Randy Gardner decided not to sleep for 11 straight days in 1964: He was 17, a high school senior and desperately wanted to win the San Diego Science Fair.
He had recently heard that a disc jockey had gone 260 hours, a shade under 11 days, without sleeping. "I could do that," Gardner told himself.
"I wanted to prove that bad things didn't happen if you went without sleep, that you wouldn't go insane," said Gardner, now 56, retired and still in San Diego. "I thought, 'I can break that record and I don't think it would be a negative experience.' "
With the help of two friends who took turns keeping him awake, marathon pinball sessions and a growing circus of television reporters, Gardner stayed awake 264 hours and achieved his record.
In doing so, he raised two seductively simple questions that continue to linger: How much sleep do people really need? And -- a matter of special interest for overworked Washingtonians -- are there ways to cheat sleep?
Those questions have taken on new urgency in recent months, as researchers have begun evaluating whether a medicine approved to treat sleepiness associated with narcolepsy can also keep healthy people awake. Military scientists in the United States and France have begun pushing the limits of the drug modafinil -- one study found that people who have not slept for two straight nights still function well while using it. A variety of studies are evaluating whether the medicine can help truck drivers and night-shift workers, since sleepy people are especially prone to traffic and industrial accidents.
While drugs like modafinil and caffeine keep people awake, no chemical is actually known to replace the restorative power of sleep. Animals deprived of sleep for long periods lose the ability to retain heat and die. But can people get by with an hour or two less every night? Does the fact that most people seem able to sleep eight or more hours every night mean they need to sleep that long?
Researchers -- who agree that a core amount of sleep is needed for survival -- disagree on whether people can reduce their sleep somewhat and still function well and remain healthy.
"There is no simple answer" to how much sleep people need, said Col. Gregory Belenky, director of neuropsychiatry at the Walter Reed Army Institute of Research in Silver Spring, who has conducted extensive research into the effects of sleep deprivation: "Individual by individual, there is a lot of variability."
A very small number of people appear to need four hours of sleep or less a night. Some experiments in the United Kingdom suggest that people can be trained to get all the rest they need in about 6 1/2 hours, without the use of stimulants.
Advocacy groups like the National Sleep Foundation recommend that Americans get more sleep, preferably seven to nine hours a night. Experts at the foundation warn that Americans have been steadily cutting back on their sleep, and point out that significant numbers of traffic and industrial accidents are caused by sleep deprivation and fatigue. Most adult Americans curtail their sleep: The National Coffee Association says 100 million drinkers of caffeinated coffee consume more than three cups each per day, the equivalent of 400 milligrams of caffeine.
The Need for Z's
Among the difficulties in finding definitive data about how much sleep people need is that the central purpose of sleep is shrouded in mystery and paradox: For example, if sleep is physically restorative, why do the inactive need as much as athletes? If the primary benefit of sleep is to the brain, as most scientists believe, why do dullards need it as much as chess grandmasters? Moreover, why does the brain remain active during sleep? In fact, some "restful" parts of sleep occur during periods of heightened brain activity.
Researchers have theorized that sleep may restore some chemical that is drained during periods of wakefulness. Scientists have isolated chemicals that fluctuate during sleep, like adenosine, which affects metabolism and fatigue -- but no one has pinned down a definitive chemical explanation for sleep. Some experts have speculated that sleep may be an evolutionary mechanism to conserve energy. Still others say sleep may give the brain an opportunity to process experiences or even to exercise neural pathways that have lain dormant during the day.
Whatever the explanation, sleep is so important that animal brains "remember" the amount and quality of sleep they get, and compensate for sleep deprivation by falling asleep faster and staying asleep longer. Jerry Siegel, a UCLA and Veterans Administration sleep researcher, said the drive for sleep is so powerful among mammals that when dolphins, whose brains sleep one hemisphere at a time, are deprived of sleep, the hemispheres independently keep track of how much sleep they have lost, and build up separate accounts of their "sleep debt."
Seemingly contradictory results from a variety of studies may also suggest that different amounts of sleep are necessary for day-to-day performance, highly creative tasks and for long-term health.
People who are sleep-deprived react more slowly and in less creative ways. After prolonged sleep deprivation, most people forget words and become unable to come up with "out-of-the-box" solutions to complex problems, experiments have found. Sleep-deprived people also take longer to accomplish tasks, making it doubtful that reducing the need for sleep would actually make people more productive.
Researcher Eve Van Cauter at the University of Chicago has also found physical consequences: Volunteers restricted to four hours of sleep a night show hormonal changes that make them show signs of pre-diabetes.
Van Cauter is exploring the idea that insufficient sleep may be partly behind the obesity epidemic in the United States.
Animals deprived of sleep at crucial stages of development can lose cognitive abilities like depth perception, said Thomas Scammell, a neurologist at Harvard Medical School. Sleep, he said, may be important for memory and learning, and affect cognition in subtle ways.
Deprived and Alive Simultaneously, however, according to a study conducted by Nancy Wesensten and others at the Walter Reed institute, volunteers deprived of sleep for two straight nights can maintain performance when given high doses of modafinil or caffeine.
Psychophysiologist Jim Horne at Loughborough University in the United Kingdom has found he can train people to sleep less without negative effects by gradually reducing the amount they sleep -- 20 minutes less every three to four nights -- apparently forcing the volunteers to sleep more "efficiently." Restricting sleep among some people with depression has also been found to lift their moods, he said.
