The Autonomic Nervous System: Sympathetic and Parasympathetic Control Explained

The autonomic nervous system acts before you decide

The autonomic nervous system runs a full-body alarm and a full-body recovery without asking you. A horn goes off behind you at a light, and your heart has already jumped before any thought arrives.

Notice how much happened in that half second without your permission. Your heart sped up. Your pupils widened. The small vessels in your skin narrowed, which is why a real fright leaves you pale. Your palms went damp. Your gut, which had been quietly working on lunch, stopped working on lunch. You did not decide any of it.

Then notice the second half of the moment, which almost nobody notices. A minute later you are fine. The heart has come back down. The hands are dry. The gut has resumed. You did not decide that either.

Both halves of that minute are the subject of this page. One nervous system inside you ran the alarm and then ran the recovery, and it did both while you were busy being annoyed at the driver behind you.

The simplest way to hold the idea is this. Your nervous system does two kinds of work. There is the part you command, which moves your hand when you decide to move it. And there is the part that runs itself, which handles the heart, the vessels, the gut, the sweat glands, the pupil, the bladder and the airways. That second part is the autonomic nervous system. The word autonomic shares a root with autonomous, meaning self governing, and it was chosen deliberately.

What the autonomic nervous system actually is

The autonomic nervous system regulates the internal organs, the blood vessels, the glands and the smooth muscle, without conscious command. Its three divisions are sympathetic, parasympathetic and enteric.

Its output reaches heart rate and the force of each beat, blood pressure, the calibre of the airways, digestion and gut movement. It also reaches sweating, the size of the pupil, the bladder, sexual function, body temperature and the activity of immune cells.

The sympathetic division mobilizes. The parasympathetic division restores. The enteric division sits inside the wall of the gut and runs digestion largely by itself. The neurologist Eugene Benarroch has spent a career mapping how the brain controls the organs. He sets out this three-part organization and its clinical consequences in a 2020 overview written for practicing neurologists.

One more thing belongs in the definition, and most descriptions leave it out. The autonomic nervous system is not only an output. A large fraction of the traffic runs the other way, upward, carrying reports from the organs to the brain. This system is a conversation between the brain and the body, and the body does much of the talking. That fact arrives in full in section eleven, and it changes what the whole system is for.

Every autonomic path is a two-neuron relay

The first cell hands its message to a second inside a ganglion, and that single handover is why autonomic responses arrive as patterns across the body rather than as one muscle twitching.

Start with what a nerve cell is. A neuron is a living wire. It has a body, a long fiber, and an end that releases a chemical onto whatever it is speaking to. The chemical is the message. What receives it decides what the message means.

When your brain tells a muscle in your arm to contract, the wiring is direct. One nerve cell runs from the spinal cord all the way to the muscle. One hop, one cell, one target.

The autonomic nervous system is built differently. The first cell leaves the spinal cord or the brainstem and travels a short distance to a relay station called a ganglion, which is a cluster of nerve cell bodies sitting outside the cord. There it hands the message to a second cell. That second cell carries it the rest of the way to the organ.

How Langley found the handover, and named the system

We know this because of John Newport Langley, a Cambridge physiologist who spent the 1890s working out where autonomic signals change hands. He had a clever tool. Nicotine, applied locally, stimulates and then blocks the cell bodies inside a ganglion while leaving fibers that merely pass through untouched.

So if painting a ganglion with nicotine abolished a response, the relay had to be in that ganglion. If the response survived, the fibers were only passing through. Working through the neck in his 1898 study of how cranial autonomic fibers join the cells of the superior cervical ganglion, he mapped where the handovers happen and gave the field the term autonomic.

A switchboard does something a direct wire cannot. One incoming fiber can contact many outgoing cells, so a single instruction from the cord is distributed across a whole region at once. That is the anatomical reason autonomic output is organized rather than local.

The sympathetic division mobilizes the body

Its fibers leave the spinal cord from the chest and low back, and at the organ they speak mostly through noradrenaline.

Anatomists call that exit the thoracolumbar outflow, from thoracic, meaning chest, and lumbar, meaning low back. The fibers arrive in a chain of ganglia running down each side of the spine like a string of beads. From there the second cells fan out to almost every organ, where they release noradrenaline, also called norepinephrine. The word means next to adrenaline, and the two molecules are close relatives.

What follows is the pattern you already know from the car horn. The heart beats faster and each beat is stronger. The pupil widens. Blood is redirected away from the gut and toward working muscle. Sweating rises. Digestion is put on hold. The airways widen as well, though that one works differently. Human airway muscle carries very few sympathetic fibers, and the widening comes mostly from adrenaline in the blood. None of this is damage. It is a body preparing for effort.

