Central Sensitization and How the Nervous System Sets Its Own Volume

Gain is the volume setting between input and output

A public address system shows every property of gain that matters. A microphone stands at the front of a small hall. It picks up sound. That is the input.

The microphone feeds a box called an amplifier. The amplifier has one job. It decides how much of what arrives comes back out. That decision has a name, and the name is gain. The speaker on the wall is the output, the part you actually hear.

Now turn the gain up. At first the room simply sounds louder. Turn it further and something else happens. A cough at the back of the hall comes out as a bang. The scrape of a chair comes out as a crash. Turn it further still and the system starts to scream at itself, a rising squeal that feeds its own input and will not stop.

Nothing in the room has changed. The cough is the same cough. The chair is the same chair. Every disturbance that produced a scream would have produced a rustle an hour earlier. Only one thing moved, and it was the setting.

Two features of that scene matter for the rest of this page. The first is that the setting is invisible. Photograph the room and you will not find the squeal. Examine the air and you will not find it. The squeal lives in a relationship between what went in and what came out, and a relationship cannot be photographed.

The second is that the fault runs both ways. A gain set too low is also a failure. A smoke alarm with a dying battery is not a peaceful house. It is a house that will not be told it is burning. The alarm has to be loud enough to matter and quiet enough to live with, and it has to move between those states as conditions change.

Your nervous system runs this arrangement everywhere, in every sense, in every reflex, in every organ. Something arrives. Something decides how much of it comes through. Something comes out. This page is about the middle step.

What gain is, and what the field already calls it

Central sensitization is the established term for an increase in the responsiveness of pain-signaling neurons in the central nervous system, meaning the spinal cord and the brain. Gain is the Unified Model of Tone's word for the setting behind it: how loudly the nervous system answers what arrives. The established term comes first because it is the one written in charts and studies.

The neuroscientist Clifford Woolf, who first demonstrated central sensitization, later set out its clinical implications in plain terms. Pain can be generated by a change in the state of the nervous system, with no ongoing damage to explain it. A detailed account of the underlying cellular changes was assembled by Alban Latremoliere and Woolf, who described central sensitization as a generator of pain hypersensitivity by central neural plasticity.

Gain is a ratio, and a test measures a value

Gain is how loudly a system responds relative to what arrived. Written as arithmetic it is output divided by input. That makes gain a ratio rather than a quantity, and the difference is the whole point. You cannot measure gain by measuring an input. You cannot measure it by measuring an output. You have to change the input and watch what the output does.

This is why gain hides so well in clinical medicine. A test measures a quantity. Gain is a slope. A blood panel, an image, a nerve conduction study, each returns a value from a single moment. None of them returns a relationship between two moments.

Gain is the setting, not the input and not the output. It is the only one of the three that never appears in a picture.

Four fields measure gain under four names

The word is not an import. When Woolf and the neuroscientist Michael Salter reviewed the field for Science in 2000, they titled the paper Neuronal plasticity: increasing the gain in pain. They sorted the plasticity into activation, modulation, and modification, and described each as a way of producing hypersensitivity by increasing gain. Pain researchers say gain. Cardiologists say baroreflex sensitivity and report it as a slope. Vestibular scientists say vestibulo-ocular reflex gain and report it as a ratio. Auditory scientists say central gain.

The Unified Model of Tone makes a claim about that coincidence, and the claim is the model's own. Those four fields are measuring one setting of the body's state at four sites with four instruments, and nobody has said so out loud. Gain is that setting, named once so it can be used everywhere. The model contributes the unification, and with it the fact that a single condition can be described as a few such settings in combination rather than as an isolated disease.

The warning signal is not the pain

The warning signal and the pain are two different things, and gain lives between them. Central sensitization stays invisible until that distinction is in place, and almost nobody is taught it.

A nerve is a living wire, a bundle of fibers that carries messages as tiny electrical pulses. Scattered through your skin, muscle, joints, and organs are specialised nerve endings called nociceptors. They fire when tissue is being stressed, stretched too far, burned, cut, or starved of blood. Their traffic is called nociception, and it is a warning line reporting a condition.

Pain is something else. Pain is what the brain produces after weighing that traffic against everything else it knows. Where you are. What you are doing. What happened last time. How safe you are, how tired you are, and what the signal means.

