Dorsal motor nucleus of the vagus

In brief

What it is, and where it sits

The dorsal motor nucleus of the vagus is a thin, paired column of nerve cells in the medulla oblongata, the lowest part of the brainstem, running lengthwise just under the floor of the fourth ventricle. It is the third and lowest member of a stack: the area postrema at the very back, which has an incomplete blood-brain barrier and can sample the blood directly; the nucleus tractus solitarius just below it, where everything the vagus hears from the organs arrives; and this nucleus below that, where the answer is sent back out. Anatomists call the three together the dorsal vagal complex, and they occupy a space a few millimeters across ^1,2.

The cells in it are preganglionic parasympathetic motor neurons. Preganglionic means the cell body sits in the brain or spinal cord and its axon runs out to a second nerve cell in a ganglion or plexus near the target organ, which then does the final work; parasympathetic refers to the division of the autonomic nervous system usually summarized as rest-and-digest, as opposed to the sympathetic fight-or-flight division. These particular neurons release acetylcholine and send thin, unmyelinated axons — slow-conducting fibers without the fatty insulating sheath — out of the skull in the vagus nerve, the tenth cranial nerve, whose name is Latin for wandering ^1,14.

What they do not do is touch the stomach itself. The axons end on neurons of the myenteric plexus, the sheet of nerve cells sandwiched between the two muscle layers of the gut wall, and those enteric neurons are the ones that contract the muscle and drive the glands. A 2026 review puts it precisely: the enteric neurons behave as function-specific pattern generators running particular motility and secretion programs, and what the brainstem does is choose among them ¹. This is why a single nucleus can make the stomach either squeeze or relax, as the next section describes.

The nucleus is not uniform. Cells projecting to one organ differ from cells projecting to another, both in where they sit within the column and in what they respond to. In rat brainstem slices in which neurons had first been labeled according to where their axons went, the hypothalamic peptide orexin depolarized 70 percent — 30 of 43 — of the neurons supplying the gastric fundus or corpus, the upper reservoir part of the stomach, but only 3 of 13 supplying the antrum and pylorus, 4 of 18 supplying the duodenum and 1 of 13 supplying the cecum ⁷. The modern reading of the whole vagal system, from recent transcriptomic surveys of which genes individual cells switch on, is that it works largely as a set of labeled lines rather than a single channel ¹.

Nor is it only neurons. Astrocytes — the star-shaped supporting cells of the brain, once assumed to be passive scaffolding — sit densely around these motor neurons and gate their glutamate signaling. Silencing dorsal vagal complex astrocytes chemogenetically in rats abolished the normal rise and fall of food intake across the estrous cycle, the four-to-five-day reproductive cycle of a female rat, while activating them lowered intake whatever the cycle stage ³². Almost all of this cellular detail comes from rats and mice; there is no human equivalent of these experiments and there is unlikely to be one.

What it innervates: the organs below the diaphragm, and the heart question

The great bulk of this nucleus's output goes below the diaphragm. Its axons reach the esophagus, the stomach in all its parts, the pancreas and the intestine, and that subdiaphragmatic bias is the single most useful fact about it ¹. Even the esophagus, which lies above the diaphragm, is reached by it: in male mice, optogenetically driving a chain from the hypothalamic paraventricular nucleus through this nucleus and out along the vagus — optogenetically meaning that the neurons were engineered to carry a light-sensitive protein, so that a pulse of light switches just those cells on — raised the frequency and the amplitude of the electrical signals recorded from the lower esophageal sphincter, and cutting the vagus or blocking acetylcholine receptors abolished the effect; the rise in resting sphincter pressure in those animals came from the chronic stress itself, not from the light ³. It reaches the airways too. The vagal complex of the cat was traced branch by branch in 1980 — laryngeal, tracheobronchial and pulmonary branches alongside the cardiac and gastrointestinal ones ³⁷ — and in 2026 the airway-projecting cells of this nucleus were characterized directly in rats: TWIK-2 potassium leak channels set their resting voltage and firing rate, acidosis both excites them and reduces the inhibition reaching them, and chemogenetically activating them made the airways secrete mucus and the animals sigh more often, without changing ventilation, oxygen consumption or heart rate ⁴.

The heart is the contested case, and it matters because a popular theory, discussed later on this page, is built on what this nucleus is supposed to do to it. The relevant evidence is a matched pair of tracing studies from one laboratory, done in rats with the same dye and the same counting method, one year apart. Injecting the tracer into this nucleus labeled 68 to 96 cardiac fibers from the left side and 67 to 115 from the right; those fibers reached 709 and 494 cardiac principal neurons respectively, which is at least 18 percent and 12 percent of them ⁵. The next year the same group injected the nucleus ambiguus, a separate motor nucleus further forward and toward the belly surface of the medulla: 82 to 151 fibers on the left and 60 to 122 on the right, reaching 2,248 and 1,784 principal neurons, at least 56 percent and 45 percent ⁶.

Read carefully, those numbers say something more interesting than one nucleus winning. The fiber counts are comparable; the coverage is not. Each nucleus ambiguus axon branched to about 27 to 30 cardiac neurons, roughly three times the divergence of an axon from this nucleus, and the nucleus ambiguus sent more large, fast, myelinated B-type fibers where this nucleus sent fine slow C-type ones. This nucleus also innervated the small intensely fluorescent cells of the cardiac ganglia, which the nucleus ambiguus did not touch ^5,6. The authors' own conclusion was that cardiac vagal control involves two distinct projections converging in the cardiac plexus — not that one of them is irrelevant.

The belief that the cardiac vagal fibers come from this nucleus was itself overturned, and by a single clean experiment. In 1976 McAllen and Spyer recorded in cats from medullary neurons that could be activated backward along the cardiac vagal branches. Of 57 such neurons, with conduction velocities between 3 and 15 meters per second, not one lay in the region of the dorsal motor nucleus despite repeated sampling there; all lay in the region of the nucleus ambiguus, and the 46 of them that were marked and examined histologically belonged to its principal column, rostral to the obex. Of 33 dorsal motor neurons sampled, only 3 could be activated at all, and only at high stimulus intensity, and the authors argued from their thresholds and conduction velocities that these were unlikely to be cardio-inhibitory. Their conclusion — that the preganglionic cardio-inhibitory neurons arise not in the dorsal motor nucleus but in the principal column of the nucleus ambiguus — is where the modern position comes from, and it is also the historical premise the polyvagal theory's two-nuclei claim was built on ³⁸.

The honest summary is therefore narrower than either slogan, but it is not nothing. This nucleus is not the main source of beat-to-beat vagal braking of the heart; the nucleus ambiguus is, and that is the settled position among vagal physiologists ^25,38. Its own cardiac projections, however, are real and have now been tested directly. In awake mice, optogenetically activating its cholinergic neurons slowed the heart from 605 plus or minus 18 beats per minute before stimulation to 271 plus or minus 17 during it (13 mice, p below 0.0001), an effect blocked by peripheral muscarinic and ganglionic nicotinic acetylcholine receptor antagonists but not by beta-1 adrenergic blockade; photoinhibiting the same neurons left resting heart rate unchanged (629 plus or minus 45 against 627 plus or minus 38 beats per minute, 3 mice, p equal to 0.8950), so in that preparation these neurons were sufficient to slow the heart without being necessary for its resting rate ³⁹. A 2026 rat study found that destroying this nucleus's cholinergic neurons blunted the fall in blood pressure and heart rate produced by stimulating vagal sensory fibers, and proposed a loop running between the nodose ganglion — the swelling on the vagus nerve in the neck that holds the cell bodies of its sensory fibers — and this nucleus; the authors call the pathway putative, it is one study, and it has not been replicated ⁴⁰. A 2026 preprint, not yet peer reviewed, reported that electrical stimulation of this nucleus slows the heart in control mice and loses that power with age in a mouse model of Alzheimer's disease ⁴¹.

The practical consequence for reading the literature is this: when a paper, a device or a therapy claims to measure or engage the dorsal vagal system through the heart, it is using the weaker of two projections as a proxy for the stronger, and almost always without saying so.

Gastric motility and secretion: how one nucleus both squeezes and relaxes

The circuit this nucleus belongs to is called the vagovagal reflex: stretch and chemical signals from the stomach travel up vagal sensory fibers to the nucleus tractus solitarius, that nucleus talks to this one, and this one sends the adjustment back down the vagus ². The whole loop is short enough that it can be completed without anything above the brainstem being involved, and because it sits next to the area postrema, where the blood-brain barrier is incomplete, circulating gut hormones can lean on it directly ².

Two opposite commands leave the same nucleus. The excitatory one is cholinergic: acetylcholine onto enteric neurons that contract the stomach. The inhibitory one is what physiologists call non-adrenergic, non-cholinergic — a line that uses neither acetylcholine nor norepinephrine, relying instead on vasoactive intestinal peptide and nitric oxide released by enteric neurons to relax the muscle ⁸. Which line is recruited depends on what arrives at the nucleus. In rats, microinjecting thyrotropin-releasing hormone or the peptide PACAP38, short for pituitary adenylate cyclase-activating polypeptide, into the nearby raphe obscurus excites the stomach, while substance P and vasoactive intestinal polypeptide relax it ⁹.

The best recent demonstration of how conditional this is came in 2026, and it is a sex difference. In female rats, gastric emptying was slower when estrogen levels were high. Oxytocin, the so-called anti-stress peptide, suppressed inhibitory GABA input onto neurons of this nucleus under ordinary unstressed conditions regardless of estrogen level — which is not what happens in male rats, where that suppression requires prior stress or activation of corticotropin-releasing factor receptors. Yet when oxytocin was injected into the nucleus in the living animal, the gastric effect it produced depended on the stage of the estrous cycle and on whether the animal had been stressed before, because different peripheral pathways were recruited ⁸. The central synapse behaved one way and the organ another, which is a warning about inferring what the stomach will do from what the nucleus is doing.

Stress reaches the stomach through here, and in two directions. Thyrotropin-releasing hormone microinjected into this nucleus in rats produced gastric erosions with bleeding closely resembling those caused by cold restraint stress, and the lesions were prevented by neutralizing the peptide with an antibody ²⁹. But a subthreshold dose of the same peptide, too low to raise acid secretion, protected the stomach against a strong irritant, and vagotomy or antibody against the peptide abolished the protection that a mild irritant pretreatment normally gives ²⁹. The same pathway, at different intensities, injures and protects. In male mice a chain from the hypothalamic paraventricular nucleus through this nucleus to the stomach sustains chronic gastritis, with HCN1 — a channel that passes current into a neuron when it is hyperpolarized, and so sets how readily it fires again — raised in the hypothalamic neurons, and silencing them stops the inflammation ³⁰.

Timing is part of the command. Restricting rats to eating only during their rest phase for four weeks disturbed gastric emptying, intestinal motility and gut hormone secretion together with the activity of the solitary nucleus, the area postrema and this nucleus ⁴². And not everything attributed to stress shows up inside the nucleus: when chronic heterotypic stress delayed gastric emptying in male but not female rats, cholinergic staining in this nucleus itself did not change with sex or with stress — the measurable changes were in the catecholamine cell groups around it ⁴³. That is a useful negative result, and it is the kind that rarely makes a headline.