The adrenal gland makes the surge outlast the alarm

One piece of sympathetic anatomy behaves unlike the rest. Sitting on top of each kidney is the adrenal gland, and the core of that gland is a modified sympathetic ganglion. Its cells never grew the long outgoing fiber. Instead of releasing their chemical onto one organ, they release adrenaline directly into the bloodstream.

That single design difference explains something you have felt. Nerve traffic stops the instant the brainstem stops sending it. A hormone in the blood has to be cleared, which takes minutes. This is why the shaky feeling after a near miss in traffic outlasts the near miss by a long way. The nerves have already gone quiet. The chemical has not.

Sweat glands break the transmitter rule

Textbooks teach that sympathetic means noradrenaline. Sweat glands are the exception. They are supplied by sympathetic nerves, and those nerves release acetylcholine, the transmitter you are about to meet on the parasympathetic side. A 2018 review of how nerves control human sweat secretion sets out this cholinergic sympathetic supply and how the gland responds to it.

This exception carries weight. Every skin conductance measurement, from a laboratory recording to a lie detector, rests on it. Those devices read sweat, sweat is driven by sympathetic nerves, and so a sweat measure is a sympathetic measure that says nothing about what the vagus is doing to your heart.

The parasympathetic division restores the body

Its fibers leave from the brainstem and the sacrum and speak through acetylcholine, and its great cable is the vagus nerve.

The classical name for that arrangement is the craniosacral outflow. The brainstem sits at the base of the skull where the brain becomes the cord, and the sacral fibers leave from the very bottom of the spine. At the organ, the second cell releases acetylcholine.

The effects run opposite to the sympathetic accelerator in most places. The heart slows. The airways narrow. The pupil constricts. Digestion and glandular secretion increase. Blood returns to the gut.

The vagus nerve carries most of this traffic. The name is Latin for wandering, and it earns it. The vagus leaves the medulla, which is the lowest part of the brainstem, then travels down through the neck and chest into the abdomen. It reaches the heart, the lungs and most of the digestive tract. When popular writing says parasympathetic, it usually means vagus.

The craniosacral picture is under genuine challenge

An argument about parasympathetic anatomy is still running in the literature, and it shows how anatomy actually gets settled.

A team of developmental biologists in Paris led by Isabelle Espinosa-Medina and Jean-Francois Brunet asked a question nobody had asked cleanly. Cells carry molecular signatures during development that mark which lineage they belong to.

So do the sacral autonomic neurons carry the parasympathetic signature or the sympathetic one? They compared 15 developmental and phenotypic features in mice and reported in 2016 that by every one of those features the sacral outflow matched the thoracolumbar sympathetic outflow. On their reading the parasympathetic outflow is cranial only, and the sympathetic outflow runs from the thoracic cord all the way down.

The claim was contested immediately. The neurophysiologist John Horn replied in 2018 with a paper arguing that Langley had it right the first time. His case is that the sacral outflow behaves parasympathetically in function, and that developmental markers are not the right arbiter. Textbooks still teach craniosacral.

State the position plainly. The developmental evidence is real and the functional objection is real, and the question is open. The vagus, the brainstem origin of the cranial outflow and acetylcholine as the transmitter are settled. Nothing else on this page depends on how the sacral argument resolves.

The enteric division runs digestion on its own

The enteric division is the third part of the autonomic nervous system, and it runs digestion with its own complete circuits. It is the cleanest demonstration that regulation is distributed rather than owned by the brain.

The wall of your intestine contains its own nervous system. These are complete circuits rather than stray endings from the brain. Sensory cells detect stretch and chemistry, interneurons process what those cells report, and motor neurons drive muscle and secretion. Hundreds of millions of neurons, arranged in networks, sitting in a tube.

The first hard evidence came in 1899 from two physiologists at University College London. William Bayliss and Ernest Starling were studying how the intestine moves food along and wanted to know whether the brain organized the movement. The intestine did it itself.

Distend the gut at a point and the muscle above the point contracts while the muscle below relaxes, propelling the contents forward. They published this as a study of the movements and innervation of the small intestine, and the rule became known as the law of the intestine.

A century of work confirmed the gut's independence

The Melbourne physiologist John Furness has spent his career on the gut's own circuitry. In a 2012 review of the enteric nervous system he sets out two findings. The extent of local versus central control differs markedly along the digestive tract. And the enteric system generates and coordinates the movements of the intestine through its own reflex pathways. Sever the connections to the brain and much of digestion continues.

The gut has complete reflex circuits of its own. Much of digestion continues when its connections to the brain are cut.

This is why the Unified Model of Tone does not call the nervous system the body's master regulator. Regulation is distributed. It is happening in the gut wall, in the smooth muscle of every vessel, in the lining of the airways, in immune cells and in the kidney.

What the nervous system does is integrate that distributed regulation at the highest density of any tissue in the body. It is the highest-density integrator of a regulation that does not belong to it alone. The enteric division demonstrates that distinction anatomically, which is why it is taught here rather than in a footnote.