The International Association for the Study of Pain revised its official definition in 2020, and the revision, described by the pain physician Srinivasa Raja and an international task force, states the point without ambiguity. Pain is always a personal experience, and activity in sensory neurons is not itself pain.

Tissue damage and pain do not track each other

The cleanest demonstration is eighty years old. Henry Beecher was an anaesthesiologist attached to American forces in Italy during the Second World War, and at the Anzio beachhead he did something nobody had thought to do. He asked badly wounded men, as they came in, how much pain they were in and whether they wanted morphine.

He published what he found in a 1946 report on pain in men wounded in battle. A large share of the severely wounded reported slight pain or none, and did not ask for morphine when it was offered. The tissue damage was enormous. The output was small.

The same gap shows up in modern imaging, from the other direction. A team of radiologists led by Waleed Brinjikji at the Mayo Clinic pooled scans from 3,110 people who had no back pain at all, and reported what a pain-free spine actually looks like.

Disc degeneration was present in 37 percent of twenty-year-olds and 96 percent of eighty-year-olds. Disc bulges were present in 30 percent at twenty and 84 percent at eighty. These are the findings that get circled on a report and blamed for a symptom, and they are the ordinary condition of people who feel nothing.

None of this makes tissue irrelevant. A fracture is real, an infection is real, a tumour is real, and finding them is urgent and sometimes life-saving. The evidence supports something narrower and more useful. The state of the tissue is one input to a system that decides the output, and the decision has a setting. That setting is the gain.

Central sensitization begins in the spinal cord

The first amplifier in the pain pathway sits inside the spinal cord, and central sensitization was discovered there. Most people picture the warning line as a wire running from the sore place to the brain. It is not one wire.

The fibers coming in from your body stop in the back of the spinal cord, in a region called the dorsal horn, and hand their message to a second cell. That handover happens at a synapse, which is a narrow gap where the first cell releases a chemical and the second cell decides whether to fire. Deciding is the operative word. A synapse is not a solder joint. It is adjustable, and it can be adjusted in minutes.

Wind-up: the cord climbs on a steady input

An adjustable synapse means an amplifier before anything reaches the brain at all. The claim was not obvious, and two experiments established it.

In 1965 the neurophysiologists Lorne Mendell and Patrick Wall were recording from single cells in the back of the cat spinal cord. They were asking a technical question. What do the slow, thin, unmyelinated fibers from the body actually do to those cells? They delivered identical electrical stimuli at a steady rate and watched the response.

The responses grew progressively larger across the series. Nothing about the stimulus changed. The cord was climbing. The phenomenon was named wind-up, and it is the first published picture of a spinal cord raising its own volume on a repeated signal.

Eighteen years later Clifford Woolf, then a physiologist studying injury, took on the question that wind-up made unavoidable. After an injury, the area around it becomes tender.

Does that tenderness come from the skin's own receptors becoming touchier, or from the nervous system becoming more excitable? He built an animal model of the flexion withdrawal reflex, the automatic pull-away from a noxious touch, and measured its threshold and its responsiveness before and after a peripheral injury. He reported in Nature that the increase in excitability arose in part from changes in the spinal cord itself. The injury had altered the amplifier.

Hyperalgesia, allodynia, and spread beyond the injury

Three words follow from that, and a reader who holds them can read almost any pain report.

Hyperalgesia means something that should hurt a little hurts a lot. Allodynia means something that should not hurt at all does hurt: a bedsheet on the shin, a shirt collar, a breeze across the face, a hand on the shoulder. Secondary hyperalgesia means the tenderness has spread into tissue that was never injured.

That third one is the tell, and it deserves a moment. Uninjured skin has no injured receptors in it. If the skin two inches from a burn becomes tender, no local explanation is available. The amplification has to be happening in the cells that receive from both places, which is to say centrally. This is the everyday clinical fingerprint of central sensitization, a change in gain that can be checked with a cotton swab.

The nervous system has a brake, so high gain has two causes

An amplifier climbing is only half of central sensitization. The nervous system also has a brake, a real anatomical system with its own cells and its own chemistry, and gain rises when either half fails.

The idea arrived in the same year as wind-up. The psychologist Ronald Melzack and the neurophysiologist Patrick Wall were trying to explain something ordinary: why rubbing a banged shin helps. Their answer, published in Science as a new theory of pain mechanisms, proposed that the dorsal horn contains a gate, and that other incoming signals and messages from the brain can close it. Many specifics of the 1965 circuit were later revised. The principle survived every revision. The cord modulates. It does not merely relay.

