Sleep and the Nervous System

What actually happens when you fall asleep

Sleep is a mode the brain constructs, with its own architecture, chemistry, and autonomic behavior. For most of human history the account was an absence: you go quiet, you stop responding, hours later you return. The instruments proved that picture wrong.

A nerve is a living wire. It is a bundle of fibers that carries messages through the body as tiny electrical pulses. Your brain is made of roughly eighty-six billion of these signaling cells, and they do not fire at random. They fire in rhythms, and those rhythms are loud enough to be heard from outside the skull.

The instrument that hears them is called an electroencephalogram, or EEG. Small metal discs are taped to the scalp and they pick up the summed electrical hum of the cells underneath. Think of it as a microphone held over a crowd.

When everyone in the crowd is talking at once about different things, the microphone records a small, fast, jagged signal. When the crowd starts chanting in unison, the microphone records a wave that is tall and slow. The height and the speed of the wave tell you how much of the crowd is doing the same thing at the same moment.

In the early 1950s, two physiologists at the University of Chicago pointed that microphone at sleeping people. Nathaniel Kleitman had founded the modern laboratory study of sleep, and Eugene Aserinsky was his graduate student. Aserinsky was chasing a small question. He wanted to know whether the eyes of a sleeping person move in any regular pattern, so he rigged a machine to record eye movement and brain waves together across an entire night.

What he found was that periods of rapid, darting eye movement arrived on a schedule, and that they came with a brain wave pattern that looked active rather than quiet. Their 1953 report of regularly occurring periods of eye motility during sleep split the night in half. The night held at least two different kinds of sleep.

Quiet sleep and active sleep

The night alternates between two states that share almost nothing except the name sleep.

The first is called NREM, non-rapid-eye-movement sleep. It runs on a gradient from light to deep. In light NREM the EEG wave is still smallish and fast, punctuated by two signatures worth knowing. Spindles are short bursts of fast rhythm that look like a spool of thread on the paper, and K-complexes are single tall spikes that often appear when a sound reaches a sleeping person. Both are the brain doing active work at the doorway of sleep, screening what gets through.

Go deeper and the crowd starts to chant. Deep NREM is called slow-wave sleep because millions of cells fall into unison and the recorded wave grows tall and slow. This is the state that is hardest to wake a person from, the state you feel cheated of after a broken night, and the state that carries most of what sleep is for physically.

The second state is REM, the one Aserinsky found. Here the EEG looks close to waking. The wave is small and fast, the eyes dart under closed lids, the breathing and heart rate become irregular, and dreaming is vivid and narrative. At the same time the muscles of the body are actively held still.

A circuit in the brainstem clamps down on the motor system so that the body does not act out what the brain is generating. Dreaming sleep is a brain running at nearly waking intensity inside a body that has been temporarily disconnected from it.

The sleeping brain keeps working. It changes jobs, twice, on a schedule.

Keep the pairing in mind. Deep NREM is a brain that has gone quiet and synchronous while the body is at its most settled. REM is a brain that has gone loud while the body has been immobilized. These are two different projects, and the night gives each of them its own hours.

Cycles, and why the last two hours are not like the first

The night runs in repeating cycles of roughly ninety minutes, each descending into NREM and rising into REM. William Dement, working in Kleitman's laboratory, mapped that structure; he later became a physician and founded clinical sleep medicine as a field.

He wanted to know how the brain's electrical pattern changes across an entire night, and whether those changes line up with eye movements, body movement, and what people report when woken. He published the answer with Kleitman in 1957 as cyclic variations in the sleep EEG. It is the shape every sleep clinic still uses: the descent into NREM and the rise into REM, repeated cycle after cycle until morning.

The cycles are not identical copies. The mix shifts as the night goes on. Deep slow-wave sleep is loaded into the first half of the night. REM expands across the second half, so the final cycles before waking are dominated by dreaming sleep with very little deep sleep left.

That asymmetry has practical consequences you have probably felt. Waking at five in the morning feels different from waking at midnight because you are being pulled out of a different state. Going to bed two hours late and waking at your usual time does not cost you an even slice of the night.

It costs you mostly REM, because REM was scheduled for the hours you removed. Drinking alcohol at bedtime does the reverse at the front end. A night is not a uniform substance you can trim from either end without consequence.

The shape also changes across a lifetime, and this matters enormously for how people judge their own sleep. A sleep epidemiologist named Maurice Ohayon wanted to establish what normal actually looks like at each age, so he pooled laboratory recordings of healthy people from childhood to old age and calculated the values.

His meta-analysis of quantitative sleep parameters across the lifespan found that deep slow-wave sleep declines steadily with age, while lighter sleep and the number of awakenings increase. A sixty-year-old's night is legitimately different from a twenty-year-old's. A changing night is not automatically a disordered one, and a great deal of unnecessary alarm comes from measuring an older night against a younger standard.

