Sleep Across the Life Cycle

Transcription

1 SECTION II Anatomy and Physiology CHAPTER 3 Sleep Across the Life Cycle IOURI KREININ L E A R N I N G O B J E C T I V E S On completion of this chapter, the reader should be able to 1. Describe the elements of the EEG signal and development of the EEG from infancy through adulthood. 2. Identify normal pediatric through adult wake and sleep EEG patterns. 3. Describe characteristic aspects of normal sleep and waking in different age groups. 4. Describe the changes seen in the different sleep stages and waking activity from infancy to old age. K E Y T E R M S Sleep Waking Ultradian rhythm Circadian rhythm Two-process model of sleep Gestational age Chronologic age Conceptional age Waveforms Tracé discontinu Tracé alternant High voltage slow activity Low voltage irregular activity Dominant posterior rhythm Vertex sharp transients K-complexes Sleep spindles Slow-wave sleep Sleep stages Hypnagogic hypersynchrony Rhythmic anterior theta activity INTRODUCTION Sleep organization and structure are regulated by the interaction of biologic and environmental factors, making sleep a biopsychosocial phenomenon. Descriptions of developmental and age-related sleep patterns include circadian and ultradian rhythms, sleep wake distribution and sleep stages distribution, non-rapid eye movement rapid eye movement (NREM REM) cycles, types of sleep onset, arousals and awakenings, total sleep time duration and duration of a sleep episode, naps and circadian sleep phase preferences, developmental and age-related EEG parameters, and so on. Sleep behaviors demonstrate great variability between individuals, cultures, societies, and geographical regions (1). They can be affected by biologic, physiologic, and environmental factors such as age, gender (1), body mass index (BMI), parental beliefs, cultural traditions, socioeconomic status and social regulations, work and school schedules, use of media, and education about sleep hygiene. This variability does not allow the establishment of an optimal amount of sleep or optimal sleep schedule for the entire population. Nevertheless, the extensive amount of data accumulated in sleep research and in developing conceptual frameworks makes it possible to understand and predict sleep wake patterns in different periods of the life cycle (2). The most noticeable age-related developments in sleep and sleep wake rhythms take place around birth and during the first year of life and include consolidation of sleep, reductions in number and duration of naps, and changes in arousals. DETERMINING SLEEP AND WAKEFULNESS States of sleep and wakefulness are often described as a constellation, an amalgam, or a combination of behavioral, physiologic, and EEG features (3, 4). Sleep is classified into two states: REM sleep and NREM sleep. A classification of sleep states including REM sleep and NREM sleep was proposed in a 1957 article by Dement and Kleitman (5). Although the presence or the absence of REMs is a distinguishing feature for these two states of sleep, there are many other physiologic variables and electroencephalographic parameters specific to a certain state of sleep and wakefulness. 15

2 16 SECTION II ANATOMY AND PHYSIOLOGY Wakefulness is characterized by higher sympathetic activity and higher body and brain metabolism. Parasympathetic tone is higher in sleep compared with wake, and sympathetic tone is lower. Body metabolism and body temperature is lower in sleep. Brain metabolism is lower in NREM and higher in REM. Muscle tone is lower in NREM, and muscle atonia is present in REM in most muscles. Heart rate and respiratory rate are more regular in NREM becoming more irregular in REM. Responsiveness to stimuli decreases in sleep. There are differences in hormonal secretion in wake, NREM, and REM. Sleep and wakefulness rhythmicity in humans is modulated by the biologic clock or pacemaker located in the hypothalamic suprachiasmatic nucleus (SCN). Circadian (i.e., approximately 24-hour, from Latin circa [about] and dies [day]) rhythms generated by the SCN are endogenous and persist even in the absence of environmental cues. Circadian rhythms can also be generated by the autonomously rhythmic circadian clock cells in most organs and tissues (6). The interaction between the SCN and rhythms in the peripheral tissues influences behavioral and physiologic rhythms. The master circadian clock located in the SCN is entrained by the environmental stimuli including light dark cycles, meal times (food intake), social activities, work schedule and sleep schedule, and traveling across time zones. The circadian system is integrated with the sleep wake system. The pineal gland neurohormone melatonin, secreted only during the night in the darkness (7), regulates entraining of the SCN and possibly plays an important role in the integration between circadian and sleep wake systems (8). In the two-process model of sleep, developed by Borbély (9), a circadian clock-like mechanism (Process C) modulating propensity and time of sleep is considered to be independent of preceding sleep and wakefulness (10). Process C sometimes is regarded as promoting wake more than sleep. The mechanism dependent on preceding sleep wake is called Process S or sleep homeostasis. Homeostatic Process S represents sleep homeostatic mechanisms increasing sleep drive during wakefulness or following shortage of sleep, or decreasing sleep drive during sleep or after excessive sleep. The two-process model of sleep can be viewed as an interaction of two opponent processes: Process S, representing the drive for sleep, and Process C, representing the drive for wakefulness and alertness. An increase in sleep promoting Process S above a certain threshold, without an increase in alertness promoting Process C, triggers sleep. A rise in the wake promoting Process C even on the background of Process S may cause an increased alertness. A decrease in sleep promoting Process S results in wakefulness. A decrease in alertness promoting Process C leads to sleep. Recently several other models of sleep regulation were developed including the three-process model of sleepiness alertness regulation (which adds sleep inertia or Process W to the mechanisms of sleep wake regulation) and the model of ultradian variation of slow-wave activity (10). Ultradian (meaning less than a day) rhythmicity is another type of sleep