Sympathetic and parasympathetic are not a seesaw

The single-dial picture fails in three specific ways, each with an experiment behind it.

Almost every popular account of the autonomic nervous system draws one dial. Slide it toward fight and flight, or slide it toward rest and digest. High sympathetic means low parasympathetic. The picture is memorable and the physiology does not support it.

First, the branches are two axes, not one

In 1991 three psychophysiologists, Gary Berntson, John Cacioppo and Karen Quigley, examined why autonomic studies kept producing results that made no sense on the seesaw model. Their answer was that the model has the wrong geometry. The two branches do not lie along a single line from parasympathetic to sympathetic dominance. They distribute in a two-dimensional space, and the authors named this the doctrine of autonomic space.

The practical consequence is large. Both branches can rise together, which is coactivation. Both can fall together, which is coinhibition. One can move while the other holds still. On a single dial none of these is possible. In a two-dimensional space they are ordinary. Their paper showed the model accounts for much of the error variance that had plagued the field.

Second, the diving response does both at once, on purpose

Put your face in cold water and hold your breath. Your heart slows sharply, which is a parasympathetic effect. At the same time the vessels in your limbs clamp down, which is a sympathetic effect. The neuroscientist W Michael Panneton studies the brainstem circuitry behind this. In a 2013 review of the mammalian diving response he describes the pattern of apnoea, slowed heart and vasoconstriction as shared across vertebrates and neurally mediated. It overrides ordinary homeostatic reflexes.

That combination is the correct answer to a specific problem. Slow the heart to save oxygen, and squeeze the peripheral vessels to keep the remaining blood going to the brain. A seesaw cannot produce it. A two-axis system produces it easily.

The physiologist Julian Paton and colleagues reviewed how often this happens in a 2005 survey of vagal and sympathetic interactions at the heart. Coactivation occurs during chemoreceptor reflexes, diving, the oculocardiac reflex and painful stimulation. Even at rest, the heart receives ongoing drive from both. They also argue that simultaneous coactivation can be more efficient than sympathetic drive alone, giving longer filling time with a stronger contraction.

Third, sympathetic outflow is not one signal

There is no single sympathetic volume knob. This is the correction that surprises people most.

The physiologist Shaun Morrison spent years recording from separate sympathetic nerves at the same time in animals, asking whether they moved together. They did not. In a 2001 review of the differential control of sympathetic outflow he sets out the resulting picture: an extensive array of functionally specific output channels that can be activated or inhibited simultaneously in different combinations.

The outflow to brown fat, to skin vessels, to muscle vessels and to the kidney are separately controlled. Cooling an animal drives the channel to brown fat up while other channels do something else entirely.

So there is no such thing as being sympathetic. There is a pattern of drive across many channels, and that pattern is the state. The cardiologist Dwain Eckberg made the same argument from the human side in a 1997 critical appraisal of the term sympathovagal balance. The assumptions behind the phrase do not hold up when the underlying physiology is examined.

Most blood vessels receive sympathetic supply only

One last piece of anatomy dissolves the seesaw for good. Most blood vessels receive sympathetic supply only. No parasympathetic nerve waits to widen them. So how does a vessel relax?

The sympathetic nerve is always firing at some rate. That ongoing traffic is called resting sympathetic tone, and vessel calibre is set by turning that single drive up or down from its standing baseline. Relaxation is a reduction in an ongoing signal rather than the flipping of a switch.

Gunnar Wallin and Nisha Charkoudian reviewed four decades of direct recording from human sympathetic nerves in a 2007 survey of the method and its findings. The resting level varies widely between healthy people, and within a person it is a stable individual characteristic. Sweat glands work the same way, from a baseline that rises and falls.

Hold onto that. A high resting sympathetic level in a healthy person is not a diagnosis. It is one number in a pattern, and the pattern carries the meaning.

The central autonomic network, not a command post

No single autonomic center composes the pattern. Autonomic output is assembled by a network of brain regions that Eugene Benarroch named in 1993.

Benarroch, a neurologist at the Mayo Clinic, assembled the evidence and called it the central autonomic network. His argument was that autonomic control is carried out by a set of reciprocally interconnected brain regions rather than by a command post. The regions he named are worth setting out plainly, together with the cingulate cortex his later review adds, because each is a place where something enters the machinery.

The word reciprocally is doing real work in Benarroch's description. Each level sends and receives. That is why a memory, a posture, a difficult conversation and a signal from the gut all enter the same machinery, and why each can change the same autonomic output.

Stimulating the insula moves the heart

There is a striking human demonstration at the cortical level. During surgery for epilepsy, patients are sometimes awake while parts of the cortex are electrically stimulated to map function. The neurologist Stephen Oppenheimer and colleagues stimulated the insula in five such patients and recorded changes in heart rate and blood pressure from cortical stimulation in 1992. Stimulating a small patch of cortex moved the heart.