What slow-wave sleep is for

Three separate projects run in deep sleep, and rest is the wrong word for all of them.

The first is memory. Two German neuroscientists, Susanne Diekelmann and Jan Born, spent years testing what happens to newly learned material when a person sleeps versus stays awake for the same interval. Their review of the memory function of sleep assembled the evidence that slow-wave sleep supports the consolidation of recently acquired memories. The hippocampus replays the day's patterns and the cortex takes a more durable copy. Learning happens awake. Keeping it happens asleep.

The second is a kind of housekeeping on the learning itself. Giulio Tononi and Chiara Cirelli, sleep researchers at the University of Wisconsin, asked what the price of a day of learning actually is. Every waking hour strengthens connections between cells, and strengthening cannot continue indefinitely without saturating the system and consuming more and more energy.

Their synaptic homeostasis account of sleep and the price of plasticity proposes that slow-wave sleep scales those connections back down toward a sustainable baseline, keeping what mattered and shedding the rest. Deep sleep, on this reading, is what makes the next day of learning possible.

The third is hormonal. The endocrinologist Eve Van Cauter studied the timing of hormone release across the twenty-four-hour day. She found that the single largest pulse of growth hormone is bound to the first episode of slow-wave sleep rather than to the clock alone.

Her account of growth hormone secretion during sleep showed that the pulse tracks the deep sleep itself. In children growth hormone drives height. In adults it drives tissue repair and metabolic regulation. The body scheduled its largest repair signal for the deepest part of the night, and it reads the EEG to know when to send it.

What REM sleep is for

REM sleep keeps the information in an emotional memory and strips its charge, so the same memory carries less physiological reactivity the next day.

Matthew Walker, a neuroscientist studying sleep and emotion, and his colleague Els van der Helm proposed that the night does something specific to emotional memory. The idea was that sleep does not erase what happened. It keeps the information and strips the charge, so that the same memory carries less physiological reactivity the next day. They laid out the case in a review titled overnight therapy and the role of sleep in emotional brain processing.

Then they tested it directly. In their study of REM sleep and amygdala reactivity, healthy volunteers viewed emotional images twice, separated either by a night of sleep or by an equivalent stretch of waking. The researchers watched the amygdala, an almond-shaped structure deep in the brain that registers threat and drives the body's alarm response.

After a night of sleep the amygdala responded less strongly to the same images, and the size of the drop tracked a specific electrical signature during REM. Waking the same length of time did not produce it.

That is worth stating plainly, because it changes how you should read a bad week. When sleep is disturbed, the emotional load of the previous day is carried forward rather than metabolized. The next day begins with yesterday's charge still on the system. Anyone who has slept badly for a month knows this from the inside, and it has a physiological address.

Fluid washes the sleeping brain, and net clearance is contested

The brain has a disposal problem that the rest of the body does not. Everywhere else, a network of thin vessels called the lymphatic system drains the fluid that seeps out between cells and carries the waste away. The brain has no such vessels running through its tissue. Whatever the brain's cells discard has to leave by another route.

The glymphatic route, mapped in 2012

In 2012 a group led by Jeffrey Iliff and Maiken Nedergaard at the University of Rochester went looking for that route. They injected fluorescent tracers into the fluid around the brains of mice and watched where the tracer went. Their finding was published as a paravascular pathway that facilitates cerebrospinal fluid flow through the brain.

Fluid travels along the outside of the arteries that plunge into the tissue, passes through the brain, and exits along the veins. It carries dissolved waste with it. They named the system after the glial cells that line the channels.

The next question was whether that flow changes with state. Lulu Xie, working in the same laboratory, compared clearance in mice that were awake, naturally asleep, and anesthetized. The 2013 paper, sleep drives metabolite clearance from the adult brain, reported two results.

The space between brain cells expanded by roughly sixty percent during sleep. Injected amyloid beta, a protein associated with Alzheimer's disease, was cleared about twice as fast in the sleeping brain. That result travelled fast and far, and it is the source of nearly every claim you have read about sleep washing the brain.

There is real human work alongside it. Nina Fultz and colleagues at Boston University recorded EEG, blood flow, and cerebrospinal fluid movement simultaneously in sleeping people, and reported coupled electrophysiological, hemodynamic, and cerebrospinal fluid oscillations in human sleep. Large pulses of fluid washed into the brain in time with the slow waves of deep NREM. The rhythm of deep sleep and the movement of fluid are locked together in humans, and you can watch it happen.

The 2024 challenge to faster clearance

In 2024 a group at Imperial College London led by Nicholas Franks and William Wisden set out to reproduce the clearance result with a different method of measuring where dye actually goes. Their paper, brain clearance is reduced during sleep and anesthesia, reported the opposite direction. Clearance in their mice was markedly lower in sleep and under anesthesia, not higher.