regulation-related processes. The cyclic alternation between NREM sleep and REM sleep represents ultradian process within sleep. The Waveforms of Sleep and Wakefulness EEG is generated mostly by cortical nerve cell postsynaptic potentials (11). Summated postsynaptic potentials are recorded by conventionally positioned scalp electrodes and EEG is displayed in a 2-dimensional format as a graph of voltage versus time. Some main characteristics of EEG waves include frequency, amplitude, waveform morphology or shape, and distribution. The frequency of EEG waves is measured in cycles per second (cps) or Hz and is divided into four frequency bands: beta, >13 Hz; alpha, 8 to 13 Hz; theta, 4 to 7 Hz; and delta, <4 Hz. Dominant posterior alpha rhythm has a frequency of 8 to 13 Hz, a sinusoidal morphology, and usually has higher amplitude in the occipital region during relaxed wakefulness with eyes closed. Low amplitude mixed frequency (LAMF) activity, predominantly in the theta (4 to 7 Hz) range, with mixed morphology and distributed all over from the frontal to the occipital regions, is the background EEG activity commonly seen in NREM and REM states of sleep. Slow-wave activity in the range of 0.5 to 2 Hz (delta band) with high amplitude (>75 μv) is optimally recorded from the frontal regions and occupies >20% in portions of an EEG recording identified as stage N3 sleep (12). Several transient waveform events distinguishable from the background EEG play an important role in identifying sleep wake stages. Vertex sharp waves with a duration of <0.5 seconds have maximal amplitude in the central region and may indicate stage N1 sleep (12). K-complexes consisting of a sharp upward deflection followed by a downward deflection with total duration >0.5 seconds usually have maximal amplitude in the frontal region and are associated with stage N2 sleep (12). Sleep spindles, described as sinusoidal waves with frequency of 11 to 16 Hz and duration of 0.5 seconds, usually have maximal amplitude in the central region and are also involved in scoring stage N2 sleep (12). In polysomnography, different patterns recorded from eye movement electrodes (REMs, slow eye movements, reading movements, and blinks) and electromyographic recording of chin muscle tone are used in identifying sleep wake stages. The EEG of infants and children has several agerelated characteristic features of wakefulness and sleep. Dominant posterior rhythm (DPR) of wakefulness in infants and children is different from dominant posterior alpha rhythm in adults, usually includes intermixed slower EEG patterns, and changes with age. Infant DPR is typically 3.5 to 4.5 Hz by 3 to 4 months postterm and 5 to 6 Hz by 5 to 6 months of age. DPR normally becomes 8 Hz by age 3 years, 9 Hz by 9 to 10 years, and 10 Hz by age 15 years. Some EEG patterns play an important role

3 CHAPTER 3 SLEEP ACROSS THE LIFE CYCLE 17 in recognizing the transition from wakefulness to sleep in infants and children including slowing of the DPR by 1 to 2 Hz, runs of rhythmic 3 to 5 Hz activity, vertex sharp waves, high-voltage occipital delta slowing, hypnagogic hypersynchrony, postarousal hypersynchrony, and rhythmic anterior theta activity of drowsiness (12, 13). Hypnagogic hypersynchrony is represented by the high amplitude 75 to 350 μv waves in the frequency of 3 to 4.5 Hz, occurring in bursts or runs over the central, frontal, or frontocentral regions. Hypnagogic hypersynchrony is often seen in 3 months postterm infants, in almost all children between 6 months and 4 years, and becomes rare after 12 years. Rhythmic anterior theta activity is represented by the moderate voltage theta activity in the range of 5 to 7 Hz, occurring in runs in frontal derivations. Rhythmic anterior theta activity emerges at the age of 5 years, becomes maximal at 9 to 12 years age and declines at the age of 16 years in children. In full-term neonates, four EEG patterns are described by Anders and Parmelee scoring system in A manual of standardized terminology, techniques and criteria for scoring states of sleep and wakefulness in newborn infants (14). These patterns are low voltage irregular (LVI), tracé alternant (TA), high voltage slow (HVS), and mixed (M). LVI pattern with the 14 to 35 μv amplitude is characterized by predominately theta activity in the 5 to 8 Hz range with intermixed 1 to 5 Hz slow activity. TA pattern is represented by the 3 to 8 seconds long bursts of 0.5 to 3 Hz high amplitude slow waves separated by 4 to 8 seconds mixed frequency activity with lower amplitude. HVS is a moderately rhythmic pattern consisting of medium to high amplitude (50 to 150 μv) activity in the range of 0.5 to 4 Hz. A mixed (M) pattern usually has a amplitude lower than HVS and is characterized by intermingled HVS and low voltage components. Development of Sleep Wake EEG Patterns in Infants Neonate and infant EEG patterns, as well as circadian rhythmicity and sleep and waking behavioral states, depend more upon the actual developmental age of the brain following conception than on chronologic age following birth, which makes it important to define gestational age (GA), conceptional age (CA), and chronologic age (15, 16). Chronologic age is defined as the time in days, weeks, months, or years after birth (17). GA is usually expressed as completed weeks from the first day of the last normal menstrual period to the day of delivery (17). CA (13) is also expressed in weeks by adding the chronologic age to the GA (16). A normal 1-weekold infant born at 40 weeks GA and a normal 9-week-old premature infant born at 32 weeks GA are both the CA of 41 weeks and most likely will demonstrate similar EEG patterns (13). GA is often used in the literature describing premature neonates although GA and CA are not completely interchangeable terms. Infant EEG patterns are closely correlated with brain development and CA and change rapidly from neonatal to infant patterns, usually within the first 3 months of life. In premature infants <29 weeks of CA, the EEG has a continuously discontinuous pattern, called tracé discontinu, defined as brief periods of EEG activity separated by intervals of almost flat background. By the age of 34 to 37 weeks, tracé discontinu is replaced by the TA pattern, defined as bursts of moderate to high amplitude activity in the 0.5 to 3.0 Hz range with shorter 4- to 8-second interburst periods of