Spectrum Disorders and the Nervous System

Autism is diagnosed by observation because there is nothing to biopsy

No laboratory value identifies autism spectrum disorder. A clinician watches development: communication, attention, how sensation is met, how the social world is read, and writes the diagnosis from the pattern.

There is no blood test for autism. There is no scan that confirms it and no single gene that decides it. The word spectrum does honest work in the name. Autism runs along a continuum, from a light touch on daily life to a heavy one, with no natural line where it starts and no threshold a laboratory can draw.

That is a strange shape for a diagnosis. A condition common enough to touch most extended families has no lesion to point to and no number to bound it. The usual reading calls this a gap, a biomarker medicine has not found yet. There is another reading. A difference that lives in how the whole nervous system is tuned leaves nothing to biopsy, because a setting is not a part. Deciding between those two readings takes the machinery itself, starting with a single cell.

Every circuit balances an accelerator against a brake

A neuron is a living wire, and every circuit it joins runs on the contest between glutamate driving the next cell toward firing and GABA holding it back. The balance struck between them is the working state of the circuit.

A neuron carries a signal as a small electrical pulse from one end to the other. It does not touch the next cell. It reaches across a tiny gap called a synapse and releases a chemical messenger that either pushes the next neuron toward firing or quiets it. The main excitatory messenger in the brain is glutamate, the accelerator. The main inhibitory messenger is GABA, the brake.

Neuroscientists name the balance directly: the excitation to inhibition ratio, the E and the I. Turn the ratio toward excitation and the circuit grows loud and touchy, firing at signals it once let pass. Turn it toward inhibition and the circuit falls quiet. The ratio is the tone of the circuit, the setting at which it answers the world.

In 2003 John Rubenstein and Michael Merzenich proposed that some forms of autism arise from exactly this. The two neuroscientists pointed to an increased ratio of excitation to inhibition in the circuits that handle sensation, memory, emotion, and the reading of other people. Too much accelerator and too little brake in the systems that meet the world.

They argued the imbalance would come from many genetic and environmental variables landing on the same balance, no single gene among them. Their paper set the terms the field has argued over since, and it named, in plain physiological language, a setting rather than a broken part.

Hundreds of variants converge on one shared setting

The search for the autism gene found hundreds of them, and the way it found them is the finding. The variants scatter across the genome and converge on one job: building and tuning synapses.

Medicine asked of autism the question it asks of any disease. Where is the cause. For decades that meant hunting the part at fault, a gene, a region, a molecule that could carry the whole condition. The neurogeneticist Daniel Geschwind spent his career on that map, and his survey of the field reports what the hunt turned up: hundreds of rare variants, each accounting for a sliver of cases.

What unites them is a job rather than a location. They converge on the machinery that builds and tunes synapses and regulates how genes are read. They converge, in other words, on the machinery that sets the balance of a circuit.

Hundreds of separate doorways open onto the same room. The cause of autism was never one part. It is one setting, reached by many doors.

One boundary belongs in this section as fact. A share of autism does have a findable single cause. Fragile X syndrome, a specific heritable change in one gene, is the most common, and there are others. Those causes are testable and finding them matters, because each carries medical needs of its own.

In the large remainder, no single part is at fault, because hundreds of small pushes have moved one shared setting. A lesion hunt is not built to see that pattern, and naming it requires a word for the setting.

Tone is the setting the autism genetics points at

The Unified Model of Tone names the property the nervous system adjusts to hold its regulated variables: tone, the integrated organization of the body's coupled state, taken as one whole rather than any single part.

Medicine already accepts that the body guards temperature, blood sugar, and blood acidity inside narrow windows, and that illness is any of them drifting out and failing to return. The nervous system holds harder things the same way: the strength of a reflex, the reach of attention, the trust placed in a sense. Tone is the one setting behind all of that holding. In the nervous system, tone shows as gain and as precision.

Gain is how strongly a circuit answers its input, the E to I ratio written large across whole systems. Precision is how much weight the brain places on a signal. A nervous system in healthy tone runs its gain and precision high when the moment rewards it, low when it does not, and moves freely between.