Hold both. What is established is that fluid moves through brain tissue along vessels, and that this movement is coupled to the rhythms of deep sleep in humans. The field disagrees about whether sleep accelerates net clearance. What is not established is that a poor night causes dementia.

That sentence appears constantly in popular coverage and the evidence does not support it. The tone model does not need the strong version of the claim. What matters for tone is the coupling itself: fluid movement locked to the slow waves of deep sleep is one more voice moving with the night.

The two forces that decide when you sleep

Sleep is scheduled by two independent systems. Almost every ordinary sleep complaint is one of them being asked to do the other's job.

Adenosine, the pressure of being awake

The first is pressure. The longer you are awake, the more you need to sleep, and that need has a molecule. Tarja Porkka-Heiskanen is a Finnish researcher studying brain chemistry during prolonged wakefulness. She sampled the fluid in the basal forebrain of animals kept awake for long periods, looking for a substance that rose with time awake and fell with recovery.

Her 1997 paper in Science identified adenosine as a mediator of the sleep-inducing effects of prolonged wakefulness. Adenosine accumulates while you are awake, presses on the sleep-promoting machinery, and is cleared during sleep. Caffeine works by sitting in the receptor adenosine would otherwise occupy. It does not add energy. It hides the pressure.

The suprachiasmatic clock and its light sensor

The second force is the clock, and it is a physical object. Above the point where the optic nerves cross behind the eyes sits a cluster of cells about the size of a grain of rice called the suprachiasmatic nucleus. It keeps time on its own. The physician-scientist Charles Czeisler wanted to know the true intrinsic period of the human clock, because earlier work had put it near twenty-five hours.

He used a protocol called forced desynchrony, in which volunteers live on a day length so unusual that the clock cannot lock onto it, which forces the clock to reveal its own rhythm. His 1999 measurement of the stability, precision, and near-24-hour period of the human circadian pacemaker put it at about 24.18 hours, remarkably stable, and essentially the same in older and younger adults.

A clock running at 24.18 hours drifts by roughly eleven minutes a day, so it has to be reset daily. Light does the resetting, and it does so through a sensor most people have never heard of. David Berson, a neuroscientist studying the retina, was chasing a puzzle.

Animals whose rods and cones were not functioning still set their clocks by light, so something else had to be catching it. His 2002 paper reported phototransduction by retinal ganglion cells that set the circadian clock. A distinct class of cells in the retina is itself light sensitive and wires directly to the clock. Your eye contains a timekeeping light sensor that is separate from the machinery of seeing.

What the clock does with that light is send a chemical timing signal, and here a common misunderstanding needs correcting. The psychiatrist Alfred Lewy tested whether human melatonin is suppressed by light the way it is in other mammals, since earlier studies had failed to show it.

In 1980 he showed that light suppresses melatonin secretion in humans, with bright light shutting it down while ordinary indoor light did not. The intensity of the light, and not merely its presence, is the signal the clock reads. Melatonin itself is a message about what time it is. It is not a sedative, and it does not knock the body out.

The two-process model: pressure against clock

In 1982 the Swiss sleep researcher Alexander Borbély put the two forces into one rule. Looking for a simple model that could explain both when a person falls asleep and how deeply they sleep, he proposed a two process model of sleep regulation. A homeostatic pressure builds with every waking hour and discharges in deep sleep. A circadian signal rides underneath it, indifferent to how long you have been awake.

Sleep comes when pressure is high and the clock permits it. This is why you can be exhausted at four in the afternoon and wide awake at eleven at night. It is also why an irregular schedule wrecks sleep even when the total hours look fine on paper. You can have all the pressure in the world and be sitting at the wrong point on the clock.

How the brain turns itself off, and how that can fail

Pressure and clock decide when. Something still has to do the turning off, and that something is a physical switch in the hypothalamus.

For most of the twentieth century, arousal research looked for the cells that hold the cortex awake, and found several systems in the brainstem and hypothalamus that do exactly that. In 1996 Jonathan Sherin and Clifford Saper, neuroscientists studying the hypothalamus, went the other way.

They looked for cells that switch on during sleep rather than during waking. Their paper reported activation of ventrolateral preoptic neurons during sleep. A small cluster near the front of the hypothalamus fires while an animal sleeps, and it inhibits the arousal systems directly.

Saper then assembled the anatomy into a model of the whole mechanism. In the sleep switch and hypothalamic control of sleep and wakefulness, he described the arrangement as a flip-flop. The arousal side inhibits the sleep side, the sleep side inhibits the arousal side, and mutual inhibition of that kind has a specific behavior. It does not sit halfway. It commits to one state, holds it, and then flips.

