Why We Sleep

by Matthew Walker

  You may have been expecting your general brainwave activity to look beautifully coherent and highly synchronous while awake, matching the ordered pattern of your (mostly) logical thought during waking consciousness. The contradictory electrical chaos is explained by the fact that different parts of your waking brain are processing different pieces of information at different moments in time and in different ways. When summed together, they produce what appears to be a discombobulated pattern of activity recorded by the electrodes placed on your head.

  As an analogy, consider a large football stadium filled with thousands of fans. Dangling over the middle of the stadium is a microphone. The individual people in the stadium represent individual brain cells, seated in different parts of the stadium, as they are clustered in different regions of the brain. The microphone is the electrode, sitting on top of the head—a recording device.

  Before the game starts, all of the individuals in the stadium are speaking about different things at different times. They are not having the same conversation in sync. Instead, they are desynchronized in their individual discussions. As a result, the summed chatter that we pick up from the overhead microphone is chaotic, lacking a clear, unified voice.

  When an electrode is placed on a subject’s head, as done in my laboratory, it is measuring the summed activity of all the neurons below the surface of the scalp as they process different streams of information (sounds, sights, smells, feelings, emotions) at different moments in time and in different underlying locations. Processing that much information of such varied kinds means that your brainwaves are very fast, frenetic, and chaotic.

  Once settled into bed at my sleep laboratory, with lights out and perhaps a few tosses and turns here and there, you will successfully cast off from the shores of wakefulness into sleep. First, you will wade out into the shallows of light NREM sleep: stages 1 and 2. Thereafter, you will enter the deeper waters of stages 3 and 4 of NREM sleep, which are grouped together under the blanket term “slow-wave sleep.” Returning to the brainwave patterns of figure 9, and focusing on the middle line, you can understand why. In deep, slow-wave sleep, the up-and-down tempo of your brainwave activity dramatically decelerates, perhaps just two to four waves per second: ten times slower than the fervent speed of brain activity you were expressing while awake.

  As remarkable, the slow waves of NREM are also far more synchronous and reliable than those of your waking brain activity. So reliable, in fact, that you could predict the next few bars of NREM sleep’s electrical song based on those that came before. Were I to convert the deep rhythmic activity of your NREM sleep into sound and play it back to you in the morning (which we have also done for people in the same sonification-of-sleep project), you’d be able to find its rhythm and move in time, gently swaying to the slow, pulsing measure.

  But something else would become apparent as you listened and swayed to the throb of deep-sleep brainwaves. Every now and then a new sound would be overlaid on top of the slow-wave rhythm. It would be brief, lasting only a few seconds, but it would always occur on the downbeat of the slow-wave cycle. You would perceive it as a quick trill of sound, not dissimilar to the strong rolling r in certain languages, such as Hindi or Spanish, or a very fast purrr from a pleased cat.

  What you are hearing is a sleep spindle—a punchy burst of brainwave activity that often festoons the tail end of each individual slow wave. Sleep spindles occur during both the deep and the lighter stages of NREM sleep, even before the slow, powerful brainwaves of deep sleep start to rise up and dominate. One of their many functions is to operate like nocturnal soldiers who protect sleep by shielding the brain from external noises. The more powerful and frequent an individual’s sleep spindles, the more resilient they are to external noises that would otherwise awaken the sleeper.

  Returning to the slow waves of deep sleep, we have also discovered something fascinating about their site of origin, and how they sweep across the surface of the brain. Place your finger between your eyes, just above the bridge of your nose. Now slide it up your forehead about two inches. When you go to bed tonight, this is where most of your deep-sleep brainwaves will be generated: right in the middle of your frontal lobes. It is the epicenter, or hot spot, from which most of your deep, slow-wave sleep emerges. However, the waves of deep sleep do not radiate out in perfect circles. Instead, almost all of your deep-sleep brainwaves will travel in one direction: from the front of your brain to the back. They are like the sound waves emitted from a speaker, which predominantly travel in one direction, from the speaker outward (it is always louder in front of a speaker than behind it). And like a speaker broadcasting across a vast expanse, the slow waves that you generate tonight will gradually dissipate in strength as they make their journey to the back of the brain, without rebound or return.

  Back in the 1950s and 1960s, as scientists began measuring these slow brainwaves, an understandable assumption was made: this leisurely, even lazy-looking electrical pace of brainwave activity must reflect a brain that is idle, or even dormant. It was a reasonable hunch considering that the deepest, slowest brainwaves of NREM sleep can resemble those we see in patients under anesthesia, or even those in certain forms of coma. But this assumption was utterly wrong. Nothing could be further from the truth. What you are actually experiencing during deep NREM sleep is one of the most epic displays of neural collaboration that we know of. Through an astonishing act of self-organization, many thousands of brain cells have all decided to unite and “sing,” or fire, in time. Every time I watch this stunning act of neural synchrony occurring at night in my own research laboratory, I am humbled: sleep is truly an object of awe.

