Abstract

There has been great scientific interest in sleep for well over a century, with the discoveries of the electrical activity of the brain, the arousal systems, the circadian system, and rapid eye movement sleep. In spite of these discoveries, the field of sleep medicine has existed for only about 4 decades. The evolution of the field required clinical research, development of clinical services, and changes in society and public policy that recognized the impact of sleep disorders on society. The field is still evolving as new disorders are being discovered, new treatments are being delivered, and basic science is helping us understand the complexity of sleep and its disorders.

Interest in sleep and dreams has existed since the dawn of history. Perhaps only love and human conflict have received more attention from poets and writers. Some of the world’s greatest thinkers, such as Aristotle, Hippocrates, Freud, and Pavlov, have attempted to explain the physiologic and psychological bases of sleep and dreaming. However, it is not the purpose of this chapter to present a scholarly review across the ages about prehistoric, biblical, and Elizabethan thoughts and concerns regarding sleep or the history of man’s enthrallment with dreams and nightmares. This has been reviewed by others.¹ What is emphasized here for the benefit of the student and the practitioner is the evolution of the key concepts that define and differentiate sleep research and sleep medicine, crucial discoveries and developments in the formative years of the field, and those principles and practices that have stood the test of time.

Sleep as a Passive State

“Sleep is the intermediate state between wakefulness and death; wakefulness being regarded as the active state of all the animal and intellectual functions, and death as that of their total suspension.”²

The foregoing is the first sentence of The Philosophy of Sleep, a book by Robert MacNish, a member of the faculty of physicians and surgeons of Glasgow; the first American edition was published in 1834 and the Scottish edition somewhat earlier. This sentence exemplifies the overarching historical conceptual dichotomy of sleep research and sleep medicine, which is sleep as a passive process versus sleep as an active process. Until the discovery of rapid eye movements and the duality of sleep, sleep was universally regarded as an inactive state of the brain. With one or two exceptions, most thinkers regarded sleep as the inevitable result of reduced sensory input with the consequent diminishment of brain activity and the occurrence of sleep. Waking up and being awake were considered a reversal of this process, mainly as a result of bombardment of the brain by stimulation from the environment. No real distinction was seen between sleep and other states of quiescence such as coma, stupor, intoxication, hypnosis, anesthesia, and hibernation.

The passive to active historical dichotomy is also given great weight by the modern investigator J. Allan Hobson.³ In the first sentence of his book, Sleep, published in 1989, he stated that “more has been learned about sleep in the past 60 years than in the preceding 6,000.” He went on, “In this short period of time, researchers have discovered that sleep is a dynamic behavior. Not simply the absence of waking, sleep is a special activity of the brain, controlled by elaborate and precise mechanisms.”

Dreams and dreaming were regarded as transient, fleeting interruptions of this quiescent state. Because dreams seem to occur spontaneously and sometimes in response to environmental stimulation (e.g., the well-known alarm clock dreams), the notion of a stimulus that produces the dream was generalized by postulating internal stimulation from the digestive tract or some other internal source. Some anthropologists have suggested that notions of spirituality and the soul arose from primitive peoples’ need to explain how their essence could leave the body temporarily at night in a dream and permanently at death.

In addition to the mere reduction of stimulation, a host of less popular theories were espoused to account for the onset of sleep. Vascular theories were proposed from the notion that the blood left the brain to accumulate in the digestive tract, and from the opposite idea that sleep was due to pressure on the brain by blood. Around the end of the 19th century, various versions of a “hypnotoxin” theory were formulated in which fatigue products (toxins and the like) were accumulated during the day, finally causing sleep, during which they were gradually eliminated. It had, of course, been observed since biblical times that alcohol would induce a sleeplike state. More recently, these observations included other compounds such as opium. Finally, it was noted that caffeine had the power to prevent sleep.

The hypnotoxin theory reached its zenith in 1907 when French physiologists, Legendre and Pieron,⁴ did experiments showing that blood serum from sleep-deprived dogs could induce sleep in dogs that were not sleep deprived. The notion of a toxin causing the brain to sleep has gradually given way to the notion that there are a number of endogenous “sleep factors” that actively induce sleep by specific mechanisms.