That design explains something you already know from experience. Falling asleep is abrupt. You do not fade through a gradient of half-consciousness; you are awake, and then there is a gap. The switch is built for clean transitions, and clean transitions are what a body needs when it is about to give up its vigilance for eight hours.

A switch built this way is fast and decisive. It can also get stuck, and a stuck switch is a completely different problem from a tired body.

Chronic insomnia is what a switch stuck on the awake side looks like, measured across seven instruments at once.

What the heart, the vessels, and the pressure do while you sleep

The nervous system runs two lines to the heart and vessels that pull in opposite directions. The sympathetic system is the accelerator; it speeds the heart, tightens the vessels, and raises pressure. The vagal or parasympathetic system is the brake; it slows the heart and lets pressure fall. Almost every number a doctor measures on you is the running balance between those two.

Virend Somers, a cardiovascular physician-scientist, wanted to know what the accelerator actually does across the stages of sleep, so he listened to it directly. The technique is called microneurography. A fine electrode is placed into a nerve in the leg, close enough to individual fibers to hear the traffic heading out to the blood vessels.

In 1993 he reported sympathetic nerve activity during sleep in normal subjects, and the pattern is the shape of the healthy night. Sympathetic traffic falls progressively as NREM deepens, and blood pressure and heart rate fall with it. Then REM arrives and traffic surges above waking levels, with pressure spiking toward daytime values.

The night therefore holds no steady low autonomic setting. It runs a large, structured excursion, down further than waking ever goes and then back up past it, several times a night, on schedule.

Dipping: the nightly blood pressure fall

The most useful clinical trace of that excursion is called dipping. In a well-regulated person, blood pressure falls by roughly ten to twenty percent overnight compared with daytime values. Losing that fall is a known warning sign, and it can be measured with a cuff that inflates automatically through the night. Takayoshi Ohkubo and colleagues followed more than fifteen hundred residents of Ohasama, Japan, for an average of nine years.

They wanted to know whether the size of the nighttime fall carried prognostic weight of its own. Their study of the prognostic significance of the nocturnal decline in blood pressure found a linear relationship. Each five percent less of a nighttime fall was associated with roughly twenty percent greater cardiovascular mortality, and the association held even in people whose twenty-four-hour average pressure was entirely normal.

Read that last clause twice. A person can have a normal daytime number and still be carrying risk, because what predicted the outcome was the loss of the nightly excursion. The health signal was in the range, not in the average.

The people who cannot sleep are not the sleepiest people

About one adult in ten meets the criteria for chronic insomnia disorder, and roughly a third report symptoms at some point. Charles Morin and colleagues set out the clinical picture in the Nature Reviews primer on insomnia disorder. The condition has a page of its own in this library. Its literature answers a question about sleep itself: whether sleep is a quantity the body runs short of or a state the body fails to enter.

If chronic insomnia were a shortage of sleep, one prediction follows immediately. Give these patients a chance to nap in the daytime and they should drop off faster than everyone else. Sleep-deprived people fall asleep in seconds.

The nap test insomnia patients fail in reverse

Edward Stepanski and colleagues at Henry Ford Hospital ran that test. The tool is called the multiple sleep latency test. A person is placed in a dark, quiet room four or five times across a day and asked to let themselves fall asleep, while the EEG records how many minutes it takes.

This is the standard clinical measure of how sleepy someone actually is. In their comparison of daytime alertness in chronic insomnia patients and asymptomatic controls, the insomnia patients took longer to fall asleep than the controls. Not shorter. Longer.

That is the opposite of sleep deprivation. A genuinely sleep-deprived person cannot stay awake in a dark room. These patients could not fall asleep in one, at the hour of day when their complaint says they should have been most desperate for it. That inversion is where the paradox starts.

Exhaustion and sleepiness come apart. A body can be worn through and still hold too much arousal to sleep.

Seven instruments, one elevated setting

Measure the same patients around the clock and the reason surfaces. Michael Bonnet and Donna Arand, sleep researchers at the Dayton Veterans Affairs Medical Center, recorded whole-body metabolic rate continuously across a full day and night. Ten insomnia patients were compared with ten matched normal sleepers. Their study of 24-hour metabolic rate in insomniacs and matched normal sleepers found the insomnia group elevated at essentially every time point. The whole organism was running warmer, day and night.

Eric Nofzinger and colleagues at Pittsburgh reached for positron emission tomography, a scan that tracks how much glucose brain tissue consumes. They compared seven insomnia patients with twenty healthy sleepers, both awake and in NREM sleep. Their report of functional neuroimaging evidence for hyperarousal in insomnia found greater global glucose metabolism in the insomnia group during both. It also found a smaller decline into sleep in the very regions that promote arousal.

Michael Perlis and colleagues went back to the EEG and asked what the fast frequencies were doing during sleep, rather than the slow ones everyone had been scoring. Their study of beta and gamma EEG activity in primary and secondary insomnia and good sleeper controls compared three groups of nine.