flattening. TA pattern, signifying quiet sleep (QS), disappears in most infants at the end of or after the first month of life (44 to 46 weeks of CA) and is gradually replaced by more mature HVS wave activity (11, 13, 18). Two other EEG patterns in infants represent active sleep (AS): mixed pattern (M), combining low and high amplitude EEG activity, and LVI pattern, combining fast theta activity and slow activity. Sometimes HVS can be seen in AS and mixed pattern (M) in QS (14). Waking EEG of infants and children normally contains slow EEG activity with a background DPR (11, 13, 18). According to the AASM Pediatric Task Force, the DPR is 3.5 to 4.5 Hz by 3 to 4 months postterm and 5 to 6 Hz by 5 to 6 months of age in most normal infants (11, 13). In infants, sleep spindles usually appear at 2 to 3 months postterm, K-complexes at 5 to 6 months postterm, and slow-wave activity at 4 to 4.5 months postterm. Infants younger than 3 months postterm often, but not always, have AS (Stage R, REM) onset (12, 13). Development of Sleep Wake EEG Patterns in Children Sleep EEG of older infants (5 to 6 months postterm or older) and children contains sleep spindles, K-complexes, and slow-wave activity allowing differentiation of stages N1, N2, and N3 NREM sleep. Well-defined vertex sharp waves appear at 16 months postterm. Sleep EEG background consists of LAMF activity. The waking EEG of children normally contains slow EEG activity at the background DPR. DPR in most normal children is 8 Hz by age 3 years, 9 Hz by 9 to 10 years, and 10 Hz by age 15 years. Several EEG patterns play an important role in recognizing the transition from wakefulness to sleep: EEG slowing (slowing of the DPR by 1 to 2 Hz) runs of rhythmic 3 to 5 Hz activity and high voltage occipital delta slowing; and EEG patterns in mostly theta range: with vertex sharp transients, hypnagogic hypersynchrony, postarousal hypersynchrony, and rhythmic anterior theta activity of drowsiness (12, 13). In older infants and children, sleep onset is by way of NREM. Development of Sleep Wake EEG Patterns in Adolescents The EEG of adolescents is similar to the EEG of adults. The DPR in most normal children is 10 Hz by age 15 years, which is comparable to the frequency of the dominant

4 18 SECTION II ANATOMY AND PHYSIOLOGY alpha rhythm in adults. Some EEG patterns seen in children disappear or become less prominent in adolescents: occipital delta slowing is uncommon after age 6 years, rhythmic anterior theta activity of drowsiness reaches its maximum at the age of 9 to 12 years and then declines and is seen only in 15% of 16-year-olds, hypnagogic hypersynchrony becomes rare after 13 years, and postarousal hypersynchrony disappears usually after the age of 3 years. The waking EEG of adolescents may normally contain slow EEG activity at the background frequency of the DPR. Posterior slow waves of youth are prominent at 8 to 14 years and rare after age 21 years. Adolescent EEG characteristics such as vertex sharp waves, K-complexes, sleep spindles, slow-wave activity, LAMF activity, and DPR are analogous to adult EEG patterns (12, 13). EEG Patterns in Adults Dominant posterior alpha rhythm in wakefulness, LAMF activity in sleep, vertex sharp waves in NREM, K-complexes in NREM, sleep spindles in NREM, slow-wave activity in NREM, sawtooth waves in REM sleep, and shifts to faster frequencies during arousals constitute the primary sleep wake EEG patterns in normal adults. These patterns, along with patterns recorded from eye movement electrodes (REMs, slow eye movements, reading movements, and blinks) and electromyographic recording of chin muscle tone are used to identify sleep wake states and stages. Identifying Sleep Wake Stages in Infants <2 Months Postterm For scoring sleep wake stages in infants <2 months postterm, the 2007 scoring manual The AASM manual for the scoring of sleep and associated events: Rules, terminology, and technical specifications (AASM Scoring Manual) refers to the 2007 Pediatric Task Force article published in the Journal of Clinical Sleep Medicine (12, 13). This article describes several approaches offered by different authors for scoring sleep wake states in newborns and infants including A manual of standardized terminology, techniques and criteria for scoring states of sleep and wakefulness in newborn infants by Anders T, Emde R, and Parmelee A (14). The Anders and Parmelee manual is a widely recognized system used for scoring sleep in children <2 months postterm. According to Anders and Parmelee, polygraphic patterns, behavioral patterns, respiration, eye movements, and muscle tone are used in scoring states of sleep in infants (14). The manual defines active (REM) sleep, quiet (NREM) sleep, and indeterminate sleep (IS). Active REM sleep is characterized by LVI, mixed (M), and occasionally HVS EEG patterns with eyes closed, in combination with REMs, facial and body movements, irregular respiration, and low background EMG during periods of quiescence. QS is represented by HVS, TA, or mixed (M) EEG patterns with eyes closed, absence of REMs, almost no body movements and regular respiration. Episodes of EEG recording difficult to classify as AS or QS are classified as IS. Non sleep states include crying, active awake, and quiet awake and are determined using behavioral criteria (14). Identifying Sleep Wake Stages in Infants 2 Months Postterm The 2007 AASM Scoring Manual provides pediatric scoring rules that can be applied for infants 2 months postterm and older children. There is no upper age limit established for the use of these rules in pediatric patients (12). Rules for scoring stages N2, N3, and stage R are the same as adult rules. In cases when EEG patterns used to distinguish stages of NREM sleep (K-complexes and sleep spindles for stage N2 or slow-wave activity for stage N3) are not present, the 2007 AASM Scoring Manual recommends scoring epochs without these patterns as NREM sleep (stage N). The 2007 AASM Scoring Manual indicates that in most infants 5 to 6 months postterm, NREM sleep can be scored as stages N1, N2, and N3 (12). Identifying Sleep Wake Stages in Children and Adolescents Once EEG patterns distinguishing stages of NREM sleep (K-complexes and sleep spindles for stage N2 and slowwave activity for stage N3) are present in the pediatric EEG, the scoring rules are the same as for adults (12). Age-appropriate DPR frequencies