Fourteen years later Daniel Le Bars, Anthony Dickenson, and Jean-Marie Besson, working in Paris, found the brake in action. They were recording from dorsal horn cells in the rat and applied a strong painful stimulus far away from the part of the body that cell listens to. The cell went quiet. Pain applied in one place silenced pain processing in another. They named the phenomenon diffuse noxious inhibitory controls, and it works by a signal descending from the brainstem back down the cord.

Two later syntheses matter for how this page reads the clinic. The physiologist Mary Heinricher and colleagues reviewed the descending system and showed that the same pathway can inhibit or facilitate, damping the signal in one state and amplifying it in another.

The neurologist Howard Fields described the brainstem cells that do it, a population that switches on when pain increases and a population that switches off, and called the whole arrangement state-dependent. The brake is not a fixed setting. It moves with attention, expectation, stress, and sleep.

So a person with high gain has one of two problems, and they look identical from outside. The amplifier climbed, or the brake went soft.

This distinction is not academic. It is the difference between a system that is generating too much and a system that has lost its capacity to turn itself down. Both produce the same complaint. They are approached differently, and the measurement section shows how they are told apart.

Gain is set before the input arrives

Two experiments measured the setting first and the experience second, and the setting decided the experience both times.

At Oxford, a group including the neurologist Markus Ploner and the neuroscientist Irene Tracey asked whether they could predict, before delivering a stimulus, whether that stimulus would hurt. They used functional MRI to measure how strongly two regions were communicating in the seconds before an identical noxious stimulus arrived.

One was the anterior insula, which handles the felt state of the body. The other was the brainstem, which houses the descending brake. That prestimulus connectivity determined whether the stimulus was perceived as painful. The pattern also tracked personality. More anxious and more pain-attentive people showed weaker connectivity down to the modulating regions.

Identical stimulus. Different setting. Different experience. The measurement was taken before the event.

A second study makes the same point with a drug. Ulrike Bingel and colleagues gave healthy volunteers a fixed concentration of remifentanil, a potent opioid, against a constant heat pain, and varied only what the volunteers were told. With a positive expectation, the analgesic benefit roughly doubled. With a negative expectation, the analgesia was abolished. Brain imaging tracked the difference. The blood level of the drug was the same in all three conditions.

This is the sharpest published version of a claim the Unified Model of Tone makes generally, and the model states it as its own. Input meets tone. A signal does not carry its meaning with it. The organization the system is holding when the signal lands is what decides what the signal becomes.

The same dose, the same heat, the same touch, the same week at work, will land differently on two people. It will land differently on the same person in two states. That is not noise in the data. It is the variable the data is measuring.

One boundary, because this is where two parts of tone meet. Expectation is not gain. Expectation belongs to prediction: a system acting on its internal model of the world rather than on the world itself. Prediction is one of the strongest things that sets gain. Keeping them separate is what lets you say something specific about a person: whether the setting is high because the amplifier is climbing, or high because the model the system is running says danger.

When input is lost, gain rises

Gain rises when input is lost. Everything to this point treated high gain as a response to too much, and the most important fact on this page is the opposite.

In the 1990s the neuroscientist Gina Turrigiano was working on a puzzle in brain development. Individual connections between neurons are constantly strengthening and weakening as an animal learns. Left alone, that process should run away, driving a cell either silent or into permanent overdrive. Something must be holding the total steady. She wanted to know what.

She grew cortical neurons in culture and blocked all electrical activity. The cells could no longer hear anything. Over the following two days, every excitatory input onto them scaled up in amplitude. Then she ran the reverse, blocking the inhibitory chemistry so firing rates initially rose.

Across the next forty-eight hours the inputs scaled down and firing returned close to control levels. The scaling was multiplicative. Each connection changed in proportion to its own strength, so the relative pattern of what the cell had learned was preserved while the overall volume moved.

She called it synaptic scaling, and it is one form of what the field calls homeostatic plasticity. A cell defends its own activity level by adjusting its gain.

When input is lost, gain rises. That is not a malfunction. It is the system doing exactly what it is built to do.

