Many schoolchildren try to sneak naps at their desks to get through the boring parts of class. Rainer Schmitt, on the other hand, always fought hard to stay awake. He would nod off anyway--in virtually every class period. And it was never the teacher's fault. Since childhood, Schmitt has suffered from an unusual neurological disorder: narcolepsy.

Now a 56-year-old math teacher, Schmitt still struggles with overwhelming daytime sleepiness and fatigue, the hallmark symptom of narcolepsy. Instead of feeling awake for 16 hours and sleepy for eight, as most people do, Schmitt, like other narcoleptics, wants to sleep every couple of hours during the day. Fleeting urges to nod off plague him. Eventually the desire to sleep is overwhelming, and then he may actually fall asleep, usually for a few seconds or up to several minutes. He may nod off in the middle of a lecture, meeting or conversation. He could even fall asleep at the wheel.

Narcolepsy afflicts one in 2,000 Americans--at least 150,000 people in this country--and possibly many more. People with the disorder typically do not recognize their sleepiness as a sign of a disease, experts say, so they do not consult a doctor. As a result, they are often diagnosed with the disorder 10 to 15 years after their first symptoms appear.

All of that time, studies have shown, people with narcolepsy suffer immensely at work and in their social relationships. Many are unemployed, because employers or potential ones often dismiss them as lazy and fail to provide the accommodations these people need to perform well.

German psychiatrist Carl Friedrich Otto Westphal first described narcolepsy in 1877, yet only recently have researchers begun to decipher the biological causes of this strange malady. Notably, they have pinpointed a brain chemical that is conspicuously absent in the brains of narcoleptics. What is more, surprising new evidence suggests that narcolepsy may be an autoimmune disorder like type 1 diabetes or multiple sclerosis. The recent work has led to powerful theories about what has gone awry in narcoleptics' brains and to possible new remedies for this debilitating disorder.

Living in Dreamland
Doctors usually diagnose narcolepsy by listening in on a person's brain at night. In a sleep study, or polysomnogram, electrodes pick up a subject's brain waves and muscle movements throughout the night, revealing the periods when the patient is awake or in the various stages of sleep.

During normal sleep, a person first enters so-called nonrapid eye movement (NREM) sleep, starting with the first stage and then passing through stages 2, 3 and 4, which are progressively deeper and more intense. After about 90 minutes, rapid eye movement (REM) sleep begins. The eyes shift rapidly under the eyelids, and the brain is active, creating dreams.

In contrast to this pattern, narcoleptics start REM sleep within a few minutes of dozing off, a phenomenon called sleep onset rapid eye movement period (SOREMP). Doctors consider SOREMP a diagnostic red flag. Narcoleptics also spend significantly more time in REM sleep than healthy people do. That is, they dream a lot--perhaps too much.

The long periods that narcoleptics spend in REM sleep may deprive them of deep sleep, the phase that is crucial to physical recovery and cell regeneration as well as feeling alert in the morning. Thus, narcoleptics' daytime drowsiness may stem largely from a failure in the mechanism that governs REM sleep.

Consistent with this idea, many narcoleptics appear to suffer REM sleeplike episodes in the daytime. They may see or hear dreamlike images or sounds just before falling asleep or just after awakening from their midday naps, for example. Curiously, many of them also experience bizarre muscle collapse, or cataplexy, that is reminiscent of the muscle slackening that occurs during normal REM sleep.

In cataplexy, strong emotions such as elation, surprise or anger may cause a narcoleptic's knees to buckle or his or her head to drop. In the worst cases, the entire muscle system fails, causing the person to fall down, paralyzed, for a few seconds to several minutes. "You collapse like a building hit by a bomb," Schmitt says. Similarly, in normal REM sleep, while the brain and eyes are active, the body is virtually immobile thanks to reduced muscle tone, something that may prevent people from acting out their dreams.