and slow EEG activity seen on the background of the DPR are used in younger children to determine states of wakefulness and sleep. Scoring rules for adolescents are the same as the rules used for adults (12). Identifying Sleep Wake Stages in Adults The NREM sleep state is subdivided in different stages based on EEG characteristics that were described and classified by Loomis et al. (19) in 1937, by Gibbs and Gibbs (20) in 1950, by Dement and Kleitman (21) in 1957, and by Rechtschaffen and Kales (22) in In 2007, the American Academy of Sleep Medicine published The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications, where NREM sleep was characterized as three stages: N1, N2, and N3 (8). REM sleep (Stage R) is not divided into stages even though phasic and tonic patterns of REM sleep are differentiated. Scoring of sleep stages in clinical practice is performed using consecutive 30-second segments of recording called epochs and in general is based on the following principles: An epoch consisting of >50% alpha or faster frequencies is scored as Stage W (wakefulness). If more than 50% of the epoch is represented by LAMF activity in combination with slower activity, the epoch is identified as one of the stages of sleep. Dominant alpha rhythm in the range of 9 to 11 Hz (8.5 to 13 Hz) is observed in most normal healthy adults in relaxed wakefulness with eyes closed. Approximately 20% to 25% of subjects generate limited or no dominant alpha rhythm (22, 13).

5 CHAPTER 3 SLEEP ACROSS THE LIFE CYCLE 19 An epoch consisting of >20% of high amplitude (>75 μv) slow-wave activity in the range of 0.5 to 2 Hz, recorded from frontal or central regions, is scored as stage N3 (12). An epoch that does not meet the criteria for N3 but has K-complexes or sleep spindles beginning in the first half of the epoch, or has K-complexes or sleep spindles in one of the previous epochs and which are not followed by an arousal, an awakening or a transition to stage R, is scored as stage N2 (12). An epoch that does not meet criteria for stages N3 and N2 and has REMs and low chin EMG tone is scored as Stage R sleep (12). An epoch that does not meet criteria for stages N3, N2, and Stage R is scored as Stage N1 (12). If in a single epoch 2 or more stages are identified, the stage comprising the greatest part of the epoch is assigned. Sleep and Waking in the Fetus and the Premature Infant GA, CA, and chronologic age are defined above in this chapter. Prematurity is defined as CA <38 weeks (13). Ultradian rhythms can be observed during the fetal period and in the first days of postnatal life (4). A human fetus also becomes entrained in utero to maternal circadian rhythms by being exposed to changes in maternal melatonin secretion and body temperature (23). Neonates in the first days of their life have an ultradian sleep wake cycle. Circadian and homeostatic processes emerge in the first days/weeks of life with entrainment to a 24-hour cycle after 3 months and sleep consolidation through the night at around 6 to 9 month of age (4, 2). After 24 weeks, CA brain electrical activity can be described as a discontinuous pattern (18). At 27 weeks, CA NREM periods correlating with breathing movements and REM periods have been observed in utero using real-time ultrasonic scanners (24). Description of the presence of states resembling AS and QS as early as 27 weeks of CA in premature infants is another important point in understanding sleep wake development (25). Several recent studies confirm the ability to identify different states of sleep and ultradian cyclicity in infants <30 weeks of CA (26 29). At 28 to 32 weeks of CA, the EEG still shows a discontinuous pattern with the appearance of fast 10 to 20 Hz activity superimposed on slow waves that are called delta brushes (18). The 2007 AASM Pediatric Task Force article (13) indicates that at 32 weeks of CA, EEG activity is still undifferentiated, and it is possible to distinguish only behavioral states of sleep and wakefulness based on eye movements, body movements, and changes in heart rate and respiration (30, 31). At 32 to 34 weeks of CA, it is possible to see EEG reactivity to vigorous stimulation (11). At 32 to 36 weeks of CA, a continuous EEG pattern corresponding to AS and wakefulness emerges (18). The amount of undifferentiated sleep decreases after 34 weeks of CA, which indicates another turning point in sleep wake mechanism development (4). At 36 to 37 weeks of CA, a tracé alternant pattern begins replacing the tracé discontinue pattern in QS (11, 23). At 38 weeks of CA, a HVS pattern begins to replace the tracé alternant pattern (which disappears after 44 weeks of CA) (13), and four EEG patterns are present. They are tracé alternant, HVS, LVI, and mixed (11, 18). States of wakefulness, AS, and QS now can be clearly characterized based on combination of EEG and behavioral patterns. Newborns, Toddlers, and Young Children Total sleep time for newborns and infants during the first year of life averages 14 hours per day, although the range of individual variability is very high, and some newborns will sleep 10 to 11 hours whereas others will sleep 17 to 19 hours. The range of interindividual variability narrows by the age of 12 months. Newborns have an ultradian rhythm of sleep with many short and variable in duration episodes of sleep with brief awakenings between the sleep episodes (32, 33). Circadian rhythmicity evolves during the first weeks of life. Initially, it may take the form of the 25-hour free running cycle, and later by the age of 2 to 4 months, there is a clear time cues dependent entrainment to the 24-hour cycle (2). Episodes of wakefulness become longer in the afternoon and evening, and sleep is consolidated in a longer sleep episode at night. The longest sustained periods of sleep are about 4 hours in neonates and about 7 to 9 hours in the 6- to 12-month-old infants (34). The longest sleep episodes follow the periods of prolonged wakefulness (35), which may be an expression of the developing homeostatic control of sleep (2). Sustained night periods of sleep in older toddlers around the age of three will reach 10 to 12 hours, similar to durations usually seen in prepubertal children, although interindividual differences may exceed 2 hours (36). The number of daytime sleep episodes gradually decreases during the first year of life, with daytime sleep concentrating