Two Types of Sleep

Over the course of a period of sleep, NREM and REM sleep alternate cyclically (Figure 2-1). The function of alternations between these two types of sleep is not yet understood, but irregular cycling and/or absent sleep stages are associated with sleep disorders (Zepelin et al., 2005). For example, instead of entering sleep through NREM, as is typical, individuals with narcolepsy enter sleep directly into REM sleep (Carskadon and Rechtschaffen, 2005).

FIGURE 2-1

Progression of sleep states across a single night in young adult. SOURCE: Carskadon and Dement (2005).

Four Stages of NREM Sleep

The four stages of NREM sleep are each associated with distinct brain activity and physiology. Figure 2-2 shows the EEG patterns characteristic of the four NREM stages. Other instruments are used to track characteristic changes in eye movement and muscle tone.

FIGURE 2-2

Characteristic EEG activity of each of the four stages of NREM sleep. NOTE: In stage 2, the arrow indicates a K-complex, and the underlining shows two sleep spindles.

REM Sleep

REM sleep is defined by the presence of desynchronized (low-voltage, mixed-frequency) brain wave activity, muscle atonia, and bursts of rapid eye movements (Carskadon and Dement, 2005). “Sawtooth” wave forms, theta activity (3 to 7 counts per second), and slow alpha activity also characterize REM sleep. During the initial cycle, the REM period may last only 1 to 5 minutes; however, it becomes progressively prolonged as the sleep episode progresses (Carskadon and Dement, 2005). There are numerous physiological differences between NREM and REM sleep (Table 2-1).

TABLE 2-1

Physiological Changes During NREM and REM Sleep.

Dreaming is most often associated with REM sleep. Loss of muscle tone and reflexes likely serves an important function because it prevents an individual from “acting out” their dreams or nightmares while sleeping (see Chapter 3) (Bader et al., 2003). Approximately 80 percent of vivid dream recall results after arousal from this stage of sleep (Dement and Kleitman, 1957b). REM sleep may also be important for memory consolidation (Crick and Mitchison, 1983; Smith and Lapp, 1991).

Physiology During Sleep

In addition to the physiological changes listed in Table 2-1, there are other body system changes that occur during sleep. Generally, these changes are well tolerated in healthy individuals, but they may compromise the sometimes fragile balance of individuals with vulnerable systems, such as those with cardiovascular diseases (Parker and Dunbar, 2005). Physiological changes also occur in the following systems:

• Cardiovascular: Changes in blood pressure and heart rate occur during sleep and are primarily determined by autonomic nervous system activity. For instance, brief increases in blood pressure and heart rate occur with K-complexes, arousals, and large body movements (Lugaresi et al., 1978; Catcheside et al., 2002; Blasi et al., 2003; Tank et al., 2003). Further, there is an increased risk of myocardial infarction in the morning due to the sharp increases in heart rate and blood pressure that accompany awakening (Floras et al., 1978; Mulcahy et al., 1993).
• Sympathetic-nerve activity: Sympathetic-nerve activity decreases as NREM sleep deepens; however, there is a burst of sympathetic-nerve activity during NREM sleep due to the brief increase in blood pressure and heart rate that follows K-complexes. Compared to wakefulness, there is a rise in activity during REM sleep (Somers et al., 1993).
• Respiratory: Ventilation and respiratory flow change during sleep and become increasingly faster and more erratic, specifically during REM sleep (Krieger, 2000; Simon et al., 2002). Ventilation data during REM sleep are somewhat unclear, but they suggest that hypoventilation (deficient ventilation of the lungs that results in reduction in the oxygen content or increase in the carbon dioxide content of the blood or both) occurs in a similar way as during NREM sleep (NLM, 2006). Several factors contribute to hypoventilation during NREM, and possibly REM, sleep such as reduced pharyngeal muscle tone (Krieger, 2000; Simon et al., 2002). Further, during REM sleep, there is reduced rib cage movement and increased upper airway resistance due to the loss of tone in the intercostals and upper airway muscles (Parker and Dunbar, 2005). More generally, ventilation and respiratory flow show less effective adaptive responses dur ing sleep. The cough reflex, which normally reacts to irritants in the airway, is suppressed during REM and NREM sleep. The hypoxic ventilatory response is also lower in NREM sleep than during wakefulness and decreases further during REM sleep. Similarly, the arousal response to respiratory resistance (for example, resistance in breathing in or out) is lowest in stage 3 and stage 4 sleep (Douglas, 2005).
• Cerebral blood flow: NREM sleep is associated with significant reductions in blood flow and metabolism, while total blood flow and metabolism in REM sleep is comparable to wakefulness (Madsen et al., 1991b). However, metabolism and blood flow increase in certain brain regions during REM sleep, compared to wakefulness, such as the limbic system (which is involved with emotions), and visual association areas (Madsen et al., 1991a).
• Renal: There is a decreased excretion of sodium, potassium, chloride, and calcium during sleep that allows for more concentrated and reduced urine flow. The changes that occur during sleep in renal function are complex and include changes in renal blood flow, glomerular filtration, hormone secretion, and sympathetic neural stimulation (Cianci et al., 1991; Van Cauter, 2000; Buxton et al., 2002).
• Endocrine: Endocrine functions such as growth hormone, thyroid hormone, and melatonin secretion are influenced by sleep. Growth hormone secretion typically takes place during the first few hours after sleep onset and generally occurs during SWS, while thyroid hormone secretion takes place in the late evening. Melatonin, which induces sleepiness, likely by reducing an alerting effect from the suprachiasmatic nucleus, is influenced by the light-dark cycle and is suppressed by light (Parker and Dunbar, 2005).

