Behavioral Disorders and the Nervous System

What a behavioral diagnosis actually is

A behavioral diagnosis names a pattern of observed behavior. Nothing is measured in the body, and nothing has to be. That single fact organizes everything the research below keeps finding.

Start with how one is made. A person answers questions, or a parent and a teacher fill in rating scales. A clinician counts how many listed behaviors are present, how long they have lasted, and how much of ordinary life they are costing. If the count clears a threshold, the diagnosis is written down.

That is the whole procedure. There is no blood test for attention deficit hyperactivity disorder. There is no scan that shows autism, no biopsy that shows obsessive compulsive disorder, and no marker in the blood for anxiety or post traumatic stress. The name describes a pattern of behavior across time. Everyone who assigns these names knows this, and the field has said so in public.

In 2013 a psychologist and a psychiatrist at the National Institute of Mental Health, Bruce Cuthbert and Thomas Insel, published a paper arguing that the standard diagnostic categories were holding research back. They were not critics from outside. Insel directed the agency that funds most of this research, and Cuthbert ran the unit inside it that was building the alternative.

They proposed a replacement framework called Research Domain Criteria, and set out the reasoning behind it. The categories, they argued, are useful for clinical communication and poor at mapping onto biology. Three years earlier Insel and colleagues had already put the proposal on the record, calling for research organized around measurable dimensions of function rather than diagnostic labels. Genes, circuits and physiology do not respect the borders between the labels.

Half the people who qualify for one label qualify for two

The second clue is that the labels rarely arrive alone. The epidemiologist Ronald Kessler ran a nationally representative interview survey of more than nine thousand American adults, asking how many met criteria for a mental disorder in a single year. He was measuring prevalence. He also measured the shape of the categories.

Among people who met criteria for anything in twelve months, fifty five percent carried one diagnosis, twenty two percent carried two, and twenty three percent carried three or more. Nearly half of all cases were comorbid. If the categories were separate diseases, that number would be a scandal.

The third clue is the sharpest. In the small New Zealand city of Dunedin, a research unit has followed a thousand people born in 1972 and 1973 for their entire lives. The psychologists Avshalom Caspi and Terrie Moffitt used that cohort to ask a structural question.

Do mental disorders sort into separate families across two decades of a life? Three higher order families fit the data reasonably well. One single general dimension fit it better. They named it the p factor, a general tendency toward any and all psychopathology, and observed that it explains why causes and biomarkers specific to one diagnosis are so hard to find.

So the puzzle stands in full view. Five conditions with no test, no lesion, heavy overlap, and a shared underlying dimension that nobody has been able to name in physical terms. Something is varying underneath all five, and naming it in physical terms is the work ahead.

The machinery that sets a behavioral state

Every behavioral state a checklist can count is set by a short list of parts. An accelerator and a brake, an alarm, a governor, a groove cutter, an inward sense, a timekeeper, a volume knob and a switchboard. The evidence only makes sense once the parts do.

A neuron is a living wire. It sits quiet, gathers small electrical nudges from the wires feeding it, and when the nudges add up past its threshold, it fires a pulse down its length. It either fires or it does not. Its influence comes from how often it fires, when it fires relative to its neighbors, and how many other wires it can reach. Eighty six billion of them, wired into loops, are what a brain is.

The accelerator, the brake, and the number that reads them

Two of those wiring lines run out to the organs, and they pull against each other. One is the accelerator, called the sympathetic system. It speeds the heart, tightens the vessels, opens the airways and readies the muscles. The other is the brake, called the vagal or parasympathetic system, carried mainly by a single long nerve, the vagus, which runs from the brainstem down to the heart and gut. It slows the heart and lets the body settle.

Because the two pull against each other continuously, your heart rate is never metronomic. The interval between beats stretches and shortens breath by breath. That beat to beat wobble is called heart rate variability, and it is a validated index of how actively the vagal brake is working. The word validated will carry weight when the behavioral evidence arrives.

The governor, the alarm, and the groove cutter

Now the brain itself. Behind your forehead sits the prefrontal cortex, the slow deliberate governor. It holds a goal in mind, weighs a delay against a reward, and inhibits the first impulse long enough for a second one to arrive. Deep in the temporal lobe sits the amygdala, an almond of tissue that works as the fast alarm. It reaches a verdict about danger in a fraction of a second, well before the governor has finished reading the situation.

Underneath both sits the striatum, the groove cutter. It takes a sequence that got a good result and, with repetition, welds it into a single automatic unit that runs without supervision. That is how a habit is made, and it is why habits are so cheap to run and so hard to interrupt.

Folded into a cleft on each side of the brain is the insula, the place where the body reports in. Heartbeat, breath, gut, temperature, the state of the airway, all of it arrives here and becomes something you can feel. The cerebellum, the dense structure at the back of the skull, handles timing and smoothing, turning rough intentions into movements and thoughts that arrive on the beat.

The volume knob and the switchboard

Then there is the volume knob. In the upper brainstem sits a tiny bluish cluster called the locus coeruleus, no bigger than a grain of rice, whose fibers reach almost the entire brain. It releases noradrenaline, which raises the gain on everything downstream. One small structure sets the alertness of the whole system.