The pancreas: insulin, glucagon, and the response to a falling blood sugar

The pancreas has two jobs and this nucleus is involved in both. The exocrine pancreas makes digestive enzymes; the endocrine pancreas, the islets of Langerhans, makes insulin, which lowers blood sugar, and glucagon, which raises it. The parasympathetic premotor neurons for the pancreas sit principally in this nucleus, and they are electrophysiologically distinguishable from the ones controlling the stomach — the same column, different cells ⁴⁴.

That the nucleus can set insulin output has been shown directly, in anesthetized rats. Lifting the GABA brake on it by bilateral microinjection of bicuculline raised the peak glucose-stimulated insulin response from 4.0 plus or minus 0.6 to 9.2 plus or minus 0.1 nanograms per milliliter. Blocking muscarinic acetylcholine receptors with atropine cut the response to 2.1 plus or minus 1.1, confirming the pathway is cholinergic. Blocking nitric oxide synthesis, by contrast, raised it further to 15.3 plus or minus 3.0, which means a nitric oxide line normally restrains the insulin command ¹⁰. Earlier microinjection work had found that pancreatic polypeptide placed in the dorsal vagal complex did not change basal insulin or glucagon but did potentiate the glucose-stimulated response ⁹.

The counter-regulatory response — the body's defense against a falling blood sugar — runs through the same cells. When rats were given insulin at 10 units per kilogram, blood glucose fell from 4.9 plus or minus 0.3 to 1.7 plus or minus 0.2 millimoles per liter, and the immediate-early protein Fos, which marks neurons that have recently fired, appeared in many cholinergic neurons of this nucleus, concentrated in the medial and middle parts beneath and just in front of the area postrema. Control animals had essentially none ⁴⁵. The proposed chain is that low glucose silences inhibitory neurons in the ventromedial hypothalamus, which releases orexin neurons in the perifornical region, which drive this nucleus, which drives glucagon release ⁴⁴.

Glucagon release during insulin-induced hypoglycemia is impaired in type 1 diabetes, and this is one of the central unsolved problems of diabetes care, because it is what makes low blood sugars dangerous. The review that lays out this circuitry states plainly that the reason for the failure is unknown ⁴⁴. What has been shown, in mice, is that the nucleus's synapses change in diabetes: in control animals with cyclic adenosine monophosphate (cyclic AMP), the intracellular messenger many hormones act through, raised, insulin reduced inhibitory input onto these neurons, an effect that disappeared when protein trafficking or the GLUT4 glucose transporter was blocked — and in streptozotocin-diabetic mice insulin did nothing at all ²⁸.

The exocrine side has been mapped more recently. A 2026 mouse study traced a chain from oxytocin neurons of the hypothalamic paraventricular nucleus to cholinergic neurons of this nucleus and on to pancreatic acinar cells, the enzyme-making cells, and showed that silencing the hypothalamic end suppressed enzyme secretion and reduced inflammation in a model of acute pancreatitis brought on by very high blood triglycerides (hypertriglyceridemia), while stimulating it caused inflammation ³⁴. That is mouse circuit work in an artificial disease model; nothing of the sort has been done in people, and it should not yet be read as a statement about human pancreatitis. The inputs that reach this nucleus can be unexpected: in mice given a deliberately disturbed bite, excitatory input onto its cholinergic neurons rose, pancreatic acetylcholinesterase rose, insulin and glucagon rose, and postprandial blood glucose rose while fasting glucose fell — all of it reversing when the bite was corrected ⁴⁶. That is one laboratory and rodents only — the nerve tracing was done in rats and the bite model in mice — and the dental conclusion should not be carried across to people; it is cited here only as a reminder of how many doors open onto this nucleus.

Vagotomy: the ulcer operation, and what it taught about this nucleus

Two histories run through this nucleus and they are worth keeping apart, because they did not happen in the order one would expect. The anatomy was settled late. Its place in the vagal complex was set out in modern form by tracing in the cat in 1980: horseradish peroxidase and amino acid autoradiography in 32 animals labeled the motor neurons of this nucleus and of the nucleus ambiguus, showed that nucleus ambiguus axons first run dorsomedially toward this nucleus and double back in a hairpin before leaving the medulla, and found vagal sensory fibers ending around the neurons of this nucleus — an anatomical substrate, the authors proposed, for vagal reflexes with only one synapse in the brainstem ⁴⁷; a companion paper followed the vagal branches one at a time, laryngeal, tracheobronchial, pulmonary, cardiac and gastrointestinal ³⁷. The physiology had been acted on nearly forty years before that, in an operating theater, and that is where the clinical story starts.

In 1943 Lester Dragstedt, in Chicago, cut both vagal trunks in a patient with a duodenal ulcer ^11,12. The reasoning was direct: acid caused ulcers, vagal drive caused acid, so remove the drive. Truncal vagotomy, meaning section of the vagal trunks at the esophagus, became a standard ulcer operation for a generation ^11,12, and for a generation the output of this nucleus was something surgeons routinely and deliberately destroyed.

It worked, and it had a cost that taught physiology. Cutting the trunks removes the excitatory cholinergic supply to the antrum and the pylorus as well as the acid-secreting fundus, so the stomach loses its ability to grind and propel; the operation therefore had to be paired with a drainage procedure, a pyloroplasty or a gastrojejunostomy, to let food out. Diarrhea and dumping followed in a proportion of patients. The response was a more selective operation: proximal gastric vagotomy, which denervates only the acid-secreting part and leaves the antral and pyloric nerves intact. In experienced hands it cures more than 90 percent of duodenal ulcers with minimal side effects, and it needs no drainage procedure ¹¹.

That distinction — truncal, which cuts everything below the diaphragm, against selective or proximal, which spares the rest of the gut — is the reason the operation later became an epidemiological instrument. It is a natural experiment in which a defined population had the subdiaphragmatic output of this nucleus interrupted on a recorded date, with a comparison group whose output was interrupted only partly. Nothing else in human medicine offers that.

Then it vanished. The demonstration that Helicobacter pylori causes most peptic ulcers, together with effective acid-suppressing drugs, made the operation unnecessary; by 2000 a surgical review was describing elective proximal gastric vagotomy as a rare procedure for which patients should be referred to a specialist center, and noting that few surgeons doing emergency ulcer surgery had any experience of it ¹¹. A central lesson about this nucleus — that its output is not optional, and that cutting it leaves a stomach that cannot empty — was learned on a scale that will not be repeated.

The clinical residue is still with us. In a United States series covering 82.6 million patients, postsurgical causes accounted for about 15 percent of gastroparesis ²², and the postvagotomy stomach remains a recognized cause of the syndrome long after the operation that produced it stopped being performed.

Braak 2003: the claim that Parkinson's disease starts here

In 2003 Heiko Braak and colleagues published an autopsy study that changed what this nucleus meant to neurology. Examining incidental and symptomatic Parkinson cases, they proposed that the aggregates of alpha-synuclein that define the disease — Lewy bodies inside cell bodies and Lewy neurites inside processes — appear first at two induction sites: the dorsal motor nucleus of the glossopharyngeal and vagal nerves, and the anterior olfactory nucleus. From there, they wrote, the process in the brainstem pursues an ascending course with little variation between individuals, reaching in turn the pontine tegmentum, the core of the next segment of brainstem up, the midbrain including the substantia nigra, the basal forebrain, and finally the cortex, beginning with the mesocortex of the inner front part of the temporal lobe, the transitional cortex that lies between the hippocampus and the neocortex ¹³.

A companion paper the same year stated the mechanism as an explicit hypothesis. An unidentified pathogen, able to cross the lining of the gastrointestinal tract, might enter postganglionic enteric neurons and travel from there along the unmyelinated preganglionic fibers of the vagus by retrograde axonal and transneuronal transport, reaching this nucleus and then spreading upward through a chain of interconnected vulnerable regions like a falling row of dominoes ¹⁴. No pathogen was named. None has been identified since.

Two years later the same group looked for the first link in that chain. In five autopsy individuals whose brains had been staged, alpha-synuclein inclusions were found in neurons of the submucosal plexus — a second nerve net lying nearer the gut lining than the myenteric one, named after Meissner — whose axons run into the gastric mucosa and end close to the fundic glands — that is, in cells positioned where a swallowed agent could reach them ⁴⁸. Five cases is five cases, and the chain is an interpretation of still tissue, but it gave the hypothesis a concrete starting point.

It is worth being exact about what kind of evidence this is, because the whole subsequent argument turns on it. Every brain in these studies was examined once, after death. Nobody was followed. The sequence is an inference from the way cases of differing severity can be arranged into an order — a reasonable inference, and the standard method of neuropathological staging, but an inference. The claim that the pathology starts in this nucleus and moves upward is not something anyone has watched happen in a person.

The hypothesis was enormously productive. It gave a reason to look at the gut in a brain disease, it predicted that constipation and other autonomic symptoms should precede the movement disorder by many years, and it generated the two testable predictions that the rest of this page is about: that cutting the vagus should protect against Parkinson's disease, and that putting alpha-synuclein into the gut of an animal should produce it in the brain.

The staging scheme meets the brain bank, and does not entirely survive

The first serious trouble for the obligatory reading of Braak's scheme came from counting. In 2008 a group working with 71 Parkinson brains from the United Kingdom Parkinson's Disease Society Tissue Bank reported that the substantia nigra was affected in 100 percent of cases and the nucleus basalis of Meynert in 98.5 percent, but that only 53 percent showed a distribution compatible with a caudal-to-rostral spread. Forty-seven percent did not fit. In 7 percent the dorsal motor nucleus was entirely free of pathology although alpha-synuclein inclusions were present in the substantia nigra and the cortex ¹⁵. Their title says it: this nucleus is not an obligatory trigger site.

The following year a much larger series reframed the problem. Examining a standard set of ten brain regions from 417 subjects, Thomas Beach and colleagues noted that the two staging systems then in use failed to classify up to half of all cases, because so many people had Lewy pathology confined to the olfactory bulb, or took a limbic-predominant route that bypassed the brainstem altogether. Their replacement scheme begins with an olfactory-bulb-only stage and then branches into brainstem-predominant and limbic-predominant paths ¹⁶. In that map this nucleus is one possible door, not the door.

A neuropathologist's balanced reading, published in 2019, found both camps partly right. Most published series of typical Parkinson's disease do find that 80 to 100 percent of cases fit the ascending scheme; but incidental Lewy body disease and Parkinson's disease can both show pathology in the substantia nigra with no involvement of this nucleus, and the scheme fits young-onset, slowly progressive cases far better than late-onset, rapidly progressive ones ⁴⁹. The scheme is valid for a subset of patients, which is a different statement from the one it was originally taken to make.

Meanwhile a 2012 study removed one of the assumptions that had made the gut-first story feel obvious. Comparing myenteric neuron density along the length of the human gut in Parkinson patients, controls and people with incidental Lewy bodies, it found no neuron loss in the myenteric plexus at all. What it did find was that the Lewy pathology in the gut followed the distribution of parasympathetic input from this nucleus, and not the distribution of sympathetic input or of the intrinsic enteric neurons, and that it was only rarely in the same cells as the enzyme tyrosine hydroxylase ⁵⁰. The gut pathology of Parkinson's disease, in other words, is laid down along this nucleus's own projection — which is consistent with the pathology arriving from the gut, and equally consistent with it arriving from the brain.