The Two-Process Model

The sleep-wake system is thought to be regulated by the interplay of two major processes, one that promotes sleep (process S) and one that maintains wakefulness (process C) (Gillette and Abbott, 2005). Process S is the homeostatic drive for sleep. The need for sleep (process S) accumulates across the day, peaks just before bedtime at night and dissipates throughout the night.

Process C is wake promoting and is regulated by the circadian system. Process C builds across the day, serving to counteract process S and promote wakefulness and alertness. However, this wake-promoting system begins to decline at bedtime, serving to enhance sleep consolidation as the need for sleep dissipates across the night (Gillette and Abbott, 2005). With an adequate night’s rest, the homeostatic drive for sleep is reduced, the circadian waking drive begins to increase, and the cycle starts over. In the absence of process C, total sleep time remains the same, but it is randomly distributed over the day and night; therefore, process C also works to consolidate sleep and wake into fairly distinct episodes (Gillette and Abbott, 2005). Importantly, through synchronization of the circadian system, process C assists in keeping sleep-wakefulness cycles coordinated with environmental light-dark cycles.

Sleep-Generating Systems in the Brainstem

Sleep process S is regulated by neurons that shut down the arousal systems, thus allowing the brain to fall asleep. Many of these neurons are found in the preoptic area of the hypothalamus (Figure 2-3A). These neurons, containing molecules that inhibit neuronal communication, turn off the arousal systems during sleep. Loss of these nerve cells causes profound insomnia (Saper et al., 2005a,c). Inputs from other regions of the brain also greatly influence the sleep system. These include inputs from the lower brainstem that relay information about the state of the body (e.g., a full stomach is conducive to falling asleep), as well as from emotional and cognitive areas of the forebrain. In addition, as described further in the next section, there are inputs from the circadian system that allow the wake-sleep system to synchronize with the external day-night cycle, but also to override this cycle when it is necessitated by environmental needs.

FIGURE 2-3

Sleep-generating (A) and wake-generating (B) systems in the brain. NOTE: Cholinergic (ACh) cell groups; basal forebrain (BF); dopamine (DA); gamma-aminobutyric acid (GABA); galanin (Gal); histamine (His); serotonin (5-HT); locus coeruleus (LC); laterodorsal (more...)

The sleep-generating system also includes neurons in the pons that intermittently switch from NREM to REM sleep over the course of the night. These neurons send outputs to the lower brainstem and spinal cord that cause muscle atonia, REMs, and chaotic autonomic activity that characterize REM sleep. Other outputs are sent to the forebrain, including activation of the cholinergic pathways to the thalamus to activate the EEG.