In the 1920s, the University of Chicago physiologist Nathaniel Kleitman carried out a series of sleep-deprivation studies and made the simple but brilliant observation that individuals who stayed up all night were generally less sleepy and impaired the next morning than in the middle of their sleepless night. Kleitman argued that this observation was incompatible with the notion of a continual buildup of a hypnotoxin in the brain or blood. In addition, he felt that humans were about as impaired as they would get, that is, very impaired, after about 60 hours of wakefulness, and that longer periods of sleep deprivation would produce little additional change. In the 1939 (first) edition of his comprehensive landmark monograph “Sleep and Wakefulness” Kleitman⁵ summed up by saying, “It is perhaps not sleep that needs to be explained, but wakefulness, and indeed, there may be different kinds of wakefulness at different stages of phylogenetic and ontogenetic development. In spite of sleep being frequently designated as an instinct, or global reaction, an actively initiated process, by excitation or inhibition of cortical or subcortical structures, there is not a single fact about sleep that cannot be equally well interpreted as a let down of the waking activity.”

The Electrical Activity of the Brain

As the 20th century got under way, Camillo Golgi and Santiago Ramón y Cajal had demonstrated that the nervous system was not a mass of fused cells sharing a common cytoplasm but rather a highly intricate network of discrete cells that had the key property of signaling to one another. Luigi Galvani had discovered that the nerve cells of animals produce electricity, and Emil duBois-Reymond and Hermann von Helmholtz found that nerve cells use their electrical capabilities for signaling information to one another. In 1875, the Scottish physiologist Richard Caton demonstrated electrical rhythms in the brains of animals. The centennial of his discovery was commemorated at the 15th annual meeting of the Association for the Psychophysiological Study of Sleep convening at the site of the discovery, Edinburgh, Scotland.

However, it was not until 1928 when the German psychiatrist Hans Berger⁶ recorded electrical activity of the human brain and clearly demonstrated differences in these rhythms when subjects were awake or asleep that a real scientific interest commenced. Berger correctly inferred that the signals he recorded, which he called “electroencephalograms,” were of brain origin. For the first time, the presence of sleep could be conclusively established without disturbing the sleeper, and, more important, sleep could be continuously and quantitatively measured without disturbing the sleeper.

All the major elements of sleep brain wave patterns were described by Harvey, Hobart, Davis, and others^7–9 at Harvard University in a series of extraordinary papers published in 1937, 1938, and 1939. Blake, Gerard, and Kleitman^10,11 added to this from their studies at the University of Chicago. On the human electroencephalogram (EEG), sleep was characterized by high-amplitude slow waves and spindles, whereas wakefulness was characterized by low-amplitude waves and alpha rhythm. The image of the sleeping brain completely “turned off” gave way to the image of the sleeping brain engaged in slow, synchronized, “idling” neuronal activity. Although it was not widely recognized at the time, these studies were some of the most critical turning points in sleep research. Indeed, Hobson³ dated the turning point of sleep research to 1928, when Berger began his work on the human EEG. Used today in much the same way as they were in the 1930s, brain wave recordings with paper and ink, or more recently on computer screens, have been extraordinarily important to sleep research and sleep medicine.

The 1930s also saw one series of investigations that seemed to establish conclusively both the passive theory of sleep and the notion that it occurred in response to reduction of stimulation and activity. These were the investigations of Frederick Bremer,^12,13 reported in 1935 and 1936. These investigations were made possible by the aforementioned development of electroencephalography. Bremer studied brain wave patterns in two cat preparations. One, which Bremer called encéphale isolé, was made by cutting a section through the lower part of the medulla. The other, cerveau isolé, was made by cutting the midbrain just behind the origin of the oculomotor nerves. The first preparation permitted the study of cortical electrical rhythms under the influence of olfactory, visual, auditory, vestibular, and musculocutaneous impulses; in the second preparation, the field was narrowed almost entirely to the influence of olfactory and visual impulses.

In the first preparation, the brain continued to present manifestations of wakeful activity alternating with phases of sleep as indicated by the EEG. In the second preparation, however, the EEG assumed a definite deep sleep character and remained in this condition. In addition, the eyeballs immediately turned downward with a progressive miosis. Bremer concluded that in sleep there occurs a functional (reversible, of course) deafferentation of the cerebral cortex. The cerveau isolé preparation results in a suppression of the incessant influx of nerve impulses, particularly cutaneous and proprioceptive, which are essential for the maintenance of the waking state of the telencephalon. Apparently, olfactory and visual impulses are insufficient to keep the cortex awake. It is probably misleading to assert that physiologists assumed the brain was completely turned off, whatever this metaphor might have meant, because blood flow and, presumably, metabolism continued. However, Bremer and others certainly favored the concept of sleep as a reduction of activity-idling, slow, synchronized, “resting” neuronal activity.