None of this refuted Braak. What it did was convert a general law into a description of one route among several, which is where the field has stayed.

The vagotomy experiment: one striking result, and a decade of mixed replications

If a pathogen or a misfolded protein travels from the gut to the brain along the vagus, then people whose vagus was cut should be protected. In 2015 a Danish group tested exactly that, using national registries to identify everyone vagotomized in Denmark between 1977 and 1995 and following them for decades. Patients who had a full truncal vagotomy had a hazard ratio for Parkinson's disease of 0.85 (95 percent confidence interval 0.63 to 1.14) against a matched general population, falling to 0.53 (0.28 to 0.99) in those followed more than 20 years; patients who had the superselective operation, which spares the rest of the gut, were no different from the general population at all ¹⁷. A hazard ratio is the ratio of the rate of an event in one group to the rate in another; a confidence interval is the range within which the true value probably lies.

That contrast — protection from the operation that cuts everything, nothing from the operation that spares the gut — was exactly what the hypothesis predicted, and it was reported worldwide as a confirmation. The authors themselves were more careful. They wrote that the statistical precision of the risk estimates was limited, and the headline figure rests on a subgroup whose interval only just excludes 1 ¹⁷.

Others said so in print at the time. A letter in the same journal later that year asked directly whether vagotomy reduces the risk of Parkinson's disease ⁵¹, and when the Swedish replication appeared in 2017 the editorial published alongside it was titled "Preventing Parkinson disease by vagotomy: Fact or fiction?" and was co-written by one of the authors of the Danish study ⁵². Neither is data. Both are evidence that the people closest to the work did not treat the finding as established.

Two years later the Swedish replication arrived, and it was larger: 9,430 vagotomized patients, 3,445 truncal and 5,978 selective, matched forty to one against 377,200 others, with 4,930 incident Parkinson cases over 7.3 million person-years. Overall, vagotomy was not associated with Parkinson's disease, hazard ratio 0.96 (0.78 to 1.17). Truncal vagotomy gave 0.78 (0.55 to 1.09) — an interval that crosses 1 — and only the subgroup operated at least five years before diagnosis reached 0.59 (0.37 to 0.93). The authors called it suggestive evidence of a potential protective effect ¹⁸. Notably, the crude incidence was higher in the truncal group (80.4 per 100,000 person-years) than in the reference population (67.5), before adjustment.

The most recent direct test has the largest number of Parkinson cases, and on vagotomy it is negative. Using a United States nationwide medical records network and 24,624 patients with new idiopathic Parkinson's disease, a 2023 study found strong and disease-specific associations with gastrointestinal syndromes preceding diagnosis — gastroparesis, meaning a stomach that empties too slowly with nothing blocking the way, with an odds ratio of 4.64; difficulty swallowing, or dysphagia, 3.58; irritable bowel syndrome without diarrhea 3.53; and constipation 3.32. Each was present prospectively as well, when the same records were read forward, but at different strengths: relative risks of 2.43, 2.27, 1.17 and 2.38 respectively, so the irritable bowel association, strong in the case-control comparison, is slight read forward. Appendectomy lowered later risk (relative risk 0.48). Vagotomy showed no association with Parkinson's disease whatsoever ¹⁹. How much power the study had on that last point cannot be worked out from the report: the 24,624 figure is the count of new Parkinson cases, not of vagotomized people, and the paper does not say how many vagotomized patients it found in the network, so the null is uninformative about the size of any effect ¹⁹. The largest vagotomy exposures in the literature remain the Swedish 9,430, of whom 3,445 were truncal ¹⁸, and the Taiwanese 155,944 ⁵³.

Two further results fence the claim in. A Taiwanese cohort of 155,944 vagotomized patients and the same number of matched controls found no reduction in dementia: incidence 2.43 against 2.84 per 1,000 person-years, adjusted hazard ratio 1.09 (0.87 to 1.36), and 1.04 (0.87 to 1.25) for truncal vagotomy specifically ⁵³. And the appendectomy literature, often cited alongside, is itself double-edged: the 2018 study covering more than 1.6 million people and over 91 million person-years did find lower Parkinson risk and later onset after appendectomy, especially in rural residents — but it also found aggregated and truncated alpha-synuclein in the appendix of perfectly healthy people, which shows that the protein's presence in the gut is normal ⁵⁴. And the appendectomy epidemiology does not hold up when the rest of it is laid out. A Danish registry study of 265,758 people who had an appendectomy between 1980 and 2010, against 1,328,790 matched comparison subjects, found not protection but a slightly raised risk beyond ten years of follow-up — adjusted hazard ratio 1.14 (1.03 to 1.27) — consistent past twenty years and in both sexes ⁵⁵. Two large American prospective cohorts, the Nurses' Health Study and the Health Professionals Follow-up Study, found no association (pooled hazard ratio 1.08, 0.94 to 1.23), with appendectomy done for appendicitis in women if anything modestly raising risk (1.23, 1.00 to 1.50) ⁵⁶. A Korean national cohort with 16,122 appendectomies among 703,831 subjects found nothing either way (adjusted hazard ratio 1.42, 0.88 to 2.30) ⁵⁷. The protective finding is therefore one result among several that point in different directions, and it should not be offered alongside the vagotomy data as a second line of support.

Reviews published as recently as 2024 still summarize this literature as showing that truncal vagotomy may reduce the risk of Parkinson's and Alzheimer's disease ⁵⁸. Read against the primary studies, even that hedged reading is a fair description of 2015 and an unfair one of 2023.

Putting the protein in the gut: what animals did and did not do

The second prediction can be tested directly in animals, and it has been, repeatedly, with results that do not agree. The strongest positive came in 2019: preformed alpha-synuclein fibrils injected into the muscle layer of the mouse duodenum and pylorus produced pathology first in this nucleus, then in the caudal hindbrain including the locus coeruleus, and much later in the basolateral amygdala, the dorsal raphe nucleus and the substantia nigra, accompanied in the same order by dopamine neuron loss and by motor and non-motor symptoms. Truncal vagotomy prevented all of it, and so did deleting the alpha-synuclein gene ⁵⁹. As a demonstration that material can travel this route and cause disease at the far end, it is close to definitive — in mice, with an amount of pre-aggregated protein no stomach ever encounters.

A second 2019 study, in rats, both supported and complicated the picture. Fibrils into the rat duodenal wall produced pathology along the parasympathetic route and the sympathetic route at once: enteric nervous system, autonomic ganglia, the intermediolateral nucleus of the spinal cord — the column of cells from which the sympathetic outflow leaves the cord — the heart, this nucleus and the locus coeruleus, persisting at least four months, with evidence of spread back outward again to the stomach. But it happened only in transgenic rats carrying extra human alpha-synuclein. In wild-type rats, nothing propagated ⁶⁰. That particular objection has since been partly answered. In 2026 the same duodenal fibril injection was reported to produce phosphorylated alpha-synuclein in both the duodenal muscle layer and the striatum of ordinary C57BL/6J mice at four months, together with altered gut bacteria and motor deficits, and truncal vagotomy was associated with less of the protein in the brain — though these are associations rather than a controlled test of vagotomy, and the transgenic arm of the same study had no genetically matched wild-type control, as its authors say ⁶¹.

Against these sit two studies that found the opposite. A 2018 experiment in rats and monkeys produced gut dysmotility from enteric alpha-synuclein pathology, as expected — though in the rats even that was transient and had resolved after about four months — while the brain pathology was transient too: minor involvement of this nucleus and the locus coeruleus at one month, gone at later time points, and entirely absent in non-human primates at twelve months despite enteric pathology that was robust and persistent for the whole twelve. The authors concluded that sustained spread from gut to brain is a rare event and that the enteric pathology may be an epiphenomenon ⁶². In 2020, baboons given patient-derived Lewy body extracts into either the striatum or the gut developed nigrostriatal and enteric pathology either way, but no alpha-synuclein lesions were found in the vagus nerve at all; the authors state that their study does not support transmission through the vagus and this nucleus, and propose the general circulation instead ⁶³.

That circulatory proposal was itself tested and failed. Mice were surgically joined so that they shared a bloodstream, and one partner received patient-derived aggregates into the substantia nigra. The injected animal lost dopamine neurons and accumulated phosphorylated alpha-synuclein; the partner sharing its blood showed no lesion and no change ⁶⁴. Within that time frame, blood is not the route either.

Two further results cut across the direction of the whole argument. Human alpha-synuclein overexpressed in the rat midbrain traveled down and accumulated in preganglionic vagal terminals in the stomach wall, appearing in visceromotor but not viscerosensory endings — that is, in the endings that carry commands out to the organ but not in those that carry information back — so this nucleus is a conduit out of the brain as readily as into it, and peripheral pathology in a patient does not establish a peripheral origin ³⁶. And a molecular mechanism has been proposed for the gut-to-brain direction: the enzyme delta-secretase cuts alpha-synuclein at residue 103 and tau at residue 368, and the resulting fragments aggregate and travel along the vagus faster than the full-length proteins, with oral rotenone able to set the process going ⁶⁵. A mechanism for a journey whose occurrence in people is still disputed is a useful thing to have, and not a substitute for the journey.

The revision: brain-first, body-first, and the demotion of the vagus to one route of two

By 2020 the group that had produced the Danish vagotomy study had changed the question. Instead of asking whether Parkinson's disease starts in the gut, they asked whether it starts in the gut in some people and in the brain in others. Thirty-seven newly diagnosed patients were split by whether they also had rapid-eye-movement (REM) sleep behavior disorder — a condition in which the muscle paralysis that normally accompanies dreaming sleep fails, so that people physically act out their dreams, long known to precede Parkinson's disease by years — and compared with 22 people who had the sleep disorder alone, using four imaging methods at once ⁶⁶.

The results separated cleanly. Both Parkinson groups had equally reduced dopamine storage capacity in the putamen, while two-thirds of the isolated sleep-disorder group were normal there. But the patients with the sleep disorder, and the isolated sleep-disorder group, had reduced cardiac MIBG signal — a measure of sympathetic nerve endings in the heart — and reduced colonic uptake of carbon-11 donepezil, a measure of cholinergic parasympathetic nerve density in the gut. The sleep-disorder-positive group also had enlarged colons and slower colonic transit ⁶⁶. One trajectory loses the peripheral nerves first and the brain later; the other does the reverse. The authors named them body-first and brain-first.

Then in 2025 the body-first group split again. Analyzing two postmortem datasets containing 173 and 129 Lewy-positive cases with a machine-learning algorithm, the same laboratory found that early prediagnostic cases divided clearly into brain-first and body-first, and that the body-first cases divided further into a vagal-predominant subtype and a sympathetic-predominant subtype, in roughly equal proportions ²⁰. This is the most consequential revision for this page. In the 2003 formulation, the vagus and this nucleus were the gateway. In the 2025 formulation they are one of two peripheral gateways, accounting for perhaps half of the body-first cases, which are themselves a fraction of all cases.