Finally, the switchboard. Where the brain meets the spinal cord, a region called the nucleus of the solitary tract receives the incoming reports from the body. The physiologists Michael Andresen and Diana Kunze spent years working out how this nucleus handles that traffic. They described it as the gateway for neural circulatory control.

It is the first place a visceral signal is processed on its way up. The gut has its own thick line into the same region. The neuroscientists Kirsteen Browning and Alberto Travagli mapped it in their account of how the brain and the digestive tract govern each other. Traffic runs both directions on all of these lines. The body speaks as much as it is spoken to.

That is the machinery. Every behavioral diagnosis on this list is a statement about how those parts are holding together.

Brain imaging keeps finding one address, not five

If the five behavioral conditions were five diseases, imaging should have separated them by now. Across 193 structural studies and a parallel functional literature, it has done close to the opposite.

The clinical neuroscientist Madeleine Goodkind led a team that pooled 193 published imaging studies comparing the brain structure of psychiatric patients with healthy volunteers. The studies covered six diagnostic groups, including anxiety disorders and obsessive compulsive disorder. The question was whether each diagnosis owns a distinct structural signature. The answer was that they share one. Across all six, the same regions had lost gray matter, and the loss converged on the dorsal anterior cingulate cortex and both insulae.

Look at what those regions do. The insula is where the body reports in. The dorsal anterior cingulate sits directly upstream of it and handles salience, the ongoing judgment of what matters right now and how hard to work at it. Together they form the hub where the state of the body is read and turned into a decision about effort. That is the address the six diagnoses share.

The finding extends beyond structure. The clinical psychologist Lisa McTeague led a parallel analysis of functional imaging, asking whether the brain does anything different during tasks that demand cognitive control, the deliberate steering of attention and behavior. Across a wide span of diagnoses she found a common disruption in the same control network rather than a set of disorder-specific ones.

Two independent methods went looking for what separates the diagnoses and came back with what unites them.

Psychiatry reads this convergence as a puzzle about categories, and asks how the labels should be redrawn. The Unified Model of Tone reads it differently and says so plainly. The shared address is the signature of one regulated variable failing, registered in the exact circuits that regulate it.

The insula and the cingulate keep appearing because they are where the nervous system reads its own state and decides what state to move to next. When that reading loses its range, the behavior downstream can look like inattention, fear, ritual or overwhelm, depending on the register. Same failure, different octave.

One coupled range underneath five diagnoses

Tone is the range of states the nervous system can enter, hold and leave on demand. Behavioral disorders are what the loss of that range looks like from the outside, register by register.

Every part described above is a state setter. The accelerator and brake set the arousal of the body. The volume knob sets the gain of the whole brain. The governor sets how long a goal survives against a distraction. The alarm sets the threshold at which threat wins. The inward sense sets what the body's own signals are taken to mean. The groove cutter sets how quickly a repetition hardens into an automatic loop.

None of these is a switch with two positions. Each is a range, and the ranges are coupled. Move one and the others move with it, because they are all reading and writing the same underlying organization. That coupled organization, held across the whole system, is what the Unified Model of Tone calls tone.

Health, in this reading, is the width and the flexibility of that range. A healthy nervous system can climb high into alertness for a demanding hour and come all the way back down. It can hold one setting against interference, then release it when the task ends. It can sound an alarm and then quiet the alarm once the threat is gone.

The capacity to move and return is the health. Illness begins when the range narrows, sticks, or fails to hold. A system that cannot leave the alarm state, hold attention on a setting, or release a correction loop is expressing disease as lost range rather than a broken part.

Each behavioral diagnosis is one register of the same failing range

Read that way, the five diagnoses stop looking like five diseases. They look like five registers in which a single range can fail. In attention deficit hyperactivity disorder the range will not hold a setting, drifting off a chosen state and back again on its own schedule. In anxiety and post traumatic stress the range has collapsed toward threat, so that the alarm setting has become the resting setting.

In obsessive compulsive disorder the range is trapped in a correction loop, where the signal that a matter is settled never arrives and the correction runs again. In autism the incoming detail is weighted so heavily that the system cannot generalize away from it, so the world arrives at full volume and full resolution and refuses to be summarized.

This is the model's own claim, stated as such. No study proves it. What the studies do, from here forward, is behave exactly as this claim predicts they would, including in the places where they disagree with each other.

Why a single molecule never explains a behavior

The popular account of behavioral disorders is chemical: too little serotonin makes you anxious, too little dopamine makes you inattentive. The evidence has been harder on that story than the public was told.

The chemical account satisfies because a single missing molecule implies a single replaceable molecule. The psychiatrist Joanna Moncrieff led an umbrella review of the serotonin theory of depression. An umbrella review is a review of the existing reviews and meta-analyses, an attempt to see the whole literature at once rather than one study at a time.

The team gathered the major lines of evidence, from serotonin metabolites to receptor imaging to depletion experiments to genetics, and reported no consistent evidence of an association between serotonin and depression.

The paper is contested, and the argument about its methods continues. Two things follow from it precisely. It does not show that antidepressants fail to help people, which is a separate question with its own evidence. It does show that psychiatry's most famous one-molecule story was never resting on what the public was told it rested on.

The dopamine version, measured in living tissue

Attention has its own version. The psychiatrist and neuroscientist Nora Volkow turned to positron emission tomography. That is a scan which tracks an injected radioactive tracer to count how much of a molecular target is available in living tissue. She used it to measure dopamine markers in adults with attention deficit hyperactivity disorder.