Narcolepsy may not just be a REM sleep disorder, however; narcoleptics also display certain oddities in NREM sleep, according to work reported in 2005 by neurologist Raffaele Ferri and his colleagues at the Oasi Institute for Research on Mental Retardation and Brain Aging in Troina, Italy. The Oasi investigators recorded bursts of brain activity called cyclic alternating patterns (CAPs) during NREM sleep in the brains of 49 narcoleptics, comparing them with 37 normal sleepers of the same age. The researchers recorded many fewer CAPs from the narcoleptics than from the normal sleepers, who typically have several hundred CAPs, each of which lasts just a few seconds, during the night. In addition to suggesting that NREM sleep is also impaired in narcolepsy, such CAP frequency aberrations may be indicative of a sleeping process that is less restorative than a healthy sleeper's, the scientists suggest.

Java Juice
Providing some of the first clues to the biochemical cause of such abnormal sleep, new studies are focusing on hypocretin, a tiny brain molecule produced by a cluster of only 10,000 to 20,000 cells packed at the back of the hypothalamus, a cone-shaped structure at the base of the brain. When hypocretin was first discovered in 1998, it was thought to be an appetite regulator, because this was the supposed job of that part of the hypothalamus. But since then, evidence has accumulated connecting hypocretin, or the lack of it, to narcolepsy.

Emmanuel Mignot, who leads the Center for Narcolepsy at the Stanford University School of Medicine, and his colleagues reported in 1999 that a mutation in the gene for one of the receptors that binds hypocretin caused narcolepsy in dogs. Meanwhile a team led by Masashi Yanagisawa of the University of Texas Southwestern Medical Center at Dallas showed that mice that could not produce hypocretin behaved just like narcoleptics. Mignot's group soon found that 90 percent of human narcoleptics lacked hypocretin in either their brain tissues or cerebrospinal fluid (the liquid that surrounds the brain). These studies, published in 2000, suggest that a lack of this neuropeptide causes narcolepsy and the accompanying flights into dreamland.

Exactly what function hypocretin plays in the body--and thus what is missing in narcoleptics--is still controversial. Some researchers suggest that hypocretin's role is to keep people awake and that, accordingly, narcolepsy results from inadequate stimulation of wake-promoting brain regions. This idea is consistent with the pattern of hypocretin secretion in the squirrel monkey, an animal with a sleep-wake cycle similar to that of humans.

A few years ago Mignot and his colleagues measured the amount of hypocretin in squirrel monkeys' brains during a typical day. They found a low level in the morning on awakening. Then the amount gradually rose, peaking at the end of the day just before the monkeys fell asleep. When the animals dozed off, the hypocretin level fell, eventually returning to the starting level. These results, reported by the Stanford team in 2003, suggest that hypocretin, though not promoting initial wakefulness, helps to maintain alertness as the day wears on to counteract increasing fatigue. Without it, narcoleptics can remain awake comfortably only for a few hours.

A study in rats, reported in 2000, supports this view, while also helping to account for the abundance of REM sleep in narcoleptics. Biologist Luis de Lecea and his colleagues at the Scripps Research Institute in San Diego injected rats with hypocretin in a brain region ordinarily stimulated by hypocretin. The rats stayed awake 70 percent longer than usual. In addition, when the hypocretin-injected rats did sleep, they spent considerably less time in REM sleep than rats injected with saline did. Thus, hypocretin may maintain alertness at least in part by specifically blocking the excessive intrusion of dreams into a person's sleep-wake cycle.

But others say that hypocretin's job is to police the barriers between various states of consciousness so that the brain does not readily switch between them. Neurobiologist Thomas Scammell of Beth Israel Deaconess Medical Center in Boston, in collaboration with Yanagisawa and others, carefully measured and observed the sleep-wake cycles of a strain of mice that cannot produce hypocretin. Although the mutant mice spent normal amounts of time awake, asleep, and in REM and NREM sleep, like human narcoleptics they shifted with unusually high frequency among states of wakefulness, dozing, deep sleep and REM sleep.

This abnormal shifting was not because the mice were unable to maintain alertness. Scammell's team found that under a stressful condition such as being put in a new cage--which typically induces heightened arousal--the mutant mice stayed awake just as long as normal mice, suggesting that alertness mechanisms are intact even in the absence of hypocretin.