  Returning to the analogy of the microphone dangling above the football stadium, consider the game of sleep now in play. The crowd—those thousands of brain cells—has shifted from their individual chitter-chatter before the game (wakefulness) to a unified state (deep sleep). Their voices have joined in a lockstep, mantra-like chant—the chant of deep NREM sleep. All at once they exuberantly shout out, creating the tall spike of brainwave activity, and then fall silent for several seconds, producing the deep, protracted trough of the wave. From our stadium microphone we pick up a clearly defined roar from the underlying crowd, followed by a long breath-pause. Realizing that the rhythmic incantare of deep NREM slow-wave sleep was actually a highly active, meticulously coordinated state of cerebral unity, scientists were forced to abandon any cursory notions of deep sleep as a state of semi-hibernation or dull stupor.

  Understanding this stunning electrical harmony, which ripples across the surface of your brain hundreds of times each night, also helps explain your loss of external consciousness. It starts below the surface of the brain, within the thalamus. Recall that as we fall asleep, the thalamus—the sensory gate, seated deep in the middle of the brain—blocks the transfer of perceptual signals (sound, sight, touch, etc.) up to the top of the brain, or the cortex. By severing perceptual ties with the outside world, not only do we lose our sense of consciousness (explaining why we do not dream in deep NREM sleep, nor do we keep explicit track of time), this also allows the cortex to “relax” into its default mode of functioning. That default mode is what we call deep slow-wave sleep. It is an active, deliberate, but highly synchronous state of brain activity. It is a near state of nocturnal cerebral meditation, though I should note that it is very different from the brainwave activity of waking meditative states.

  In this shamanistic state of deep NREM sleep can be found a veritable treasure trove of mental and physical benefits for your brain and body, respectively—a bounty that we will fully explore in chapter 6. However, one brain benefit—the saving of memories—deserves further mention at this moment in our story, as it serves as an elegant example of what those deep, slow brainwaves are capable of.

  Have you ever taken a long road trip in your car and noticed that at some point in the journey, the FM radio stations you’ve been listening to begin dropping out in signal strength? In contrast, AM radio statio
ns remain solid. Perhaps you’ve driven to a remote location and tried and failed to find a new FM radio station. Switch over to the AM band, however, and several broadcasting channels are still available. The explanation lies in the radio waves themselves, including the two different speeds of the FM and AM transmissions. FM uses faster-frequency radio waves that go up and down many more times per second than AM radio waves. One advantage of FM radio waves is that they can carry higher, richer loads of information, and hence they sound better. But there’s a big disadvantage: FM waves run out of steam quickly, like a muscle-bound sprinter who can only cover short distances. AM broadcasts employ a much slower (longer) radio wave, akin to a lean long-distance runner. While AM radio waves cannot match the muscular, dynamic quality of FM radio, the pedestrian pace of AM radio waves gives them the ability to cover vast distances with less fade. Longer-range broadcasts are therefore possible with the slow waves of AM radio, allowing far-reaching communication between very distant geographic locations.

  As your brain shifts from the fast-frequency activity of waking to the slower, more measured pattern of deep NREM sleep, the very same long-range communication advantage becomes possible. The steady, slow, synchronous waves that sweep across the brain during deep sleep open up communication possibilities between distant regions of the brain, allowing them to collaboratively send and receive their different repositories of stored experience.

  In this regard, you can think of each individual slow wave of NREM sleep as a courier, able to carry packets of information between different anatomical brain centers. One benefit of these traveling deep-sleep brainwaves is a file-transfer process. Each night, the long-range brainwaves of deep sleep will move memory packets (recent experiences) from a short-term storage site, which is fragile, to a more permanent, and thus safer, long-term storage location. We therefore consider waking brainwave activity as that principally concerned with the reception of the outside sensory world, while the state of deep NREM slow-wave sleep donates a state of inward reflection—one that fosters information transfer and the distillation of memories.

  If wakefulness is dominated by reception, and NREM sleep by reflection, what, then, happens during REM sleep—the dreaming state? Returning to figure 9, the last line of electrical brainwave activity is that which I would observe coming from your brain in the sleep lab as you entered into REM sleep. Despite being asleep, the associated brainwave activity bears no resemblance to that of deep NREM slow-wave sleep (the middle line in the figure). Instead, REM sleep brain activity is an almost perfect replica of that seen during attentive, alert wakefulness—the top line in the figure. Indeed, recent MRI scanning studies have found that there are individual parts of the brain that are up to 30 percent more active during REM sleep than when we are awake!

  For these reasons, REM sleep has also been called paradoxical sleep: a brain that appears awake, yet a body that is clearly asleep. It is often impossible to distinguish REM sleep from wakefulness using just electrical brainwave activity. In REM sleep, there is a return of the same faster-frequency brainwaves that are once again desynchronized. The many thousands of brain cells in your cortex that had previously unified in a slow, synchronized chat during deep NREM sleep have returned to frantically processing different informational pieces at different speeds and times in different brain regions—typical of wakefulness. But you’re not awake. Rather, you are sound asleep. So what information is being processed, since it is certainly not information from the outside world at that time?