Health is that freedom of movement, the width of the range. Tone that is held at one value, unable to move when the moment asks, is what manifests as difficulty and, further out, as disease.

Read this way, a spectrum profile is a setting held high and held steady rather than a part that failed. The genetics literature made the same correction. Reviewing how the many autism risk variants act, Sacha Nelson and Vera Valakh in 2015 read the disturbance as a failure of the brain's own set-point regulation.

A fixed shift in the accelerator does not fit the data. The circuits that should re-tune themselves toward balance keep settling on a raised target and holding there. The fault sits in the regulating, not in a part.

The components here are established and credited: excitability, autonomic balance, set-point regulation each have their own literatures. The model's claim is the unification, that one organizing setting runs through every scale of the nervous system and links what look like separate problems. Autism, with its scattered causes and its whole-body reach, is the most demanding place to test that claim.

The brake begins life as an accelerator

GABA, the adult brain's main quieting signal, excites the developing brain, and it flips to a brake around birth. In two rodent models of autism, that flip failed.

The flip is real chemistry. The amount of chloride sitting inside a cell decides whether GABA quiets it or drives it. Early in development chloride runs high, and GABA excites. As the brain matures, the cell pumps chloride out, and the same GABA signal now inhibits. That handover is one of the most important switches in the building of a brain. A brain whose brake never comes online runs with the accelerator held down.

A team led by Roman Tyzio asked whether the switch fails in autism, working in two separate rodent strains bred to model it. One strain was built on a drug exposure in the womb and the other on the Fragile X gene. In both models the chloride switch failed: GABA stayed excitatory when it should have become inhibitory, and the young brains ran hot.

Giving the mothers bumetanide, a drug that lowers chloride inside the cell and lets GABA brake again, moved the offspring's brain activity and behavior back toward typical. Blocking the mother's own oxytocin, the birth hormone that helps drive the switch, produced the autistic-like pattern in otherwise healthy pups. The 2014 result reads as a failure of the developmental GABA switch, and as a setting that could be moved.

The scope of the finding is rodent work in bred models, and the human bumetanide trials that followed have run mixed. The model expects that shape of result: a drug meets each child's tone, and identical inputs land differently on differently organized nervous systems, so pooled averages blur what individual regulation determines.

What the experiment establishes is simpler and stands. The same signal can act as accelerator or brake depending on how the cell is tuned, the tuning can miss its mark, and a missed setting can in principle be moved back. That is tone in a single circuit, and it is the whole argument about autism in miniature.

Autism shows in many systems at once because they are coupled

A single broken part would show as a single broken function. Autism shows instead as a pattern running through sensing, movement, face and voice reading, sleep, the gut, and the steadiness of the heart, all at once.

That reach is the standing puzzle for a lesion account and the plainest evidence for a tonal one. The nervous system is a set of coupled systems, rhythms held in step: the fast electrical traffic of thought riding on the slower swells of breath, heartbeat, hormone, and the daily clock. Tone is the setting shared across that coupling, one organization expressed in many forms at once. When the setting runs high, it runs high across the coupled whole.

The model's canon analogy carries this exactly. Tone is the chord, not the notes. A spectrum profile is the chord these coupled systems sound together at one high, steady setting. Health is a chord free to change key as the moment asks.

The consequence for causation is large. If the profile is a chord, its origin need not live in any single voice. It can live in the tuning of the whole, in how the voices are set and how freely they move together. That is a kind of cause the search for a single part cannot find. In autism it runs through the predicting brain, the sensory circuits, the felt body, the heart, and the gut.

Perception is a prediction corrected by the senses

The brain does not receive the world like a camera. It predicts the world, checks the guess against what the senses report, and treats the disagreement, the prediction error, as the signal worth processing.

At every instant the brain holds a model of what is about to happen, the next sound, the next touch, the next look on a face, and runs the prediction forward. Where guess and report disagree, the difference is a prediction error, a small alarm that the model needs updating. Perception is the settling of that argument. Nobody sees the world raw; everyone sees the brain's best guess, corrected by the parts the senses insist on.