Death in the Brain
Scientists are also making headway in explaining the lack of hypocretin in narcoleptics. Research has shown that narcoleptics have only about 10 percent of the normal number of hypocretin-producing cells in their hypothalamus. Their lack of such cells may reflect a problem with their immune systems.

People carry different types of immune system proteins, called human leukocyte antigens (HLAs), on the surface of their white blood cells. Many autoimmune disorders, in which the immune system mistakenly attacks healthy body tissues instead of, say, infectious microbes, are associated with a distinctive set of these molecules.

In the 1990s Mignot and his colleagues discovered that about nine out of 10 patients with sleep-wake disturbances bore the HLA subtype HLA-DQB1*0602, which is less common in healthy sleepers. This finding, according to Mignot and his Stanford colleague Seiji Nishino, suggests that narcolepsy may be an autoimmune disorder: narcoleptics' immune systems may mistakenly attack and destroy their hypocretin-producing cells. Additional research supports this view. For instance, researchers have been able to induce certain features of narcolepsy in mice by injecting them with immune system proteins from narcoleptics.

Both genetic and environmental factors are thought to underlie the development of narcolepsy. One to 4 percent of children with a narcoleptic parent will develop this disorder. Their risk is 20- to 80-fold higher than the incidence in the general population, suggesting that there are one or more genes that confer a risk of disordered sleep. One of them might be an immune component such as HLA-DQB1*0602.

Genetics, however, cannot fully account for narcolepsy. About one fifth of the healthy population also carries HLA-DQB1*0602. And if an identical twin has narcolepsy, the other twin (whose genetic material is identical to that of his or her sibling) will develop the disorder only a third of the time.

Which environmental factors might tip the balance toward narcolepsy is still unknown. One possibility is stress. John Harsh of the University of Southern Mississippi, along with Dante Picchioni of the Walter Reed Army Institute of Research in Silver Spring, Md., asked patients with narcolepsy to describe what was happening in their lives just before they got sick. Compared with healthy volunteers surveyed about similar periods in their lives, a relatively high percentage of narcoleptics had experienced life-changing events--such as the birth of a child, getting a new job or moving to a new home--in the months before the onset of narcoleptic symptoms. Thus, major life changes, and the accompanying psychological stress, may alter the immune system in ways that facilitate the emergence of the disorder.

Some researchers hypothesize that prenatal infections may play a role as well. At least two research teams--one including Mignot and the other led by Harsh--have established a connection between birth month and narcolepsy: the disorder is unusually common among people born in March and disproportionately rare in people born in September. Other neurological disorders follow similar patterns; multiple sclerosis clusters among people born between March and July, whereas epilepsy springs up more often among those born between December and March. Such patterns may result if infections that spur the development of neurological disorders during a certain stage of pregnancy are more likely during some times of year than others.

Wake-up Calls
Preventing narcolepsy is probably an impractical goal. But medicine can help patients cope. Currently many narcoleptics take amphetaminelike stimulants to combat daytime sleepiness. Some require antidepressants to treat cataplexy and dreamlike hallucinations. Doctors also often prescribe behavioral interventions such as scheduling naps during a patient's sleepiest times. Such remedies target symptoms, however, rather than the cause of the disorder, and the drugs are not without side effects.

Researchers hope that the discovery of hypocretin and its role in the disorder will lead to important new medications. Some say that the best approach most likely will be the replacement of missing hypocretin, although getting the molecule--or something like it--into the brain is a formidable challenge. Hypocretins are unstable molecules. They easily disintegrate in the bloodstream and digestive tract, so simply injecting or swallowing them will not be effective.

For now, Schmitt works around his narcolepsy. He can generally stay awake through the morning to teach. Then he divides up afternoons and evenings to accommodate his unusual napping needs. As a consequence, he may find himself wide awake at 3 A.M., sitting at his desk, correcting papers. "I have made my peace with my school," he says. And the school has made peace with its math teacher's sleep.