Resetting the Circadian Clock

The 24-hour circadian clock embedded in our genes is fundamental to life on this planet, says leading sleep researcher Dr. David Dinges—who has spent the past two decades trying to understand how it works and come up with ways to beat it.

By Harry Goldstein

Sometime in the first half of the next century, we’re going to select some brave humans, stick them in a highly engineered missile and shoot them at the planet Mars. The round-trip could take up to two years. During the journey, time itself will come to mean something other than it does on Earth. Figuring out what that something might be — along with assorted other mysteries associated with how people exist in time — is what keeps Dr. David Dinges up at night. It’s what keeps his test subjects awake as well.
   Dinges is professor of psychology in psychiatry, chief of the Division of Sleep and Chronobiology and director of the Unit for Experimental Psychiatry at the School of Medicine. The laboratory he runs — call it Starship Dinges — is nominally located in a low-traffic area of the Hospital of the University of Pennsylvania (HUP), but functions as a world of its own. The floors rest on massive springs to minimize vibrations. Everyone speaks in hushed tones or whispers, so as not to disturb the sleep subjects. The lights are covered over with sheets of aluminum, resulting in an eerie semi-darkness in which less then one-twentieth as much illumination is available to the eye as in normal daylight. This lighting is kept constant, just as it would be on the first manned spacecraft heading for Mars, tentatively scheduled for 2014.
   The physiological effects of prolonged space flight are going to be formidable. NASA needs to know as much as possible about bone loss, muscle loss, radiation exposure, balance and motion problems, cardiovascular changes, immune changes, sleep-circadian changes and performance capability before any mission to Mars involving humans gets the green light. Toward that end, the National Space Biomedical Research Institute (NSBRI) formed a consortium led by the Baylor College of Medicine that includes Harvard, MIT and Johns Hopkins, as well as Dinges’ lab at Penn. Dinges is associate director of the human performance, sleep and chronobiology team, whose mission is to determine how to optimize human physiology and brain function without the geophysical light/dark cycle of Earth and in the face of 24-hour operations on the spacecraft.
   To a visitor to the lab, a few days into the protocol the crew of Starship Dinges already look like they’ve been to the moon and back. Despite having gotten about eight hours of sleep over the course of the day (broken up into periodic naps), they seem haggard, glassy-eyed, with sallow complexions. The group currently in the lab has been randomly assigned to get a nap of one hour and 20 minutes in the afternoon. Their sleep is interrupted at night to study sleep inertia, the brain’s reluctance to release the neurobiology of sleep — in lay terms, wake up — and function quickly and alertly. Dinges explains sleep inertia as the grogginess and disorientation that you feel when you’ve been awakened too soon from a deep sleep.
   Dinges’ starship troopers live in an environment that simulates the space and environmental conditions of a spacecraft heading for Mars: The three male subjects sleep in one room, the single female subject in another. There is a large common space where the ersatz astronauts eat, work on batteries of tests and are evaluated physically. They have some free time to watch movies or play video games, and they’re allowed to see newspapers — which don’t always come on the days they’re dated.
   The subjects don’t completely lose track of time, but lack the time cues that reset their physiology. One thing the researchers are looking at are circadian-mediated hormones: melatonin, cortisol and, most significantly for the purposes of the NASA study, growth hormone, three-quarters of which is secreted during slow-wave sleep. Sleep is disturbed radically in space. Every major study released in the last 10 years shows sleep periods are reduced from seven-to-eight hours to four-to-six hours. One of the major side effects of living in microgravity for prolonged periods is the loss of muscle and bone mass — which explains the interest in the connection between growth hormone and the sleep system. It might turn out that sleep can be optimized to promote the secretion of growth hormone, which will keep astronauts healthier longer.
   The subjects are wired-up continuously to a small high-tech recorder they wear that measures their brain waves, heart rate, eye movements. and muscle activity, 24 hours a day. They also wear a rectal probe connected to a wrist-mounted sensor unit that measures their core body temperature and arm movements. Dinges insists that the discomfort level is fairly low, which the subjects themselves confirm. “Core body temperature is one of the best measures of the output of the suprachiasmatic nucleus in the hypothalamus — the brain clock, the circadian clock,” explains Dinges. “That’s critical because we have to know what phase [subjects] are at in the circadian cycle to be able to understand what’s happening to the rest of their physiology and brain function.”