Karl Friston built the most complete version of this account, the free-energy principle, published in 2010. The brain works to minimize surprise, keeping the gap between prediction and sensation small. The idea that turns this into physiology, and into the reading of autism that follows, is weighting. The brain does not treat every signal as equally trustworthy.

It assigns each one a precision, a measure of how much to trust that signal against its own prediction. High precision on a sense means let it overrule the guess. High precision on the prediction means trust the model and let the noise pass. Where the brain sets that weighting decides what reaches awareness and what is smoothed away.

Autistic perception weights the senses too heavily

Three research groups, working separately between 2012 and 2014, arrived at the same reading of autism: the setting that trusts the senses over the brain's predictions is turned up and held there.

Elizabeth Pellicano, working with the vision scientist David Burr, put it in a phrase that carries the mechanism. For the autistic brain, they argued in 2012, the world becomes too real. Perception normally leans on prior expectation to smooth the flood of sensation, so a familiar scene is mostly guessed and lightly checked. Weaken the pull of the prior and weight the incoming evidence too heavily, and the smoothing fails. Every detail arrives at full strength, unsoftened by expectation.

Rebecca Lawson, with Geraint Rees and Karl Friston, named the mechanism directly in 2014: an aberrant precision in autism, an imbalance in the weight given to sensory evidence against prior belief. Prediction errors are treated as too important to ignore. A brain that cannot discount its own error signals must attend to everything the senses raise.

Sander Van de Cruys and colleagues drew the strengths and the costs from the single setting in the same year. In their account autism carries an inflexibly high precision of prediction errors, and that one setting spans the profile. It gives the fine discrimination, the eye for the exact detail others miss, the intolerance of a pattern that does not quite hold.

It also gives the overload, because a system that treats every error as urgent gets no rest from the world. One setting, turned up and hard to turn down, and the strengths and the difficulties fall out of it together.

Hyper-reactive circuits make the world arrive too strongly

One level below the precision weighting sits the raw excitability of the circuits, and the Markrams measured it: local cortical circuits in an autism model that answer too strongly and rewire too readily.

Henry and Kamila Markram, who build detailed models of cortical circuits, studied a rat strain carrying autistic-like traits, expecting to find circuits that answered too weakly. They found the reverse. Their intense world theory, published in 2010, describes local circuits that are hyper-reactive and hyper-plastic. A circuit tuned this way perceives more, attends more, remembers more, and feels more. A world met by such circuits runs painfully bright and loud and sharp, and the Markrams read autistic withdrawal as a sane response to that intensity.

This is the E to I ratio of the second section, felt from the inside. It also explains a fact that puzzles families: the same child can be flooded by a sound yet slow to register their own name. A system running near its ceiling has little range left to answer with.

Where it is already saturated it cannot rise further, so a strong input can meet a strangely flat response. Over-responsive and under-responsive sensory profiles are the two edges of one narrowed range, the signature of a setting with nowhere left to move.

Interoception is an input, and autism shifts its weighting

The sensory setting reaches inward, to interoception, the felt report of heartbeat, breath, fullness, and the low weather of the gut, and through that channel it reaches other people.

The neuroanatomist A.D. Craig mapped the pathways, tracing in 2002 how the body's inner condition reaches the brain and gathers in a deep fold of cortex called the insula into the felt sense of being a body. Interoception is the raw material of emotion.

A feeling is in large part a reading of the body, the quickened heart and tight breath the brain interprets as fear or excitement. This is input quality in the model's terms: the fidelity and the weighting of what the body reports about itself. Shift the weighting on those inner signals and the reading of one's own state shifts, which shifts every feeling built on it.

One account tied the bodily sense and the social sense to a single control. Working with Karl Friston, E. Quattrocki proposed in 2014 that oxytocin, the hormone woven through birth, bonding, and trust, works in part by tuning the precision of interoception.

The same setting that governs how clearly a person senses their own body governs how they sense another person, because reading another mind runs on the same interoceptive machinery turned outward. A shift in that one setting shows as both at once, a body sensed differently and a social world sensed differently, from a single root. In autism the two do arrive together, which is what one shared weighting predicts.