into one long nap, which will occur around noontime in toddlers and later in life in the midafternoon (37). Nap patterns differ significantly among various cultural groups and among individual children. Most children will stop napping at the age of 3 to 5 years (2). Brief arousals, possibly driven by the 50- to 90-minute rhythm of ultradian sleep cycles, and less frequent longer arousals are commonly observed in toddlers and preschoolers and may result in awakenings. They are considered a normal developmental feature, unless children cannot soothe themselves back to sleep without the intervention of caregiver, when it becomes a behavioral problem (2). Nocturnal arousals may become more frequent in infants after the age of 9 months but usually disappear after the age of 3 years (36). Cultural or lifestyle bed-sharing or roomsharing may increase the number of arousals in infants (38). The presence of a parent at bedtime was found to be a predictor of less total nighttime sleep in infants (39).

6 20 SECTION II ANATOMY AND PHYSIOLOGY Infants have ultradian (<24 hour) sleep wake organization during the first weeks of life. Infants become entrained to a circadian 24-hour rhythm around 3 months of life (4). Sleep onset in newborns is mostly into REM sleep, becoming NREM at the age of 3 to 6 months (40). Newborns sleep is around 50% active (REM) sleep and 50% quiet (NREM) sleep. The amount of quiet (NREM) sleep increases with age in children and the amount of active (REM) sleep decreases, reaching adult levels of REM at the age of 2 to 6 years. The number of REM episodes decreases in older infants and in toddlers because of the decrease in the number of naps (4). The ultradian NREM REM sleep cycle is shorter in children than in adults, gradually increasing with age. NREM REM sleep cycle is about 50 to 60 minutes in infants, about 75 minutes in toddlers, and about 80 to 90 minutes by the age of 6 years. As daytime naps are eliminated, typically there is a high percentage of slow-wave sleep (SWS) in the first third of the night especially during the first NREM sleep period (41). There is a very strong need for SWS in these years, often causing the first REM sleep episode of the night to be skipped, and the latency to REM sleep may increase to 180 minutes. After the first NREM period, the percentage of SWS declines over the night in children after infancy. In young children, sleep in the beginning of the night may become so deep that it may be nearly impossible to awaken the toddler without applying very strong stimuli. In the period from about 5 years to puberty, the longest sleep period remains quite stable at around 10 hours a night. Bed times are to a high degree under parental control. Morning rise time is fixed during the week because of school start time. However, prepubertal children often wake by themselves and rise early even on nonschool days. Sleep propensity during the day is very low (42). It appears that individual sleep patterns (short sleeper or long sleeper) becoming apparent in school-age children may start emerging in early childhood (2). In an extensive review of the literature, Ohayon et al (43). report age-related trends regarding quantitative sleep parameters from 5 years onward. Adolescents Sleepiness and alertness are direct correlates of the sleep wake system, and the circadian rhythm leads to an increased alertness in the morning and afternoon and a reduced alertness during nocturnal and lunch hours. Sleeping and waking during the second decade of life undergo profound changes (44). The most conspicuous changes in adolescents as compared with prepubertal children are a phase delay of sleep and waking, increasing sleep propensity and excessive tiredness or sleepiness in the morning, and decrease in total sleep time and sleep debt. Phase Delay of Sleep and Waking Adolescents stay up later in the evening both on school nights and on weekends. Also, they sleep more in the morning and early day, waking up later during weekends and holidays than during the week. Recent analysis of worldwide sleep pattern surveys demonstrated differences in adolescence sleep patterns in different countries and regions, but delayed sleep phase pattern was noticed worldwide (45). It has been proposed that across adolescence, the daily accumulation of sleep pressure inherent in the homeostatic process will become slower, facilitating adolescents staying awake for a longer time (46, 47). Increasing Sleep Propensity and Excessive Tiredness or Sleepiness in the Morning As children reach puberty, there is a significant change in daytime sleep propensity with increasing sleep propensity in the morning. Many teenagers demonstrate sleep latencies bordering on pathologic sleepiness, especially on morning multiple sleep latency tests (MSLTs) performed at 8:30 and 10:30. In a study on sleep patterns, circadian timing, and sleepiness at the transition from the 9th to the 10th grade, which necessitates a transition to an earlier school start time in the United States, high school students as a group fell asleep within 8.5 minutes, and a third of the adolescents studied had sleep latencies between 6 and 1.8 minutes on the morning MSLTs (46). REM sleep may be observed on one or more MSLTs in the first part of the day in this age group without having anything to do with narcolepsy (46). The occurrence of morning MSLT REM sleep episodes in adolescents may be explained by the timing inherent in their circadian rhythm; that is to say, their body is still experiencing the end of night in spite of having arisen one or more hours before the MSLT was performed. These findings are supported by survey data from the World Health Organization (WHO) reporting excessive morning tiredness or sleepiness in adolescents from 28 countries (48). Overall and for most countries, the level of feeling tired in the morning four or more times a week on school days increased with age in both girls and boys. Decrease in Total Sleep Time and Sleep Debt In addition, there is a dramatic fluctuation of total sleep time as well as bedtimes and rise times between weekdays and weekends. Across the twentieth century, a decrease in total sleep time was observed in all age groups, including adolescents. The Zurich Longitudinal Studies show an increasingly later bedtime but an unchanged wake time across cohorts across decades (49). It is difficult to characterize an overall normal sleep time; variability from person to person is high. Volitional control is one of the most important factors regulating length of sleep, and genetic factors also play an important role. About 9 hours of nightly sleep is assumed to be optimal for alertness in adolescents, but they sleep considerably less on school nights. Less than 50% of students in 10th grade in Rhode Island, United States, obtained as much as an average of 7 hours. The vast majority of students slept longer on nonschool