Wake-Generating Systems in the Brainstem

Wakefulness is generated by an ascending arousal system from the brainstem that activates forebrain structures to maintain wakefulness (Figure 2-3B). This idea, originally put forward by Morruzzi and Magoun (1949), has more recently been refined (Jones, 2005a; Saper et al., 2005c). The main source for the ascending arousal influence includes two major pathways that originate in the upper brainstem. The first pathway, which takes origin from cholinergic neurons in the upper pons, activates parts of the thalamus that are responsible for maintaining transmission of sensory information to the cerebral cortex. The second pathway, which originates in cell groups in the upper brainstem that contain the monoamine neurotransmitters (norepinephrine, serotonin, dopamine, and histamine), enters the hypothalamus, rather than the thalamus, where it picks up inputs from nerve cells that contain peptides (orexin or hypocretin and melanin-concentrating hormone). These inputs then traverse the basal forebrain, where they pick up additional inputs from cells containing acetylcholine and gamma-aminobutyric acid. Ultimately, all of these inputs enter the cerebral cortex, where they diffusely activate the nerve cells and prepare them for the interpretation and analysis of incoming sensory information.

The Suprachiasmatic Nucleus

The suprachiasmatic nucleus (SCN) is responsible for regulating circadian rhythms in all organs. It receives direct inputs from a class of nerve cells in the retina that act as brightness detectors, which can reset the clock genes in the SCN on a daily basis. The SCN then transmits to the rest of the brain and body signals that bring all of the daily cycles in synchrony with the external day-night cycle.

The main influence of the SCN on sleep is due to a series of relays through the dorsomedial nucleus of the hypothalamus, which signals to the wake-sleep systems to coordinate their activity with the day-night cycles. The SCN also coordinates cycles of feeding, locomotor activity, and hormones, such as corticosteroids (Chou et al., 2003). Under some conditions (e.g., limited food availability) when there are changes in the external temperature, or even under conditions of behavioral stress (e.g., the need to avoid a predator), animals must shift their daily cycles to survive. In such circumstances, the dorsomedial nucleus may shift to a new daily cycle, which can be completely out of phase with the SCN and the light-dark cycle, and its signals also shift the daily cycles of sleep, activity, feeding, and corticosteroid hormone secretion (Saper et al., 2005b,c).

Another major output of the SCN is to a pathway that controls the secretion of melatonin, a hormone produced by the pineal gland. Melatonin, which is mainly secreted at night, acts to further consolidate the circadian rhythms but has only limited effects directly on sleep.

Sleep and Thermoregulation

Body temperature regulation is subject to circadian system influence. An individual’s body temperature is higher during the day than at night (Figure 2-5). At night there is a gradual decline in body temperature, a decrease in heat production (called the falling phase of the body temperature rhythm), and an increase in heat loss, all which promote sleep onset and maintenance, as well as EEG slow-wave activity. Conversely, there is a gradual increase in body temperature several hours before waking. The brain sends signals to other parts of the body that increase heat production and conservation in order to disrupt sleep and promote waking (Szymusiak, 2005).

FIGURE 2-5

Body temperature in relation to time of day. SOURCE: NHBLI (2003).

Newborns and Infants

At birth, sleep timing is distributed evenly across day and night for the first few weeks, with no regular rhythm or concentration of sleeping and waking. Newborns sleep about 16 to 18 hours per day; however, it is discontinuous with the longest continuous sleep episode lasting only 2.5 to 4 hours (Adair and Bauchner, 1993; Roffwarg et al., 1966). Newborns have three types of sleep: quiet sleep (similar to NREM), active sleep (analogous to REM), and indeterminate sleep (Davis et al., 2004). Sleep onset occurs through REM, not NREM, and each sleep episode consists of only one or two cycles (Jenni and Carskadon, 2000; Davis et al., 2004). This distinctive sleep architecture occurs mostly because circadian rhythms have not yet been fully entrained (Davis et al., 2004).

Circadian rhythms begin to arise around 2 to 3 months of age, leading to sleep consolidation that manifests in greater durations of wakefulness during the day and longer periods of sleep at night (Sheldon, 2002). Circadian rhythm development in the first 3 months includes: emergence of the 24-hour core body temperature cycle (1 month of age); progression of nocturnal sleeping (2 months of age); and cycling of melatonin and cortisol hormones in a circadian rhythm (3 months of age) (Jenni and Carskadon, 2000).