And in February, a controversial study conducted at the University of California, San Diego, found that people who sleep six or seven hours a night actually live longer than people who sleep eight or more.
Daniel Kripke, the lead investigator for this study, said the results don't mean that sleep is dangerous. He said the findings only underscore that scientists don't really know whether humans need some fixed amount of sleep. People, he said, should avoid routinely using drugs either to sleep more or sleep less.
"We don't advise long sleepers to do anything at the current time, because we haven't proven if they cut back their sleep it will benefit them," he said. "We certainly wouldn't recommend stimulants."
Horne said the rationale for wanting less sleep could influence its consequences: "Someone who says they have so many pressures they have to cut into their sleep is being foolish."
"If you are restricting your sleep for pleasure, by all means do so," he said. "But if you are restricting your sleep because you are overwhelmed and can't cope, then sleep is your buffer."
11 Days Awake That distinction is perhaps best illustrated by Randy Gardner himself, who found his record-setting experience difficult but also exhilarating: The first few days, he recalled, were relatively easy. Gardner visited the jail and the airport, took long walks on the beach with two "research assistants" and played lots of basketball. He ate normally and used no stimulants.
After the first four or five days, reporters began showing up. William Dement, a Stanford University sleep researcher, came down to San Diego help monitor the experiment.
By the sixth or seventh day, Gardner's eyes began to give him "a sandpaper feeling" around sunrise. Other side effects, he noted, were "slurred words, memory wasn't working so great. Nothing was very pronounced or long-lasting. I was pretty grumpy -- I would tell people to shut up."
One night after 3 a.m., Dement played 100 games against Gardner on a baseball machine in a penny arcade. The professor lost every one: The teenager was still in good physical shape.
As the 11th day loomed, media attention turned into a circus, the phone rang off the hook and throngs of supporters helped Gardner fight off sleep. There was a brief news conference afterward: Gardner declared that mastering sleep was a question of "mind over matter." He felt he could go another day or two, but the experiment was being conducted over his Christmas holidays and he had to get back to school.
Researchers whisked him away to a sleep lab and hooked EEG monitors to his head. It took him three seconds to fall asleep. He slept for about 14 hours, woke up and said he felt fine.
Gardner's experiment revealed that missing 80 to 90 hours of sleep clearly left him with sleep debt, but he didn't need to sleep 80 hours right away. Gardner also showed that sleep debt isn't paid back uniformly: There were times during his experiment when he could easily stay awake, and other times when it was nearly impossible.
Sleep Rhythms Superimposed over the buildup of sleep debt, in other words, people have a circadian rhythm, with crests and troughs. This cycle is closely related to light and the change of seasons.
Evolutionarily, humans appear to be programmed to feel alert and sleepy at different times of the day. The greatest risk for fatal traffic accidents caused by sleepiness is in the early hours of the morning, when people's circadian rhythm is at a trough. Afternoon brings on another trough, while the late morning and early evenings appear to be peaks of alertness.
"Most theories suggest [the need for] sleep is homeostatic -- which is, it accumulates during the day and the longer you are awake the more sleepy you become," said Peretz Lavie, a researcher at the Israel Institute of Technology in Haifa. "We showed this is true, but not for the evening period. You become sleepier [in the afternoon]. Then you reach 8 or 9 o'clock and paradoxically you become alert."
Lavie has subjected volunteers to ultra-short sleep cycles -- seven minutes in bed and 13 minutes awake -- for up to 48 hours. After two days of irregular sleep, Lavie found that people couldn't be paid to fall asleep during periods of the circadian peak and couldn't be paid to stay awake during the trough.
While sleep deprivation experiments have taught scientists a lot about how much sleep people need, one researcher approached the question from the opposite direction. When Thomas Wehr at the National Institute of Mental Health had 15 people sign up for a month-long experiment, the psychiatrist placed them in the dark for 14 hours a day.
Wehr's volunteers first slept an average of 12 hours a night -- a sign that they (like most people) were probably quite sleep-deprived.
After three weeks, they gradually reduced their sleep to about eight hours. But another interesting thing happened. Instead of sleeping in one period during the night, the volunteers began sleeping in two distinct periods.
They would fall asleep around 8 in the evening and wake up around midnight. They would then fall asleep again around 2 a.m. and sleep another four hours.
"Maybe our ancestors used to sleep this way" before artificial light became cheaply available, said Wehr. "If any of us sleeps like this, we would call it a sleep disorder -- but maybe it's normal sleep."
A historian later told Wehr that before artificial light began to "shorten the night," people did indeed sleep in two cycles -- which they sometimes called "first sleep" and "morning sleep." The period in between was called "watch."
Between the sleep periods, Wehr's volunteers were neither anxious nor restless. They were also neither completely awake nor asleep, but in a quiescent period of rest.
The anxiety most people feel about waking up and not being able to go back to sleep in the middle of the night apparently is based on believing that they need to pack all their rest into one seven- to eight-hour burst, said Wehr. He speculated that quiescent rest, which is "extinct" in modern life, may explain why some people may benefit from meditation, a similarly restful state.
Wehr started the study to find out if people deprived of light would get more depressed, as some people report feeling during the winter. As it turned out, his volunteers felt better. "They felt so rested," he said.
When asked what his research said about people who wanted to cheat sleep, Wehr said, "That's kind of a Faustian wish, isn't it? Someone asked the Texas billionaire how much is enough money? He said, 'A little more.' That's how we feel about being awake."