All of the physiological information going to the digital recorder on the subjects’ waists is stored and downloaded daily. On selected days of the protocol, subjects also have an intravenous blood line in their arms from which small amounts of blood are taken every 15 minutes for 26 consecutive hours, using a state-of-the-art miniature blood pump system. A fully equipped blood center processes all the blood extracted from the subjects — which never exceeds a pint, or the amount one can volunteer to donate to the Red Cross. The blood gets spun down in one of the centrifuges, pipetted into tubes and put in a minus-70-degree freezer. The blood will be studied not only for hormone levels, but for use in other collaborative experiments the laboratory has underway with investigators at Harvard and Baylor on changes in humoral and cellular immune function in response to chronic sleep loss.

There are also infrared cameras in all the rooms hooked up to closed-circuit monitors at the entrance to the lab. Here at the control console, one of Dinges’ students oversees the protocol according to a script that maps out each minute of every day, from nap-time to dinner, lights out to waking, just the way it would be on the ship to Mars. As the subjects prepare to go down for their afternoon nap, Dinges points out that “we deliberately do not give the test subjects their own bedrooms. On a spaceship, quarters are tight, astronauts sleeping right next to each other. Their job is to try to go to sleep. If they can’t sleep, they still have to lay there in the dark” — activities which the infrared cameras allow the lab technicians to track.

“We try not to burden them with too many probes,” Dinges whispers. The subjects get an electrode holiday every four or five days. On those days, they don’t have to wear electrodes and they can shower. “Everyone looks forward to those days,” he adds, with a smile. “Just psychologically, you can’t ask people not to shower for 14 days.”
Who can withstand the rigors of a simulated space flight? All sorts of people, it seems — provided they’re willing to put up with some inconvenience. In the last three years, Dinges has studied more than 100 individuals in sleep deprivation. Subjects complete a screening process that, besides determining their overall fitness for study, also tests their commitment. They wear recording devices at home, and submit to urine tests and psychological tests. They’re very healthy people between 22 and 42 years of age, with average or above-average IQs. They’re also compensated for their time — about $100 per day, roughly the equivalent of the minimum wage for a 24-hour working day.
They earn that $100, and not just for sitting there allowing themselves to be probed. Every day they perform several different kinds of computerized neurobehavioral tests and synthetic work tasks — most developed in Dinges’ lab and used in 1998 NASA space shuttle missions involving Neurolab and Senator John Glenn’s recent flight. The tests measure frontal-lobe function and a host of other neuropsychological factors. These neurobehavioral assessment batteries are performed every two hours, for 45 minutes to an hour. Dinges’ widely used psychomotor vigilance test is a sustained attention task that measure people’s tendency to lapse or drift off. Other tests measure working memory, cognitive speed, ability to estimate time, mood states, mental manipulation and alertness levels. For an electroencephalograph (EEG) assessment, which measures the electrical activity of the brain, subjects stare at a dot for five minutes and aren’t allowed to touch their heads. While each subject wears about $15,000 worth of miniaturized technology in the lab, they are also expected to wear a wristwatch-sized recorder for two weeks before and one week after their “voyage” on Starship Dinges.