They also reported a difference between the sides, with right-sided stimulation more often producing acceleration. Read the side difference cautiously, because five patients is a small series and later work has complicated it. The main finding stands. A patch of thinking cortex has a line to the heart, which is how prediction and meaning reach autonomic output.

The brainstem patch that generates resting sympathetic drive

Resting sympathetic drive is generated by a small population of neurons in the rostral ventrolateral medulla, and silencing them collapses blood pressure.

Rostral ventrolateral medulla means the upper front-and-side portion of the lowest part of the brainstem. It is usually abbreviated to RVLM.

In 1984 a group at Cornell led by Christopher Ross and Donald Reis set out to find where in the brainstem blood pressure is generated. They stimulated the region electrically and chemically in anaesthetised rats and recorded what happened.

In their report on tonic vasomotor control by the rostral ventrolateral medulla, stimulation raised arterial pressure by more than 80 millimetres of mercury and accelerated the heart. Applying an inhibitory chemical to the same spot dropped pressure. Blocking the neurons with a nerve toxin collapsed pressure to a level that cutting the spinal cord could not lower further.

That last detail is the important one. If silencing this patch produces the same pressure as disconnecting the brain from the body entirely, then this patch was supplying essentially all of the brain's ongoing contribution to blood pressure.

The RVLM is an integrator, not a panic button

The physiologist Patrice Guyenet spent decades working out what the RVLM does with its inputs. His synthesis in a 2006 review of the sympathetic control of blood pressure describes these neurons as a common final relay. Pressure sensors, oxygen sensors, and descending signals from the forebrain all converge here, and what leaves is the resolved answer.

These neurons fire even at rest, held in check by continuous inhibition arriving from the pressure-sensing pathway. Your resting blood pressure is not the absence of autonomic activity. It is a negotiated equilibrium between drive and restraint, recalculated continuously. That equilibrium is what the tone model calls a set point, maintained by the autonomic nervous system rather than by any organ alone.

The baroreflex corrects blood pressure within a heartbeat

The baroreflex is the fastest autonomic loop you can watch, and it lifts the vagal brake on your heart within a single beat when you stand up.

In the wall of the carotid arteries in your neck and in the arch of the aorta above your heart there are stretch sensors. They do not measure pressure directly. They measure how much the vessel wall is stretched, which rises and falls with pressure. Their fibers run to the nucleus of the solitary tract in the medulla, and they report continuously.

When you stand up, gravity pulls roughly half a litre of blood into your legs within seconds. Pressure at the neck falls. The sensors stretch less and their firing drops. The medulla receives less inhibition, and the answer comes back in two stages with different speeds.

The vagal brake lifts within a single heartbeat, so the heart is already faster on the next beat. Sympathetic drive to the vessels rises more slowly, taking several seconds to build the narrowing that holds the pressure. Both limbs are working before you have finished standing.

Level and sensitivity are two different properties

Two features of this loop matter for everything that follows.

First, it corrects in both directions by design. Low pressure calls for more sympathetic drive. High pressure calls for less. The same loop handles both, without any change in the machinery. A regulator that only pushes one way is not a regulator.

Second, the responsiveness of the loop is a separate property from the level of the pressure. Clinicians measure it as baroreflex sensitivity, which is how much the interval between heartbeats changes for each unit of pressure change. You can have a normal blood pressure with a sluggish loop, or a high pressure with a brisk one. The level tells you where the autonomic nervous system is sitting. The sensitivity tells you how well it answers.

Baroreflex failure leaves pressure produced, not regulated

What happens when the loop fails is the clearest possible proof of what it was doing. The clinical pharmacologist David Robertson and colleagues at Vanderbilt described eleven patients whose baroreflex had been destroyed. The causes differed from patient to patient: familial paraganglioma syndrome, neck surgery or radiation for pharyngeal carcinoma, lesions of the nucleus of the solitary tract, and surgical division of the glossopharyngeal nerves.

In two the cause was never found. They set out what happens to those patients in a 1993 report on the diagnosis and treatment of baroreflex failure. These patients swing violently between hypertensive crisis and hypotension, and emotion or mental effort provokes responses far larger than they should be, because nothing is reporting the result back. The pressure is not regulated. It is merely produced.

The vagus is mostly a listening nerve

The vagus is predominantly sensory. Most of its fibers carry reports from the organs up to the brain, which makes the autonomic nervous system a measuring system as much as a control system.

The counting was done carefully. Two neuroscientists at Purdue, James Prechtl and Terry Powley, took the abdominal vagus of the rat and prepared complete cross-sectional electron microscope montages. They counted every axon in the trunks and all five branches. Their conclusion in a 1990 inventory of the fiber composition of the rat abdominal vagus was precise.