The 2026 critical appraisal by the same group sets out what supports the model and what does not. In its favor: some people develop autonomic dysfunction and sleep disorders years before diagnosis while others do not; in people with prodromal sleep disorder, cardiac sympathetic denervation precedes nigrostriatal degeneration — the loss of the dopamine pathway that runs from the substantia nigra to the striatum — by approximately a decade, and in people without it the order reverses; and autopsy series often show bottom-up or top-down gradients ²¹. That decade is a modeled figure, and the model is worth stating. Fitting longitudinal imaging from two cohorts, the cardiac sympathetic signal fell to 50 percent of normal at about 5 years and 25 percent at about 10, while the dopaminergic signal reached those marks at about 8 and about 15 years; patients with the isolated sleep disorder had a median cardiac signal already down to 21.3 percent of normal while their median dopaminergic signal was still 81.5 percent, which places cardiac sympathetic loss 8.6 years ahead of nigrostriatal involvement and the whole prodromal phase at more than 20 years ⁶⁷. These are curves fitted to imaging data, not observed individual histories. Against it: the review is written by the model's own proponents, which is a conflict of a kind worth naming even when the review is scrupulous, and every autopsy line of evidence remains a sequence inferred from snapshots.

What is no longer in dispute is that gastrointestinal dysfunction precedes Parkinson's disease, reproducibly and specifically. Gastroparesis carries an odds ratio of 4.64 against matched controls and 2.43 as a prospective relative risk; dysphagia 3.58 and 2.27; constipation 3.32 and 2.38 — and each of these associations held specifically for Parkinson's disease when it was tested against Alzheimer's disease and cerebrovascular disease as well as against matched controls, which is what makes them specific rather than a general marker of ill health ¹⁹. A 2026 perspective treats gut dysfunction as a core feature of the disease while cautioning that microbiome associations, though reproducible, are confounded by geography, constipation, medication, diet, disease stage and reverse causation, and that trials of probiotics, fecal transplantation and dietary change have not established disease modification ⁶⁸. A 2024 hypothesis paper extends the two-route model to environmental exposure, proposing that inhaled toxicants seed the olfactory route and ingested ones the gut route — testable, and so far untested in people ⁶⁹.

One 2026 result closes part of the loop between the pathology and the symptom. Working in several mouse models meant to mimic both genetic and sporadic disease, Yu and colleagues found that alpha-synuclein pathology markedly impairs the excitability of the cholinergic neurons of this nucleus, and that chemogenetically switching those neurons back on restores gut motility — which makes the hypoexcitability a cause of the dysmotility in those animals rather than something that merely accompanies it. Electroacupuncture at the hind-limb point ST36, which engages vagal pathways, raised the excitability of the same neurons and relieved the dysmotility through the cholinergic brain-gut axis ⁷⁰; an invited commentary in the same issue sets out why that matters ⁷¹. It is mouse work in artificial models, and nothing of the kind has been done in a person. But it is the first demonstration that the gut complaints which precede a Parkinson diagnosis by years can be produced, and reversed, by changing what this nucleus does.

Diagnostically, the practical state of play in 2026 is that the earliest cases are hard to catch anywhere. In autopsy-confirmed incidental Lewy body disease, an alpha-synuclein seed amplification assay — a test that uses a sample to seed aggregation in a tube and then detects it — was positive in 75.8 percent of cerebrospinal fluid samples but only 35.1 percent of submandibular gland and 37.5 percent of skin samples; in established Parkinson's disease all three performed well. The numbers are small and the authors call the results preliminary ⁷². The body-first prediction that peripheral tissue should light up first is, at the moment, not what the peripheral tests show.

Diabetes, gastroparesis and the GLP-1 drugs

Gastroparesis means delayed gastric emptying with no mechanical obstruction, and it is the clinical syndrome that sits closest to this nucleus. In a 2018 United States claims database the prevalence of definite gastroparesis — with documented delayed emptying — was 21.5 per 100,000 persons, and it is two to four times more common in women. Among 82.6 million patients the causes were type 2 diabetes 51.7 percent, postsurgical 15 percent, medication-induced 11.8 percent, idiopathic 11.3 percent, type 1 diabetes 5.7 percent and other 4.5 percent ²². Parkinson's disease appears on the list of other risk factors ²².

It is worth stating what delayed emptying is not. It is a finding, not a mechanism: the same measurement can arise from weak antral and duodenal contraction, from a pylorus that will not relax, from a stomach that will not accommodate a meal, or from heightened visceral sensation, and the test cannot distinguish them ⁷³. Only some of those failures implicate the brainstem.

In diabetes, the brainstem is implicated, and the animal evidence is specific. In diabetic mice, insulin's ordinary modulation of inhibitory synapses onto these neurons is lost: in control animals with cyclic adenosine monophosphate (cyclic AMP), the intracellular messenger many hormones act through, raised, insulin reduced inhibitory input, an effect blocked by disrupting protein trafficking or the GLUT4 transporter, while in streptozotocin-diabetic mice insulin had no effect at all ²⁸. That the cholinergic drive out of this nucleus falls in disease has now been shown far more convincingly in a different model: in mouse models of Parkinson's disease, alpha-synuclein pathology impairs the excitability of these cholinergic neurons, and switching them back on chemogenetically restores gut motility ⁷⁰. In diabetic mice, a much smaller study found fewer activated cholinergic neurons here than in controls, and found that cutting the subdiaphragmatic vagus largely abolished the gastric motility response — amplitude, area under the curve and frequency — evoked by electroacupuncture at the points ST36 and PC6; the arm that counted the neurons used three mice per group, so it carries little weight on its own ²⁷. The review literature frames the open question honestly: chronic glucose dysregulation remodels these circuits, and it is not known whether the remodeling drives the metabolic problem or compensates for it ⁷⁴.

The GLP-1 receptor agonists — semaglutide and its relatives — brought this physiology into tens of millions of people. Gastrointestinal side effects occur in roughly 30 to 50 percent of patients, usually at initiation and dose escalation, and delayed gastric emptying is one of several mechanisms invoked, alongside central activation of emetic pathways, altered intestinal motility and the effects of rapid weight loss itself ²³. The honest position on how much of that delay runs through this nucleus, rather than through vagal sensory endings, the area postrema or the gut directly, is that it has not been separated in people.

The human emptying data are more interesting than the headlines. In 30 people with obesity given semaglutide 1.0 mg weekly in a crossover design, first-hour gastric emptying was slowed — area-under-curve ratio 0.73 (0.61 to 0.87) — but overall gastric emptying was not statistically different from placebo ⁷⁵. In 15 people with type 2 diabetes on oral semaglutide escalated to 14 mg, emptying in the first postprandial hour fell by 31 percent on the acetaminophen measure ⁷⁶. By contrast, eight weeks of the short-acting agonist lixisenatide in 30 patients raised gastric retention of a glucose drink 2.19-fold (1.82 to 2.64, p below 0.001), and postprandial glucose lowering tracked how much emptying slowed ⁷⁷. The reason for running that last trial is stated in its own abstract: tachyphylaxis for slowing of gastric emptying is seen with continuous exposure to GLP-1, so a short-acting drug was tested to see whether the effect could be sustained ⁷⁷. The long-acting drugs' effect on the stomach is front-loaded and partly fades; the short-acting one's does not.

Since 2024 the same question has been asked a different way, in fasted patients about to have an operation, with an ultrasound probe on the stomach. In a prospective study of 220 patients, 43 of 107 who had taken weekly semaglutide within ten days had increased residual gastric content against 3 of 113 unexposed patients — a propensity-weighted odds ratio of 36.97 (16.54 to 99.32) — with no case of aspiration ⁷⁸. Across three hospitals, 113 of 316 patients taking these drugs, 35.8 percent, had high residual content, and the thresholds that separated them were withholding the injection for 7.5 days or less and fasting from solid food for 21.3 hours or less ⁷⁹. A matched case-control study found a full stomach in 21 of 43 patients on semaglutide against 8 of 44 matched controls (odds ratio 4.29, 1.63 to 11.29) and solid content in 18 of 43 against 3 of 44 (9.85, 2.57 to 37.76) ⁸⁰. A two-center study of 93 patients, 52 of them on a weekly drug held for at least seven days, found increased residual content in 42 percent of those against 24 percent of controls — a difference its own authors call clinically relevant but statistically inconclusive in an underpowered sample, adjusted prevalence ratio 1.65 (0.7 to 3.7) ⁸¹. The 2026 review states the position plainly: increased residual gastric volume is demonstrated, confirmed aspiration events are not ²³. What the ultrasound studies themselves now argue for is guiding anesthetic management by what the stomach looks like on the day rather than by stopping the drug in everyone ^79,80.

Three commercial disclosures belong with those numbers. The two semaglutide studies were conducted by Novo Nordisk, which makes the drug, with four of the six authors on each paper being company employees ^75,76; the 2026 review of gastrointestinal adverse effects includes a co-author affiliated with the same company ²³. None of that makes the data wrong. It is simply part of what a reader should know.

The polyvagal theory's dorsal vagal state, measured against what this nucleus does

Outside neurology, the best-known claim about this nucleus comes from the polyvagal theory, set out by Stephen Porges and widely taught in trauma therapy, somatic practice and popular psychology. In its author's own 2025 summary, the theory proposes a hierarchical organization of autonomic states mediated by the vagus, with the ventral vagal complex uniquely supporting social engagement and physiological flexibility through mechanisms the theory names neuroception, co-regulation and dissolution ²⁴. In the clinical teaching derived from it, the dorsal vagal state — attributed to this nucleus — is the evolutionarily older, bottom tier: shutdown, collapse, freeze, dissociation.

In February 2026 a paper appeared that is unusual in the literature. Thirty-nine acknowledged experts in the physiology and evolution of the vagus nerve and in vertebrate social behavior, many of them cited in the polyvagal literature as supporting it, were invited by Paul Grossman to evaluate the theory; all but one accepted, the exception being a scholar unfamiliar with it. They appraised five elements: whether respiratory sinus arrhythmia is a direct measure of central vagal drive to the heart; the theory's characterizations of the neuroanatomy and function of the nucleus ambiguus and this nucleus; its account of vagal evolution; its claims about the specificity of mammalian social behavior; and its reading of the older physiological literature. All co-authors agreed that major tenets are not supported by past or current knowledge and in several instances are inconsistent with the broader evidence base, and concluded that the theory is untenable ²⁵. Grossman had made the same arguments single-handed in 2023 ⁸².

Set beside this page, the anatomical objection is the easy one to check. The theory requires two vagal nuclei with distinct cardiac effects, with this nucleus producing the shutdown response. The rat tracing data say that this nucleus reaches at least 18 and 12 percent of cardiac principal neurons against the nucleus ambiguus's 56 and 45 percent, with roughly a third of the branching and finer fibers ^5,6; this nucleus's territory is overwhelmingly subdiaphragmatic ¹. The clinical signal almost always used to index a dorsal vagal state, respiratory sinus arrhythmia, is cardiac, and cardiac vagal control is predominantly the nucleus ambiguus's job. Reading a state of this nucleus off a heart-rate measure is, on the available anatomy, reading the wrong organ.

Porges replied in the same issue. His argument is not that the anatomical counts are wrong but that the critique evaluates a reconstructed proxy of the theory, conflating neuroanatomy with neurophysiology, reducing theory to measurement, and substituting phylogenetic continuity for functional organization; an appendix argues that several of the objections were raised two decades earlier and already answered ²⁶. Readers can judge whether that is a defense of the theory's claims or a redescription of their scope. What the reply does not do is produce physiological data showing that this nucleus produces the cardiac and behavioral effects attributed to it.

