Published On Sep 22, 2007
Yardsticks developed in 1968 can’t explain many disorders. New tools may reveal what really happens when one’s head hits the pillow.
WHEN ROBERT THOMAS, A SLEEP RESEARCHER, approached Ary Goldberger, a cardiologist, Thomas was looking for a way to analyze sleep patterns that didn’t require patients to spend an entire night in a hospital or clinic for a traditional sleep study. Thomas thought that while most sleep research focused on the brain, signals from other parts of the body might also yield crucial information. “Sleep is a brain function, but it’s also a full-body state,” says Thomas, who wanted to see what could be learned by looking at information from the heart.
As it happened, Goldberger’s lab at Beth Israel Deaconess Medical Center in Boston, which specializes in extracting hidden data from complex signals, had already developed a computer algorithm for analyzing sleep patterns using an electrocardiogram, or ECG. (Goldberger had wanted to help fellow cardiologists employ a familiar instrument to diagnose sleep apnea, a disorder that has been found to significantly increase heart disease and mortality. He also thought patients would prefer wearing a heart monitor at home to participating in an overnight sleep study in the clinic, and using ECGs meant doctors could follow patients over several nights to monitor the effects of treatments.) But when the researchers compared ECG-analyzed sleep studies with results from the conventional approach, the two didn’t seem to match up, so they abandoned the effort.
That mismatch didn’t bother Thomas (also at Beth Israel), who was hunting for something different: a heart-generated signal that might correspond to the cyclic alternating pattern, or CAP, a somewhat controversial brain-based indicator of restless sleep. CAP episodes may signal microarousals: frequent, fleeting awakenings that seem to underlie many sleep problems. But CAP patterns are extremely difficult to measure, and a technician who hasn’t been specially trained won’t detect them on an electroencephalogram (EEG), the traditional sleep study’s primary tool.
When Goldberger and his team dusted off their algorithm and compared ECG analyses of sleep studies with those of CAP from EEGs, it was almost a perfect fit. “The CAP patterns just fell out of the data,” Goldberger recalls about seeing ECG patterns that showed up just when CAP episodes occurred. “It was miraculous.”
That may seem like a small advance, and its validity remains to be confirmed. Yet as one of several recent efforts to improve on long-standing ways of gauging sleep disorders, the effort could be significant. Chronic sleep problems rank among the top complaints patients bring to doctors, affecting as many as 70 million Americans, according to the Institute of Medicine. Breathing disorders, including sleep apnea, are among the most serious, yet as many as nine of 10 sleep breathing concerns remain undiagnosed and untreated.
Though one problem may be that some of these patients aren’t seen by doctors, another problem may be how sleep is measured. In 1968 the American Academy of Sleep Medicine (AASM) pooled observations about EEG patterns in the normal sleep of young adults and settled on a system of several distinct sleep stages. During the nearly four decades since then, that early consensus about what happens during sleep has proven to be extraordinarily useful, and it remains the gold standard for understanding sleep problems. Yet it seems increasingly likely that it tells only part of the story of what happens when the head hits the pillow.
MODERN SLEEP RESEARCH DAWNED in the 1930s, after the arrival of the electroencephalogram. The sleeping brain, long thought to be at rest, as if in a reversible coma, turned out not only to be active but to have two distinct patterns of activity, rapid eye movement (REM) and non-REM (NREM). Patterns of electrical activity in the brain show up on EEG graphs as waves and vary according to whether the subject is undergoing REM or NREM sleep. In REM, brain waves are “activated,” appearing on an EEG readout as rapid, random, low-amplitude waves that somewhat resemble those of waking. In NREM, much larger, slower and more rhythmic delta waves gradually appear, giving the deepest stage of sleep the name slow-wave sleep, or SWS. Before reaching SWS, however, a sleeper passes through a phase of NREM, punctuated by fast, erratic frequencies.
According to the original 1968 formulation, NREM had four parts. We enter shallow stage 1 from wakefulness and pro-gress through stage 2, from which it is easy to awaken. Next we slip into slow-wave sleep in stage 3 and then descend deeper in stage 4. (The AASM recently collapsed stages 3 and 4 into stage 3, concluding that research didn’t support the idea that they were two discernibly separate periods.)
These sleep stages, gauged by brain-wave activity, remain the cornerstone of sleep studies, which also monitor changes in heart rate, temperature and muscle tone. Besides the six EEG electrodes attached to a patient’s head to measure brain waves, two electrooculograms detect the rapid eye movements of REM. Three electromyogram leads at the chin register the decreased muscle tone of REM, and two on each leg check for periodic limb movements in sleep (PLMS), a disorder characterized by repeated jerky movements that can mar the quality of sleep. Two electrocardiogram leads on the chest track the slow, even heartbeats of NREM and the more rapid, irregular ones of REM. For detecting potential breathing disorders, there’s a microphone to record snores, a pneumotachograph in the nose to measure airflow, two piezoelectric bands around the chest and abdomen to register the force of inhalation, and a meter taped to a finger to record whether oxygen saturation falls following breathing disruptions.