The primary insomnia group showed more high-frequency activity across NREM sleep than the good sleepers or the group whose insomnia came with depression. Fast activity is the signature of a crowd talking rather than chanting. Waking-type cortical activity sits inside hours that standard staging scores as sleep.

Alexandros Vgontzas and colleagues at Penn State sampled ACTH and cortisol, the pituitary and adrenal stress hormones, repeatedly across a full twenty-four hours. Eleven chronic insomnia patients were compared with thirteen controls. They found nyctohemeral activation of the hypothalamic-pituitary-adrenal axis in chronic insomnia, meaning elevation across both the day and the night. Twenty-four-hour ACTH was significantly higher in the insomnia group.

Cortisol ran higher too, and at that sample size the difference fell short of the conventional threshold. The elevation was greatest in the evening and the first half of the night, and cortisol ran highest in those with the most objective sleep disturbance. The stress axis was already running high before these patients got into bed.

The blunted dip arrives before the diagnosis

The next two measurements read the same nightly excursion that dipping measures. Paola Lanfranchi and colleagues monitored overnight blood pressure in thirteen people with chronic insomnia who were not hypertensive, alongside thirteen matched good sleepers. Their study of nighttime blood pressure in normotensive subjects with chronic insomnia found the nocturnal systolic dip roughly halved, eight percent against fifteen. The nightly fall whose loss predicts cardiovascular death is blunted here before any hypertension diagnosis exists.

Vgontzas then examined more than seventeen hundred people from a general population sample with overnight recordings. He asked whether insomnia tracked hypertension regardless of how much these people actually slept. His finding on insomnia with objective short sleep duration and the risk for hypertension was that it did not.

The association concentrated in those whose recordings showed the least sleep, and the highest odds fell on people under five hours. Insomnia with normal recorded sleep duration carried far less. The study was cross-sectional, so it establishes association rather than prospective risk. Even so, the pattern is the phenotype speaking: the cardiovascular signal concentrates where the recording, not the complaint, shows the least sleep.

Dieter Riemann and colleagues in Freiburg gathered these strands into a formal position. Their review of the hyperarousal model of insomnia argued that the disorder is elevated arousal spanning the physiological, cortical and cognitive levels, present around the clock rather than only at night. That is now the dominant framework in the field.

The sample sizes stay small: most of these results rest on a dozen or two people, and none of them would carry an argument alone. They were collected by different groups, in different decades, with methods that share nothing but the patients, and they converge.

Now read that back as a statement about sleep. A body can be worn out and not sleepy. A brain can spend eight scored hours asleep with waking-type activity running through them. The recording can say sleep while every other instrument in the room says alert. That is the paradox these findings produce. Sleep is a state the whole system enters together rather than a quantity a body accumulates by lying still, and the entering can fail while the hours accrue.

The insomnia page carries the condition forward. It sets out the three kinds of factor that turn a bad fortnight into a bad decade, and the counterintuitive treatment they point to. It weighs the evidence for and against sleeping tablets, and it explains why the night a patient lives and the night on the machine so often disagree.

Regularity and range predict health better than hours

Almost every public message about sleep is a target number. The mortality evidence points at timing and range instead: a U-shaped duration curve, regularity beating quantity, and the nightly dip carrying its own risk.

Francesco Cappuccio and colleagues at Warwick pooled prospective studies following large populations over years, to see how self-reported sleep duration related to death from any cause. Their systematic review of sleep duration and all-cause mortality found risk elevated at both ends. Short sleepers carried higher mortality, and so did long sleepers, with the long end showing the stronger association. The relationship is a U, not a slope, which is what you expect from a regulated variable rather than a resource.

The sharper result came later, and it came from watches rather than questionnaires. Daniel Windred and colleagues analysed accelerometer recordings from around sixty thousand adults in the UK Biobank. They computed how long each person slept, and also how consistent that person's sleep timing was. Their finding was that sleep regularity is a stronger predictor of mortality risk than sleep duration. Going to bed and waking at the same times mattered more than the total.

Put the three results in a row. Both too little and too much predict harm. Consistency of timing predicts more strongly than quantity. And the loss of the nightly blood pressure excursion predicts cardiovascular death even when the twenty-four-hour average is normal.

Every one of those is a statement about range and rhythm rather than about amount. That is the shape of a regulated system being read correctly.

Sleep read through the Unified Model of Tone

The Unified Model of Tone reads the night as one variable moving through its full range. The claim is the model's own, stated on its own authority rather than as a summary of any one study cited here.

Tone is the coupled organization the nervous system holds across the whole body. The word does not mean muscle tension here, and it does not mean a mood. More precisely, tone is the integrated organization of the body's interacting state transitions, read at once across the cortex, the brainstem, the heart, the vessels, the hormones and the tissues.