The resting heart reads the autism setting

The body's oldest control system carries the same balance as the cortex: a sympathetic accelerator and a vagal brake. In autistic children, the measurable swing of that brake runs low.

The autonomic nervous system runs the organs without asking permission. Its sympathetic branch speeds the heart, sharpens the senses, and readies the body to act. Its parasympathetic branch, carried mostly by the vagus nerve, slows the heart and turns the body toward rest and repair.

The neurologist Eduardo Benarroch described in 1993 how a central autonomic network governs both, running from the insula and cingulate cortex down to the hypothalamus and brainstem. The same insula that carries the felt body governs the heart. Sensing, feeling, and autonomic control share an address.

Heart rate variability opens the window

A healthy heart does not beat like a metronome. The gap between beats shifts as the vagal brake is applied and eased, and that beat-to-beat flexibility is heart rate variability.

Julian Thayer built a model of neurovisceral integration around it in 2000: the variability reflects how well the higher brain and the heart are coupled, and the flexibility itself is the mark of health. Heart rate variability is a validated index of autonomic state. Reading it as a window onto tone is the model's interpretation of that real and repeatable signal.

Opened on autistic children, the window shows what the setting predicts. In 2020 a team led by Catharine Lory measured resting heart rate variability in children with autism and found it significantly lower than in other children. The lower it ran, the more atypical the child's attentional response to sensation. A weaker brake, a body held closer to arousal, and a sensory world met with less room to steady.

Stephen Porges's polyvagal theory, published in 2001, gives the wider frame, linking the strength of the vagal brake to the capacity for calm social engagement. The autonomic indices are the measured fact. Reading the vagal brake as the bodily ground of social ease is the interpretive lens the model shares, and it ties the steadiness of an autistic child's heart to the reach toward another person.

The insula writes and reads the body's immune state

Digestive trouble, disrupted sleep, and a braced nervous system trail autism so consistently that a purely mental account cannot place them. A 2021 experiment shows the loop that holds them together.

A team led by Tamar Koren, in the laboratory of the neuroimmunologist Asya Rolls, marked the neurons active in the insula while a mouse had gut inflammation, then later switched those exact neurons back on. The insula neurons had stored the immune state and could recall it, driving the inflammation to return on command. The brain holds a representation of the body's inner condition, including its immune weather, and it writes to that condition as well as reading it.

This is coupling with the volume turned up. The felt body, the autonomic balance, the state of the gut, and the tone of the immune system are wired into one loop through the insula, each shaping and shaped by the rest.

A setting held high across that loop shows at once as sensory intensity, a strained autonomic balance, a reactive gut, and an immune system that runs hot. The conditions that travel with autism are the same chord, sounding in the organs that share the coupling.

A braced nervous system defends its own high setting

A setting turned high once would matter little if the system relaxed it. The nervous system learns its settings and then defends them, and in autism the defense is the persistence.

A brain that has run at high gain and high precision comes to expect the world it built at that setting, and it braces to meet it. Bruce McEwen named the cost in 1998: allostatic load, the wear that gathers when stability is bought through constant, expensive compensation. Friston's predicting brain explains why the bracing persists. A system whose model expects intensity spends its resources preparing for intensity, and that preparation has a physiology: a heart held near arousal and senses held near saturation.

This is where the model explains what an average cannot. The same noisy classroom lands lightly on one child and floods another, because the input does not act alone. It meets a nervous system already tuned a particular way, and the outcome belongs to the meeting.

Two children in the same room are not having the same event. The high setting in autism is one the body is actively holding, for reasons written early and defended since, which is why it behaves like a set point and not like a passing state.

Widening the range differs from damping the system

Two approaches move a nervous system running hot, and both appear in autism support. One damps the output, one restores the range, and the two differ even when the surface calms by the same amount.

A sedating drug lowers arousal by damping the whole system. It quiets the accelerator across the board, reliably and sometimes necessarily, and it pushes in one direction whether or not the underlying regulation has changed. For a child in genuine distress that relief has a real place.