Outgoing fibers may represent over a quarter of the total, a higher proportion than the widely accepted estimates of the time allowed. The popular figure of 90 percent sensory is too strong. The large majority sensory finding stands, and it comes from rat abdominal branches rather than from a human count.

What those incoming fibers carry is mapped in a 2000 anatomy of the afferent vagal system by Hans-Rudolf Berthoud and Winfried Neuhuber. Stretch of the stomach wall, the chemical composition of what arrived, the state of the airways, the pressure in the great vessels, the presence of inflammation. The organs file continuous reports.

Vagal reports become feeling in the insula

The neuroanatomist Arthur Craig gave this traffic its modern framing. He worked out where these signals land in the human brain. In a 2002 synthesis on interoception, the sense of the physiological condition of the body, he argued that this pathway is a genuine sense, distinct from touch. Its destination in the insula is where bodily state becomes a feeling. Hunger, breathlessness, nausea, the flutter of dread in the chest: these are readings from a sense organ that happens to be your own interior.

The inflammatory reflex connects the vagus and immune system

The two-way traffic has a consequence that surprised immunology. Kevin Tracey's group at the Feinstein Institutes was studying inflammation. Stimulating the vagus in rats during lethal endotoxin exposure suppressed the release of tumour necrosis factor and prevented shock, reported in their 2000 paper on vagus nerve stimulation and the inflammatory response. They named the circuit the inflammatory reflex, and it has an incoming arm too. Vagal fibers detect inflammation and report it, and the outgoing arm damps it.

The human follow-through came in 2016, when Frieda Koopman and colleagues implanted vagus stimulators in patients with rheumatoid arthritis and reported inhibited cytokine production and reduced disease severity. The study was small and open-label, meaning everyone knew who was being stimulated, and several authors had financial ties to the device company. It establishes that the circuit exists in humans. It is not a cure, and it is not standard care.

Sympathetic drive supplies range, not survival

Two experiments 70 years apart, one removing the sympathetic chain in cats and one following transplanted human hearts, gave the same answer.

Neither experiment removed the whole autonomic nervous system. The first took the sympathetic chain and left the parasympathetic and enteric divisions working. The second cut the nerves to one organ.

Cannon removes the sympathetic chain

Walter Cannon was an American physiologist at Harvard who had named homeostasis and wanted to know how much of it the sympathetic system was responsible for. In 1929 he and his colleagues removed the entire sympathetic chain from cats, then studied what those animals could and could not do.

The work appeared as observations on animals surviving complete exclusion of sympathetic nerve impulses. The result was not what a modern reader expects. The animals lived. Kept warm, fed and sheltered, they were unremarkable. What they had lost was the ability to meet a demand: cold, exertion, blood loss, emotional stress. Sheltered they were fine. Challenged they had nothing to give.

The transplanted heart repeats the finding in humans

When a heart is transplanted, its nerves are cut. The new heart beats, because heart muscle generates its own rhythm, and for a time no brain is speaking to it. Some recipients slowly regrow sympathetic nerves into the graft and some do not, which creates a natural experiment.

Frank Bengel and colleagues in Munich measured reinnervation directly with positron emission tomography and a tracer that accumulates in sympathetic nerve endings, then tested the same patients on a bicycle. In their 2001 study of sympathetic reinnervation and cardiac performance, 16 of 29 recipients showed reinnervation. At rest, the two groups were indistinguishable. During exercise the reinnervated group reached a higher peak heart rate, exercised longer, and had a contractile response resembling healthy controls.

At rest, no difference. Under demand, all the difference. Sympathetic drive is not what keeps a sheltered body alive. It is what gives that body range.

One further number makes the point concrete. Block both autonomic branches pharmacologically in a healthy adult and the heart neither stops nor idles slowly. It settles at what physiologists call the intrinsic rate.

Jose and Collison measured that rate across a range of ages in a 1970 study of its normal range and determinants in man. In young adults it sits near 100 beats per minute and it declines steadily with age. Your resting pulse of 60 is the heart's own rhythm with the vagal brake applied continuously.

Six instruments that measure autonomic state, and their limits

Autonomic state can be measured six ways, and each instrument reads one channel of the pattern. Knowing the boundary of each is what separates a useful number from a misleading one.

Heart rate variability

Your heart does not beat metronomically. The interval between beats varies from beat to beat, and it varies most in time with your breathing, lengthening as you breathe out. That beat-to-beat variation is heart rate variability, and it exists because the vagus can change the timing of the very next beat. The measurement standards were set in 1996 by a joint task force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology, and those definitions are still the reference.

It carries real prognostic weight. In the Framingham Heart Study, Hisako Tsuji and colleagues followed 2,501 people free of clinical heart disease. They found that reduced heart rate variability predicted new cardiac events after adjusting for the usual risk factors. The hazard ratio was about 1.47 for a one standard deviation drop in overall variability.

Four limits define what the number means.