Its defining property is range. A healthy system can move that organization a long way in either direction and return. Health is the width of that regulated range. Illness is its collapse toward a stuck value.

Sleep, read this way, is the largest single excursion of tone the body performs. Every night the system has to leave the vigilant, outward-facing configuration it holds all day. It travels down into a state where the cortex chants in unison, the accelerator eases, and pressure falls by a tenth or more.

It holds that for a while. Then it surges back up past waking levels in REM, and climbs out on schedule in the morning. The night is the daily demonstration that the range still exists.

A sleep stage is a chord, not a note

That is why the night cannot be read as one measurement. A sleep stage is a chord, not a note. Sounding it requires the cortical rhythm, the sleep-promoting cluster in the hypothalamus, and the withdrawal of sympathetic traffic. It also requires the vagal brake, the fall in core temperature, the hormonal pulse, and the movement of fluid through the tissue. All of them are held in step.

None of those rhythms is tone by itself. Oscillation is the medium tone is written in, and the organization of the whole set is what the night keeps or loses. These are coupled voices carrying the same organization in different mediums: electrical, chemical, thermal, mechanical, and fluid. What a polysomnogram scores as stage three is the sound of those voices agreeing.

Chronic insomnia is a system that cannot leave the alert state. Every around-the-clock measurement is a different voice reporting the same stuck chord.

Look again at what those measurements are. Elevated metabolic rate is the whole organism's tone. Elevated brain glucose use is the cortex's tone. Fast EEG inside sleep is the cortical rhythm refusing to synchronize. A stress axis running high across the day and the night is the endocrine tone.

The blunted dip is the cardiovascular tone. Not one of them is a lesion. Every one of them is the same setting expressed in a different medium. That is exactly what the model predicts when the organization itself is stuck rather than a part being broken.

This also answers the question that hangs over every insomnia clinic. Why is there no lesion to find? Because nothing is broken. The switch works, the clock works, the adenosine builds, the machinery is intact.

What has changed is the setting the whole system is holding, and a setting leaves no shadow on a scan. That is why a chronic insomnia patient's overnight study so often reads close to normal, and why being told the study was normal is one of the most demoralizing sentences these patients ever hear.

The model also explains why the same precipitant produces such different outcomes. No input acts on an empty body. A bereavement, a newborn, or a stretch of night shifts meets a nervous system that already has a tone, and the same event becomes a different event depending on what it lands on.

In one person the arousal spikes and settles within weeks. In another, whose baseline range was already narrow, the arousal is absorbed into the standing setting and stays. What the insomnia literature calls a predisposing factor and what this model calls a narrow baseline range are one variable named twice.

Causation, unification, and the test the model sets itself

Which direction does causation run? A bad night could raise cortisol as easily as high cortisol could ruin a night. In a coupled system the question is malformed. There is no upstream part issuing orders to downstream parts. The cortex, the autonomic outflow, and the endocrine axis are voices tuned to each other, each reshaping the others, and asking which one caused the chord is like asking which instrument caused the key.

What the word tone adds to six existing measurements is unification. The same variable that appears here as hyperarousal appears in the cardiovascular literature as sympathetic overdrive, in chronic pain as central sensitization, and in autonomic medicine as reduced variability.

Each field measures it in its own units and treats it as its own local phenomenon. The model's claim is that these are one variable read at different sites. That is why the same person so often carries several of them at once, and why an intervention aimed at one moves the others.

One result separates an input that restores that shared organization from one that pushes a single output: a genuine restoration of tone moves dysregulated values toward the healthy middle from either side.

The treatment that resolves insomnia barely adds sleep

Two things can be done to a bad night. You can add sleep to it, or you can restore the system's ability to produce sleep. The insomnia literature holds the cleanest natural experiment in medicine for telling those apart.

The first-line treatment for chronic insomnia is behavioral: a structured programme called cognitive behavioral therapy for insomnia, usually shortened to CBT-I. What the programme contains, how the guideline panels graded it against medication, and what the sleeping tablets are measured to do and to cost are all set out on the insomnia page. Three of its results say more about sleep itself than about treatment.

CBT-I resolves the disorder while adding eight minutes

Start with the result that should stop you. James Trauer and colleagues in Melbourne pooled twenty randomized trials of face-to-face CBT-I to quantify what the therapy actually changes on the clock. In their systematic review and meta-analysis of CBT for chronic insomnia, time to fall asleep improved by about nineteen minutes.

Time awake after falling asleep improved by about twenty-six minutes, and sleep efficiency improved by about ten percentage points. Total sleep time improved by about eight minutes, and that estimate carried a confidence interval crossing zero. Crossing zero means the result includes the possibility of no gain at all.