She found reduced dopamine markers in the reward pathway compared with controls. That result is real and it has been influential. It is also a group difference with overlapping distributions, which is why no scan of this kind diagnoses anybody. It names one voice, and one voice does not name the chord.

The model's reading is blunt. A behavior is a chord sounded by coupled voices held in step: arousal, prediction, interoception, threat, timing and metabolism. Change any one voice and every other voice shifts, because they are not independent channels running in parallel. They are aspects of one organization.

Which is why a single molecule never accounts for a behavior, and why every attempt to make it do so ends in a literature full of modest effects and open disagreement. You cannot name a chord by naming one note, and the chord is the thing the patient lives inside.

Arousal has an optimum, and it fails in both directions

The centerpiece of the range argument is over a century old, and it did not come from this model. In 1908 two psychologists measured how fast mice learn under different strengths of stimulus, and found a curve instead of a line.

Robert Yerkes and John Dodson set out to answer a narrow question. Does a stronger punishment teach an animal faster? They put mice in an apparatus with a white chamber and a black chamber, trained them to choose one, and delivered a small electric stimulus when the animal chose wrong. Then they varied the strength of that stimulus and timed how many trials the learning took.

They expected a straight line. Stronger stimulus, faster learning. What they got was a curve with a peak in the middle. Learning was fastest at an intermediate strength. Weak stimuli taught slowly. Strong ones taught slowly too. They then made the discrimination harder, by reducing the difference in brightness between the chambers, and found the peak had moved. The optimum strength was lower for the harder task.

Two lessons live in that curve. Performance is not a function of how much arousal you have but of whether the amount matches the task. And the right amount is a moving target, set by what is being asked.

The volume knob has two modes

The neuroscientist Gary Aston-Jones spent decades recording from the locus coeruleus, the volume knob described earlier, in monkeys working through target detection tasks. He wanted to know what this one small structure contributes to attention. He found two modes. In phasic mode, the cluster fires in crisp bursts locked to the target, and the animal is engaged and accurate.

In high tonic mode, its baseline firing is elevated and steady, and the animal becomes distractible, restless and prone to sampling alternatives. With Jonathan Cohen he set the finding out as an adaptive gain theory in which the relation between tonic firing and performance traces the same inverted U.

The neurobiologist Amy Arnsten studied the other end of the same curve. She asked what stress does to the governor behind the forehead, and traced the mechanism in detail. Stress-level release of the alerting chemicals engages an internal signaling cascade that rapidly weakens prefrontal connections and takes that region offline, while strengthening the alarm circuits underneath. The prefrontal cortex needs a middle dose of the very chemicals that, at higher doses, disable it.

The system that holds your attention, your plan and your restraint has an optimum, and it fails in both directions. Too little arousal and it never engages. Too much and it is switched off from the inside. Established science, not this model's invention.

What the model adds is the next sentence. If arousal has an optimum that must be moved to and held, then that capacity is a real property of the nervous system. Its failure is a real behavioral disorder, whether or not it leaves a lesion.

The same drug at the same dose produced opposite effects

In 1997 a single dose of a dopamine drug improved some healthy volunteers and worsened others, split cleanly by where each nervous system started. That experiment is the sharpest in the behavioral literature, and it is almost never quoted to patients.

The cognitive neuroscientists Daniel Kimberg, Mark D'Esposito and Martha Farah gave healthy volunteers a single dose of bromocriptine, a drug that mimics dopamine at its receptors. They then tested working memory and the ability to switch between tasks. The hypothesis was straightforward. More dopamine activity, better performance on tasks known to depend on it.

The average result was unremarkable. The individual results were not. When the team split the volunteers by their baseline working memory capacity, the drug's effect reversed across the split. Volunteers with lower capacity improved. Volunteers with higher capacity got worse. The effect of the drug depended on the state of the person receiving it, on the same dose, on the same day, in the same lab.

Stability against flexibility, on one curve

The cognitive neuroscientist Roshan Cools, working with D'Esposito, spent the following years generalizing that result. Their synthesis states the principle directly. Dopamine acts on working memory and cognitive control along an inverted U shaped curve, so there is an optimum rather than a maximum. Effects are baseline dependent and can be paradoxical.

Cognitive control is a moving balance between stability, holding a state against interference, and flexibility, letting the state go when the world changes. Push the balance toward stability and you buy focus at the cost of rigidity. Push it toward flexibility and you buy adaptability at the cost of distraction.

The drug did not have an effect. The meeting had an effect.

This is the principle the model calls input meets tone, and it reorganizes how every study of behavioral disorders should be read. No input acts on an empty body. An input acts on a nervous system that is already somewhere in its range, and the outcome is a product of the two.

That is why the same medication helps one child and agitates another. Why the same meditation practice settles one person and unsettles another. Why an intervention that works beautifully in a trial produces a modest average, because averaging across people who started at opposite ends of the range cancels the effect out.

A model that predicts uniform responses is refuted by this literature on the first page. A model built on input meeting tone predicts precisely the pattern the literature actually shows.

Heart rate variability moves where the brake carries the signature

Across 140 case-control studies, heart rate variability reads reduced in every psychiatric group, medicated or not. In attention deficit hyperactivity disorder it reads flat. Both results confirm the same model.