Gain comes down as well, and the everyday version has a name. A working group of neuroscientists led by Catharine Rankin revisited habituation and set out its defining characteristics. A repeated harmless stimulus produces a progressively smaller response. The response recovers if the stimulus stops for a while. Habituation is your nervous system turning its own volume down on something that turned out not to matter. It is why you stop hearing the refrigerator.

So a healthy gain has a shape. It rises when the moment demands it. It falls when the alarm proves false. It climbs to compensate when a channel goes quiet, and it settles again when the channel comes back. That capacity to move and return is the same thing the rest of this library calls the width of the range. Gain is one of the axes along which that range can be wide or narrow.

Tinnitus with a normal hearing test is central gain compensation

The case is so clean it is almost a proof, and it is not about pain: less signal leaves the ear, the same amount arrives upstairs, and the ringing is the compensation made audible.

Some anatomy first. Sound moves the eardrum, which moves three small bones, which drive a coiled fluid-filled tube called the cochlea. Inside the cochlea, hair cells convert that motion into nerve traffic. The auditory nerve carries the traffic to the brainstem, which passes it upward to the cortex, where it becomes hearing.

Tinnitus is the perception of sound with no sound present. The traditional account is cochlear damage, and for many people that fits. It does not fit the large group of people who hear a constant ringing and then pass a standard hearing test.

Less signal leaves the ear, the same amount arrives upstairs

Roland Schaette and David McAlpine at University College London went after exactly that group. They recorded auditory brainstem responses, which are a series of waves generated at successive stations along the pathway. Wave I comes from the auditory nerve as it leaves the cochlea. Wave V comes from further up the brainstem. In tinnitus patients with normal audiograms, wave I was significantly reduced while wave V was normal. Less signal left the ear. The same amount arrived upstairs.

The output was renormalised. Something in between had turned itself up.

They then built a computational model of the auditory pathway and showed that a tinnitus percept emerges from precisely that homeostatic response to reduced input. The auditory researchers Benjamin Auerbach, Paulo Rodrigues, and Richard Salvi reviewed the wider evidence and named the phenomenon central gain enhancement.

Cochlear output falls, and activity in central auditory structures is paradoxically increased at louder intensities. They connect the same mechanism to hyperacusis, in which ordinary sound becomes intolerable. The auditory neuroscientist Arnaud Noreña assembled it into an integrative model of tinnitus built on central gain control.

Read that as the model reads it. The ringing is the sound of a system compensating correctly for a loss it was never told about. The percept is a symptom of the setting, and the setting is the thing to address. This is developed further on the tinnitus page and in the wider discussion of the senses.

Give the input back and the compensation never happens

If gain rises because information was lost, then returning the information should keep gain where it belongs. That is a prediction, and it has been run.

Arnaud Noreña and the neuroscientist Jos Eggermont exposed cats to a traumatic noise. One group was then housed in a quiet environment for several weeks, which is the standard protective instinct after an ear injury. A second group was placed in an enriched acoustic environment, sound deliberately matched to the frequency range where the hearing loss was expected, played at a moderate level above it.

The quiet group developed hearing loss across a broad span from 6 to 32 kHz, averaging around 40 dB at the worst frequencies, together with the expected reorganization of the cortical frequency map. The enriched group came out with loss restricted to a narrow 6 to 8 kHz band, with normal thresholds from 16 to 32 kHz, and no demonstrable cortical map reorganization.

Nothing was suppressed. Nothing was blocked. The system was given real information back, matched to what it had lost, and the gain compensation it would otherwise have made was never needed.

The edges of the result are part of the result. It is an animal experiment. It is prevention rather than reversal, delivered immediately after the trauma, and the human translation is unsettled, as the measurement and intervention sections show. As a demonstration of a principle it is exactly what the Unified Model of Tone means by restoring rather than masking, and the model did not have to invent it. Somebody published it in 2005.

Phantom limb pain re-tunes gain, in a direction still disputed

Two capable research groups hold two live answers about which direction the re-tuning runs. After an arm is amputated, most people still feel the missing hand. Many feel pain in it, sometimes severe, sometimes for decades. The limb is gone. The map of the limb is not.

The psychologist Herta Flor and colleagues asked whether phantom pain tracked what the cortex did after the input stopped. Using magnetic recordings, they measured how far the face representation had shifted into the territory that used to belong to the hand. The size of that shift correlated with the amount of phantom limb pain. Maladaptive reorganization became the textbook explanation and stayed there for close to twenty years.