  As is the case when you are awake, the sensory gate of the thalamus once again swings open during REM sleep. But the nature of the gate is different. It is not sensations from the outside that are allowed to journey to the cortex during REM sleep. Rather, signals of emotions, motivations, and memories (past and present) are all played out on the big screens of our visual, auditory, and kinesthetic sensory cortices in the brain. Each and every night, REM sleep ushers you into a preposterous theater wherein you are treated to a bizarre, highly associative carnival of autobiographical themes. When it comes to information processing, think of the wake state principally as reception (experiencing and constantly learning the world around you), NREM sleep as reflection (storing and strengthening those raw ingredients of new facts and skills), and REM sleep as integration (interconnecting these raw ingredients with each other, with all past experiences, and, in doing so, building an ever more accurate model of how the world works, including innovative insights and problem-solving abilities).

  Since the electrical brainwaves of REM sleep and wake are so similar, how can I tell which of the two you are experiencing as you lie in the bedroom of the sleep laboratory next to the control room? The telltale player in this regard is your body—specifically its muscles.

  Before putting you to bed in the sleep laboratory, we would have applied electrodes to your body, in addition to those we affix to your head. While awake, even lying in bed and relaxed, there remains a degree of overall tension, or tone, in your muscles. This steady muscular hum is easily detected by the electrodes listening in on your body. As you pass into NREM sleep, some of that muscle tension disappears, but much remains. Gearing up for the leap into REM sleep, however, an impressive change occurs. Mere seconds before the dreaming phase begins, and for as long as that REM-sleep period lasts, you are completely paralyzed. There is no tone in the voluntary muscles of your body. None whatsoever. If I were to quietly come into the room and gently lift up your body without waking you, it would be completely limp, like a rag doll. Rest assured that your involuntary muscles—those that control automatic operations such as breathing—continue to operate and maintain life during sleep. But all other muscles become lax.

  This feature, termed “atonia” (an absence of tone, referring here to the muscles), is instigated by a powerful disabling signal that is transmitted down the full length of your spinal cord from your brain stem. Once put in place, the postural body muscles, such as the biceps of your arms and the quadriceps of your legs, lose all tension and strength. No longer will they respond to commands from your brain. You have, in effect, become an embodied prisoner, incarcerated by REM sleep. Fortunately, after serving the detention sentence of the REM-sleep cycle, your body is freed from physical captivity as the REM-sleep phase ends. This striking dissociation during the dreaming state, where the brain is highly active but the body is immobilized, allows sleep scientists to easily recognize—and therefore separate—REM-sleep brainwaves from wakeful ones.

  Why did evolution decide to outlaw muscle activity during REM sleep? Because by eliminating muscle activity you are prevented from acting out your dream experience. During REM sleep, there is a nonstop barrage of motor commands swirling around the brain, and they underlie the movement-rich experience of dreams. Wise, then, of Mother Nature to have tailored a physiological straitjacket that forbids these fictional movements from becoming reality, especially considering that you’ve stopped consciously perceiving your surroundings. You can well imagine the calamitous upshot of falsely enacting a dream fight, or a frantic sprint from an approaching dream foe, while your eyes are closed and you have no comprehension of the world around you. It wouldn’t take long before you quickly left the gene pool. The brain paralyzes the body so the mind can dream safely.

  How do we know these movement commands are actually occurring while someone dreams, beyond the individual simply waking up and telling you they were having a running dream or a fighting dream? The sad answer is that this paralysis mechanism can fail in some people, particularly later in life. Consequentially, they convert these dream-related motor impulses into real-world physical actions. As we shall read about in chapter 11, the repercussions can be tragic.

  Finally, and not to be left out of the descriptive REM-sleep picture, is the very reason for its name: corresponding rapid eye movements. Your eyes remain still in their sockets during deep NREM sleep.fn3 Yet electrodes that we place above and below your eyes tell a very different ocular story when you begin to drea
m: the very same story that Kleitman and Aserinsky unearthed in 1952 when observing infant sleep. During REM sleep, there are phases when your eyeballs will jag, with urgency, left-to-right, left-to-right, and so on. At first, scientists assumed that these rat-a-tat-tat eye movements corresponded to the tracking of visual experience in dreams. This is not true. Instead, the eye movements are intimately linked with the physiological creation of REM sleep, and reflect something even more extraordinary than the passive apprehension of moving objects within dream space. This phenomenon is chronicled in detail in chapter 9.

  Are we the only creatures that experience these varied stages of sleep? Do any other animals have REM sleep? Do they dream? Let us find out.

  Chapter 4

  Ape Beds, Dinosaurs, and Napping with Half a Brain

  Who Sleeps, How Do We Sleep, and How Much?

  WHO SLEEPS

  When did life start sleeping? Perhaps sleep emerged with the great apes? Maybe earlier, in reptiles or their aquatic antecedents, fish? Short of a time capsule, the best way to answer this question comes from studying sleep across different phyla of the animal kingdom, from the prehistoric to the evolutionarily recent. Investigations of this kind provide a powerful ability to peer far back in the historical record and estimate the moment when sleep first graced the planet. As the geneticist Theodosius Dobzhansky once said, “Nothing in biology makes sense except in light of evolution.” For sleep, the illuminating answer turned out to be far earlier than anyone anticipated, and far more profound in ramification.