The psychophysiologist Julian Thayer, with the psychiatrist Richard Lane, proposed a model tying the vagal readout to the brain. Their account of neurovisceral integration argues that the prefrontal circuits which regulate emotion and the brainstem circuits which regulate the heart are one system, so heart rate variability indexes the flexibility of both at once.

Thayer later argued with the clinical psychologist Theodore Beauchaine that the measure is a transdiagnostic marker, tracking a capacity that runs underneath the diagnostic labels rather than belonging to any one of them.

The pooled data are substantial. The researcher Gail Alvares led a systematic review and meta-analysis that pooled 140 case-control studies of heart rate variability in psychiatric populations. Two results matter here. Heart rate variability was reduced across every patient group compared with controls, and the reduction remained significant in patients taking no medication at all. Several psychotropic drugs reduced it further. The narrowed range is not a side effect of treatment.

The condition-specific work agrees where the model expects agreement. A meta-analysis led by the psychiatrist Ying-Chih Cheng found lower baseline heart rate variability in autistic individuals, together with blunted responsiveness of the vagal brake to challenge. The psychologists Martha Schneider and Andreas Schwerdtfeger pooled the post traumatic stress literature and found the same reduced signature.

The ADHD null, and why the model needs it

The psychologist Julian Koenig led a meta-analysis of resting vagal tone in attention deficit hyperactivity disorder, pooling eight studies and 587 participants. The result was no significant difference between patients and controls. The authors state the contrast explicitly. Unlike the internalizing disorders, this one is not associated with altered short-term resting vagal tone.

Separately, the researcher Alessio Bellato reviewed 55 studies of autonomic function in the same condition, across skin conductance, heart rate and pupil measures. The dysfunction he found tilted toward hypo-arousal rather than hyper-arousal, most clearly at rest and during tasks demanding sustained attention. Almost half of the reported findings were null.

Those are careful studies, and this model expects both of their results.

Two predictions are being confirmed at once. The first is the coupled chord. Heart rate variability samples a single voice, the vagal brake, out of a chord sounded by many. Where the dysregulation is carried on that channel, as in the internalizing conditions, the measure moves and moves consistently. Where it is spread across the other coupled voices, the measure comes back flat while the person is plainly dysregulated. Silence one voice and you learn about one voice.

The second is that range can fail in either direction. Bellato's tilt toward hypo-arousal is the range argument itself. A system stuck too low on the inverted U underperforms exactly as a system stuck too high does. Drifting attention and a hunger for stimulation are what a volume knob set below the task would produce.

ADHD is a collapse of range in a different register than anxiety, so it fails a test built for anxiety's register. A model that expected one uniform autonomic signature across five diagnoses would be dead on this evidence.

One discipline point, stated plainly. That heart rate variability is a validated index of autonomic function is established fact. That it offers a partial window onto tone is this model's interpretation, laid on top of the measurement. The two are different kinds of statement, and they stay separate here.

How the body is read, and misread

Your brain reads your own body continuously, and most of what you call a feeling is made out of that reading. The sense has a name, interoception, an anatomy centered on the insula, and a measurable accuracy that varies person to person.

The neuroanatomist Bud Craig spent his career tracing where signals from inside the body actually go. He was mapping a pathway that had been assumed to be crude and diffuse. He found it precise, and he followed it to a specific destination. In a landmark review of interoception he argued that the insula holds a detailed representation of the body's physiological condition. Subjective feeling, he proposed, is built there.

The neuroscientist and psychiatrist Hugo Critchley then tested the human version directly. He put volunteers in a scanner and gave them a simple task. Listen to a train of notes and judge whether it is in time with your own heartbeat. Some people are strikingly accurate at this. Most are not. Critchley wanted to know what the accurate ones have.

He found that activity in the right anterior insular and opercular cortex predicted accuracy. The amount of gray matter in that same right anterior region tracked accuracy too, and it tracked how aware of their own insides people said they were. Accuracy, separately, tracked how intensely they reported experiencing emotion. The volume of a patch of cortex marks how well a person reads their own body.

Anxiety as a forecast, autism as a mismatch

The psychiatrists Martin Paulus and Murray Stein took that anatomy and made a claim about anxiety with it. In their reading, anxiety is an insula that over-forecasts a bad bodily future. The region does not merely report the body's current state, it predicts the next one, and in anxiety the prediction runs consistently worse than what arrives. The alarm fires against a forecast rather than an event.

The neuroscientist Sarah Garfinkel added the distinction that makes this usable. Interoceptive accuracy, how well you actually detect your own heartbeat, is not the same as interoceptive confidence, how attuned to your body you believe yourself to be. In autistic adults she found reduced accuracy alongside heightened subjective sensitivity, and the gap between the two predicted anxiety symptoms. The problem was neither too little body awareness nor too much. It was a mismatch between the signal and the confidence placed in it.

The psychiatrist Sahib Khalsa gathered a large group of researchers to write a roadmap placing interoception at the center of mental health research, which is roughly where Goodkind's imaging meta-analysis had independently pointed. Both insulae, shared across six diagnoses. The inward sense is the address the behavioral diagnoses share, and a mis-set inward sense is a mis-set body state, because in a coupled system reading and setting are the same act.