The Ascending Reticular System

After World War II, insulated, implantable electrodes were developed, and sleep research on animals began in earnest. In 1949, one of the most important and influential studies dealing with sleep and wakefulness was published: Moruzzi and Magoun’s classic paper “Brain Stem Reticular Formation and Activation of the EEG.”¹⁴ These authors concluded that “transitions from sleep to wakefulness or from the less extreme states of relaxation and drowsiness to alertness and attention are all characterized by an apparent breaking up of the synchronization of discharge of the elements of the cerebral cortex, an alteration marked in the EEG by the replacement of high voltage, slow waves with low-voltage fast activity” (p. 455).

High-frequency electrical stimulation through electrodes implanted in the brainstem reticular formation produced EEG activation and behavioral arousal. Thus, EEG activation, wakefulness, and consciousness were at one end of the continuum; sleep, EEG synchronization, and lack of consciousness were at the other end. This view, as can be seen, is hardly different from the statement by MacNish quoted at the beginning of this chapter.

The demonstration by Starzl and coworkers¹⁵ that sensory collaterals discharge into the reticular formation suggested that a mechanism was present by which sensory stimulation could be transduced into prolonged activation of the brain and sustained wakefulness. By attributing an amplifying and maintaining role to the brainstem core and the conceptual ascending reticular activating system, it was possible to account for the fact that wakefulness outlasts, or is occasionally maintained in the absence of, sensory stimulation.

Chronic lesions in the brainstem reticular formation produced persisting slow waves in the EEG and immobility. The usual animal for this research was the cat because excellent stereotaxic coordinates of brain structures had become available in this model.¹⁶ These findings appeared to confirm and extend Bremer’s observations. The theory of the reticular activating system was an anatomically based passive theory of sleep or an active theory of wakefulness. Figure 1-1 is from the published proceedings of a symposium entitled Brain Mechanisms and Consciousness, which published in 1954 and is probably the first genuine neuroscience bestseller.¹⁷ Horace Magoun had extended his studies to the monkey, and the illustration represents the full flowering of the ascending reticular activating system theory.

Figure 1-1 Lateral view of the monkey’s brain, showing the ascending reticular activating system in the brainstem receiving collaterals from direct afferent paths and projecting primarily to the associational areas of the hemisphere.

(Redrawn from Magoun HW: The ascending reticular system and wakefulness. In Adrian ED, Bremer F, Jasper HH [eds]. Brain mechanisms and consciousness. A symposium organized by the Council for International Organizations of Medical Sciences, 1954. (Courtesy of Charles C Thomas, Publisher, Springfield, Ill.)

Early Observations of Sleep Pathology

Insomnia has been described since the dawn of history and attributed to many causes, including a recognition of the association between emotional disturbance and sleep disturbance. Scholars and historians have a duty to bestow credit accurately. However, many discoveries lie fallow for want of a contextual soil in which they may be properly understood and in which they may extend the understanding of more general phenomena. Important early observations were those of von Economo on “sleeping sickness” and of Pavlov, who observed dogs falling asleep during conditioned reflex experiments.

Two early observations about sleep research and sleep medicine stand out. The first is the description in 1880 of narcolepsy by Jean Baptiste Edouard Gélineau (1859-1906), who derived the name narcolepsy from the Greek words narkosis (a benumbing) and lepsis (to overtake). He was the first to clearly describe the collection of components that constitute the syndrome, although the term cataplexy for the emotionally induced muscle weakness was subsequently coined in 1916 by Richard Henneberg.

Obstructive sleep apnea syndrome (OSAS), which may be called the leading sleep disorder of the 20th century, was described in 1836, not by a clinician but by the novelist Charles Dickens. In a series of papers entitled the “Posthumous Papers of the Pickwick Club,” Dickens described Joe, a boy who was obese and always excessively sleepy. Joe, a loud snorer, was called Young Dropsy, possibly as a result of having right-sided heart failure. Meir Kryger¹⁸ and Peretz Lavie^19,20 published scholarly accounts of many early references to snoring and conditions that were most certainly manifestations of OSAS. Professor Pierre Passouant²¹ provided an account of the life of Gélineau and his landmark description of the narcolepsy syndrome.