7 CHAPTER 3 SLEEP ACROSS THE LIFE CYCLE 21 nights, but on the average much <9 hours (46). In a longitudinal sleep camp study, children between the ages of 10 and 12 years were followed each summer for 4 to 6 years. When in the sleep center, it was observed that the sleep amount did not decrease across pubertal (Tanner) stages. The more mature adolescents were the ones who often had to be awoken even after 10 hours of sleep (50). Carskadon and coworkers (51) observed a very rapid sleep onset at 6 p.m. after being awake for only 10 hours before bedtime in a group of adolescents who had slept 10 hours a night for a whole week. This may suggest that teenagers have a chronic sleep debt. There may be several reasons for these changes. Psychosocial issues may be more pressing in these years of life. Academic and social demands usually increase, and many adolescents experience psychological intra- and interpersonal problems associated with the process of growing up. There is an increasing degree of autonomy; parental control is more or less lost during these years (50). Also, biologic changes in the brain processes regulating sleep and wakefulness, that is, the homeostatic and the circadian mechanisms, appear linked to pubertal development (52). Self-reports of pubertal development correlate with the circadian phase preference changes seen in adolescents. In addition, physical measures of pubertal development correlate with the time point at which the hormone melatonin starts to increase in the evening (53). Effect of Development and Insufficient and Irregular Sleep Across adolescence, from early to late pubertal development, there is a substantial decrease in SWS of around 40% even when total sleep time is constant and sleep is not restricted by scheduling constraints. When nocturnal sleep is restricted, SWS may increase slightly. A longitudinal sleep camp study demonstrated that REM sleep as a percentage of total sleep time is maintained if total sleep time is constant, and that REM latency moves forward from about 150 to 100 minutes around this age. When adolescents are on a self-selected schedule, however, their sleep patterns may give rise to peculiar EEG changes that are otherwise encountered only in pathologic conditions. Mentioned above is the intrusion of REM sleep on MSLTs. Also, premature episodes of REM sleep (sleep onset REM periods or SOREMPs) may be seen as a consequence of chronic insufficient nocturnal sleep and an irregular sleep schedule. On recovering from sleep deprivation, several changes may occur, with some variations depending on whether it is the first, second, or third recovery night. In general, on the first recovery night, sleep latency is shorter, sleep is deeper and longer, and there is a higher threshold to arousal. Usually recovery sleep has a higher amount of SWS and stage N2 sleep, which accumulates early, and a lower amount of wakefulness. The NREM REM cycle is intact, but there may be a lengthening of REM latency in the first sleep cycle because of the pressure to catch up on the SWS. On the second or subsequent nights, usually there is an REM rebound after the recuperation of SWS. An irregular sleep wake rhythm may also lead to difficulties falling asleep until late at night, or lead to fragmented sleep with frequent awakenings and sometimes an increase of jerks in the musculature or hypnogenic myoclonia (54). Adult Ideal Sleepers Reported night sleep durations demonstrate a high variability from person to person and from weekday to weekend nights, appearing to be about an hour longer in young adults compared with middle-aged adults and an hour longer on weekend nights compared with weekday nights (3, 55 57). In general, approximately 7.5 hours of sleep on weekdays and approximately 8.5 hours of sleep on weekends are reported by the majority of young adults. Night sleep duration is also determined by genetic sleep need, circadian rhythms, and volitional factors (3). Normal sleep in healthy young adults demonstrates sleep states and stages with a certain progression, sequence, distribution, and cyclic ultradian NREM REM alternating patterns. In a normal young adult, sleep begins with a brief, several minutes long, period of stage N1 sleep represented mostly by LAMF in the EEG. Stage N1 is followed by stage N2 with K-complexes and/or sleep spindles appearing. Stage N3 follows stage N2 when the amount of high amplitude slow-wave EEG activity progresses and meets the 20% scoring criteria for N3 sleep. Arousal threshold, or the intensity of stimulus required to produce arousal, is the lowest during stage N1 and the highest during stage N3 sleep. Stage R follows stage N3 in the normal young adult. This sequence of sleep stage progression is called the NREM REM cycle, and during the whole night sleep period, there are several (four to six) ultradian NREM REM cycles with an average length of 90 to 110 minutes. In the first third of the night, SWS predominates, and in the last third, REM sleep predominates. NREM sleep (N1+N2+N3) constitutes 70% to 80% of the total night s sleep duration. REM sleep (Stage R) occupies 20% to 25% of the total night s sleep duration, and wakefulness accounts for <5% of the night (3). Some recent changes in terminology and technical, recording, and scoring specifications introduced in 2007 in the AASM Scoring Manual may require a reassessment of the normative data for sleep stages duration (3). Primary Sleep Disorders and Sleep Wake Patterns in Adulthood Sleep wake pattern and sleep stage distribution in adulthood are affected by many factors including primary sleep disorders, other medical disorders, medications, alcohol intake, sleep hygiene, sleep history, and age. As we age, the prevalence of sleep-disordered breathing and periodic limb movements in sleep (PLMS) increases, affecting sleep wake patterns in the adult population.