AWAKENING TO SLEEP AS A SUBJECT
   It’s tempting to imagine that Dinges’ research on sleep is motivated by some personal need (“I can’t sleep; therefore I will study it”). But no. “I was always regarded as a good sleeper in my family — going all the way back to infancy,” Dinges reports. According to a family story, when he was about nine, his father cut his hair while he slept. “I awoke long after the haircut, none the wiser until my sister’s giggling prompted me to look in a mirror — in those days crew cuts were standard issue for kids in Kansas,” he adds. In fact, up through college, “my interest in sleep was mostly as an art form — something to be enjoyed.”
   Not that he had much chance. Dinges received his undergraduate degree from Benedictine College, a modest-sized liberal-arts college on the cliffs overlooking the Missouri River in Atchison, Kansas, about 350 miles east of his home in western Kansas — which “seemed quite far at the time.” Academics were strongly emphasized by the Benedictine monks and “civilian” Ph.D.s who made up the faculty, and it was during those years, Dinges says, that he developed “an intense fascination with psychology, biology, philosophy and mathematics” — majoring in the first, and minoring in the other three (a course load of 20-24 credits a semester).
   From the age of about 10, Dinges remembers wanting to grow up to be a scientist. “For some reason, I always had a sense that this is what I would do,” he says. “It must have come from books or television or great teachers, because where I grew up there were no scientists that I knew. However, my parents stressed the importance and value of education, and the Catholic schools I attended rewarded academic rigor.”
   Graduate study in experimental psychology at Saint Louis University — another 350 miles to the east — led to Dinges’ “obsession with sleep.” He studied with Professor Donald Tepas, “a first-rate physiological psychologist,” in whose laboratory he undertook his first research involving sleep and the brain, and also worked four summers as the research assistant to Dr. Hallowell Davis at Central Institute for the Deaf and Washington University School of Medicine. “The late Hallowell Davis was among the most inspiring and decent human beings I have had the privilege of knowing and working with. He was internationally renowned for many discoveries and was among a small group of investigators who first recorded the human EEG from scalp electrodes in the 1930s, reporting the first evidence of systematic EEG changes when people went to sleep,” Dinges recalls. “The sleep research I performed with Hal Davis while in graduate school served to cement my interest in the topic. Both Professor Davis and Professor Tepas set me on the course I pursue to this day.”
   Dinges’ first post-graduate position was at George Washington University in Washington (completing his phased move to the East Coast), where he set up and operated an EEG laboratory to study infants born to mothers being treated with methadone, under a federally funded grant to determine if methadone was more harmful to infant brain development than street drugs. During this study, Dinges discovered that “the sleep EEG was a far more sensitive marker of drug effects and developmental alterations in brain function” than the evoked brain response measures he had initially intended to use. “Thus, my first ‘real’ scientific challenge as a Ph.D. resulted in my turning to sleep physiology as the primary focus of investigation,” he says, the results of which were published in the journal Science in 1980.
   Dinges joined Penn as a post-doctoral fellow in 1977, attracted by the opportunity to work and learn more about sleep in the laboratory of Dr. Martin T. Orne, now emeritus professor of psychiatry, and the availability of affordable housing for himself and his wife Christine. “Both reasons for coming to Penn proved correct,” Dinges says. His tenure with the lab he now directs “has lasted more than 20 years, and has been highly productive scientifically. Christine and I fell in love with Philadelphia, with Penn, and with southern New Jersey beaches. I have never regretted these decisions a day in my life.”
   Penn today boasts one of the country’s largest, most productive and internationally recognized communities of scientists and clinicians studying sleep, sleep disorders and circadian biology, Dinges says, crediting both his “outstanding colleagues” and the “bold and visionary leadership” of the medical school. “From a biomedical-research perspective in general, and a sleep-research perspective in particular, the University of Pennsylvania School of Medicine has been the place to be in the 1990s. There is no sign that this will change as we collectively cross into the next century.”

THE PROFESSOR AND PUBLIC RELATIONS
   If Dinges is eager to credit his mentors and colleagues for their roles in his intellectual development, he’s doing his best to perform the same function for the generation of scholars who’ll come of age in the next century. Over the course of the last three years, Dinges has been able to employ more than 200 undergraduates in his sleep research lab, which is in full-run mode 340 days a year, conducting a variety of studies. Twenty-five undergrads currently work there, most of whom are pre-med, along with two graduate students, a postdoctoral fellow and eight full-time staff — protocol managers; EEG, hardware and software people; blood-sample technicians and nurses.

“The full-time staff and pre- and postdoctoral fellows are vital to the lab’s success, but we also give the undergraduate students an enormous amount of responsibility,” Dinges says with obvious pride. “We train them in the technical procedures, electrode applications, managing the protocol. There’s a whole sequence they have to follow, and they’re trained to deal with emergencies.”