A fair statement of what this nucleus actually does with the heart sits between the two positions, and it is narrower than either — but it is not an absence. It has real cardiac projections, counted and published ⁵. More than that, a peer-reviewed 2024 mouse study showed that optogenetically activating its cholinergic neurons in awake animals slows the heart hard: from 605 plus or minus 18 beats per minute before stimulation to 271 plus or minus 17 during it, in 13 mice, p below 0.0001, through peripheral muscarinic and ganglionic nicotinic acetylcholine receptors and not beta-1 adrenergic ones; photoinhibiting the same neurons left resting heart rate unchanged in 3 mice; retrograde tracing from the cardiac fat pad labels these cells; and chemogenetically activating them produced bradycardia alongside a correlated reduction in anxiety-like behavior ³⁹. A 2025 review by the same group describes the function of these cardiac-projecting neurons as a controversy now being resolved rather than a settled absence ⁸³, and a 2025 study found that fifteen days of a high-fat diet selectively increased the tonic GABA inhibition reaching them ⁸⁴. A 2026 rat study found that destroying this nucleus's cholinergic neurons blunted vagally evoked falls in blood pressure and heart rate, through a pathway the authors themselves call putative, in a single unreplicated study ⁴⁰; an unreviewed 2026 preprint reported that electrical stimulation of the nucleus slows the heart in mice ⁴¹. So the honest statement is both stronger and narrower than the theory's: this nucleus is not the main source of beat-to-beat vagal braking, but in mice it is sufficient to slow the heart — and the behavioral sign that came with that slowing was anxiolysis, not shutdown, the opposite valence to the one the theory assigns it. It is still rodent evidence about a rodent heart, and not a basis for a hierarchy of human emotional states. The dorsal vagal complex page treats the theory in full; this page is confined to the question of whether the structure can bear the weight.

One more point of fairness. The clinical practices built on the theory — attention to breathing, to safety, to co-regulation — may well help people, and whether they do is a separate empirical question that the anatomical dispute does not settle. The 2026 evaluation says this itself: its authors note that alternative explanations of the relations between vagal function and psychological processes exist, and that the paper does not address them ²⁵.

• Gastric emptying scintigraphyA meal labeled with a tiny amount of radioactive tracer is followed by camera as it leaves the stomach; the standard measure of the one output of this nucleus that can be quantified in a living person ²².Suspected gastroparesis; before and after vagotomy; measuring what a GLP-1 drug is doing to the stomach ^22,77The criterion standard in the 2022 American College of Gastroenterology and 2025 American Gastroenterological Association guidelines: gastric emptying scintigraphy with more than 10 percent retention at four hours, in a patient with symptoms and no mechanical obstruction on upper endoscopy or imaging. The severity bands — mild 10 to 15 percent, moderate 16 to 35 percent, severe above 35 percent — come from the 2022 American Gastroenterological Association clinical practice update, and a carbon-13 spirulina breath test is approved by the United States Food and Drug Administration as an alternative measure ²². Sensitive enough to show a 2.19-fold rise in gastric retention after eight weeks of lixisenatide ⁷⁷.It measures a result, not a nucleus. The same delay can come from weak antral contraction, a pylorus that will not relax, a stomach that will not accommodate the meal, or heightened visceral sensation, and the test cannot tell them apart: it measures retention, not the mechanism producing it ⁷³.
• Acetaminophen (paracetamol) absorption testAcetaminophen is absorbed in the small intestine but not the stomach, so how fast it appears in the blood after a meal is a proxy for how fast the stomach emptied ^75,76.Drug trials, where repeated scintigraphy would be impracticalSensitive to short, early effects: a 31 percent fall in the acetaminophen area-under-curve in the first postprandial hour on oral semaglutide ⁷⁶, and a first-hour ratio of 0.73 on injected semaglutide where overall emptying was not statistically different from placebo ⁷⁵.An indirect measure that also reflects intestinal absorption and liver metabolism, and it is weighted to the first hour, which is exactly why the two semaglutide studies can show an early delay and no overall delay in the same subjects ⁷⁵. Both were run by the manufacturer.
• Parasympathetic and sympathetic nerve imaging (carbon-11 donepezil positron emission tomography, metaiodobenzylguanidine scintigraphy)Carbon-11 donepezil binds the cholinesterase enzyme and so tracks cholinergic parasympathetic nerve density in the gut on positron emission tomography (PET); metaiodobenzylguanidine (MIBG) is taken up by sympathetic nerve endings in the heart and imaged by scintigraphy ⁶⁶.Research classification of Parkinson patients into body-first and brain-first trajectoriesSeparated the groups in the study that proposed the split: patients with rapid-eye-movement (REM) sleep behavior disorder and those with the isolated sleep disorder had reduced cardiac MIBG (p below 10 to the minus 5) and reduced colonic donepezil uptake (p equal to 0.008) against patients without it ⁶⁶.Neither tracer sees this nucleus. They image the nerve endings many synapses downstream, in a cross-sectional design, with 13 to 24 patients per group ⁶⁶. The trajectory is inferred from a snapshot.
• Postmortem alpha-synuclein immunohistochemistry and stagingBrain and gut tissue stained for the protein that aggregates in Lewy bodies, then scored against a published staging scheme ^13,16.The only way the pathology in this nucleus is actually seen; the entire Braak argument rests on itReproducible enough that most series of typical Parkinson disease find 80 to 100 percent of cases fitting the ascending scheme ⁴⁹, and good enough to show that gut Lewy pathology follows the distribution of this nucleus's parasympathetic input rather than sympathetic or intrinsic enteric neurons ⁵⁰.It cannot show direction or time. Every brain is seen once, at the end; the sequence is an inference from how cases of differing severity are arranged. The two older staging systems left up to 50 percent of cases unclassifiable, which is why a new one was needed ¹⁶.
• Heart rate variability and respiratory sinus arrhythmiaThe beat-to-beat quickening and slowing of the heart with breathing, widely offered in clinical and popular writing as a readout of vagal tone ⁸².Research and commercial devices claiming to measure parasympathetic stateA real and reproducible physiological signal, and a reasonable approximate index of cardiac vagal influence under controlled breathing.It is not a measure of this nucleus. Cardiac vagal output comes predominantly from the nucleus ambiguus, which in rats reaches at least 56 percent of cardiac neurons against this nucleus's 18 percent ^5,6. Its own cardiac neurons are not inert — optogenetically activating them in awake mice produces a robust slowing of the heart ³⁹ — but they are not the source of the beat-to-beat signal, and the 2026 expert evaluation concluded that treating respiratory sinus arrhythmia as a direct index of central vagal drive to the heart, and reading a dorsal vagal state off it, is not defensible ²⁵.

What is still unknown, and what would settle it

The first gap is that almost nothing on this page was measured in a living human being. The cardiac fiber counts that settle the polyvagal anatomy are rat counts ^5,6. The organ-by-organ tuning is rat slice electrophysiology ⁷. The insulin and glucagon circuitry is rat and mouse ^10,44,45. The propagation experiments are mouse, rat and baboon, and they disagree ^59,60,62,63. The only human evidence about this nucleus comes from autopsy tissue, from surgical registries recording operations that are no longer performed, and from imaging that sees its nerve endings several synapses away ^15,17,66. A method for assessing this nucleus in a living person would change the field, and none is on the horizon. The nerve, as opposed to the nucleus, has become measurable: a 2024 systematic review and meta-analysis of twelve case-control studies found the vagus nerve modestly thinner on neck ultrasound in Parkinson's disease — a pooled effect size of minus 0.79 standard deviations (minus 1.34 to minus 0.25, p equal to 0.004) — more marked on the right and smaller in women, and correlating with cardiac parasympathetic decline and with motor ratings. The authors are careful: the reduction is small, individually variable and of unclear clinical significance, and it tells you about the nerve, not about the cells the nerve comes from ⁸⁵.

The second is the direction of travel. Gut pathology is real, it is specific, and it precedes the diagnosis by years ¹⁹. But the protein moves out of the brain along this nucleus's own axons as readily as it moves in ³⁶, healthy appendixes contain the aggregates ⁵⁴, and in the species closest to us the gut-to-brain journey has repeatedly failed to happen ^62,63. The experiment that would settle it — following people prospectively from before any pathology, with a test sensitive enough to catch the earliest deposits, and seeing where they appear first — is exactly the experiment that current peripheral tests cannot support, since in autopsy-confirmed incidental Lewy body disease the skin and submandibular gland are positive in only about a third of cases ⁷².

The third is why these particular cells. The leading mechanistic proposal is that their unusual resting voltage and the calcium they handle with every spike make them metabolically fragile, which comes from mouse recordings and modeling ³³. It is a plausible account of selective vulnerability and it has not been shown in human tissue. If it is right, it predicts that the vulnerability is a property of the cell rather than of what arrives at it — which would undercut the whole seeding story for this nucleus, and nobody has tested that prediction directly. What has been tested is the therapeutic corollary, and it failed. In STEADY-PD III, 336 patients with early Parkinson's disease took immediate-release isradipine 5 mg twice daily or placebo for 36 months, and the drug did not slow progression: the adjusted change in the Unified Parkinson's Disease Rating Scale parts I to III was 2.99 points against 3.26, a treatment effect of minus 0.27 (minus 3.02 to 2.48), p equal to 0.85, with the investigators themselves noting that the dose may have been too low to engage the channels in question ⁸⁶. That is a negative trial of a mechanism aimed mainly at the midbrain rather than at this nucleus, and it does not refute the cell-fragility proposal; it does mean the proposal has had its one clinical test and did not pass it.

The fourth is therapeutic, and it is the one patients ask about. If the body-first route is real in some people, interrupting it early should prevent or delay the disease; no such intervention exists, and the gut-directed trials conducted so far — probiotics, fecal microbiota transplantation, dietary change — have not established disease modification ⁶⁸. On the other side, the drugs that act hardest on this nucleus's output are now taken by tens of millions of people for weight and for diabetes, and how their long-term effect on vagal gastric control interacts with any of this is simply unknown ²³. Stimulating the nucleus, rather than cutting its output, is now being tried in animals. In rats given streptozotocin into the cerebral ventricles to model Alzheimer-like degeneration, fifteen days of chemogenetic activation of its cholinergic neurons brought a raised cardiac sympathetic tone back to control levels, improved recognition memory and lowered hippocampal amyloid density in males; female rats showed no autonomic, memory or amyloid abnormality at the same time point, so the effect was demonstrated in one sex only ⁸⁷. Rats, an artificial model, one laboratory — and the first suggestion that this nucleus might be a target rather than only a casualty.