8 22 SECTION II ANATOMY AND PHYSIOLOGY Sleep-disordered breathing may result in a decrease in the amount of SWS and/or REM sleep caused by the sleep apnea-related problems. The amount of stages N1 and N2 may increase as well as the frequency of arousals and sleep fragmentation. Successful treatment of sleep apnea syndromes may decrease the frequency of arousals and sleep fragmentation, also leading to the rebounds of REM and SWS. Periodic limb movements (PLM) may be associated with sleep fragmentation and high frequency of arousals. It appears that PLMS can be associated with the significant decrease in sleep duration and poor sleep quality; however, the variability of the amount of PLMS from night to night, the PLM index, and the association of PLMS with the restless leg syndrome (RLS) should be taken into consideration (58). RLS is associated with insomnia. RLS is a clinical condition defined by the presence of an irresistible urge to move the legs, accompanied by uncomfortable sensations (such as pain, ache, tingling, vibration), which are worse during periods of rest or inactivity, or at certain times of day, such as the late afternoons, early evenings, or at bedtime, and are noticeably less severe upon awakening in the morning. Movement of the legs, stretching, or massaging brings about transient relief. The prevalence of RLS increases with age with peak prevalence, according to some studies, between 50 and 59 years for men and between 60 and 69 years for women (58). Narcolepsy affects approximately 0.05% of general population. Its prevalence in different countries is variable because of genetic factors (59). Narcolepsy symptoms include excessive daytime sleepiness and irresistible sleep attacks, which may or may not be associated with cataplexy, hypnagogic or hypnopompic hallucinations, and sleep paralysis. The peak age of the onset of the symptoms is between the ages of 15 and 25 years with another smaller peak between 35 and 45 years. Sleep fragmentation and frequent awakenings may also be associated with narcolepsy. On the MSLT, patients with narcolepsy have short sleep latency and usually two or more sleep-onset REM episodes. The most common sleep disorder is insomnia. It is necessary to underline that sleep complaints or symptoms of insomnia may indicate one of primary or comorbid insomnia disorders, a symptom, or a combination of both. Estimates of the prevalence of insomnia in general population depend on the variety of definitions of insomnia and range (60) from 4% to 48%. AASM has not recommended the routine use of the attended polysomnography for the evaluation of transient or chronic insomnia without sleeprelated breathing disorder symptoms or atypical clinical presentation (61), but in the practice of a sleep center, many patients undergoing sleep studies for the evaluation of sleep-disorder breathing have symptoms of insomnia. Owing to the variety of clinical forms of insomnia, PSG measures in patients with insomnia symptoms may differ significantly. In psychophysiologic insomnia patients, sleep architecture can be relatively normal with some increase of stage N1 and a variable amount of SWS. In cases of psychophysiologic insomnia, sleep latency may be increased and total sleep time and sleep efficiency may be decreased, which can help to differentiate it from another type of insomnia sleep state misperception, in which normal sleep latency and sleep duration are observed with normal number of arousals and awakenings. In cases of idiopathic insomnia, sleep latency and the number of awakenings increase, sleep efficiency is reduced, and the amount of stage N1 may be increased with the major reduction of SWS (62). Sleep stage distribution in subjects with primary or comorbid sleep disorders may be affected also by the use of drugs prescribed for the treatment of sleep disorders for the treatment of comorbid conditions or by the use of recreational drugs. Sleep in Old Age In a recent review, Stepnowsky and Ancoli-Israel (63) described normal and abnormal age-related changes in sleep in elderly. In middle-aged and old-aged individuals, the subjective evaluations of sleep typically show large variations as age-related sleep changes are perceived by some as normal changes because of aging, whereas others perceive the same changes as sleep problems. Subjective changes in sleep may include longer time needed to fall asleep, frequent awakenings, less time spent asleep and spending more time in bed, low satisfaction with the quality of sleep, daytime sleepiness, and napping. Women more often present with complaints about their sleep, whereas men often experience serious problems without complaining. Many elderly show an altered timing of sleep and waking; they wake early in the morning and go to bed very early. Total time in bed often increases. Some spend as long as 12 hours or more in bed per day. This is a common strategy used by older people who think their sleep has become worse. Sleep efficiency is often reduced to 70% to 80% because actual sleep time is much shorter than time in bed. Most studies report that older people sleep less, but total sleep time over the 24-hour day and night may be unchanged when naps are included (64). The large variability in sleep among older subjects probably reflects some of the larger variability in health status as well as the larger variability of lifestyle in these age groups as compared with the young (65). Several external factors associated with retirement, changes in social interaction, an inactive sedentary lifestyle with very little time outdoors or in bright daylight, institutionalization, use of multiple medications, caffeine and alcohol intake, and a wide range of physical and mental health issues common in elderly people may be more important for sleep than the processes associated with aging per se. Sleep Duration Generally, elderly people have reduced sleep efficiency and frequently reduced total night sleep time, which is