Restoring tone aims at something else: widening the range the system can move through. The same child can then meet a loud room and come down from it, rise to a demand and then rest. The surface may look similar on a calm afternoon. The difference shows in the range that has been returned.

The drug-free approaches to autonomic regulation carry mixed trial evidence. The model reads that mixture the way it reads every input: an approach meets each child's tone, and group averages flatten what individual regulation decides. The model's own claim is narrower and testable. Whatever genuinely helps a dysregulated nervous system will work by widening its range rather than pinning it at one value, and the next section states the test that separates the two.

Restored tone moves both sensory edges toward the middle

A model that explains an over-responsive child and an under-responsive child owes a prediction that separates restoring regulation from masking a behavior. The Unified Model of Tone stakes one, and no sedative can imitate it.

The claim is bidirectional restoration. A correction that genuinely restores tone moves a dysregulated system toward the healthy middle from either side. The child whose sensory gain runs high trends down toward ease. The child whose response runs flat, slow to register the world, trends up toward engagement. What is restored is the capacity to reach the middle. A sedative does the opposite by design, lowering arousal in both children alike, because it overrides the regulator rather than returning it.

Restore the tone and different children move toward one center from opposite sides. Mask it and everyone slides the same way. That divergence is the signature, and it is measurable.

The test is straightforward to state. Take children who begin over-responsive and children who begin under-responsive on the same measured setting, whether sensory response, heart rate variability, or the balance of the autonomic branches. Apply an approach that aims to restore regulation, and watch which way each group moves. Convergence toward the middle from both sides confirms the claim.

A uniform shift in one direction marks the approach as a push on the output, helping whichever group it points at and carrying the other group further from the middle. That contrast is the sharpest discrimination the model draws between restoring and masking. Autism, with its two directions of sensory dysregulation and its growing set of autonomic measures, is a place the test can be run.

Autism has no biomarker because regulation leaves no lesion

The missing cause has an address. Autism is a difference of regulation rather than of structure, and regulation leaves nothing to biopsy, nothing to resect, nothing for a blood panel to catch.

The cause was never too subtle to see. It was the wrong category of thing to look for: a change in how a coupled system is tuned rather than a break in one of its parts. Read this way, the scattered facts of autism gather into one. The genetics runs to hundreds of variants because they are hundreds of doorways onto a single setting.

The condition is a spectrum because a setting has a range and every value along it is possible. The sensory intensity, the social difference, the strained autonomic balance, and the reactive gut appear together because they are one setting read across a coupled body. And the same input floods one child and not another because an input meets each child's tone, and the outcome belongs to that meeting.

Two boundaries hold, stated as fact. A share of autism has a specific findable cause, Fragile X above all, and searching for it is not optional, because those conditions carry needs of their own. And the model addresses dysregulation and distress, the sensory overload and autonomic strain that can make a life harder, never a person's identity or worth.

Many autistic people know their perception as difference and as strength, and the precise, detailed mind the intense-world and precise-minds accounts describe is real. What the Unified Model of Tone explains is the dysregulated edge of the profile, as a setting of tone that can in principle be moved toward ease, and it is built to be tested.

How autism relates to the rest of the library

Autism is where the library's account of regulation faces its hardest case, a condition with no lesion, and the neighboring pages carry the pieces this one builds on.

Behavioral disorders is the umbrella: every diagnosis made by observed behavior, ADHD and anxiety alongside autism, meets the same absence of biomarkers. That page carries the precision evidence at the level of the whole family.

• Pediatrics holds the developmental windows in which the settings behind autism are first written.
• The autonomic nervous system is the anatomy behind section eleven, the accelerator, the vagal brake, and the network that governs them.
• Heart rate variability is the instrument the 2020 Lory finding was made with, and its page works through what a personal number does and does not mean.
• Gut health follows the insula loop into the digestive tract, where the enteric nervous system runs its own regulation.
• Anxiety is the braced, high-arousal state of section thirteen experienced from the inside, in autistic and non-autistic nervous systems alike.
• Mental health examines what diagnosis by behavior can and cannot say about a nervous system, the question autism poses most sharply.

Frequently asked