The fifth is one this page has no answer to at all. Several of the most striking recent findings about this nucleus — the estrogen-dependent sex difference in how oxytocin acts on it ⁸, the astrocytes that gate its signaling across the estrous cycle ³², the hypothalamic chain that drives gastritis through it ³⁰ — came from 2026, from a handful of laboratories, in rodents. Two more arrived in September 2026, both on a function this page has otherwise left aside, which is the nucleus's reach into the immune system. In mice, chronic restraint stress before an endotoxin challenge worsened mortality and inflammation, and the pathway identified was a circuit from the medial prefrontal cortex to this nucleus: chemogenetically activating it restored splenic nerve output, normalized splenic norepinephrine and acetylcholine, reduced inflammation and improved survival, and the benefit largely disappeared when the splenic sympathetic nerves were chemically destroyed ⁸⁸. Against that sits a more sobering result from a group working on lupus in mice, who report that directly stimulating the parasympathetic neurons of this nucleus did not change blood pressure or autoantibody levels at all, and improved only the kidney pathology ⁸⁹. Both are mouse work, the second reported within a review by its own authors, and neither has a human counterpart. The nucleus looks less like a simple output relay every year, and the gap between what is known about it in mice and what can be said about it in people is widening rather than closing.

Where it connects

Sources

89 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.

1. 1
   Berthoud HR, Munzberg H, Yu S, Morrison CD, Kim A, Neuhuber WL. Anatomy and function of efferent and afferent vagal innervation of the stomach.Auton Neurosci · 2026 · 267:103465doi:10.1016/j.autneu.2026.103465 · PMID 42628243

   The most recent full review of the wiring. States that the cholinergic parasympathetic efferent neurons sit in the dorsal motor nucleus, that they do not touch gastric effector cells directly but synapse onto myenteric plexus neurons acting as function-specific pattern generators, and that transcriptomic work points to a largely labeled-line organization. A review, not new data, and the detailed anatomy it summarizes is overwhelmingly rodent.

2. 2
   Travagli RA, Anselmi L. Vagal neurocircuitry and its influence on gastric motility.Nat Rev Gastroenterol Hepatol · 2016 · 13(7):389-401doi:10.1038/nrgastro.2016.76 · PMID 27251213

   The standard review of the vagovagal reflex loop: afferents to the solitary nucleus, solitary nucleus to the dorsal motor nucleus, dorsal motor nucleus out to the gut, with circulating hormones able to reach the circuit. Also covers plasticity after stress and gastrointestinal dysfunction in Parkinson disease. A review; its mechanistic detail is from rats.

3. 3
   Zhang K, Song D, Zhang Y, et al. Vagal control of the brain-esophagus axis ameliorates stress-induced esophageal motility dysfunction in male mice.Mol Metab · 2026 · 109:102384doi:10.1016/j.molmet.2026.102384 · PMID 42203107

   Chronic mild stress raised lower esophageal sphincter pressure in male mice; optogenetically driving the hypothalamus to dorsal motor nucleus to vagus pathway reproduced the abnormal contractions and vagotomy or acetylcholine receptor blockade abolished them. Extends the nucleus's reach above the diaphragm, to the esophagus. Male mice only.

4. 4
   de Souza DP, Martins Sá RW, de Souza LH, Ramalho FS, Moraes DJA. Electrophysiological properties of the dorsal motor nucleus of the vagus motoneurones controlling airway mucus secretion in rats.J Physiol · 2026 · online 14 Sep 2026doi:10.1113/JP291653 · PMID 42734251

   Rats. The airway-projecting parasympathetic motoneurons of this nucleus have potassium-dominated leak currents carried by two-pore-domain TWIK-2 channels, which set their resting membrane potential, firing frequency and their excitation by acidosis; acidosis also reduces the inhibition reaching them and activates local astrocytes, whose optogenetic activation drove action potentials in them through NMDA glutamate receptors. Chemogenetically activating them evoked airway mucus secretion and more sighing, without changing pulmonary ventilation, oxygen consumption or heart rate. Establishes a third territory, the airways, alongside the gut and the sparse cardiac projection. Rats.

5. 5
   Cheng Z, Powley TL, Schwaber JS, Doyle FJ 3rd. Projections of the dorsal motor nucleus of the vagus to cardiac ganglia of rat atria: an anterograde tracing study.J Comp Neurol · 1999 · 410(2):320-41PMID 10414536

   Counted the dorsal motor nucleus's own cardiac output in rats: 68-96 fibers from the left side and 67-115 from the right, innervating 709 (left) and 494 (right) cardiac principal neurons, at least 18 percent and 12 percent of them. Shows the nucleus does reach the heart. Rat anatomy; no human equivalent exists. The PubMed record carries no DOI.

6. 6
   Cheng Z, Powley TL. Nucleus ambiguus projections to cardiac ganglia of rat atria: an anterograde tracing study.J Comp Neurol · 2000 · 424(4):588-606PMID 10931483

   The companion study, same tracer and same laboratory, for the nucleus ambiguus: 82-151 fibers on the left and 60-122 on the right, innervating 2,248 and 1,784 principal neurons, at least 56 percent and 45 percent, with divergence of about 1:27 to 1:30 against the dorsal motor nucleus's roughly threefold lower figure, and more large (B-type) fibers where the dorsal motor nucleus sends fine C-type ones. This pair of papers is the cleanest quantitative comparison available, and it is in rats. No DOI on the record.

7. 7
   Grabauskas G, Moises HC. Gastrointestinal-projecting neurones in the dorsal motor nucleus of the vagus exhibit direct and viscerotopically organized sensitivity to orexin.J Physiol · 2003 · 549(Pt 1):37-56doi:10.1113/jphysiol.2002.029546 · PMID 12679367

   Whole-cell recordings from rat brainstem slices after labeling neurons by where they project. Orexin depolarized 47 of 102 neurons overall but 70 percent (30 of 43) of those going to the gastric fundus or corpus, against 3 of 13 to the antrum and pylorus, 4 of 18 to the duodenum and 1 of 13 to the cecum. The clearest demonstration that the nucleus is organized by target organ rather than uniformly. Rat slices.

8. 8
   Bhagat R, Travagli RA, Browning KN. Oestrogen modulates stress-induced vagally-dependent gastric responses to oxytocin.J Physiol · 2026 · 604(18):7570-7593doi:10.1113/JP291603 · PMID 42727053

   Female rats. Gastric emptying was slower when estrogen was high; oxytocin suppressed inhibitory GABA input to dorsal motor nucleus neurons under unstressed conditions, which does not happen in male rats; and oxytocin injected into the nucleus slowed the stomach through non-adrenergic, non-cholinergic vagal pathways using vasoactive intestinal peptide and nitric oxide. Rat electrophysiology and in vivo recording; the sex difference is the point and it has not been shown in humans.

9. 9
   Krowicki ZK. Role of selected peptides in the vagal regulation of gastric motor and endocrine pancreatic function.J Physiol Pharmacol · 1996 · 47(3):399-409PMID 8877896

   Review of the author's microinjection work: thyrotropin-releasing hormone and PACAP38 into the raphe obscurus excite the stomach while substance P and vasoactive intestinal polypeptide relax it, and pancreatic polypeptide into the dorsal vagal complex potentiates glucose-stimulated insulin secretion without changing basal levels. Rat work, summarized by one of its authors; no DOI on the record.

10. 10
    Mussa BM, Sartor DM, Rantzau C, Verberne AJM. Effects of nitric oxide synthase blockade on dorsal vagal stimulation-induced pancreatic insulin secretion.Brain Res · 2011 · 1394:62-70doi:10.1016/j.brainres.2011.04.015 · PMID 21530944

    Anesthetized rats. Chemically activating the nucleus raised the peak glucose-stimulated insulin response from 4.0 plus or minus 0.6 to 9.2 plus or minus 0.1 nanograms per milliliter; blocking muscarinic receptors cut it to 2.1 plus or minus 1.1; blocking nitric oxide synthesis raised it to 15.3 plus or minus 3.0. Direct evidence that the nucleus can set insulin output, in rats under anesthesia.

11. 11
    Johnson AG. Proximal gastric vagotomy: does it have a place in the future management of peptic ulcer?World J Surg · 2000 · 24(3):259-63doi:10.1007/s002689910042 · PMID 10658058

    States that Dragstedt introduced truncal vagotomy for duodenal ulcer in 1943, that proximal gastric vagotomy in experienced hands cures more than 90 percent of duodenal ulcers with minimal side effects, and that the discovery of Helicobacter pylori reduced the operation to a rarity. A surgeon's review; the cure rate is the author's summary of the literature, not a pooled estimate.

12. 12
    O'Leary JP. Dragstedt's first vagotomy in the treatment of duodenal ulcer disease.Am Surg · 1994 · 60(5):376-7PMID 8161091

    A short historical note on the first operation. The PubMed record carries no abstract and no DOI, so it is cited only for the existence and attribution of the event, not for any number.

13. 13
    Braak H, Del Tredici K, Rüb U, de Vos RAI, Jansen Steur ENH, Braak E. Staging of brain pathology related to sporadic Parkinson's disease.Neurobiol Aging · 2003 · 24(2):197-211doi:10.1016/s0197-4580(02)00065-9 · PMID 12498954

    The paper that put this nucleus at the center of Parkinson research. From incidental and symptomatic cases it proposes that lesions begin in the dorsal motor nucleus of the glossopharyngeal and vagal nerves and in the anterior olfactory nucleus, and ascend through the brainstem with little variation between individuals. It is a cross-sectional autopsy series read as a sequence; no one was followed over time.

14. 14
    Braak H, Rüb U, Gai WP, Del Tredici K. Idiopathic Parkinson's disease: possible routes by which vulnerable neuronal types may be subject to neuroinvasion by an unknown pathogen.J Neural Transm (Vienna) · 2003 · 110(5):517-36doi:10.1007/s00702-002-0808-2 · PMID 12721813

    The companion paper that states the mechanism as a hypothesis: an unidentified pathogen crossing the gut lining and traveling by retrograde axonal and transneuronal transport along unmyelinated preganglionic vagal fibers into the brain. Explicitly a hypothesis, with the falling-dominoes image supplied by the authors; no pathogen was identified then and none has been since.

15. 15
    Kalaitzakis ME, Graeber MB, Gentleman SM, Pearce RKB. The dorsal motor nucleus of the vagus is not an obligatory trigger site of Parkinson's disease: a critical analysis of alpha-synuclein staging.Neuropathol Appl Neurobiol · 2008 · 34(3):284-95doi:10.1111/j.1365-2990.2007.00923.x · PMID 18053026

    Seventy-one Parkinson brains from a single bank. The substantia nigra was affected in 100 percent and the nucleus basalis of Meynert in 98.5 percent; 53 percent fit the proposed caudal-to-rostral spread and 47 percent did not, and in 7 percent the dorsal motor nucleus was clean although the nigra and cortex were not. The strongest single counterexample to the obligatory-trigger reading of Braak.

16. 16
    Beach TG, Adler CH, Lue L, et al. Unified staging system for Lewy body disorders: correlation with nigrostriatal degeneration, cognitive impairment and motor dysfunction.Acta Neuropathol · 2009 · 117(6):613-34doi:10.1007/s00401-009-0538-8 · PMID 19399512

    Ten standard brain regions from 417 subjects. Notes that the two major staging systems then in use failed to classify up to 50 percent of cases, and proposes stages that include an olfactory-bulb-only stage and a limbic-predominant route that bypasses the brainstem. The alternative map in which this nucleus is one possible entrance rather than the entrance.