Not only are the students exposed to crucial lab time and expected to pull their weight as part of the research team, but Dinges is a huge believer in sharing the expertise that landed his lab so many projects. His students who take on honors thesis projects go to scientific meetings and present their work. “I want them on the podium,” he says. “I want them to see how science plays out in a more public forum, among physicians and scientists.”
   He expects even more from graduate students. Not only do they work intensively in the lab, they’re the ones sent to perform cockpit or truck simulator studies, or sometimes to conduct field studies, where they’re monitoring pilots on transpacific flights, for example. But it doesn’t end there. “We expect them to go with us to the annual review in front of some of our funding agencies, to present our results to our external scientific-advisory committee,” says Dinges. “We make an effort to show they’re actively involved and to show the funding agency that we’re training the next generation of researchers to deal with these kinds of issues.”
   Dinges thinks that if a scientist takes public money, he or she should be able to explain what will be done with it and why it’s worth doing. Funding from the National Institutes of Health (NIH) for investigating the consequences of acute and chronic sleep loss has been “our mainstay,” says Dinges, but other research support has come from the Department of Defense, “which has had a long-standing interest in ways to maintain the biological limits of human function, as well as NASA and more recently from the Department of Transportation, which is concerned with preventing drowsy driving crashes.”

BEING AND SLEEPLESSNESS
   “I desperately want to know what wakefulness is and what sleep is and what controls them,” says Dinges. “What is biology trying to tell us about why brains need to sleep?” Normal sleep is fundamental to achieve normal wakefulness and effortless consciousness. “I am obsessed with consciousness, what people call ‘being awake.’ Most people know that there are different stages of sleep — REM and non-REM. But most people think of wakefulness as a single state: ‘I’m awake!’ We know that is not so.”
   Dinges’ experiments show that people’s ability to integrate information, remember, be alert, is dynamically changing across the day and night. As the pressure for sleep increases due to experimental sleep loss, or a sleep disorder, or night-shift work, the wake state — as reflected in responses during sustained attention and brain waves — becomes unstable, drifting in and out of full functioning second by second. When sleep pressure is elevated, our sense of wakefulness as a continuous state is illusory. It is a “story” that Dinges says the brain weaves together by filling in attention and memory gaps with a coherent theme. It’s this unstable apparent alertness that hides the actual neurobehaviorial impairments of sleepy patients and subjects. And it is this lack of appreciation for the impairment and likelihood of a sudden sleep attack that makes the excessively sleepy person so dangerous when driving or engaged in other safety-sensitive activities.

“If you saw one of our subjects who’d been awake 88 hours, you’d observe that — even though when he sits down to perform he’s having uncontrolled sleep attacks, his reaction times are slow, he can’t remember even one out of six words he’s just seen, he can’t estimate time, his cognitive throughput is very slow — when you interact socially with him, he appears to be normal and he is likely to tell you he is only a little tired but doing OK! How is that possible? What’s going on with the brain? What role does body posture, social stimulation, environmental stimulation play in this paradox? Understanding how and why humans so often misjudge their alertness and waking performance capabilities are major scientific goals for our laboratory.”
   Achieving these goals is important at the applied level to figure out why people don’t stop driving or performing other tasks even when they’re severely impaired from sleepiness. We’ve all had it happen: You’re behind the wheel driving down a dark, empty road. Your eyelids are heavy and your head’s nodding. When the nodding starts, the neck muscles are giving out. They can’t hold up your head any longer. Something deep in the primitive brain is shutting off the lights and switching you into sleep mode. And yet, some people will continue to drive in that stuporous semi-awake state — often straight off the road. “What we see in our studies over and over again is that people’s ability to judge their capability and their alertness level is completely divorced from their physiology, from their EEG and their performance,” says Dinges. “And the more impaired you become, the more dissociated it gets, which may explain why some severely sleepy patients with sleep disorders do not report sleepiness.”
   Over the years, Dinges and his colleagues have been able to figure out why this dissociation between perceived capabilities and actual performance happens so frequently. The objective judgment of performance capability is influenced by body posture and social interaction. The very act of asking someone, “Are you all right?” and the expectation the question carries with it, changes that person’s ability to judge it. Dinges’ psychomotor vigilance test is a very sensitive assay for unstable wakefulness due to sleepiness. In lab experiments, in tests of cockpit-napping among commercial-airline pilots flying transpacific routes, and in tests done on physicians on-call at Stanford and at Penn, Dinges has found that although sleepiness occurred in everyone, a subgroup of people tended to fail severely after only a modest amount of sleep deprivation, while another subgroup was remarkably resilient to sleep loss up to 40 hours of continuous wakefulness. This means that certain people can stand being awake for longer periods without performance degradation than can other people, and that some of the latter are impaired much sooner than they realize. Preliminary evidence from Dinges’ lab suggests these different vulnerabilities to sleep loss may be stable, enduring characteristics of people — a question he hopes to definitively answer in a newly funded research project.
   These findings give rise to ethical and policy dilemmas: Should people get safety-sensitive jobs based on their performance vulnerability to sleep deprivation? If, for example, we can objectively detect the sleepiness of a driver using invisible light reflecting off the driver’s eyes — something Dinges’ research has demonstrated — should such devices become mandatory in motor vehicles? Dinges feels a certain responsibility for resolving these dilemmas, which in a way he helped create through his research. He is a strong advocate for the need to establish the scientific validity of alertness-monitoring technologies before they become available to the public, because his studies for DOT have demonstrated that most of the current technologies do not reliably detect sleepiness impairment. He also believes that companies and most government agencies should not deploy such screening and monitoring technologies without the voluntary consent of the workers to prevent a nasty Big Brotherish future.
   