Prediction fails in a different register in autism, obsessive compulsive disorder, and anxiety

The nervous system does something more interesting than reacting. It guesses ahead, constantly, and checks the guess. In autism, obsessive compulsive disorder and anxiety, that forecasting engine fails in three distinct registers.

The neuroscientist Karl Friston formalized the guessing into a single principle. A brain carries a model of its world and its body, generates predictions from that model, and spends its energy on the gap between what it forecast and what arrived. That gap is prediction error, and the free energy principle holds that perception, action and learning are all ways of reducing it. Perception adjusts the model to fit the world. Action adjusts the world to fit the model.

One control setting governs the whole arrangement. Precision is how much weight the system gives to incoming error against its own prior expectation. Set precision high and the world overrides your model constantly. Set it low and your model overrides the world. There is no correct fixed value. There is a range that has to be moved through as circumstances change, which is tone, in the vocabulary of this model.

Autism: precision stuck high and rigid

Autism research arrived at that control setting from the inside. The psychologists Elizabeth Pellicano and David Burr proposed that autistic perception involves weaker priors, so less of the brain's own expectation is imposed on incoming sensation. In their phrase, the world becomes too real, arriving at a resolution the rest of us smooth away.

The psychologist Sander Van de Cruys and colleagues sharpened it into a proposal they called high and inflexible precision of prediction errors. Their detailed predictive coding account holds that the setting is both too high and too rigid to be adjusted by context. Rebecca Lawson, Geraint Rees and Friston published an aberrant precision account of the same territory.

Read those three papers together and what they describe is not a deficit. It is a control parameter stuck outside the range where it could be adjusted. A stuck parameter is a different kind of thing from a broken part, and it explains capability and difficulty in the same breath. High precision on incoming detail is what produces both the overwhelm in a supermarket and the accuracy that a smoothing brain would have averaged away.

The perceptual evidence fits. The neuroscientists Caroline Robertson and Simon Baron-Cohen reviewed the sensory findings in autism and argued that altered sensory perception is a core feature rather than an accompaniment to a social one.

The occupational scientist Ayelet Ben-Sasson pooled the symptom literature and found large differences in sensory modulation, with the largest gap in under-responsivity, followed by over-responsivity and sensation seeking. Note that shape. The difference runs in both directions at once, which is what a range problem looks like and not what a single deficit looks like.

The psychologist Shulamite Green imaged the mechanism. She scanned autistic and non-autistic adolescents while delivering mildly aversive touch and sound. Youth with sensory over-responsivity showed stronger responses in the sensory cortices and the amygdala, and decreased habituation to the repeated stimulus.

Habituation is the ordinary fading of a signal that has proven harmless. Its absence means every repetition arrives as new information. Green also found that a subgroup without over-responsivity showed prefrontal down-regulation of the amygdala, which is the brake working. Same diagnosis, two settings, two behaviors.

Beneath the psychology sits a claim about the tissue itself. The neuroscientists John Rubenstein and Michael Merzenich proposed that many features of autism follow from an increased ratio of excitation to inhibition in key circuits. Excitation pushes neurons toward firing and inhibition holds them back, and the balance between them sets how noisy, how reactive and how stable a circuit is. That balance is tone written at the level of the cell.

Obsessive compulsive disorder: the check that never closes

Now take the same prediction engine and jam it in a different place. In obsessive compulsive disorder the forecast is checked and the check never resolves. The neuroscientist Lara Menzies pooled the imaging and neuropsychological literature and revisited the orbitofronto-striatal model, the loop running from the evaluative front of the brain into the groove cutter and back. The psychiatrist Jonathan Posner scanned unmedicated adults with the condition, avoiding the confound of treatment, and found reduced functional connectivity within that limbic loop.

The neuroscientist Ann Graybiel spent a career on what the striatum does with repetition. In her account of habits and rituals, a repeated sequence gets chunked into a single automatic unit with a marked beginning and end. A groove, cut deeper by every pass. When the signal that closes the loop fails to arrive, the sequence has a start and no finish, and it runs again. Ritual is a groove with no exit.

One engine, one control parameter, three failures. Precision too high and too rigid, and the world will not summarize. The closing signal missing, and the correction loop will not end. The forecast biased dark, and the alarm answers a future that never comes. The model reads this as one variable failing in three registers of behavioral disorder, and states the reading as its own.

Fear is released by a circuit, and trauma fails toward either end

The amygdala reaches its verdict about danger in a fraction of a second. Release takes a separate act, infralimbic neurons firing the memory of safety, and in post traumatic stress that release fails in two opposite directions.

The neuroscientist Joseph LeDoux traced the circuit by which a neutral signal comes to trigger a defensive response, working out where the association is stored and how it reaches the body. His synthesis of the emotion circuits in the brain put the amygdala at the center. It receives a fast crude route from the senses and a slower detailed one. It drives the body's response before the slower route has finished.

The question that matters clinically is not how fear is learned. It is how it is released.

The neuroscientists Mohammed Milad and Gregory Quirk answered it with a rat, an electrode and a patient experiment. Rats learned that a tone predicted a shock. Then the tone was played repeatedly with no shock, until the fear response faded, a process called extinction. The next day the researchers tested whether each rat remembered the safety.