17. 17
    Svensson E, Horváth-Puhó E, Thomsen RW, Djurhuus JC, Pedersen L, Borghammer P, Sørensen HT. Vagotomy and subsequent risk of Parkinson's disease.Ann Neurol · 2015 · 78(4):522-9doi:10.1002/ana.24448 · PMID 26031848

    Danish registry cohort of everyone vagotomized between 1977 and 1995. Truncal vagotomy carried a hazard ratio of 0.85 (95 percent confidence interval 0.63 to 1.14) against the general population, falling to 0.53 (0.28 to 0.99) beyond 20 years of follow-up, while superselective vagotomy showed nothing. The authors state plainly that the statistical precision of the estimates was limited; the headline result rests on a subgroup whose interval only just excludes 1.

18. 18
    Liu B, Fang F, Pedersen NL, et al. Vagotomy and Parkinson disease: A Swedish register-based matched-cohort study.Neurology · 2017 · 88(21):1996-2002doi:10.1212/WNL.0000000000003961 · PMID 28446653

    The larger replication: 9,430 vagotomized patients (3,445 truncal, 5,978 selective) matched 40 to 1 against 377,200 others, with 4,930 incident Parkinson cases over 7.3 million person-years. Overall vagotomy was not associated with Parkinson disease (hazard ratio 0.96, 0.78 to 1.17); truncal vagotomy gave 0.78 (0.55 to 1.09), and only the subgroup operated at least five years before diagnosis reached 0.59 (0.37 to 0.93). A partial, weaker replication, not a confirmation.

19. 19
    Konings B, Villatoro L, Van den Eynde J, et al. Gastrointestinal syndromes preceding a diagnosis of Parkinson's disease: testing Braak's hypothesis using a nationwide database for comparison with Alzheimer's disease and cerebrovascular diseases.Gut · 2023 · 72(11):2103-2111doi:10.1136/gutjnl-2023-329685 · PMID 37620120

    24,624 patients with new idiopathic Parkinson disease in a United States records network. Gastroparesis, dysphagia, irritable bowel syndrome without diarrhea and constipation were specifically associated with later Parkinson disease (odds ratios 4.64, 3.58, 3.53 and 3.32), and appendectomy lowered risk (relative risk 0.48) - but vagotomy showed no association at all. The most recent direct test of the vagotomy claim and the largest set of Parkinson cases, and on vagotomy it is null. The abstract reports no vagotomy-exposed cohort size — the 24,624 figure counts new Parkinson cases — so the null says nothing about effect size.

20. 20
    Andersen KB, Krishnamurthy A, Just MK, et al. Sympathetic and parasympathetic subtypes of body-first Lewy body disease observed in postmortem tissue from prediagnostic individuals.Nat Neurosci · 2025 · 28(5):925-936doi:10.1038/s41593-025-01910-9 · PMID 40082617

    Two autopsy datasets with 173 and 129 Lewy-positive cases. Early prediagnostic cases separated cleanly into brain-first and body-first, and a machine-learning algorithm split the body-first group again into a vagal-predominant and a sympathetic-predominant subtype in roughly equal numbers. The paper that demotes the vagus from the peripheral route to one of two. Postmortem and cross-sectional; the sequence is inferred, not observed.

21. 21
    Borghammer P, Van Den Berge N, Berg D, Takahashi R, Postuma RB, Horsager J. Brain-first and body-first subtypes of Lewy body disease.Nat Rev Neurol · 2026 · 22(9):564-580doi:10.1038/s41582-026-01241-8 · PMID 42463844

    The current critical appraisal by the group that proposed the model. Summarizes the supporting lines - prodromal autonomic and sleep disorder, cardiac sympathetic denervation preceding nigrostriatal degeneration by roughly a decade in people with prodromal sleep disorder, and bottom-up or top-down gradients at autopsy - and sets out the model's limitations. A review by its own proponents, which is a conflict of a kind worth naming.

22. 22
    Camilleri M. Gastroparesis: A Review.JAMA · 2026 · 336(7):587-602doi:10.1001/jama.2026.12181 · PMID 42485161

    Current clinical reference. Prevalence of definite gastroparesis 21.5 per 100,000 persons in a 2018 United States claims database; among 82.6 million patients the causes were type 2 diabetes 51.7 percent, postsurgical 15 percent, medication-induced 11.8 percent, idiopathic 11.3 percent, type 1 diabetes 5.7 percent, other 4.5 percent; scintigraphy with more than 10 percent retention at four hours is the criterion standard. A narrative review of guidelines and claims data, not a prospective study.

23. 23
    Singhani V, Ehsan M, Valencia S, et al. Gastrointestinal adverse effects of GLP-1 receptor agonists and dual GIP/GLP-1 receptor agonists: A narrative expert review of approved therapies, oral formulations, and pipeline agents.Nutr Clin Pract · 2026 · 41(5):1318-1333doi:10.1002/ncp.70180 · PMID 42808923

    States that nausea, vomiting, diarrhea and constipation occur in roughly 30 to 50 percent of patients, usually at initiation and dose escalation, and attributes them to delayed gastric emptying, central activation of emetic pathways, altered intestinal motility and rapid weight loss itself. A narrative review, not a pooled analysis; one co-author is affiliated with Novo Nordisk.

24. 24
    Porges SW. Polyvagal Theory: Current Status, Clinical Applications, and Future Directions.Clin Neuropsychiatry · 2025 · 22(3):169-184doi:10.36131/cnfioritieditore20250301 · PMID 40735382

    The theory's author stating it in his own words, including the hierarchical organization of autonomic states and the role assigned to the ventral vagal complex. Cited here so the claims under dispute are taken from their source rather than from critics. A narrative review by the theory's originator.

25. 25
    Grossman P, Ackland GL, Allen AM, et al. Why The Polyvagal Theory Is Untenable: An international expert evaluation of the polyvagal theory and commentary upon Porges, S.W. (2025).Clin Neuropsychiatry · 2026 · 23(1):100-112doi:10.36131/cnfioritieditore20260110 · PMID 41768017

    Thirty-nine invited experts in vagal physiology and vertebrate evolution, all but one of whom accepted, appraise five core claims of the polyvagal theory, among them its characterization of the neuroanatomy and function of the nucleus ambiguus and the dorsal motor nucleus. All co-authors agree the major tenets are unsupported. It is an expert consensus statement and a review of existing evidence, not new experimental data.

26. 26
    Porges SW. When A Critique Becomes Untenable: A Scholarly Response To Grossman Et Al.'s Evaluation Of Polyvagal Theory.Clin Neuropsychiatry · 2026 · 23(1):113-128doi:10.36131/cnfioritieditore20260111 · PMID 41768026

    The reply published alongside the expert evaluation. Argues the critique attacks a reconstructed proxy of the theory and conflates neuroanatomy with neurophysiology. It does not present new physiological data and does not dispute the anatomical findings themselves.

27. 27
    Liu Y, Liu Y, Wang Y, Sun H, Zhu B, Gao X. [Neural mechanism of acupuncture-driven vagus nerve-gastric reflex regulating gastric motility in mice with diabetic gastroparesis].Zhongguo Zhen Jiu · 2026 · 46(5):773-781doi:10.13703/j.0255-2930.20251124-k0001 · PMID 42116782

    In streptozotocin-and-high-fat-diet diabetic mice, the number of activated cholinergic neurons in the dorsal motor nucleus was lower than in controls, and subdiaphragmatic vagotomy largely abolished the gastric motility response — amplitude, area under the curve and frequency — evoked by electroacupuncture at ST36 and PC6. Cited only as weak secondary support for reduced cholinergic drive in diabetic gastroparesis; the arm that counted the neurons used three mice per group, and the stronger evidence for impaired drive from this nucleus is [72].

28. 28
    Blake CB, Smith BN. cAMP-dependent insulin modulation of synaptic inhibition in neurons of the dorsal motor nucleus of the vagus is altered in diabetic mice.Am J Physiol Regul Integr Comp Physiol · 2014 · 307(6):R711-20doi:10.1152/ajpregu.00138.2014 · PMID 24990858

    In control mice with cyclic AMP raised, insulin reduced inhibitory synaptic input onto these neurons, an effect blocked by disrupting protein trafficking or the GLUT4 glucose transporter. In streptozotocin-diabetic mice insulin did nothing. A concrete synaptic lesion in diabetes, demonstrated in mouse brain slices only.

29. 29
    Kaneko H, Taché Y, Kusugami K. Importance of medullary thyrotropin-releasing hormone in brain-gut circuits regulating gastric integrity: preclinical studies.J Gastroenterol · 2002 · 37 Suppl 14:128-32doi:10.1007/BF03326431 · PMID 12572880

    Review of rat experiments showing the two faces of drive through this nucleus: thyrotropin-releasing hormone microinjected into it produced vagally mediated gastric erosions resembling those of cold restraint stress, while a subthreshold dose protected the stomach against a strong irritant, and vagotomy abolished that protection. Rats only; a review of the authors' own studies.

30. 30
    Liu E, Xin X, Shi M, et al. Hcn1-dependent engram neurons in the PVN encode gastric inflammatory sensitization.Sci Adv · 2026 · 12(20):eaeb6961doi:10.1126/sciadv.aeb6961 · PMID 42139357

    Mouse work tracing a hypothalamus to dorsal motor nucleus to stomach circuit that drives gastritis, with the HCN1 ion channel raised in the hypothalamic neurons and stress able to reactivate them. Shows the nucleus as the final common path by which a brain state reaches the stomach wall. Mice; the engram labeling method and chemogenetics have no human counterpart.

31. 31
    Gao S, Feng X, Zhang N, et al. The Serotonergic and Cholinergic Signaling Pathways Drive Gastric Precancerous Lesions and Its Potential Mechanism.Cancer Sci · 2026 · online 17 Oct 2026doi:10.1111/cas.70563 · PMID 42888890

    In a chemically induced mouse model of gastric precancerous change, cholinergic neurons in the dorsal motor nucleus were overactive, and chemogenetically silencing them lowered gastric choline acetyltransferase — the enzyme that makes acetylcholine — and serotonin, and reduced the mucosal lesions. Mouse model work, published two days before this page was reviewed, with no human replication; it is evidence that too much drive through this nucleus can be harmful, not that it causes cancer in people.

32. 32
    Ozkaya KS, Yalcin C, Haar EE, Bhagat R, Browning KN. Brainstem Astrocytes Regulate Estrus-Dependent Oscillations in Food Intake.J Neurochem · 2026 · 170(6):e70498doi:10.1111/jnc.70498 · PMID 42313475

    Chemogenetic silencing of dorsal vagal complex astrocytes abolished the normal rise and fall of food intake across the rat estrous cycle, and activating them lowered intake regardless of cycle stage, with the effect on dorsal motor nucleus neurons running through NMDA (N-methyl-D-aspartate) glutamate receptors. Shows the non-neuronal cells around this nucleus are part of the circuit. Rats.

33. 33
    Cooper G, Lasser-Katz E, Simchovitz A, Sharon R, Soreq H, Surmeier DJ, Goldberg JA. Functional segregation of voltage-activated calcium channels in motoneurons of the dorsal motor nucleus of the vagus.J Neurophysiol · 2015 · 114(3):1513-20doi:10.1152/jn.00432.2014 · PMID 26156385

    Mouse slice recordings with modeling. These motor neurons sit at an unusually depolarized resting voltage, their Cav1.3 calcium currents are more than an order of magnitude smaller than their high-voltage-activated currents, and constitutively open Kv4 potassium channels keep their firing slow. Offered as a reason the cells carry a heavy calcium and mitochondrial oxidant load. Mouse; the link to human disease is inference.