WHO NEEDS SLEEP, AND HOW MUCH?
   If Dinges is successful in the other half of his work — developing pharmacological and technological ways to level the sleep-need playing field — the inability to function because of lack of sleep or need to sleep will be less of a factor in performance and, therefore, in deciding who gets a job or who gets fired. A key to these technologies will be diagnostics that will tell who is more and less needy when it comes to sleep. “If we can find the predictors of this we can use them, not to select people for jobs, but rather to make sure that people who are using the alertness gauges or wake-promoting substances are the ones who most need them, including the millions of people who become sleepy for medical reasons,” Dinges says.
   But the secrets to sleep’s mysteries aren’t confined to the exhausted head-nodding and burning eyelid-drooping indicative of the need to sleep. There is also a variation in individuals’ ability to sleep and wake, which can determine how much sleep you need and when you need it. Dinges suspects that some people have a strong sleep-drive and a weak wake-drive, making them very vulnerable to sleepiness. Others have a strong sleep-drive and a strong-wake drive, or a weak sleep-drive and a strong wake-drive, making them able to withstand sleep deprivation. Still others could have a weak sleep-drive and a weak wake-drive, which might predispose them to insomnia. These drives constitute the sleep-homeostatic side of the brain’s alertness control system.
   The other half of the system is circadian, the clock running in our brains. We all recognize “morning people” and “night owls,” but there are individuals whose circadian systems are so robust that they fall asleep at the same time every night and can in no way stay up past a certain hour. Dinges believes that all individual differences in sleep patterns and waking alertness levels may be explained by neurobiological interactions of the sleep-wake drives and circadian clock in the brain — something he is helping to model mathematically with collaborators at Harvard Medical School. If an accurate mathematical algorithm of human alertness and performance can be produced, it can be used for computer prediction of safe work-rest schedules and for anticipating when humans will require sleep or other alertness-promoting interventions.