Sigmund Freud and the Interpretation of Dreams

By far the most widespread interest in sleep by health professionals was engendered by the theories of Sigmund Freud, specifically about dreams. Of course, the interest was really in dreaming, with sleep as the necessary concomitant. Freud developed psychoanalysis, the technique of dream interpretation, as part of his therapeutic approach to emotional and mental problems. As the concept of the ascending reticular activating system dominated behavioral neurophysiology, so the psychoanalytic theories about dreams dominated the psychological side of the coin. Dreams were thought to be the guardians of sleep and to occur in response to a disturbance in order to obviate waking up, as exemplified in the classic alarm clock dream. Freud’s concept that dreaming discharged instinctual energy led directly to the notion of dreaming as a safety valve of the mind. At the time of the discovery of rapid eye movements during sleep (circa 1952), academic psychiatry was dominated by psychoanalysts, and medical students all over America were interpreting one another’s dreams.

From the vantage point of today’s world, the dream deprivation studies of the early 1960s, engendered and reified by the belief in psychoanalysis, may be regarded by some as a digression from the mainstream of sleep medicine. On the other hand, because the medical–psychiatric establishment had begun to take dreams seriously, it was also ready to support sleep research fairly generously under the guise of dream research.

Chronobiology

Most, but not all, sleep specialists share the opinion that what has been called chronobiology or the study of biologic rhythms is a legitimate part of sleep research and sleep medicine. The 24-hour rhythms in the activities of plants and animals have been recognized for centuries. These biologic 24-hour rhythms were quite reasonably assumed to be a direct consequence of the periodic environmental fluctuation of light and darkness. However, in 1729, Jean Jacques d’Ortous de Mairan described a heliotrope plant that opened its leaves during the day even after de Mairan had moved the plant so that sunlight could not reach it. The plant opened its leaves during the day and folded them for the entire night even though the environment was constant. This was the first demonstration of the persistence of circadian rhythms in the absence of environmental time cues. Figure 1-2, which represents de Mairan’s original experiment, is reproduced from The Clocks That Time Us by Moore-Ede and colleagues.²²

Figure 1-2 Representation of de Mairan’s original experiment. When exposed to sunlight during the day (upper left), the leaves of the plant were open; during the night (upper right), the leaves were folded. De Mairan showed that sunlight was not necessary for these leaf movements by placing the plant in total darkness. Even under these constant conditions, the leaves opened during the day (lower left) and folded during the night (lower right).

(Redrawn from Moore-Ede MC, Sulzman FM, Fuller CA. The clocks that time us: physiology of the circadian timing system. Cambridge, Mass: Harvard University Press; 1982. p. 7.)

Chronobiology and sleep research developed separately. The following three factors appear to have contributed to this:

The Discovery of REM Sleep

The characterization of rapid eye movement (REM) sleep as a discrete organismic state should be distinguished from the discovery that rapid eye movements occur during sleep. The historical threads of the discovery of rapid eye movements can be identified. Nathaniel Kleitman (Fig. 1-3, Video 1-1), a professor of physiology at the University of Chicago, had long been interested in cycles of activity and inactivity in infants and in the possibility that this cycle ensured that an infant would have an opportunity to respond to hunger. He postulated that the times infants awakened to nurse on a self-demand schedule would be integral multiples of a basic rest-activity cycle. The second thread was Kleitman’s interest in eye motility as a possible measure of “depth” of sleep. The reasoning for this was that eye movements had a much greater cortical representation than did almost any other observable motor activity, and that slow, rolling, or pendular eye movements had been described at the onset of sleep with a gradual slowing and disappearance as sleep “deepened.”²³

Figure 1-3 Nathaniel Kleitman (circa 1938), Professor of Physiology, University of Chicago, School of Medicine.

In 1951, Kleitman assigned the task of observing eye movement to a graduate student in physiology named Eugene Aserinsky. Watching the closed eyes of sleeping infants was tedious, and Aserinsky soon found that it was easier to designate successive 5-minute epochs as “periods of motility” if he observed any movement at all, usually a writhing or twitching of the eyelids, versus “periods of no motility.”

After describing an apparent rhythm in eye motility, Kleitman and Aserinsky decided to look for a similar phenomenon in adults. Again, watching the eyes during the day was tedious, and at night it was even worse. Casting about, they came upon the method of electrooculography and decided (correctly) that this would be a good way to measure eye motility continuously and would relieve the human observer of the tedium of direct observations. Sometimes in the course of recording electrooculograms (EOGs) during sleep, they saw bursts of electrical potential changes that were quite different from the slow movements at sleep onset.

