studies comparing depressed children and and wake-up times throughout the week, and the adolescents and controls, objective (EEG) studies frequent occurrence of delayed sleep phase syn- have revealed a somewhat inconsistent pattern of drome in adolescents must also be considered. Al- though anecdotal evidence suggests use of melatonin findings. In childhood depression, defined here as may have some positive impact, controlled studies at Tanner Stage 1 or 2, increased sleep latency has been this time indicate that cognitive and behavioral in- the most reliable finding associated with depression. terventions around sleep schedules are probably Although reduced REM latency, an impressive more effective. marker of adult depression, has been found in one controlled study , the majority of childhood stud- ies have been negative for REM latency differences [38,41– 43]. However, in a follow-up study, in which Sleep and Anxiety Disorders no baseline REM latency differences had been seen, a The clinical picture for sleep changes in anxiety cholinergic challenge (with arecoline) resulted in disorders largely parallels that described for depres- reduced REM latency in depressed children, com- sion, albeit with much less available data. Anxiety pared with normal controls studied in the same disorders are quite common in children and adoles- environment. cents yet often go undiagnosed. Subjective com- Studies of depressed adolescents (aged 12 to 18 plaints of difficulty falling asleep and frequent night- years and Tanner Stages 3, 4, and 5) have revealed a time awakening are common in children and somewhat intermediate pattern between child and adolescents with anxiety disorders. Increased fear adult studies. Some studies have reported reduced and vigilance, the hallmarks of anxiety disorders, is REM latency in inpatient samples of depressed ado- in many ways antithetical to sleep because sleep is in lescents [40,45,46], although most outpatient studies essence a behavioral cessation of vigilance. Feelings have found negative or inconsistent results [47–50]. of danger or threat, emotionally as well as physically, With careful attention to sleep/wake schedule issues appear to interfere with the context of safety neces- and adaptation to setting, reduced REM latency and sary to promote normal sleep onset and continuity. prolonged sleep latency were detected in depressed Adequate treatment of anxiety disorders including adolescents [38,51]. Increased REM density has also cognitive behavioral interventions focused on the been seen in some studies including a study that period preceding sleep onset can be an important indicated increased REM density associated with component of treatment for these problems. The increased recurrence of depressive episodes. issues regarding sleep/wake schedules, improved Although these studies raised interesting questions bedtime habits, and treatment of the primary anxiety regarding the pathophysiology and links between disorder are very similar to those discussed under sleep changes and depression, it is important to depression. emphasize that, from a clinical perspective, EEG sleep studies have not been found to contribute to diagnosis or treatment issues. Sleep and ADHD Treatment of sleep complaints and problems in- The significant overlap with difficulty with self- cluding regularizing the sleep/wake schedule, cog- control of behavior, attention, and impulses has nitive behavioral therapy for insomnia and/or short- already been discussed in the previous section. Suf- term medication treatment for severe insomnia can fice it to say that sleep deprivation causes behavioral have a positive impact on depressive symptoms in problems in younger children and sometimes in some cases. On the other hand, effective treat- adolescents that are similar to those of ADHD. The ment of depression can also be a critical aspect of effects of stimulant medication, the treatment of improving sleep problems. choice for ADHD, transiently improving daytime Adolescents’ tendency to have distressing or neg- symptoms of