Although often less than pleasant for subjects—some swear they didn’t get a lick of sleep, though the monitors prove otherwise—polysomnography provides valuable diagnostic evidence. But some disorders don’t register. For example, people are frequently referred for sleep studies because of fatigue and excessive daytime sleepiness, explains Amit Verma, director of the divisions of neurophysiology and sleep disorders at the Methodist Neurological Institute in Houston. These aren’t trivial complaints—they make people prone to accidents and reduce quality of life. The first guess at a cause is usually sleep apnea, in which oxygen loss due to missed breaths rouses the sleeper as many as 100 times per hour. Apnea typically leaves a dramatic footprint on an EEG, showing patterns of wakefulness, and there’s usually a clear drop in oxygen saturation, among other telltale signs. Yet the sleep studies of some patients who feel perpetually sleepy look relatively normal.
ONE BIG PROBLEM WITH SLEEP RESEARCH based on electroencephalography, according to Terrence Sejnowski, a Howard Hughes Medical Institute investigator at the Salk Institute for Biological Studies in La Jolla, Calif., is that even today sleep technicians must score EEGs manually, looking at one 30-second period, or epoch, at a time to determine that, say, a certain epoch looks like NREM stage 2. Sejnowski thinks this manual system misses many nuances that an automated system could catch. For example, Verma suspects that upper airway resistance syndrome, or UARS—essentially a milder form of sleep apnea—may be the real problem for some patients who complain of daytime sleepiness. Yet the brain-wave and breathing disturbances caused by UARS, though they may cause complications, are fleeting and don’t leave the telltale footprint that apnea does on most EEGs and other traditional sleep study measures. They may go undetected during standard scoring of EEGs.
Scientists have tried for years to automate EEG scoring, but there has always been a fatal flaw, Sejnowski says. “Previous computer systems have relied on expert EEG scoring,” he says, “but experts only agree 75% of the time.”
So Sejnowski, a neuroscientist who develops computer models that teach themselves to recognize patterns in complex signals, and graduate student Philip Low created an algorithm for analyzing low- and high-frequency EEG patterns during sleep, building in much shorter epochs for finer resolution. “When we started, we had no idea what we’d find,” Sejnowski says. “But we were amazed to see three clear clusters emerge. They’re different from traditional sleep stages, though they actually correspond pretty well to those stages.”
The first pattern is an intermediate stage between wakefulness and sleep that corresponds to traditional NREM 2. In the second, slow-wave sleep occurs as in the newly consolidated NREM 3 (now called N3), and in the third, REM appears. (They found no NREM 1, which Sejnowski believes has been a catch-all for epochs that contain abrupt transitions between stages.)
In addition to giving a clearer picture of sleep stages, the computer-analyzed EEGs also revealed information that manually scored tests miss entirely. To his surprise, looking at the new graphs, Sejnowski found that both slow-wave sleep and REM are fragmented, with the brain shifting between different wavelengths every two to three seconds. These shifts occur in the low-frequency range during REM and in high frequencies during SWS. That fragmentation could prove to be clinically important, providing diagnostic clues.
A CLOSER READING OF EEGS WAS also how researchers, working in Parma, Italy, more than 20 years ago, discovered the cyclic alternating pattern. These scientists noticed spontaneously occurring “unstable” periods in NREM during which brain waves alternate rapidly between frequencies. A CAP repeats every 20 to 60 seconds or so and alternates with more stable periods of NREM. The researchers think that during the times of unstable CAP, the sleeper undergoes frequent sleep disturbances and even microarousals that are so subtle they may not be readily visible on conventional sleep studies.
CAP is part of the normal fabric of sleep, but it becomes more pronounced in sleep disorders. During the two decades since they first identified CAP, the Parma researchers have sought to show connections between unusual CAP activity and sleep problems. One study, which they reported in 1996 in the Journal of Clinical Neurophysiology, compared 12 sleepers with periodic limb movements to 12 control subjects. Those with PLMS showed above-normal CAP activity, and almost all of their jarring movements occurred during CAP. More recently, a May 2007 study in the journalSleep found a link between excessive CAP and subjective complaints of daytime sleepiness in patients with previously undiagnosed upper airway resistance syndrome.