Read that pairing carefully. It is the insomnia paradox restated in the units of treatment. The therapy reorganizes the night. It compresses the fragmented, effortful, half-aroused hours into a consolidated block, and it lowers the arousal that was producing the fragmentation. The patient recovers, and the total is almost unchanged. If insomnia were a deficit of sleep quantity, this result would be impossible.

The treatment that resolves chronic insomnia adds roughly eight minutes of sleep. It changes the shape of the night rather than the size of it.

What sedatives do to the recording

Sedatives change the shape of a night too, and that has been visible since the earliest spectral studies. Borbély and colleagues gave healthy volunteers single bedtime doses of benzodiazepine hypnotics and analysed the whole night's EEG frequency by frequency. Splitting a recording that way shows which rhythms a drug suppressed and which it amplified, rather than only how long each stage lasted.

Their study of the effect of benzodiazepine hypnotics on all-night sleep EEG spectra found reduced low-frequency activity, which is the slow waves of deep sleep, and increased activity in the spindle frequency range. Some of the changes persisted into the following drug-free night. The recording shows a night that has been chemically reshaped.

The head-to-head comparison closes the argument. Børge Sivertsen and colleagues in Bergen randomized older adults with chronic insomnia to CBT, to the sedative zopiclone, or to placebo, with overnight recordings at six weeks and six months.

In their randomized controlled trial of cognitive behavioral therapy versus zopiclone, the behavioral treatment beat the drug on three of four outcomes at both time points. Zopiclone did not differ from placebo on most measures. The behavioral group spent substantially more time in slow-wave sleep. Total sleep time was similar in all three groups.

Relief has real value, and a hypnotic prescribed by a physician who knows the patient is a legitimate clinical option. The guideline grades sit on the insomnia page, and a weak grade there describes the certainty of the evidence rather than proof that a drug fails. Stopping one abruptly can bring the insomnia back harder than it arrived.

This is the restore-versus-mask distinction with recordings attached, and it is drawn by aim rather than by instrument. A sedative adds a chemical to one node of the system, moves the output, and alters the architecture of the night while it does so. It manages the night. A treatment that lowers arousal restores the system's own capacity to make one, and the recording changes shape even though the total does not move.

Either aim can be pursued with either tool. A behavioral programme handed over as a rulebook becomes one more score to chase, and in a disorder of arousal that feeds the thing it was meant to lower. A hypnotic can be aimed the other way, at returning the reserve a system needs before it can reorganize at all. The line runs across the instruments, not between them.

What a restoring input does in sleep that a sedative cannot

The model makes a claim a drug cannot make, and it can be checked. A genuine restoration of tone should move a dysregulated value toward the healthy middle from either side. In sleep that means three specific things. The same intervention should extend the night of the person who sleeps too little, and shorten the night of the person who sleeps excessively.

It should raise a blunted blood pressure dip toward the normal ten to twenty percent, without driving it into the extreme dipping that carries its own risk. And it should lengthen a pathologically short daytime nap latency while shortening a pathologically long one. Convergence toward the middle from opposite starting points is the signature. A sedative cannot do this. A sedative pushes one direction on everyone who takes it, which is precisely why the same dose that helps one patient flattens another.

The trial is ordinary. Phenotype a group of patients before treatment, split them by whether each measure starts high or low, treat them all the same way, and look at where they land. Convergence toward the middle with a narrowing spread marks a restoration. A uniform shift in one direction marks a mask, which helps whichever group it happens to point at and carries the other group further from the middle.

That study has not been done, and the CBT-I results do not substitute for it. The meta-analysis reports pooled averages for the whole treated group. It never splits patients by whether a given measure started high or low, and a uniform shift in the group mean is precisely the outcome this prediction rules out as evidence. The convergence claim is open. It is the model's strongest prediction about sleep and its least tested.

Trackers, melatonin, hygiene, and the causes that must be found

Six boundaries keep the tone reading of sleep inside what measurement supports: wearables, melatonin, sleep hygiene, heart rate variability, findable medical causes, and which numbers carry information about range.

What a wrist tracker can and cannot see

The first concerns wearables, because most people now arrive at this subject holding a number from a wrist. Massimiliano de Zambotti and colleagues reviewed the validation evidence for consumer sleep trackers against laboratory polysomnography in their assessment of wearable sleep technology in clinical and research settings.

These devices are reasonably good at telling sleep from wake in aggregate and at tracking night-to-night patterns, and poor at classifying stages. Your ring does not know how much deep sleep you had. It is inferring stages from movement and heart rate with limited accuracy, and the trend line over weeks is far more meaningful than any single night's stage breakdown.

There is a second-order problem with the number as well. Kelly Baron and colleagues described a pattern they named orthosomnia in their report on patients taking the quantified self too far. Patients arrived distressed by tracker data, and the pursuit of a perfect score was itself feeding the arousal that was wrecking their sleep. In a disorder of hyperarousal, a device that invites nightly self-scrutiny is not a neutral instrument.