Then a team at Oxford led by the neuroscientist Tamar Makin re-tested it with structural and functional MRI. They found something close to the opposite. Phantom pain was associated with preserved structure and preserved function in the former hand area, along with reduced functional connectivity between that area and the rest of sensorimotor cortex. On their reading, the persistent experience is what maintains the representation, rather than the representation collapsing and causing the experience.

Two capable groups asked one question and got two answers, and the disagreement is live.

What both results share is the deeper statement, and it is the one this page needs. After input is lost, the system re-tunes its own gain around the loss, and the pain tracks the re-tuning rather than the missing tissue. Which direction the re-tuning runs is precisely what the field has not settled. The model does not require an answer, and it would be in trouble if the re-tuning turned out to be irrelevant to the pain. It has not.

How gain fails

Gain fails in five characteristic ways, and central sensitization is only the first of them. Every part of tone fails in its own way. These are gain's.

1. Gain too high: sensitization

Fibromyalgia is where gain set too high has been documented most thoroughly. The rheumatologist Roland Staud and colleagues measured wind-up in people rather than in cells, delivering identical heat pulses to the hand at a fixed rate. Patients with fibromyalgia rated the first stimulus higher than controls, summed more steeply across the series, and reported after-sensations that were larger, longer, and more often painful.

The pain researcher Richard Gracely and the rheumatologist Daniel Clauw then did the experiment that settles the question of whether the reports are exaggerated. They applied pressure to the thumbnail during functional MRI. Pressure sufficient to produce equal subjective pain in both groups produced similar patterns of brain activation. Equal physical pressure produced far more activation in patients. Clauw's later clinical review describes fibromyalgia as a disorder of pain and sensory processing rather than of the painful tissues.

High gain is not confined to pain, which is the point of giving it a name of its own. In migraine, the neuroscientists Rodrigo Noseda and Rami Burstein traced a physical route from light-sensitive retinal cells to thalamic neurons that also carry pain signals from the coverings of the brain. That route is why light makes a headache worse in a body with no eye disease.

Burstein had earlier tested pain thresholds on the skin around the eye and on the forearm, during attacks and between them. He found that 79 percent of patients developed cutaneous allodynia during migraine, so brushing hair or resting glasses on the nose becomes painful. In anxiety, the same setting is measured with a startle probe. The psychophysiologist Christian Grillon reviewed decades of that work and describes potentiated startle under threat as a laboratory readout of a system answering neutral events loudly.

2. Gain too low: the warning line goes quiet

This is the failure mode nobody discusses. In three families in northern Pakistan, children were found who could not feel pain at all. A team led by the geneticist James Cox and the clinical geneticist Geoffrey Woods mapped the trait and found loss-of-function mutations in SCN9A, the gene for a sodium channel that nociceptors need to fire. The channel was dead and the warning line carried nothing. The result was not a blessing. It was burns, unnoticed fractures, and self-injury.

The quieter version is a blunted read of the body's own interior. The neuroanatomist Bud Craig described interoception as a distinct sense, a dedicated pathway reporting the physiological condition of the tissues to the brain.

A large group led by the neuroscientist Sahib Khalsa mapped how blunted or distorted interoception runs through depression, anxiety, eating disorders, and addiction. Numbness, disconnection, and not knowing you are exhausted until you fall over are gain failures too. They are simply the failures nobody brings to a clinic as a symptom.

3. Gain that will not come down

This one has a measurement you can watch happen. The neurologist Jean Schoenen and colleagues sat healthy volunteers and migraine patients in front of a reversing checkerboard and averaged the brain's electrical response in six successive blocks of a hundred. Healthy volunteers showed the response getting smaller across blocks, which is habituation. Migraine patients between attacks, on days when they felt entirely well, showed potentiation instead of habituation, the amplitude climbing to a peak in the second to fourth blocks. Same stimulus, opposite direction.

The objection travels with the finding, because this one is contested rather than settled. A blinded study by Petter Moe Omland and colleagues in Norway found no confirmation of abnormal habituation in migraine between attacks. The standing of the effect as a hallmark of the condition is disputed in the field. What the original result describes, if it holds, is gain stuck in the wrong regime in a person with no symptoms at that moment. Whether it holds is an open question.

4. Gain that is slow to return

Staud's after-sensations are the version you can feel: the stimulus stops and the signal keeps going. A system whose gain rises appropriately and then takes hours to come back down is not the same problem as one that never rises. How a failure ages is a subject of its own, treated on the time course page.