Meanwhile they recorded from neurons in a strip of medial prefrontal cortex called the infralimbic region. They found that these neurons fired during the recall of extinction, and that the animals whose neurons fired most were the ones who stayed calm. Then they went further. Stimulating those neurons during the tone reproduced the calm, in animals that had not been extinguished.

Safety is not the absence of alarm. It is a signal that a specific circuit has to send.

That is the brake, in the brain, doing a positive act. The psychologist Lisa Shin and the psychiatrist Israel Liberzon assembled the human version in a review of the fear, stress and anxiety circuitry. The pattern they describe across the anxiety and trauma disorders is an over-responsive amygdala with insufficient prefrontal regulation.

Trauma collapses toward either end

Now the finding that ends any account of trauma as simple over-arousal. The psychiatrist Ruth Lanius studied what happens in the brain when people with post traumatic stress disorder recall the trauma. She found two opposite responses. Most showed the expected pattern, reliving the event with high arousal and reduced prefrontal activity.

A substantial minority showed the reverse. They dissociated, describing themselves as detached from their own body, and their scans showed increased medial prefrontal activity with reduced autonomic arousal. Lanius argued for a dissociative subtype defined by overmodulation rather than undermodulation.

The hormone story runs the same way. The textbook prediction for a chronically stressed person is high cortisol, the body's main stress steroid. The neuroscientist Rachel Yehuda measured cortisol excreted in the urine over twenty four hours in combat veterans with post traumatic stress disorder and in controls. She found lower cortisol in the patients, and a narrower range of values. The opposite of the prediction, and the range detail is the one to keep.

The psychiatrist Kasia Kozlowska and colleagues described the fuller repertoire in their account of the defense cascade. That is the ordered sequence a body runs under threat, from arousal and fight or flight through freezing and on to collapse. Shutdown is one of the defenses, not the failure of defense.

So the same diagnosis produces opposite physiologies. A paradigm that reads a behavioral disorder as a lesion has to treat that as a subtyping problem, two diseases sharing a name. The Unified Model of Tone reads it as one finding. When a regulated range collapses, it can collapse toward either end.

Braced at the top, flooded with alarm, or pinned at the bottom, numb and offline. Both are the loss of the middle. Both are the same disorder of range, and the model would be in trouble if trauma only ever went one way.

How a temporary setting becomes the resting one

A nervous system that goes high and comes back down has lost nothing. Behavioral disorders begin when the return fails, and the mechanisms of that failure have names: allostatic load, critical periods, and the nightly maintenance of sleep.

The neuroscientist Bruce McEwen gave the answer its vocabulary. Allostasis is stability achieved through change, the body altering itself to meet a demand. It is exactly what a healthy system should do. His argument about allostatic load is that the same machinery becomes corrosive when the demand never lifts. The protective response, run continuously, becomes the damage. The cost is not in the surge. It is in the failure to shut the surge off.

That cost is visible in tissue. The neuroscientist Jason Radley put rats through repeated restraint stress and then examined the pyramidal neurons of the medial prefrontal cortex, the governor. He found apical dendrites that had retracted and reorganized, meaning the branching structures through which those cells gather their inputs had physically shrunk. The governor loses reach. A behavioral disorder does not require a lesion, and yet chronic loss of range leaves structural traces of its own.

The dose-response of early adversity

The human dose-response has been measured. The physician Vincent Felitti and colleagues mailed a questionnaire to more than thirteen thousand adult members of a health plan, and more than nine thousand answered. It asked about their childhoods across seven categories of adversity, covering abuse, violence against their mother, and a household marked by addiction, mental illness or imprisonment. They then compared those counts against adult health.

The Adverse Childhood Experiences study found a graded relationship. More categories of early adversity, worse adult outcomes, across mental health, behavior and physical disease alike. The psychologists Robin Nusslock and Gregory Miller later proposed a mechanism for the physical half of that finding. Early adversity, they argue, sets up a self-reinforcing loop between threat circuits and the immune system that carries forward across the lifespan.

Timing changes what an input becomes. The neuroscientist Takao Hensch studies critical periods, the windows in development during which a circuit is unusually open to being shaped by experience and after which it is far harder to reshape. His work on critical period plasticity shows that the windows themselves open and close under the control of local inhibition. The same experience delivered inside a window and outside it produces different nervous systems. Input meets tone, and tone has a calendar.

Sleep restores the range nightly

Then there is sleep, which is where a range is restored nightly. The neuroscientist Lulu Xie and colleagues measured the movement of fluid through the brains of sleeping and waking mice. They found that sleep drives metabolite clearance. The space between cells expands during sleep and waste products flush out far faster. Sleep does maintenance the waking brain cannot do.

Take it away and the brake fails first. The neuroscientists Seung-Schik Yoo and Matthew Walker kept healthy volunteers awake for a night and then showed them emotionally aversive images in a scanner. Compared with rested controls, the sleep-deprived brains showed a hyper-reactive amygdala that had lost its functional connection to the prefrontal cortex. Alarm amplified, governor disconnected, after one night.

And the effect is not the same for everyone. The neuroscientist Andrea Goldstein and colleagues deprived volunteers of sleep and then measured brain activity while people anticipated a possibly unpleasant image. Sleep loss increased anticipatory activity in the emotional circuits, and it did so most strongly in the volunteers who were already anxious by trait. The same lost night, a much larger effect on the person with less range to spare. Input meets tone, again, on a Tuesday, in every household with a teenager.