sleep deprivation and yet further inter- ative ruminations at bedtime further complicates the fering with nighttime sleep, is also a complex issue in interaction of insomnia, depression, and anxiety. need of further study. The same principles Helping these patients actively focus on positive regarding ADHD symptoms also apply to conduct December 2002 SLEEP REGULATION AND BEHAVIOR 183 disorder and aggressive behavior. However, there One specific area in which there is critical need for are virtually no controlled studies in these areas with study is in relation to early starting times among careful measures of sleep. some high schools. This is undoubtedly an important factor exacerbating sleep deprivation among some adolescents. Although a number of school districts Critical Questions and Future Directions have already implemented delays in the starting time Although the past decade has seen research progress (and many more are considering such changes), in these areas, major gaps in existing knowledge there is a great need for better empirical data exam- continue. There is clearly a paucity of well-controlled ining the underpinnings for, and consequences of, studies to guide specific health policy decisions and such changes at both a biologic and psychosocial recommendations regarding sleep in adolescence. In level. School districts that have established year- particular, there is need for improved understanding round schooling might be important participants in of the acute and chronic effects of inadequate sleep in these types of studies. adolescents. Currently there are not even clear guidelines for defining adequate sleep in adoles- cents. A much better delineation of the links among References sleep, behavior, and affect regulation is needed. This 1. Roth T, Roehrs TA, Carskadon MA, Dement WC. Daytime will likely require new methodologies and ap- sleepiness and alertness. In: Kryger MH, Roth T, Dement WC proaches to these complex and overlapping domains. (eds). Principles and Practice of Sleep Medicine 2nd Edition. The implications of sleep deprivation in adoles- Philadelphia: WB Saunders & Co., 1994:40 –9. cence are broad. They are certainly associated with 2. Bonnet M. Sleep deprivation. In: Kryger MH, Roth T, Dement WC (eds). Principles and Practice of Sleep Medicine 2nd increased risk of psychopathology. Increased stress, Edition. Philadelphia: WB Saunders & Co., 1994:50 – 68. poorer social adaptation, impaired attention and 3. Pearlman CA. Sleep structure variation and performance. In: motivation, and sleepiness are in the majority of Web WB (ed). Biological Rhythms, Sleep and Performance. cases subclinical sequelae of sleep deprivation that New York: John Wiley & Sons, 1982:143–73. are likely to have a societal cost that manifests in 4. Puca FM, Livrea P, Genco S, et al. REM sleep deprivation in normal humans. Changes in anxiety, depression and aggres- various domains. For example, there is likely to be a siveness, and HVA and 5-HIAA levels in the lumbar cerebro- measurable negative impact on performance in spinal fluid. Boll Soc Ital Biol Sper 1976;52:782–7. school. Some percentage of teen substance abuse, as 5. Rechtschaffen A, Bergmann BM, Everson CA, et al. Sleep a method of heightening arousal or decreasing the deprivation in the rat: Integration and discussion of the findings. Sleep 1989;12:68 –87. effects of anxiety and depressions, may be attribut- able to the direct and indirect effects of sleep depri- 6. McKenna J, Mack J. Origins and paleoecology of hominid sleep: An evolutionary perspective. Sleep Res 1992;21. vation. Some percentage of automobile and work- 7. Carskadon MA, Dement WC. Normal human sleep: An over- place accidents is most definitely attributable to view. In: Kryger MH, Roth T, Dement WC (eds). Principles sleepiness. Finally, establishing poor habits in ado- and Practice of Sleep Medicine 2nd Edition. Philadelphia: WB lescence may extend to adulthood. Saunders & Co., 1994:16 –26. Although the costs of adolescent sleep deprivation 8. Anders TF, Sadeh A, Appareddy V. Normal