Sleep cycles also change because of the emergence of the circadian rhythm and a greater responsiveness to social cues (such as breast-feeding and bedtime routines). By 3 months of age, sleep cycles become more regular: sleep onset now begins with NREM, REM sleep decreases and shifts to the later part of the sleep cycle, and the total NREM and REM sleep cycle is typically 50 minutes (Anders et al., 1995; Jenni and Carskadon, 2000). By 6 months of age, total sleep time reduces slightly and the longest continuous sleep episode lengthens to approximately 6 hours (Anders et al., 1995; Jenni and Carskadon, 2000). As sleep cycles mature, the typical muscle paralysis of REM sleep replaces the propensity for movement in what was called “active sleep” as a newborn. By 12 months old, the infant typically sleeps 14 to 15 hours per day with the majority of sleep consolidated in the evening and during one to two naps during the day (Anders et al., 1995).

Young Children

There are a limited number of studies that address normal sleep architecture in young children; however, one trend that appears to be consistent is that sleep amounts decrease as a child gets older. The reduction cannot be attributed solely to physiologic requirements, because cultural environments and social changes also influence changing sleep characteristics in young children. Total sleep time decreases by 2 hours from age 2 to age 5 (13 hours to 11) (Roffward et al., 1966). Socially, the decrease in time asleep may be a result of decreased daytime napping, as most children discontinue napping between 3 and 5 years old (Jenni and Carskadon, 2000). Other social and cultural factors that begin to influence sleep include how, with whom, and where children sleep and the introduction of school time routines (Jenni and O’Connor, 2005).

Physiologically, it has been suggested that by the time children enter school (typically 6 years old) they begin to manifest circadian sleep phase preferences—a tendency to be a “night owl” or “morning bird” (Jenni and Carskadon, 2000). Older children, however, are significantly more likely to experience challenges in initiating and maintaining sleep than younger children. In addition, older children are more likely to have nightmares, which usually disrupt sleep, making it discontinuous (Beltramini and Hertzig, 1983). One study found that children appear to have longer REM sleep latencies than adolescents and consequently spend a greater percentage of sleep time in stages 3 and 4 (Gaudreau et al., 2001).

Adolescents

A complex and bidirectional relationship exists between pubertal development and sleep. Studies underscore the importance of using pubertal stage, rather than chronologic age as the metric in understanding sleep, as has been found for other physiologic parameters in the second decade of life. It has been determined that adolescents require 9 to 10 hours of sleep each night (Carskadon et al., 1993; Mercer et al., 1998), though few adolescents obtain adequate sleep. In the United States, the average total sleep time in a sample of eighth-grade students was found to be 7.9 hours (Wolfson et al., 2003). Over a quarter of high school and college students were found to be sleep deprived (Wolfson and Carskadon, 1998).

SWS and sleep latency time progressively declines with advancing pubertal development (Carskadon et al., 1980); however, time spent in stage 2 increases (Carskadon, 1982). These changes are likely in part due to pubertal and hormonal changes that accompany the onset of puberty (Karacan et al., 1975). For instance, at midpuberty, there is significantly greater daytime sleepiness than at earlier stages of puberty. Afternoon sleepiness is greater than that in late afternoon and evening in more mature adolescents than in younger subjects. With increasing age, the total time spent sleeping decreases, as does REM sleep. However, if bedtime is fixed, the duration of REM sleep remains constant (Carskadon, 1982; Carskadon et al., 1983).

Adults

Sleep architecture continues to change with age across adulthood. Two major attributes of age-related sleep changes are earlier wake time and reduced sleep consolidation (Dijk et al., 2000). A hallmark change with age is a tendency toward earlier bedtimes and wake times. Older adults (approximately ages 65 to 75) typically awaken 1.33 hours earlier, and go to bed 1.07 hours earlier, than younger adults (approximately ages 20 to 30) (Duffy et al., 1998). There are no conclusive studies that demonstrate why older adults experience earlier wake times, despite decreased sleep efficiency, but one hypothesis may be an advanced circadian pacemaker that accompanies age (Dijk et al., 2000). It is unclear if this is due to older adults experiencing an increased sensitivity to light (Dijk et al., 2000; Ancoli-Israel, 2005). Nonetheless, the consequences of an advanced circadian rhythm are a 1-hour advance in body temperature increase in the early morning and misaligned melatonin and cortisol secretion rhythms with the circadian clock (Dijk et al., 2000).