9 CHAPTER 3 SLEEP ACROSS THE LIFE CYCLE 23 often interpreted as changes in ability to sleep with aging, while the need for sleep remains unchanged. In studies comparing healthy elderly with younger subjects, it was demonstrated that the healthy elderly slept less than their younger counterparts. In the recent study by Dijk et al. (66), young adults slept approximately 23 minutes more than middle-aged adults, who slept approximately 20 minutes more than older adults. The authors of this study found a reduction in daytime sleep propensity and a decrease in sleep continuity and SWS in healthy elder adults. This suggests that there may be a slight change in the overall need for sleep as we age, which can be attributed to the possible attenuation in homeostatic sleep requirement, meaning that older adults maybe accumulate less of a sleep debt during wakefulness. It was hypothesized that healthy elderly adults need less sleep to maintain alertness. Elderly people often take daytime naps, which should be considered in evaluating 24-hour total sleep time in elder adults. Sleep Fragmentation Difficulty maintaining sleep continuity is the most commonly reported sleep problem in this age group and is reported by about one-third or more of elderly individuals in cross-sectional surveys (67). The age-related increase in wakefulness is most prominent during the second half of the night, with more full arousals as well as more transient arousals. The number of nocturnal arousals can amount to a considerable disruption in sleep. Arousals are frequently related to illnesses or sleep disorders, however, and sleeprelated respiratory disturbances such as apneas and hypopneas show a high prevalence in the elderly. Phase Advance Sleep habits tend to be more stable in the elderly. For instance, in a study among healthy centenarians, the majority reported a fixed early sleep time (8:30 p.m.), a fixed final wake time, a regular afternoon nap, and the casual use of chamomile tea as part of their bedtime ritual. The centenarians reported an average of three interruptions of sleep per night to use the bathroom, and their opinion of their sleep and waking pattern was good (49). The sleep wake patterns of these centenarians exhibit characteristic aspects frequently seen in older age groups, such as a tendency toward a phase advance, an increase of time in bed, and a general need for napping. Circadian Rhythm and Disturbed Sleep Circadian amplitude is attenuated with aging (68). The reduced amplitude of temperature, especially seen in men, leads to an increase in daytime sleepiness and a decrease in daytime alertness. The attenuation of the circadian amplitude may also contribute to the increased fragmentation and the reduced sleep quality observed with increasing age. Finally, reduced entrainment of sleep in the elderly may cause difficulties with travels across time zones, that is, jet lag and shift work. Levels of melatonin have been shown to decrease with aging (69), and it was suggested that a decrease in the level of melatonin is directly related to the decreased sleep levels seen in the elderly. In other studies, it was demonstrated that the decrease in melatonin secretion occurs in good sleepers as well as in poor sleepers (7). On the other hand, in several studies on insomnia patients, administration of exogenous melatonin resulted in sleep quality improvement and improved morning alertness. Modification of Sleep Stages The quantitative decrease observed in SWS across adolescence continues across adulthood. The physiologic background is poorly understood, however, and may be a consequence of brain neuronal maturity rather than a function of age only. This decrease in SWS may explain why middle-aged and older people often experience lighter sleep that is more easily disturbed. Men are subject to a more pronounced decrease, and at 60 years, some men no longer have SWS, whereas women may preserve SWS longer. In particular, the EEG delta amplitude is reduced and frequency may be increased. Decrease of delta in the elderly was demonstrated by visual scoring as well as computer analysis. In one of the studies, it was observed that the average all-night NREM power density in the range 0.3 to 4 Hz in the elderly was 60% lower than in young adults, and in young adults, it was 70% lower than in children (70). Reduction of delta in the elderly is often interpreted as attenuated homeostatic drive although the proportions of delta increments following waking are similar in the elderly and the young (71). Stage N3 sleep may be completely obliterated in the 90-year-old person. In contrast, REM sleep as a percentage of total sleep time is maintained at 20% to 25% in the healthy elderly. In the demented, age-related changes become more pronounced and there is a reduction of REM sleep, which has been correlated with intellectual functioning (72). In general, changes in the following polysomnographic parameters may be observed in the elderly: sleep latency, time in bed, the number of arousals, the number of awakenings, and the percentage of stages N1 and N2 increase, whereas sleep efficiency, total sleep time, and SWS decrease. Primary Sleep Disorders in Old Age Advanced Sleep Phase Syndrome Circadian rhythm sleep disorders in the elderly may be attributed to the age-related changes in the SCN, to the reduction in melatonin secretion with age, and to the weak environmental factors entraining circadian rhythms (63). Little time outdoors and institutionalization reduce the exposure to bright daylight, which is one of the strongest circadian rhythms entraining exogenous cues. Bright light exposure is known to stabilize the circadian sleep wake cycle.