sleep in neonates and children. In: Ferber R, Dryger M (eds). Principles and have not been calculated, there are several steps that Practice of Sleep Medicine in the Child. Philadelphia: WB might be taken to begin to correct the problem. Saunders & Co., 1995:7–18. However, many of these involve rather sweeping 9. Wolfson AR. Sleeping patterns of children and adolescents: social policy changes. For example, broader aware- Developmental trends, disruptions, and adaptations. In: Dahl RE (ed). Child and Adolescent Psychiatric Clinics of North ness of the problem among parents, pediatricians, America: Sleep Disorders. Philadelphia: WB Saunders & Co., and educators is a first step. Limitations on work 1996:549 –68. hours for adolescents may have a complex series of 10. Dahl R, Carskadon MA. Sleep and its disorders in adoles- implications but is likely to help. Scheduling of cence. In: Kryger MH, Roth T, Dement WC (eds). Principles and Practice of Sleep Medicine in the Child, 2nd Edition. Vol. events and programming that is targeted at adoles- 2–3. WB Saunders & Co., 1995:19 –27. cents may serve to decrease the amount of stimulat- 11. Busby K, Pivik RT. Failure of high intensity auditory stimuli to ing activities that are available during prime sleep affect behavioral arousal in children during the first sleep hours. Changes in school curriculum and policy cycle. Pediatr Res 1983;17:802–5. could also be beneficial. For example, guidelines for 12. Carskadon MA, Harvey K, Duke P, et al. Pubertal changes in daytime sleepiness. Sleep 1980;2:453–60. scheduling certain types of classes within the school 13. Carskadon MA, Dement WC. Sleepiness in the normal ado- day might maximize performance when attention lescent. In: Guilleminault C (ed). Sleep and Its Disorders in and concentration are at their best. Children. New York: Raven Press, 1987:53–66. 184 DAHL AND LEWIN JOURNAL OF ADOLESCENT HEALTH Vol. 31, No. 6S 14. Carskadon MA. Patterns of sleep and sleepiness in adoles- 34. Schlesinger A, Redfern MS, Dahl RE, Jennings JR. Postural cents. Pediatrician 1990;17:5–12. control, attention and sleep deprivation. Neuroreport 1998;9: 15. Carskadon MA, Rosekind MR, Galli J, et al. Adolescent 49 –52. sleepiness during sleep restriction in the natural environment. 35. Dahl RE. The impact of inadequate sleep on children’s day- Sleep Res 1989;18:115. time cognitive function. Semin Pediatr Neurol 1996;3:44 –50. 16. Thorpy MJ, Korman E, Spielman AJ, Glovinsky PB. Delayed 36. Horne JA. Human sleep, sleep loss and behaviour: Implica- sleep phase syndrome in adolescents. J Adolesc Health Care tions for the prefrontal cortex and psychiatric disorder. Br J 1988;9:22–7. Psychiatry 1993;162:413–9. 17. Kryger MH, Roth T, Carskadon MA. Circadian rhythms in 37. Ryan ND, Puig-Antich J, Ambrosini P, et al. The clinical humans: An overview. In: Kyger MH, Roth T, Dement WC picture of major depression in children and adolescents. Arch (eds). Principles and Practice of Sleep Medicine, 2nd Edition. Gen Psychiatry 1987;44:854 –61. Philadelphia: WB Saunders & Co., 1994:301– 8. 38. Dahl RE, Ryan ND, Matty MK, et al. Sleep onset abnormalities in depressed adolescents. Biol Psychiatry 1996;39:400 –10. 18. Carskadon MA, Mancuso J. Daytime sleepiness in high school adolescents: Influence of curfew. Sleep Res 1988;17:75. 39. Dahl RE, Ryan ND. The psychobiology of adolescent depres- sion. In: Cicchetti D, Toth S (eds). Adolescence: Opportunities 19. Graber JA, Brooks-Gunn J, Warren MP. The antecedents of and Challenges. Vol. 7. Rochester, NY: University of Rochester menarcheal age: Heredity, family environment, and stressful Press, 1996:197–232. life events. Child Dev 1995;66:346 –59. 40. Emslie GJ, Rush AJ, Weinberg WA, et al. Children with major 20. Ferber R. Clinical assessment of child and adolescent sleep depression show reduced rapid eye movement latencies. Arch disorders. In: Dahl RE (ed). Child and Adolescent Psychiatric Gen Psychiatry 1990;47:119 –24. Clinics of North America. Philadelphia: WB Saunders & Co., 1996:569 –80. 