Younger adults may experience brief awakenings, but they are usually minor and occur close to an REM sleep transition; thus, sleep remains relatively consolidated. Arousal occurring mostly from REM sleep in young adults suggests that there is a protective mechanism to keep from awakening during NREM sleep; however, this protective effect appears to also decline with age (Dijk, 1998). As an individual ages (between the ages of 20 to 60), SWS declines at a rate of about 2 percent per decade (Figure 2-6) (Dijk et al., 1989; Astrom and Trojaborg, 1992; Landolt et al., 1996; Ancoli-Israel, 2005). Because arousal thresholds are typically highest during SWS, and because SWS declines with age, older adults experience more frequent awakenings during a sleep episode. Another important variable may be an age-related reduction both in homeostatic sleep pressure and circadian pacemaker effectiveness during the night (Dijk et al., 2000).

Gender Differences

Although there have been few systematic studies, there appear to be gender-based differences in sleep and circadian rhythms. Available evidence is strongest in adults; however, gender differences have also been observed in infancy (Bach et al., 2000; Moss and Robson, 1970; Hoppenbrouwers et al., 1989), childhood (Meijer et al., 2000; Sadeh et al., 2000; Acebo et al., 1996), and adolescence (Giannotti et al., 2002; Laberge et al., 2001). In adults, men spend greater time in stage 1 sleep (Bixler et al., 1984) and experience more awakenings (Kobayashi et al., 1998). Although women maintain SWS longer than men, they complain more often of difficulty falling asleep and midsleep awakenings. In contrast, men are more likely to complain of daytime sleepiness (Ancoli-Israel, 2000).

In women, the menstrual cycle may influence sleep-wake activity; however, methodological challenges have limited the number of conclusive findings (Metcalf, 1983; Leibenluft et al., 1994). There have been a number of studies that suggest that women’s sleep patterns are greatly affected during pregnancy and the postpartum period (Karacan et al., 1968; Hertz et al., 1992; Lee and Zaffke, 1999; Driver and Shapiro, 1992). For example, women often experience considerable daytime sleepiness during pregnancy and during the first few postpartum months, and as will be discussed in greater detail in Chapter 3, they are also at a higher risk of developing restless legs syndrome (Goodman et al., 1998; Lee et al., 2001).

Elderly People

Problematic sleep has adverse effects on all individuals, regardless of age; however, older people typically show an increase in disturbed sleep that can create a negative impact on their quality of life, mood, and alertness (Ancoli-Israel, 2005; Bliwise, 2005). Elderly individuals sleep 36 percent less than children at age 5 (Figure 2-6). Although the ability to sleep becomes more difficult, the need to sleep does not decrease with age (Ancoli-Israel, 2005). Difficulty in initiating and maintaining sleep is cited in 43 percent of the elderly (Foley et al., 1995), although these problems are more commonly among adults suffering from depression, respiratory symptoms, and physical disability, among others (Ancoli-Israel, 2005). However, declining sleep efficiency and quality has also been observed in healthy older people (Dijk et al., 2000).

Changes in sleep patterns affect males and females differently. The progressive decrease in SWS is one of the most prominent changes with aging; however, it appears to preferentially affect men. The gender difference is unclear, but it has been suggested that older women have “better-preserved” SWS than men (Reynolds et al., 1985). Women ages 70 and older spend around 15 to 20 percent of total sleep time in stages 3 and 4; men of the same age spend only around 5 percent of total sleep time in stages 3 and 4 (Redline et al., 2004). Another gender contrast is that older women go to bed and wake up earlier than older men, which suggests that body temperature rhythms are phase-advanced in elderly women (Campbell et al., 1989; Moe et al., 1991; Monk et al., 1995). However, both men and women have increased stage 1 and decreased REM sleep.

Older people also experience a decrease in melatonin levels, which may be due to the gradual deterioration of the hypothalamic nuclei that drive circadian rhythms (Ancoli-Israel, 2005). The inability to maintain long sleep episodes and bouts of wakefulness may reflect, in addition to other medical factors, a continuously decreasing sleep homeostasis (Dijk et al., 2000; Bliwise, 2005). Other prominent factors are the continuous increase in sleep latency and nighttime awakenings and inconsistency of external cues such as light exposure (which tends to be low), irregular meal times, nocturia, and decreased mobility leading to a reduction in exercise (Dijk et al., 2000; Ancoli-Israel, 2005; Bliwise, 2005).