The high-frequency component does track vagal traffic to the heart. The low-frequency component does not track sympathetic traffic, despite 30 years of software labelling it that way. Gustavo Reyes del Paso and colleagues reanalysed the studies behind that assumption. Their conclusion in a 2013 review of low frequency variability as an index of sympathetic cardiac tone is that the evidence does not support the interpretation. Any app showing you a sympathetic score derived from low frequency power is showing you an unvalidated number.

Second, respiratory sinus arrhythmia and vagal tone come apart when breathing changes. Paul Grossman and Edwin Taylor set out in a 2007 review of what respiratory sinus arrhythmia actually indexes that breathing rate and depth alter the measure independently of vagal traffic. Breathe slowly during a recording and your number improves whether or not anything about you has changed.

Third, wrist devices have real accuracy constraints. A guideline group led by Benjamin Nelson reviewed the use of wrist-worn consumer wearables in biobehavioral research in 2020. They set out where optical sensing degrades, particularly during movement. They also show how the proprietary processing inside these devices limits comparison between brands.

Fourth, and most important for a reader with an app: heart rate variability is a validated index of cardiac autonomic state, largely vagal. The Unified Model of Tone reads it as one channel of the body's whole organization, which is a broader claim than the instrument alone makes. The instrument page on heart rate variability works through what a personal number does and does not mean.

Skin conductance

Two electrodes on the fingers, a tiny current, and a measure of how easily it passes. Sweat carries current, so conductance rises when sweat glands activate. Those glands are driven by sympathetic cholinergic nerves, which makes this a direct sympathetic readout. Measurement standards were published in 2012 by a Society for Psychophysiological Research committee led by Wolfram Boucsein. The clinical version is called the sympathetic skin response, reviewed by Roberto Vetrugno and colleagues in a 2003 account of its mechanisms and clinical applications.

The limits: it is fast and it habituates, so a repeated stimulus produces smaller responses regardless of state. It reads sudomotor traffic only. It says nothing about the heart, nothing about the gut, and nothing about the vagus.

Pupillometry

Pupil size reflects both autonomic branches and tracks arousal moment to moment. Margaret Bradley and colleagues tested this in a 2008 study of the pupil as a measure of emotional arousal. Dilation tracked how arousing a picture was, in both pleasant and unpleasant directions, and it covaried with skin conductance. The limits are obvious once stated. Light level dominates the signal, and mental effort changes it independently of emotion, so pupil data outside controlled conditions is difficult to interpret.

Blood and urine catecholamines

You can measure noradrenaline in plasma. The difficulty is what the number means. David Goldstein, Graeme Eisenhofer and Irwin Kopin worked out the sources and fates of these molecules. In a 2003 review of the sources and significance of plasma catechols they show that a plasma level reflects release, reuptake and clearance together.

Only a small fraction of what a nerve releases ever reaches the general circulation. Regional spillover methods can approximate traffic to a particular organ, and they are research procedures rather than clinic tests.

Microneurography

The most direct method available in humans. A fine tungsten electrode is placed through the skin into a nerve, usually the peroneal nerve at the knee, until it records the traffic in sympathetic fibers going to muscle vessels. Wallin and Charkoudian's review of human sympathetic nerve recording describes what four decades of it have established.

It is the reference standard, and its limit is severe: one nerve, one target class, a trained operator, and a laboratory. Outflows to different organs are separately controlled, so one nerve is one channel of a much larger pattern.

Orthostatic and reflex testing

The clinical bedside version. Lie the patient down, stand them up or tilt them, and watch what pressure and heart rate do. Add deep breathing at 6 breaths per minute, then a Valsalva manoeuvre, which is a forced exhale against a closed airway.

That is the standard autonomic reflex screen, described by the neurologist Phillip Low in a 2003 practical account of testing the autonomic nervous system. The diagnostic thresholds are set by a 2011 consensus statement on orthostatic hypotension, neurally mediated syncope and postural tachycardia syndrome, produced by an international panel led by Roy Freeman.

This is the family of tests that matters if symptoms are real and disabling. It is also where the ceiling of clinical autonomic testing sits. These tests are excellent at finding failure. They are far less able to characterise a system that passes every threshold and still does not answer demand well.

What a high or low autonomic number means

A single autonomic number carries no verdict. What high and low mean depends on whether the value moves when the day demands it and returns afterward.

A high resting sympathetic level, measured directly, varies severalfold between healthy young adults and stays stable within a person over years. Wallin and Charkoudian's four decades of recordings show that the same level is offset in different people by differences in cardiac output and vascular responsiveness. Two people with very different sympathetic traffic can have identical blood pressure.

Low heart rate variability carries population-level risk, as the 2,501-person Framingham cohort showed. It also drops with age, with poor sleep, with acute illness, with alcohol the night before and with a hard training session. A single low reading is information about this morning.