GETTING WIRED A DOSE AT A TIME
   One commonly used way to counter the sleep drive is good old caffeine. Dinges directs Penn’s Air Force-funded Center for Countermeasures for Jetlag and Sleep Deprivation. The Center has subcomponent projects at Harvard under the direction of Dr. Charles A. Czeisler and at Dr. Dale M. Edgar’s lab at Stanford. The group is studying the effects of low-dose caffeine and other wake-promoting therapeutics.
   Dinges is convinced that even though caffeine is the most widely used wake-promoting compound in the world, people exposed to job-induced sleepiness are not optimizing its capability. Center studies at Penn and Harvard are investigating whether it is possible to “promote physiological alertness and performance capability in people who must be awake for prolonged periods, by giving them, not a big dose of caffeine like I might have in this cup” — Dinges hoists a beaker filled to the brim with black coffee — “but a small amount every hour.”
   At Harvard, low-dose caffeine is being studied in a forced desynchrony protocol, where people live in complete time-isolation for about 50 days. Their period of wakefulness is 28 hours, followed by 14 hours of sleep — throwing the normal 24-hour circadian cycle totally off-kilter. Penn’s role in the Center is studying the need for sleep (the sleep homeostatic drive), using protocols in which people are kept awake for prolonged periods. The Air Force protocol represents the outer boundary of how far the Penn group — or any other laboratory in the world — has gone with sleep deprivation in a large number of subjects. To date in this placebo-controlled trial of low dose caffeine, 51 subjects have spent 10 grueling days in the lab. After getting about eight hours of sleep a night for the first two days, an IV goes in and stays in for the next five days. Small amounts of blood are drawn regularly. Brain activity and heart rate are continuously recorded via several electrodes on the face, head and body, and a rectal probe measures core body temperature to monitor the circadian cycle. An accelerometer is strapped to the wrist to detect motion. The subjects are tested quasi-continuously on all 10 days.
   And every hour they get a pill — either a sugar pill (placebo) or a low dose of caffeine (0.3 mg of caffeine per kilogram of body weight). This continues for 88 hours. Depending on the luck of the draw, subjects are assigned either to no sleep at all during that 88 hours (3.7 days of continual wakefulness) or they get two-hour sleep opportunities every 12 hours — four hours of sleep per day if you take full advantage of each opportunity. Afterward, they get either seven or 14 hours of recovery sleep.
   The study, which is in its final year of data acquisition, “is going to tell us a great deal about what the limit is on human functioning in severe sustained-operation scenarios,” says Dinges. Should sustained low-dose caffeine prove to be an effective and safe way to keep people functional during emergencies when sleep is rarely possible — such as military conflicts — Dinges envisions that ultimately people may wear an electrochemical patch, like a nicotine patch, with timed doses of caffeine (or some other safe wake-promoting substance) that they can put on when they have to work late at night and they can pull off anytime — and that ideally can also monitor blood levels for toxicity.
   “I talk about this wake-up chemical patch, about unobtrusively monitoring truck driver or pilot alertness with infrared technology, and about sleep logistics for people going to Mars, and I often see listeners shake their heads in disbelief,” says Dinges. “But developments like these will come to pass in the next century. Our laboratory’s research on them reflects our desire to help not only NIH discover ways to better direct and manage pathologically sleepy patients, but also to help federal agencies, policy makers, industries and the public at large safely and effectively cope in a world that demands more time flexibility and more wakefulness than biology naturally permits. We must find ways to prevent [disasters like] the Exxon Valdez and Three Mile Island — events where humans working with highly lethal systems made critical mistakes on the night shift because they were tired. We also must stop the carnage on our highways from fall asleep crashes. You just can’t say, ‘Let’s go back to an agrarian economy and everyone will sleep when it gets dark outside.’ We’ve changed our world.”
   For the United States and the developed world, the transition from an agrarian to an industrial society and now to the information age has been paralleled by the development of ever-more sophisticated ways to permit humans to circumvent time and space, from electric lights at the century’s beginning, to telephones and jet planes, and now the Internet and space travel. The Internet doesn’t operate by a circadian clock, or even a digital one. On the Internet, there is only the perpetual Now. Thanks in part to the Internet and e-mail, the 40-hour work week is just a fading memory for many people. In the information age, you never have to stop working. That is, until you fall face down on your keyboard. This is the world Dinges is working to help us cope with — as organisms biologically engineered to live on the cycles of Earth time.
   “I hope we never find a way to do away with sleep,” Dinges observes, “although I’m certainly working on the front lines for us to cope safely with less sleep. Because sleep connects us in a fundamental way to life on this planet. Our brain’s circadian clock and our need to sleep are daily reminders to us of our place in the universe — we hail from a planet called Earth, with a 24-hour light-dark cycle. We need to sleep.”

Harry Goldstein is a freelance writer based in New York.