5. Loss and gain in the same person

This is the most clinically confusing pattern, and the most instructive. A large European consortium led by the neurologist Ralf Baron sorted 902 patients with peripheral neuropathic pain by their sensory profiles, then confirmed the sorting in a further 233. They found three distinct subgroups. The largest, 42 percent, was sensory loss: a limb gone quiet to touch and temperature. The second, 33 percent, kept its sensory function largely intact and was hypersensitive to heat and cold.

The third, 24 percent, had lost small-fiber function and was hypersensitive to pinprick and to a brush drawn across the skin. Loss and amplification, side by side, in one sample of 1,135 people. Numbness and hypersensitivity in the same limb are not a contradiction. They are what a system looks like when it has lost a channel and turned up what remains.

High gain is a reading, not a diagnosis

One caution belongs here and it is not optional. High gain must never be assumed when something findable is on the table. The neurologist Anne Louise Oaklander biopsied skin from patients carrying a fibromyalgia label. She found that 41 percent had biopsies diagnostic for small-fiber polyneuropathy, against 3 percent of controls. Some of those cases had specific and treatable causes. A sensitized system and a real peripheral driver are not alternatives. Look for the driver.

How gain is measured, and what the numbers cannot tell you

Gain is a slope, so every instrument that measures it works the same way. Deliver a controlled input. Measure the output. Change the input and measure again. This is how central sensitization is read in a living person.

Instruments that read the pain system

Quantitative sensory testing. A standardised battery of thermal and mechanical stimuli that produces a sensory profile for one region of the body. The German Research Network on Neuropathic Pain, in a paper led by the physician Roman Rolke, standardised the protocol and published reference values from 180 healthy subjects tested over face, hand, and foot.

The profile takes about thirty minutes and is designed so that each finding can be read as a plus sign or a minus sign, which is to say as gain or as loss.

Temporal summation. Wind-up measured in a living person. Identical stimuli at a fixed interval, with the rating expected to climb slightly. A sensitized system climbs much more, as Staud's fibromyalgia work showed. This is the closest thing in the clinic to a direct reading of the spinal amplifier.

Conditioned pain modulation. The brake, measured. Rate a test pain. Apply a second painful stimulus somewhere else entirely. Rate the test pain again. In a healthy system the second pain lowers the first. The neurologist David Yarnitsky set out how this human protocol relates to the animal work on diffuse noxious inhibitory controls and why it matters clinically.

The physiologist Serge Marchand's group, in a study led by Nancy Julien, tested it across three groups of thirty: fibromyalgia, chronic low back pain, and healthy volunteers. Healthy subjects and low back pain patients showed the inhibitory effect. The fibromyalgia group did not show it at all. Two chronic pain groups, one with a working brake and one without.

Evoked potentials. Averaged brain responses to a repeated stimulus, with the change in amplitude across blocks read as habituation. Schoenen's checkerboard is the canonical example.

Gain measured outside pain

Two measurements make gain impossible to dismiss as a metaphor, because they come from fields that never discuss pain.

Baroreflex sensitivity is reported in milliseconds of heart-period change per millimetre of mercury of blood pressure change. That is output divided by input, written in clinical units. The ATRAMI study, led by the cardiologist Maria Teresa La Rovere, followed 1,284 patients after a heart attack. It found that a baroreflex sensitivity below 3.0 ms per mmHg carried a significant independent risk of cardiac mortality. A slope predicted death.

Vestibulo-ocular reflex gain is eye velocity divided by head velocity. Turn the head and the eyes must counter-rotate at the same speed to keep the world still. A team led by the vestibular scientist Hamish MacDougall validated a video method against the scleral search coil standard. They found closely comparable recordings, though in a small sample of eight healthy subjects and eight patients. That ratio is now measured in clinics daily and nobody calls it a metaphor.

What the instruments cannot tell you

Each instrument has limits, and the limits are specific. Quantitative sensory testing measures the person, not the cord. It tells you the slope is steep; it cannot tell you where along the chain the steepening lives. Values overlap heavily between patients and healthy controls, so no single number diagnoses an individual. Conditioned pain modulation varies with protocol, time of day, and mood, and it is a group-level tool more than a personal one.