High variability is not automatically better either. Very high values appear in athletes with excellent regulation and also in some pathological rhythms. The number without context is not a verdict.

The dynamic reading is the useful one

Ask whether the value moves when the day asks it to move, and whether it comes back afterward. An autonomic system that rises with a demand and returns is doing its job. A system that sits high and never comes down, or sits low and never rises, has lost range in one direction.

That reading is why postural tachycardia syndrome is defined by a response rather than by a level. The diagnostic criterion is an excessive rise in heart rate on standing, as set out in the 2021 National Institutes of Health expert consensus on that condition.

Range is the bridge to how the Unified Model of Tone reads all of this, and it is a claim about movement rather than about any single number.

What moves autonomic state, and what does not

Several inputs move autonomic state reliably, and the reliable ones are breathing, exercise and sleep. Each is measured here by how far it moves an autonomic number.

Breathing

The most reproducible input available. Slowing the breath to roughly 6 breaths a minute lengthens the vagal effect on each exhale and brings breathing into step with the natural rhythm of the blood pressure control loop. Sylvain Laborde and colleagues pooled 223 studies in a 2022 systematic review and meta-analysis of voluntary slow breathing. Vagally mediated variability rose during the breathing, immediately after a single session, and after multi-session training.

Read the finding precisely. It shows the input reaches the autonomic nervous system. It is a change in an autonomic measure rather than a demonstrated change in disease outcomes.

Heart rate variability biofeedback

The trained version of the same input. A person breathes at their individual resonance frequency, usually close to 6 per minute, while watching their own rhythm on a screen. Paul Lehrer and Richard Gevirtz set out the proposed mechanism in a 2014 account of how and why heart rate variability biofeedback works. Their argument is that repeated exercise of the pressure-control loop increases its responsiveness over time. The mechanism has supportive evidence and the question is open.

Exercise

Youssra Amekran and Abdelkader Jalil El Hangouche pooled 16 randomised trials totalling 623 healthy adults in a 2024 meta-analysis of exercise training and heart rate variability. Training improved overall variability and the vagally related measures. Effects varied with sex, age and exercise type, and the authors note how few trials exist.

Sleep

The input that moves the autonomic nervous system most and gets the least attention. A meta-analysis by Suling Zhang and colleagues of 11 randomised trials in 549 participants found that sleep deprivation reduced the vagally related variability measure and raised low-frequency measures. Some indices moved and others did not reach significance, and the authors call for standardised protocols. The direction is consistent. The magnitude is open.

Mindfulness does not reliably raise resting vagal variability

Lydia Brown and colleagues pooled 19 randomised trials in a 2021 meta-analysis of mindfulness interventions and vagally mediated heart rate variability. Resting vagal variability did not improve significantly compared with controls. Removing an outlier study left the estimate smaller still. Heterogeneity was high.

The model predicts exactly this kind of result. An input meets a system already in some state, and identical inputs land differently on differently organized people. Resting vagal variability is not where a mindfulness intervention reliably shows up, and a pooled average across mixed populations is where that person-to-person variation disappears.

Ear-clip vagus stimulation fails its central claim

A large market now sells vagal stimulation: ear clips, cold plunges, humming, gargling, devices. The specific testable claim is that transcutaneous auricular vagus nerve stimulation, meaning electrical stimulation of the ear branch of the vagus, raises vagal heart rate variability.

Vinzent Wolf and colleagues tested it with a living Bayesian meta-analysis of 16 single-blind sham-controlled studies in healthy participants. The result was strong evidence for the null hypothesis, with an effect size of 0.014 and a Bayes factor of about 25 favoring no effect. Acute ear stimulation did not change vagal heart rate variability compared with sham.

Be precise about what that overturns. The vagus remains central. Implanted vagus nerve stimulation is an approved therapy in epilepsy and depression, and it produced the rheumatoid arthritis result above. In heart failure the large device trials disappointed.

Inder Anand and colleagues compare the ANTHEM-HF, INOVATE-HF and NECTAR-HF results in a 2020 analysis of symptomatic and functional responses to vagus nerve stimulation. The null result is narrow and useful. Clipping a stimulator to your ear has not been shown to move the number people buy it to move.

Autonomic state as a face of tone

The Unified Model of Tone reads autonomic state as the most measurable expression of tone, the integrated organization the nervous system maintains across the body. Everything above is established physiology, credited to the people who established it.

Tone includes the body's ability to move where the moment demands and return to balance afterward. Health is the width of that range. Autonomic measures read that range more directly than any other instrument set, which is why this page is where the rest of the library goes for autonomic state.

Three consequences follow, and each is testable.

Autonomic state is a chord, not a note

Morrison's separately controlled output channels and Berntson's two-dimensional space say the same thing from two directions. There is no single autonomic value. There is a pattern across many channels, and the pattern is the state. That pattern is the organization, and the organization is tone. What a clinician or a wearable measures is one voice out of many, and a single voice can read normal while the chord is badly out of tune.