Melatonin is a timing signal, and hygiene is not treatment

The second correction concerns melatonin. Because the clock uses it as a timing signal, it works best as a timing tool, taken in small doses and at the right hour to shift the clock. Its effect on ordinary insomnia is modest.

Eduardo Ferracioli-Oda and colleagues pooled nineteen randomized trials in their meta-analysis of melatonin for primary sleep disorders. They found a reduction in time to fall asleep of roughly seven minutes, and a small increase in total sleep. It is a real effect and a small one, and it is not a sedative.

The third concerns sleep hygiene. The standard advice list, meaning a dark cool room, no screens, no late caffeine, and a regular schedule, is sensible and it is not a treatment for chronic insomnia. Leah Irish and colleagues examined the empirical basis for each recommendation in the general population.

Their review of the role of sleep hygiene in promoting public health reported the individual items broadly supported as influences on nocturnal sleep. Public health is a different question from treating a disorder, and the insomnia page sets out what hygiene achieves when it is tested on its own as a treatment. Being told to try sleep hygiene when you have had chronic insomnia for four years is not treatment.

Heart rate variability has not confirmed the reading

The fourth is a limit on this model's own measurement anchor. Heart rate variability, the beat-to-beat variation in the timing of the heartbeat, is a validated index of autonomic state. Reading it as a window on tone is this model's interpretation, and the two should never be blurred. The insomnia literature has been unkind to the assumption. Kirsty Dodds and colleagues in Sydney reviewed every study of heart rate variability in insomnia patients.

They found the results too inconsistent to establish any reliable impairment. The insomnia page carries that null in full. The model reads it as expected, since a distortion may sit more in the cortical or the endocrine voice than in the cardiac one. The cardiac measure does not currently support a simple claim about insomnia, and the model makes none here.

Red flags: the causes that must be found

The fifth boundary outranks the other five: findable causes must still be found. A sleep complaint can be the visible edge of obstructive sleep apnea, in which the airway closes repeatedly through the night. The American Academy of Sleep Medicine's guideline for diagnostic testing for adult obstructive sleep apnea addresses it directly, because it is common, treatable, and carries serious cardiovascular consequences when missed.

It can also be restless legs syndrome, thyroid disease, uncontrolled pain, or a medication side effect. It can be a circadian rhythm disorder such as delayed sleep phase, alcohol, or depression, which travels with insomnia in both directions.

Loud snoring with witnessed pauses, morning headaches, severe daytime sleepiness, or a sudden change in sleep with no obvious trigger all deserve a physician's attention rather than a framework. Not every sleep disorder reduces to poor tone, and the model does not claim otherwise. A closed airway, a thyroid, a drug, and a diagnosable circadian disorder are real causes, and each is treated on its own terms.

The claim the model actually makes is narrower. Every one of those conditions has a tonal expression, and many sleep problems are initiated, maintained, or amplified by failures of tonal regulation. A page about regulation must never delay a diagnosis, and reading everything as tone would be exactly the error this model is meant to correct.

The sixth is what remains worth measuring. The numbers worth tracking are the ones that carry information about range. The size of the nighttime blood pressure fall is one. The regularity of sleep timing across weeks is another. So is how long it takes to fall asleep during the day when given the chance, and how far the brain's arousal systems power down between waking and sleep.

The last is the direction of change when the same intervention is given to people who start on opposite sides of a value. Those are the numbers that describe a system's ability to move and return. The number of hours describes almost nothing by comparison.

Health is a wide range that the system can travel and come home from. The night is where that range is exercised in full. Chronic insomnia is what it looks like when the system can no longer make the trip.

How sleep relates to the rest of the library

Sleep exercises every foundation of tone in a single night, and two disorders of the night have pages of their own. Each neighboring page carries one piece of this one.

Two disorders of the night border this page. Insomnia is the stuck switch in full: the factors that install it, the treatment evidence, and the tablets. Sleep apnea is the structural neighbor, an airway that closes repeatedly and shreds the architecture this night is built from, with cardiovascular consequences when missed.

Four more pages read the same organization at other hours.

• Anxiety is the aroused configuration of insomnia experienced in daylight.
• Mental health holds the two-way traffic between disturbed nights and disturbed mood, including REM's overnight work on emotional memory.
• Heart rate variability is the instrument that reads the autonomic night beat by beat, and the one that has not confirmed hyperarousal in insomnia.
• Blood pressure is where dipping and its loss are read in full.

Two final pages hold the machinery. The vagus nerve is the brake whose engagement deepens the night's descent. Dysautonomia is what the nightly excursion looks like when autonomic regulation itself is the presenting problem.

Frequently asked