The damage tracks the lost range

If tone is range, the most informative number in behavioral science should be a measure of variation rather than a mean. Across attention, emotion and general physiology, it is.

Ask what the most reliable cognitive finding in attention deficit hyperactivity disorder is and most people will guess slowness. The psychiatrist Xavier Castellanos analyzed how response times are distributed across a task rather than merely averaged, and found markedly increased moment-to-moment variability. The characteristic pattern is mostly normal responses punctuated by occasional very slow ones. The average is dragged up by lapses. The lapses are the finding.

The psychologist Edmund Sonuga-Barke, working with Castellanos, proposed a mechanism. The brain has a default mode network, a set of regions that becomes active during rest and inward thought and normally quiets when a task begins. Their hypothesis is that this resting activity intrudes into task performance at intervals, producing exactly the periodic lapses observed. The system cannot hold one mode against the other. That is a range that will not hold a setting, described in the language of networks.

Stuck emotion predicts illness before it arrives

The same principle holds in emotion. The psychologist Peter Kuppens measured how strongly a person's emotional state at one moment predicts their state at the next moment, a quantity he called emotional inertia. High inertia means the state changes sluggishly and resists updating. Kuppens found that higher inertia tracked lower psychological well-being.

A later prospective study of adolescents took the measure forward in time and found that inertia predicted the onset of depressive disorder. Not which emotion, how stuck it was. The psychologists Todd Kashdan and Jonathan Rottenberg reviewed the wider literature and argued that psychological flexibility is a fundamental aspect of health. Their evidence is that the ability to shift approach as circumstances change predicts outcome better than any fixed strategy.

Then the general law, which comes from outside psychiatry entirely. The physicians Lewis Lipsitz and Ary Goldberger studied the fluctuations in physiological signals, heart rhythm, gait, hormone release, and asked what happens to them with aging and disease.

Their answer, published as the loss of complexity hypothesis, is that healthy physiology is irregular in a specific structured way, and that aging and illness make it more regular, more predictable, more monotonous. Goldberger later extended the analysis, showing that the fractal fluctuation patterns of healthy physiology break down with disease and aging.

Read that across the whole behavioral literature. Reduced heart rate variability. Absent habituation. Emotional inertia. Precision stuck too high to be adjusted. A cortisol range narrowed as well as lowered. Attention that cannot hold, and a fear response that cannot release. Every one of them is a loss of controlled variation, arriving under a different name in a different subfield. The model's claim is that they keep arriving because they are the same finding, and that the variable losing its range is tone.

Managing the output is not the same as widening the range

A medication holds one voice of the behavioral chord at a chosen value while it is present. Restoring regulation widens the range the system can move through. The two aims are different, both are legitimate, and the data can tell them apart.

A medication acts on the output. A stimulant raises the gain on the alerting systems, and attention improves while the drug is present. A selective serotonin reuptake inhibitor changes the availability of a transmitter, and the intrusive loop loosens for many people who take it. A benzodiazepine pushes inhibition up and the alarm quiets. These effects are real, reliable and, for a great many people, the difference between a functioning life and a lost one.

What a drug does, mechanically, is hold one voice of the chord at a chosen value for as long as it is present. That is a genuine and often necessary intervention. It is also, by its own design, a one-directional push.

Bellato's review noted that stimulants increased autonomic activity in the studies reviewed, which is the point. Alvares found that several psychotropics reduced heart rate variability further, which is also the point. The measure of range moved down while the symptom improved. Those two facts can both be true because they are answers to different questions.

Range is trainable, and one loop shows it

The question this model asks is different. Can the system get its range back, so that it moves to the setting the moment requires and returns on its own?

There is precedent for the idea that range is trainable rather than fixed. The psychologists Paul Lehrer and Richard Gevirtz set out the mechanism of heart rate variability biofeedback. A person breathes at a slow rate near the natural resonance of the baroreflex loop. Over sessions, the amplitude of their heart rate oscillations increases. The intervention does not clamp a value.

It exercises a loop until the loop swings wider. A meta-analysis led by the psychologist Vera Goessl pooled the trials and found a large reduction in self-reported stress and anxiety with the training. The trials behind that number are small, and the finding stands here as one illustration of a widening loop rather than proof of anything larger.

Hold the two aims apart and a lot of confusing clinical experience becomes readable. Masking manages the output while the intervention is present, and the output returns when it stops. Restoring widens the range the system can move through, and the symptom falls because the regulator has recovered its capacity to regulate.

The first is often necessary. The second is what the word healing was supposed to mean. They are complements, and most people are best served by having both available and by knowing which one they are receiving.

The prediction a drug cannot make

The model forecasts bidirectional restoration: a genuine tonal correction moves a dysregulated measure toward the healthy middle from either side. A one-directional intervention cannot produce that pattern, so the two mechanisms leave different fingerprints in the data.

That difference is what the split-group trial reads. The person whose measure sits too high should trend down. The person whose measure sits too low should trend up. Same intervention, opposite directions, converging on the middle.

An intervention that manages output cannot produce that pattern. It pushes one way by design, so everyone who receives it moves the same direction, and the person who was already low goes lower.