The migraine habituation finding is a difference between groups rather than a test that identifies a particular patient, and blinded work has failed to confirm it. And none of these instruments measures tone. Reading them together as measurements of one organization is the model's interpretation, and it is established when those measures turn out to load on one common factor.

The conditions that are mostly gain

Gain carries most of the weight in a specific set of conditions, and naming what combines with what is what stops a description from being generic. The parts of tone compose. Any condition in this library is two or three of them in combination.

Gain also carries real weight, in combination, on several other pages.

• Long COVID and other post-viral pictures combine gain with load and time course, which is why the same activity is tolerated one week and not the next.
• Unexplained symptoms and idiopathic diagnoses are frequently a gain reading that no test was built to see, since every standard test returns a value and gain is a slope.
• Trauma and mental health involve a threat-weighted gain that raises the volume on the whole interior.
• And neck pain and low back pain that persist long past tissue healing are the everyday version of a slope that never came back down.

What gain is not

Gain is one of the foundational dimensions of tone, and an ontology is only useful if its terms stay distinct. Each of the following is a separate dimension with its own page, and the line matters.

Set point is the value the system is defending. Gain is how steeply it answers when that value is disturbed. A thermostat set to twenty degrees and a thermostat that blasts the furnace at a one degree drop are two independent settings, and a person can have either problem, or both.

Oscillation is the rhythm and range a single system moves through over time. Gain describes the size of an answer. A rhythm can be intact while every response inside it is too large.

Coupling is whether separate systems stay in step with each other. Oscillation is within a system. Coupling is between systems. Gain is neither: it is the steepness of one system's response, whatever the others are doing.

Prediction is the system acting on its internal model rather than on the world. Expectation is among the strongest things that sets gain, as the remifentanil study showed, which makes prediction an input to gain rather than a synonym for it.

Load is what holding a state costs and what accumulates when the system cannot stop paying. Running at high gain is expensive. The gain is the setting. The bill is the load.

Constraint and slack describe where the system has room to move, mechanically and neurally. A system with no slack has fewer ways to answer, which often shows up as a steeper answer, though the two are separate readings.

Input quality is the fidelity of the signal coming in, meaning what the system actually knows about itself. Gain is how loudly it answers what it received. The tinnitus story is the two working together, and it is the reason they must be named separately: the input degraded, and the gain compensated.

Time course describes how any of these change as a problem ages, from acute to adaptive to entrenched. It is deliberately different in kind from the others, which describe a moment.

Gain is not tone

Tone is the integrated, coupled organization the nervous system maintains across the whole body, and its capacity to move where the moment demands and return afterward. Gain is one dimension of that organization, one of the axes along which it can be described. Confusing the part with the whole is how a useful idea turns into a slogan.

What moves gain, including the null results

Gain moves, and the interventions that move it divide by what they aim at: the input, the meaning of the signal, or the output. The evidence for each is carried here with its null results, which the model reads rather than hides.

Movement changes gain, and the direction depends on the state it meets

The pain researchers Kelly Naugle, Roger Fillingim, and Joseph Riley pooled the experimental literature on what a single bout of exercise does to pain perception. In healthy adults, aerobic, isometric, and resistance exercise all raised pain thresholds and lowered pain ratings, with effects ranging from small to large depending on how pain was induced and how the exercise was done.

In chronic pain populations the same review found the magnitude and direction highly variable, hypoalgesic in some conditions and hyperalgesic in others, depending on the population and the intensity.

That is usually reported as a limitation. Read through this model it is a result. Input meets tone. Where the system retains range, movement is information and gain comes down.

Where the system is already at its ceiling and has no slack, the identical dose is one more demand it cannot afford, and the answer gets louder. A framework that expected one input to do one thing to everyone would call this literature inconsistent. A framework built on the state of the receiver expects exactly this shape.

Returning real input

The enriched acoustic environment result is the cleanest version, and the human translation is unsettled. A Cochrane review led by the hearing scientist Magdalena Sereda examined sound therapy for tinnitus. It found eight trials with 590 participants, no usable data for the main comparisons, and no evidence favoring one device over another, at low certainty.

That is a null and it stays on the page. The model reads it precisely. The animal work matched the enrichment to the specific deprived frequency range, while generic sound delivers sound. Delivering sound and delivering the missing information are different acts, and the trials have largely tested the first.