41. Puig-Antich J, Goetz R, Hanlon C, et al. Sleep architecture and REM sleep measures in prepubertal major depressives. Stud- 21. Carskadon MA, Vieira C, Acebo C. Association between ies during recovery from the depressive episode in a drug-free puberty and delayed phase preference. Sleep 1993;16:258 –62. state. Arch Gen Psychiatry 1983;40:187–92. 22. Dahl RE. Introduction: Sleep and child psychiatry. In: Dahl RE 42. Dahl RE. Child and adolescent sleep disorders. In: Kaufman (ed). Child and Adolescent Psychiatric Clinics of North Amer- DM (ed). Child and Adolescent Neurology for Psychiatrists. ica: Sleep Disorder. Vol. 3. Philadelphia: WB Saunders & Co., Baltimore: Williams and Wilkins Publishers, 1992:169 –94. 1996:543–8. 43. Young W, Knowles JB, MacLean AW, et al. The sleep of 23. Wahlstrom KL, Frederickson J, Wrobel G. School Start Time childhood depressives: Comparison with age-matched con- Study. Minneapolis: University of Minnesota; 1997:1–17. trols. Biol Psychiatry 1982;17:1163–8. 24. Wolfson AR, Carskadon MA. Early school times affect sleep 44. Dahl RE, Ryan ND, Perel J, et al. Cholinergic REM induction and daytime functioning in adolescents. Sleep Res 1996;25:117. test with arecoline in depressed children. Psychiatry Res 25. Allen RP. Social factors associated with the amount of school 1994;51:269 –82. week sleep lag for seniors in an early starting suburban high 45. Kutcher S, Williamson P, Marton P, Szalai J. REM latency in school. Sleep Res 1992;21:114. endogenously depressed adolescents. Br J Psychiatry 1992;161: 399 –402. 26. Strauch I, Meier B. Sleep need in adolescents: A longitudinal approach. Sleep 1988;11:378 –86. 46. Dahl RE, Ryan ND, Williamson DE, et al. Regulation of sleep and growth hormone in adolescent depression. J Am Acad 27. Carskadon MA. Adolescent sleepiness: Increased risk in a Child Adolesc Psychiatry 1992;31:615–21. high-risk population. Alcohol Drugs Driving 1990;6:317–28. 47. Gershoni-Baruch R, Epstein R, Tzischinsky O, et al. Acti- 28. Petta D, Carskadon MA, Dement WC. Sleep habits in children graphic home-monitoring of the sleep patterns of in vitro aged 7–13. Sleep Res 1984;13:86. fertilization children and their matched controls. Dev Med 29. Dahl R. The regulation of sleep and arousal: Development and Child Neurol 1994;36:639 –45. psychopathology. Dev Psychopathol 1996;8:3–27. 48. Goetz RR, Puig-Antich J, Ryan N, et al. Electroencephalo- 30. Dahl RE, Pelham WE, Wierson M. The role of sleep distur- graphic sleep of adolescents with major depression and nor- bances in attention deficit disorder symptoms: A case study. mal controls [see comments]. Arch Gen Psychiatry 1987;44: J Pediatr Psychol 1991;16:229 –39. 61–8. 31. Walters A, Mandelbaum D, Lewin DS, et al. Dopaminergic 49. Dahl RE, Puig-Antich J. Sleep disturbances in child and therapy in children with restless legs/periodic limb move- adolescent psychiatric disorders. Pediatrician 1990;17:32–7. ments in sleep and ADHD. Pediatr Neurol 2000;22:182–6. 50. Khan AU, Todd S. Polysomnographic findings in adolescents with major depression. Psychiatry Res 1990;33:313–20. 32. Picchietti DL, Walters AS: Restless legs syndrome and peri- odic limb movement disorder in children and adolescents: 51. Da Silveira Neto O, Da Silveira OA. Kleine-Levin syndrome: Comorbidity with attention-deficit hyperactivity disorder. In: Report of a case. Arq Neuropsiquiatr 1991;49:330 –2. Dahl RE (ed). Child and Adolescent Psychiatric Clinics of 52. Rao U, Dahl RE, Ryan ND, et al. The relationship between North America: Sleep Disorders. Philadelphia: WB Saunders longitudinal clinical course and sleep and cortisol changes in & Co., 1996:729 –40. adolescent depression. Biol Psychiatry 1996;40:474 –84. 33. Dahl RE, Carskadon MA. Sleep and its disorders in adoles- 53. Dahl RE. Sleep in behavioral and emotional disorders. In: cence. In: Ferber R, Kryger M (eds). Principles and Practice of Kryger MH, Ferber R (eds). Principles and Practice of Sleep Sleep Medicine in the Child. Philadelphia: WB Saunders & Medicine in the Child. Philadelphia: WB Saunders & Co., Co., 1995:19 –27. 1995:147–54.