This is why chasing one number rarely works, and why two people with identical resting heart rate variability can live in different bodies. It is also why the model expects mixed results whenever a study pushes one channel and measures one channel.

The nervous system integrates a regulation it does not own

The enteric division makes this concrete. Digestion continues when the connections to the brain are cut. Vascular smooth muscle regulates locally. Immune cells make their own decisions and then report them upward through the vagus. Regulation is distributed through the whole organism. The nervous system integrates that regulation at the highest density of any tissue, which is a real and central role. It is a different thing from being in charge.

Bidirectional restoration is the signature of a restored regulator

Here the model draws its sharpest practical line.

A drug moves a value one way. A beta blocker lowers heart rate in everyone who takes it, including the person whose heart rate was already low. Beta blockers save lives and have a real place. The point is mechanical. The drug substitutes for the regulation rather than restoring it.

The model claims something different for an input that restores tone. It should move a dysregulated loop toward the middle from either side. The same input should bring a high value down and a low value up, because it is restoring the system's ability to find its own value rather than pushing the value itself. The baroreflex is the natural place to look, because that loop already corrects in both directions by design.

State the test plainly. It needs a prospective design and a sham arm matched for contact and attention. The target must be chosen in advance, from a measure taken before the outcome is known. The direction of the predicted change must be stated before the data arrive.

Convergence toward the middle from both starting sides is the signature. If the treated group converges no more than the sham group, the signature is gone. If the input moves everyone the same direction regardless of where they started, it pushed the output rather than restoring the regulation, and it carried half the group further from the middle.

Restoring regulation versus masking the output

The distinction the model draws is between two aims rather than between good medicine and bad medicine.

Some interventions change the output. A drug that slows the heart, dries a secretion, blocks a receptor or quiets a symptom changes what the autonomic nervous system produces. Symptomatic relief has genuine value, and in acute and life-threatening situations it is the correct and sometimes the only answer.

Other interventions aim at the organization producing the output, so that the system finds its own value again. The model's contention is that the second aim has been under-pursued, because tone has never been named as its own regulatory system. A practitioner cannot deliberately work on a variable that has no name in their framework. That is a gap in vocabulary, and naming the variable is what closes it.

Three objections to the model, answered

The first objection is that this is relabeling. Sympathetic drive, vagal tone, baroreflex sensitivity and allostatic load are all established, so gathering them under one heading adds nothing. The components are not claimed as new, and every one is credited above to whoever established it. What is claimed is the unification and the composition.

These measures are read as faces of one organization rather than as independent variables, and any condition is described as a combination of a small number of dimensions. A relabeling makes no new prediction. The bidirectional test above makes one, and existing instruments can run it.

The second objection is that a framework which explains everything predicts nothing. The answer is the bidirectional test just stated, plus a second prediction. If autonomic measures, reflex responsiveness, recovery time and rhythm coupling are faces of one organization, they must behave like faces of one organization. Recorded together in the same people they should share a common underlying factor. Loading on that common factor is the finding that confirms the claim.

The third is the standard point about causation, and it applies fully. Almost everything in the measurement section is observational, and observation cannot establish which way the arrow points. Low variability may reflect disease rather than produce it. That is precisely why the model does not rest on those correlations. It rests on a prospective test with the direction of the result named in advance.

Secondary causes hold throughout. Not every autonomic abnormality is a regulation problem. Diabetes, amyloidosis, Parkinson disease, spinal cord injury and autoimmune neuropathy disturb autonomic function through mechanisms of their own. So do medication effects, thyroid disease and blood loss. Each needs finding and treating on its own terms. What this frame addresses is the large remainder, where the workup is clean, the hardware is intact, and the autonomic nervous system has lost range.

How this page relates to the rest of the library

The autonomic nervous system is anatomy, and the dimensions of the Unified Model of Tone are properties expressed through that anatomy. The library keeps them separate because collapsing them is what makes explanations vague.

Two further distinctions are worth stating outright, because readers conflate them constantly.

• Heart rate variability is an instrument, and this page is the autonomic state that instrument samples.
• The vagus nerve is one cable inside the autonomic nervous system, and it is not the system.
• The wiring underneath this page, meaning the two-neuron chain, the intermediolateral column, the receptor families and the segmental map, is taught on the neurology lesson on autonomic circuitry.

Six pages in the library lean on this reading of autonomic state most directly.

• Dysautonomia is where autonomic regulation itself is the presenting problem.
• Blood pressure is the baroreflex read as a number.
• Gut health is where the enteric division and the vagal conversation meet.
• Sleep is where autonomic state and circadian timing become inseparable.
• Anxiety is the sympathetic pattern experienced from the inside.
• And long COVID is where autonomic testing has become central to a condition medicine is still mapping.

Frequently asked