When they were observing infants, Aserinsky and Kleitman had not differentiated between slow and rapid movements. However, on the EOG, the difference between the slow eye movements at sleep onset and the newly discovered rapid motility was obvious. Initially, there was a great deal of concern that these potentials were electrical artifacts. With their presence on the EOG as a signal, however, it was possible to watch the subject’s eyes simultaneously, and when this was done, the distinct rapid movement of the eyes beneath the closed lids was extremely easy to see.

At this point, Aserinsky and Kleitman made two assumptions:

The basic sleep cycle was not identified at this time, primarily because the EOG and other physiologic measures, notably the EEG, were not recorded continuously but rather by sampling a few minutes of each hour or half-hour. The sampling strategy was done to conserve paper (there was no research grant) and because there was not a clear reason to record continuously. Also, the schedule made it possible for the researcher to nap between sampling episodes.

Aserinsky and Kleitman initiated a small series of awakenings, both when rapid eye movements were present and when rapid eye movements were not present, for the purpose of eliciting dream recall. They did not apply sophisticated methods of dream content analysis, but the descriptions of dream content from the two conditions were generally quite different with REM awakenings yielding vivid complex stories and non-REM (NREM) awakenings often yielding nothing at all or very sparse accounts. This made it possible to conclude that rapid eye movements were associated with dreaming. This was, indeed, a breakthrough in sleep research.^24,25

The occurrence of the eye movements was quite compatible with the contemporary dream theories that dreams occurred when sleep lightened in order to prevent or delay awakening. In other words, dreaming could still be regarded as the “guardian” of sleep. However, it could no longer be assumed that dreams were fleeting and evanescent.

All-Night Sleep Recordings and the Basic Sleep Cycle

The seminal Aserinsky and Kleitman paper was published in 1953. It attracted little attention, and no publications on the subject appeared from any other laboratory until 1959. Staying up at night to study sleep remained an undesirable occupation by anyone’s standards. In the early 1950s, most previous research on the EEG patterns of sleep, like most approaches to sleep physiology generally, had either equated short periods of sleep with all sleep or relied on intermittent time sampling during the night. The notion of obtaining continuous records throughout typical nights of sleep would have seemed highly extravagant.

However, motivated by the desire to expand and quantify the description of rapid eye movements, then graduate student William Dement and Kleitman²⁶ did just this over a total of 126 nights with 33 subjects and, by means of a simplified categorization of EEG patterns, scored the paper recordings in their entirety. When they examined these 126 records, they found that there was a predictable sequence of patterns over the course of the night, such as had been hinted at by Aserinsky’s study but entirely overlooked in all previous EEG studies of sleep. Although this sequence of regular variations has now been observed tens of thousands of times in hundreds of laboratories, the original description remains essentially unchanged.

The usual sequence was that after the onset of sleep, the EEG progressed fairly rapidly to stage 4, which persisted for varying amounts of time, generally about 30 minutes, and then a “lightening” took place. Whereas the progression from wakefulness to stage 4 at the beginning of the cycle was almost invariable through a continuum of change, the lightening was usually abrupt and coincident with a body movement or series of body movements. After the termination of stage 4, there was generally a short period of stage 2 or stage 3 which gave way to stage 1 and rapid eye movements. When the first eye movement period ended, the EEG again progressed through a continuum of change to stage 3 or 4, which persisted for a time and then lightened, often abruptly, with body movement to stage 2, which again gave way to stage 1 and the second rapid eye movement period (see p. 679 of Dement and Kleitman²⁶).

Dement and Kleitman found that this cyclical variation of EEG patterns occurred repeatedly throughout the night at intervals of 90 to 100 minutes from the end of one eye movement period to the end of the next. The regular occurrences of REM periods and dreaming strongly suggested that dreams did not occur in response to chance disturbances.

At the time of these observations, sleep was still considered to be a single state. Dement and Kleitman characterized the EEG during REM periods as “emergent stage 1” as opposed to “descending stage 1” at the onset of sleep. The percentage of the total sleep time occupied by REM sleep was between 20% and 25%, and the periods of REM sleep tended to be shorter in the early cycles of the night. This pattern of all-night sleep has been seen over and over in normal humans of both sexes, in widely varying environments and cultures, and across the life span.