Changing what the signal means

The psychologist Rilana Cima and colleagues randomised 492 people with tinnitus to a stepped-care programme built on cognitive behavior therapy or to usual care. Over twelve months the specialised group improved on quality of life, tinnitus severity, and tinnitus impairment, with effect sizes of 0.24, 0.43, and 0.45. Note what did and did not change. The percept was not reported as eliminated. What moved was the response to it, which is a gain change operating at the level of meaning rather than at the synapse.

Autonomic training

A pilot led by the psychologist Afton Hassett taught twelve women with fibromyalgia to breathe at their resonant frequency across ten sessions of heart rate variability biofeedback. Pain, depression, and functioning improved by three-month follow-up, and heart rate variability rose. Twelve people, open label, no control group. That is a signal worth testing properly rather than evidence of efficacy.

Medication acts on the output

Drugs that strengthen descending inhibition or damp dorsal horn excitability genuinely lower gain for many people. For someone in severe pain that is valuable and sometimes indispensable. They act on the output reliably and in one direction. The counterweight sits inside the same literature.

The anaesthesiologists Martin Angst and David Clark reviewed the evidence for opioid-induced hyperalgesia, in which a drug given to reduce pain can leave a person more sensitive to pain than before. An intervention aimed at the output moved the setting, in the wrong direction. Nobody involved did anything wrong. There was simply no framework in which the setting was a thing to watch.

Gain, read as one dimension of tone

Gain is one dimension of tone: how steeply the nervous system answers what arrives. Tone is the integrated, coupled organization the nervous system maintains across the whole body, and health is the width of the range that organization can move through and return. Read this way, three things follow that the standard framing of central sensitization cannot say.

The symptom is often the compensation

The ringing, the tenderness, the flinch at light, the startle at a closing door. Each is a system doing the sensible thing given the gain it is holding. Chasing the symptom means arguing with a correct answer to a badly posed question.

Restoring and masking are different aims

A clinician who lowers a person's pain has done something real, and a person in severe pain deserves relief. The distinction is what the intervention is for. Suppressing an output leaves the setting exactly where it was, so the next input meets the same slope. Restoring means the setting itself moves back into range, and the person's response to the next input differs because the organization changed.

That is the aim being described, and the aim runs ahead of the evidence. No intervention has been shown to restore gain in this sense in humans, here or anywhere else. The distinction is rarely drawn because tone has never been recognized as its own regulatory system, so there has been no framework in which a setting is a thing to aim at.

The bidirectional test that separates restoring from masking

A drug pushes in one direction. It lowers what is high, and it lowers what is already low. The opioid-induced hyperalgesia literature reviewed in 2006 shows an output-directed agent can even move the setting the wrong way. The Unified Model of Tone predicts something a one-directional agent cannot do.

An intervention that genuinely restores tone should move a dysregulated gain toward the middle from either side. By tonal input this page means any intervention aimed at the organization the system is holding rather than at the symptom it is producing. What follows is a study design, not a finding.

Take a mixed group. Some with steep temporal summation and an absent brake. Some with blunted thresholds and a poor read of their own interior. Deliver the same tonal input. The model says both move toward normative values. A one-directional agent moves both the same way.

If both groups move in the same direction, the input is pushing the output rather than restoring gain, and it belongs in the masking column. It helps whichever group it happens to point at and carries the other group further from the middle.

That trial has not been run. Nobody has tested whether any intervention moves gain toward normative values from both sides at once. Until somebody runs it, the design is a proposal rather than a result, and nothing in it is evidence that any treatment does what the design describes.

Most of the evidence on this page is correlational, which is why the prestimulus designs matter. They measure the setting before the event and predict what the event becomes. It is also why the model's own test is interventional and bidirectional rather than observational. The components are credited to the people who found them. Central sensitization, synaptic scaling, baroreflex sensitivity, and vestibulo-ocular gain were all described by others, and the model's claim is the unification: four instruments reading one setting of one state.

That setting composes with the other foundations of tone to describe any condition, and the composition makes predictions the separate literatures do not. A name that predicts is more than a name. The composition claim carries a test of its own. If the conditions this page calls gain-dominant do not share gain measurements when those are recorded together, the ontology does not hold.

What is left is a question worth carrying out of here. When something in your body answers too loudly, the useful question is rarely how loud the signal was. It is what setting your system was holding when the signal arrived, and whether that setting can still move.

Frequently asked