In their work with electrocardiograms, Thomas and Goldberger also found two sleep patterns, unstable and stable sleep, “as different from each other as Metallica and Beethoven,” Thomas says. These patterns are similar to the CAP (unstable) and non-CAP (stable) patterns visible to the trained eye on an EEG, but instead of just reflecting a brain state, they represent the coupling of many biological systems, including the autonomic nervous system, respiration and heart rate. Stable and unstable patterns oscillate along a recurring time scale, much as CAP episodes do, and are independent of standard sleep stages. It seems everyone has some unstable sleep, but sleep disorders produce more of it.
Hard to glean from EEG squiggles, stable and unstable sleep patterns are easily recognized on the new readouts produced by Goldberger’s algorithm. These graphs, called sleep spectrograms, resemble 3-D mountain ranges, with the landscape of stable sleep hovering over the unstable ranges. By showing the impact of arousals or breathing disruptions, the landscapes may essentially map how well someone is sleeping.
The stable sleep of a healthy subject produces much higher peaks on a spectrogram than does unstable sleep, whereas in a patient with sleep apnea, for example, the stable landscape almost disappears, and the unstable ranges dominate. Yet after an apnea patient is aided with a special breathing device, a healthy landscape profile will immediately appear on the spectrogram. According to Goldberger and Thomas, this method also makes it easier to distinguish obstructive sleep apnea, caused by mechanical blockage at the throat, from the rarer central sleep apnea, caused by faulty brain signals or by heart failure. That’s an important distinction, because central sleep apnea often requires different treatments, including drugs to modulate breathing control, and may be worsened by conventional therapies for obstructive apnea.
IN ADDITION TO HELPING PHYSICIANS diagnose disorders, stable and unstable patterns reveal something about sleep quality that doesn’t involve measuring age-variable slow-wave sleep. While healthy 60-year-olds show little slow-wave sleep on an EEG—apparently, older people get much less SWS—their ECG-derived spectrograms show plenty of stable sleep. Indeed, Thomas thinks that stable patterns, which aren’t captured at all in standard sleep studies, could be an effective measure of sleep quality. In stable sleep, the brain, the heart and respiration become calm, whereas in unstable sleep, everything fluctuates as on a choppy sea—the brain undergoes more frequent microarousals, or CAP, and heart and respiration rates rise and fall, for instance—and this produces a less restorative slumber.
Because neither CAP nor stable-unstable sleep patterns correspond with standard sleep stages, many sleep physicians are skeptical of their importance. But others, including Verma, welcome this research as an attempt to explain what’s going on with patients who don’t meet the criteria for sleep apnea but whose sleep isn’t refreshing. And Virend Somers, a consultant cardiologist at the Mayo Clinic in Rochester, Minn., likes having an alternative approach for evaluating the quality and nature of sleep and identifying its pathologies. “What happens to the heart during sleep, as measured by an ECG, is an important emerging area,” Somers says.
Yet Verma doesn’t think the sleep spectrograms will soon replace the EEG as a diagnostic tool. Despite the statistical association between CAP and ECG-measured unstable sleep, directly gauging CAP activity on an EEG is much further along, in terms of scientific validity and widespread acceptance, than the new ECG-based approach, Verma says. Still, the ECG research, plus several other independent lines of inquiry into alternating brain patterns, has finally begun to modify our 40-year-old conception of what sleep is. As Sejnowski says of his powerful new model for recognizing previously invisible EEG patterns, “We’re zooming in on the nuances of sleep and finding things no one has ever seen.” It could take years before the average sleep clinic is equipped to uncover such nuances. Racing against the almost-daily news about the ill effects of poor sleep, it can catch up none too soon.
1. “An Electrocardiogram-Based Technique to Assess Cardiopulmonary Coupling During Sleep,” by Robert Thomas, Joseph Mietus, Chung-Kang Peng and Ary Goldberger, Sleep, Volume 28, 2005. An introduction to the new method of “listening” to sleep by analyzing patterns from heart signals, then displaying those patterns in a visual format called a sleep spectrogram.
2. “The Nature of Arousal in Sleep,” by Péter Halász, Mario Terzano, Liborio Parrino and Róbert Bódizs, Journal of Sleep Research, Vol. 13, 2004. The authors theorize that, one, microarousals result from the cyclic alternating pattern (CAP) of brain waves; and, two, contrary to conventional views, arousals are normal elements of sleep that become dangerous only in sleep disorders.
3. “The Cyclic Alternating Pattern Demonstrates Increased Sleep Instability and Correlates With Fatigue and Sleepiness in Adults With Upper Airway Resistance Syndrome,” by Christian Guilleminault, Cecilia Lopes, Chad Hagen and Agostinho da Rosa,Sleep, May 1, 2007. This small but intriguing study found that patients complaining of excessive daytime sleepiness have a mild breathing disorder that conventional sleep studies do not detect, but that CAP analysis can.