The split-group design

The test is straightforward to run. Take a measure of regulation that varies in both directions, such as heart rate variability, resting autonomic arousal, or the response-time variability Castellanos described in ADHD. Phenotype people before the intervention and split them into those who start high and those who start low on that measure. Deliver the intervention. Then look at the shape of the change.

Convergence toward the midpoint from both directions, with the spread narrowing, marks a restoration. A uniform shift in one direction, or no change in the low group, marks an intervention that manages the output and leaves the regulation where it was. Report the split, never the pooled average, because a pooled average across two groups moving toward each other is a null.

The same design answers the objection this evidence invites. Everything described here is a correlation. Reduced heart rate variability accompanies anxiety, and shrunken prefrontal dendrites accompany chronic stress, and none of that establishes which came first. Correlation is answered two ways in this model.

First by the coupled reading, which predicts that no single voice will ever behave as a clean cause because the voices are aspects of one organization. Second by the prediction above, which forecasts what a specific intervention will do to people who start on opposite sides of the healthy band.

Two objections, answered in the open

The first objection is that tone is a relabeling. Autonomic tone, allostasis, excitability, precision, set points, all of these already exist, and calling them tone adds a word rather than a discovery. The answer is that the contribution is the unification. Eight separate literatures which barely cite each other have independently converged on the same shape. An optimum with failure in both directions.

A baseline that determines what an input does. A loss of controlled variation that tracks the illness. Naming that shared shape as one regulated property, and giving it a measurable handle and a split-group test, is a claim about how the parts fit. Each component science remains its own. The synthesis is this model's, and it is stated as this model's.

The second objection is that a theory this broad predicts nothing. The answer is the split-group design above, plus an observation about ambition. A unifying model is supposed to unify. Judge it the way a unifying theory is judged. The unification is coherent across domains that were previously unrelated, it generates a prediction nobody was making before, and the model would rather be tested than admired.

Why the cause was hiding in the regulation

Five behavioral conditions, no test for any of them, heavy overlap, and a shared general factor nobody could name in physical terms. A collapsed range explains all four facts at once, because a range that collapses leaves no broken part behind.

The search for a cause has been a search for a lesion, meaning a broken part that could be pointed at. That search has genuine successes elsewhere in medicine, and it must always be run first. Some behavior changes come from a thyroid, a seizure disorder, a sleep apnea, a medication, a nutritional deficiency, a tumor or a genetic syndrome. Those causes are findable and they must be found, because finding them changes everything about what should happen next. Nothing here argues against looking.

In the great majority of cases the search comes back empty, and the empty result is read as a failure of detection. Read through tone, the empty result is the finding. A regulated range that has collapsed leaves no broken part behind, because nothing broke.

The parts are intact and the organization holding them together has lost its width. That is precisely the kind of disorder a search for lesions is built to miss. It is also the kind that would produce five names, endless overlap, a shared imaging address, and a single general factor underneath all of it.

Tone leaves tracks a checklist cannot see

The strength of reading behavior this way is that tone leaves tracks. Heart rate variability reads the flexibility of the vagal brake. Response-time variability reads whether attention can hold a setting. Habituation reads whether a harmless repetition is allowed to fade. Interoceptive accuracy, and the gap between accuracy and confidence, reads how the body is being interpreted. Sleep continuity reads whether the range gets restored at night. Each is a window onto the regulation itself, and each can be measured before and after anything.

A nervous system regaining its range shows it, and it shows it in ways a symptom checklist cannot see. Variability rises. The alarm still fires and then releases. A repeated sound stops arriving as news. Attention holds through a task and lets go at the end. The person still goes high when the day demands it, and comes back down when it does not.

The definitive question has not been answered. Nobody yet knows how much of the burden carried under these five names could be relieved by restoring regulation rather than managing its output. The split-group design above is the study that would ask it. What is already clear is where to look. A nervous system that can move to meet the moment and return to rest is more than a lower score on a checklist. It is a person with their range back.

How behavioral disorders relate to the rest of the library

Behavioral disorders form a condition page, and the readings it rests on each have a page of their own. The claims below are the specific connections, one per page.

• Tone is the pillar this page reads through: the organization the nervous system holds across its coupled state setters, and the range that organization can move through.
• Prediction owns the forecasting engine that fails three ways here, from precision stuck high in autism to the check that never closes in obsessive compulsive disorder.
• Gain carries the inverted U, the locus coeruleus modes, and the reason arousal can miss its optimum from either side.
• Time course explains why an alarm learned in a developmental window, or held for years, answers inputs differently than a fresh one.

• Spectrum disorders takes the precision and sensory-modulation evidence summarized here into a full account of the autism literature.
• Mental health holds the wider psychiatric picture, including the p factor and the transdiagnostic markers this page could only introduce.
• Trauma carries the defense cascade and the dissociative subtype in the depth they deserve, including what overmodulation means for care.
• Dysautonomia is where the accelerator and brake become the presenting problem rather than the background, the same autonomic machinery read from the body's side.
• Brain activity explains the imaging methods behind Goodkind's 193-study shared address and what a functional network finding can and cannot certify.
• And pediatrics holds the developmental half of this story, because critical periods mean most behavioral diagnoses are made while the windows Hensch described are still open.

Frequently asked