34. 34
    Fang H, Liu A, Yang JF, et al. PVN regulates acinar cell-mediated inflammatory response via DMV projections to the acinar cell.J Neuroinflammation · 2026 · 23(1)doi:10.1186/s12974-026-03870-8 · PMID 42186091

    Traces a hypothalamic oxytocin neuron to dorsal motor nucleus cholinergic neuron to pancreatic acinar cell chain in mice, and shows that silencing the hypothalamic end suppresses digestive enzyme secretion and inflammation in a hypertriglyceridemia-associated pancreatitis model while stimulating it causes inflammation. Mouse circuit work in a disease model; no human data.

35. 35
    Li R, Liu M, Zhang Z, et al. High ambient temperature activates a neural circuit for gut glucose uptake in male mice.Nat Commun · 2026 · 17(1)doi:10.1038/s41467-026-75522-7 · PMID 42448680

    Heat raised intestinal glucose absorption through the sodium-glucose co-transporter 1 (SGLT1), the protein that carries glucose across the intestinal lining, in male mice, and blocking vagal motor transmission abolished the effect, with tracing showing a preoptic to hypothalamic to dorsal motor nucleus chain. An example of the nucleus setting absorption, not only motility. Male mice.

36. 36
    Ulusoy A, Phillips RJ, Helwig M, Klinkenberg M, Powley TL, Di Monte DA. Brain-to-stomach transfer of α-synuclein via vagal preganglionic projections.Acta Neuropathol · 2017 · 133(3):381-393doi:10.1007/s00401-016-1661-y · PMID 28012041

    Human alpha-synuclein overexpressed in the rat midbrain reached the stomach wall, accumulating in preganglionic vagal terminals, and was found in visceromotor but not viscerosensory gastric endings. The traffic runs both ways: the nucleus is a relay out of the brain as well as into it, which weakens the inference that peripheral pathology proves a peripheral origin. Rat, viral overexpression.

37. 37
    Kalia M, Mesulam MM. Brain stem projections of sensory and motor components of the vagus complex in the cat: II. Laryngeal, tracheobronchial, pulmonary, cardiac, and gastrointestinal branches.J Comp Neurol · 1980 · 193(2):467-508doi:10.1002/cne.901930211 · PMID 7440778

    The second half of the 1980 cat tracing study, following the vagal branches one at a time: laryngeal, tracheobronchial, pulmonary, cardiac and gastrointestinal. The PubMed record carries no abstract, so it is cited here only for the existence and the scope of that branch-by-branch mapping, which is where the modern organ-by-organ account of the vagal complex begins.

38. 38
    McAllen RM, Spyer KM. The location of cardiac vagal preganglionic motoneurones in the medulla of the cat.J Physiol · 1976 · 258(1):187-204doi:10.1113/jphysiol.1976.sp011414 · PMID 940054

    Cat electrophysiology. Of 57 neurons activated backward from the cardiac vagal branches, with conduction velocities of 3 to 15 meters per second, none lay in the region of the dorsal motor nucleus despite repeated sampling and all lay in the region of the nucleus ambiguus, the 46 marked ones in its principal column rostral to the obex; only 3 of 33 dorsal motor neurons responded, at high intensity, and were judged unlikely to be cardio-inhibitory. The experiment that moved cardiac vagal control from this nucleus to the nucleus ambiguus. Cats, 1976, electrophysiological localization only.

39. 39
    Strain MM, Conley NJ, Kauffman LS, et al. Dorsal motor vagal neurons can elicit bradycardia and reduce anxiety-like behavior.iScience · 2024 · 27(3):109137doi:10.1016/j.isci.2024.109137 · PMID 38420585

    Awake mice. Optogenetically activating cholinergic neurons of this nucleus lowered heart rate from 605 plus or minus 18 to 271 plus or minus 17 beats per minute (13 mice, p below 0.0001) through peripheral muscarinic and ganglionic nicotinic but not beta-1 adrenergic receptors, while photoinhibition left resting heart rate unchanged (629 plus or minus 45 against 627 plus or minus 38, 3 mice, p equal to 0.8950); retrograde tracing from the cardiac fat pad labeled these cells, which were more hyperexcitable than their nucleus ambiguus counterparts, and chemogenetic activation produced bradycardia with a correlated fall in anxiety-like behavior. The peer-reviewed demonstration that this nucleus is sufficient, though not necessary, to slow the murine heart. Mice.

40. 40
    Masukawa D, Nakao Y, Bito H, et al. Cholinergic neurons in the dorsal motor nucleus of the vagus contribute to vagally mediated cardiovascular responses through a putative pathway involving the nodose ganglion.J Pharmacol Sci · 2026 · 162(3):197-205doi:10.1016/j.jphs.2026.08.003 · PMID 42810776

    Rat study arguing the dorsal motor nucleus has a cardiovascular role after all, but an indirect one: killing its cholinergic neurons blunted the fall in blood pressure and heart rate evoked by vagal afferent stimulation, and carbachol put into the nodose ganglion produced those responses, abolished by atropine. The authors themselves call the pathway putative; it is one rat study and has not been replicated.

41. 41
    Falvey A, Palandira SP, Chaudhry S, et al. Cholinergic impairment in the dorsal motor nucleus of the vagus during experimental Alzheimer's disease.bioRxiv · 2026 · preprint, posted 25 Sep 2026doi:10.64898/2026.09.19.752906 · PMID 42818216

    In 5xFAD mice, a strain carrying five Alzheimer-related mutations, the density of cholinergic neurons in this nucleus fell by six and ten months, and electrical stimulation of the nucleus lost both its heart-rate-slowing and its anti-inflammatory effect with age. This is a preprint and has not been peer reviewed; it is cited only to mark an open line of work, and nothing in the page rests on it.

42. 42
    Salgado-Delgado RC, Espinosa-Tanguma R, Valdés Abadía B, Ramírez-Plascencia OD, Escobar C, Saderi N. Feeding during the resting phase causes gastrointestinal tract dysfunction and desynchronization of metabolic and neuronal rhythms in rats.Neurogastroenterol Motil · 2023 · 35(12):e14687doi:10.1111/nmo.14687 · PMID 37815021

    Four weeks of feeding rats only during their rest phase disturbed gastric emptying, intestinal motility, gut hormone secretion and the activity of the solitary nucleus, area postrema and dorsal motor nucleus together. Evidence that the circuit is set by time of day as well as by meals. Rats.

43. 43
    Zaman A, Özçelik H, Yücel E, et al. Effect of sex on chronic stress induced alterations in hindbrain catecholaminergic system and autonomic dysfunction resulting in gastrointestinal dysmotility.Brain Res · 2024 · 1842:149112doi:10.1016/j.brainres.2024.149112 · PMID 38969083

    Chronic stress delayed gastric emptying in male but not female rats, yet cholinergic staining in the dorsal motor nucleus itself did not change with sex or with stress - the changes were in the catecholamine cell groups around it. A useful negative control: not every gastric effect of stress shows up as a change inside this nucleus. Rats.

44. 44
    Verberne AJM, Mussa BM. Neural control of pancreatic peptide hormone secretion.Peptides · 2022 · 152:170768doi:10.1016/j.peptides.2022.170768 · PMID 35189258

    Review setting out that pancreatic parasympathetic premotor neurons sit principally in the dorsal motor nucleus, are electrophysiologically distinguishable from the ones controlling the stomach, and are driven during low blood sugar by a hypothalamic chain ending in orexin neurons. A review of mostly animal work; the failure of glucagon release in type 1 diabetes is stated as unexplained.

45. 45
    Senthilkumaran M, Bobrovskaya L, Verberne AJM, Llewellyn-Smith IJ. Insulin-responsive autonomic neurons in rat medulla oblongata.J Comp Neurol · 2018 · 526(16):2665-2682doi:10.1002/cne.24523 · PMID 30136719

    Insulin at 10 units per kilogram dropped blood glucose from 4.9 plus or minus 0.3 to 1.7 plus or minus 0.2 millimoles per liter and switched on the immediate-early protein Fos in many cholinergic dorsal motor nucleus neurons, concentrated beneath and just in front of the area postrema, while control animals had none. Maps which part of the nucleus answers to hypoglycemia. Rats; Fos marks recent activity, not function.

46. 46
    Liu X, Shi M, Ren H, et al. Excitatory Impact of Dental Occlusion on Dorsal Motor Nucleus of Vagus.Front Neural Circuits · 2021 · 15:638000doi:10.3389/fncir.2021.638000 · PMID 33776655

    Tracing was done in rats and the bite model in mice. Cholera toxin B injected into the rat inferior alveolar nerve labeled trigeminal axon endings around the cholinergic neurons of this nucleus; mice with a deliberately disturbed bite showed more excitatory input onto those neurons, higher pancreatic acetylcholinesterase, higher insulin and glucagon and a higher postprandial blood glucose, all reversed when the bite was corrected. An unexpected input to the nucleus; a single laboratory, rodents only, and the dental conclusion should not be carried to people.

47. 47
    Kalia M, Mesulam MM. Brain stem projections of sensory and motor components of the vagus complex in the cat: I. The cervical vagus and nodose ganglion.J Comp Neurol · 1980 · 193(2):435-65doi:10.1002/cne.901930210 · PMID 7440777

    Horseradish peroxidase histochemistry and amino acid autoradiography in 32 cats. Labeled motor neurons in the dorsal motor nucleus, the nucleus ambiguus and neighboring nuclei; showed that nucleus ambiguus axons first run dorsomedially toward the dorsal motor nucleus, loop back in a hairpin and then leave with its axons; and found vagal afferent labeling around the neurons of the dorsal motor nucleus, which the authors propose as the substrate for monosynaptic vagal reflexes. The modern anatomical foundation. Cats; a tracing study only.

48. 48
    Braak H, de Vos RAI, Bohl J, Del Tredici K. Gastric alpha-synuclein immunoreactive inclusions in Meissner's and Auerbach's plexuses in cases staged for Parkinson's disease-related brain pathology.Neurosci Lett · 2005 · 396(1):67-72doi:10.1016/j.neulet.2005.11.012 · PMID 16330147

    Five autopsy cases whose brains had been staged. Alpha-synuclein inclusions were found in submucosal Meissner plexus neurons whose axons run into the gastric mucosa, which the authors offer as the first link in an unbroken chain from gut to brain. Five cases, no controls reported in the abstract, and the chain is an interpretation of static tissue.

49. 49
    Jellinger KA. Is Braak staging valid for all types of Parkinson's disease?J Neural Transm (Vienna) · 2019 · 126(4):423-431 (online 25 Jun 2018)doi:10.1007/s00702-018-1898-9 · PMID 29943229

    A neuropathologist's balanced reading: most series of typical Parkinson disease find that 80 to 100 percent of cases fit the scheme, but incidental Lewy body disease and Parkinson disease can both show pathology in the substantia nigra with no involvement of the dorsal motor nucleus, and the scheme fits young-onset slow cases better than late-onset rapid ones. A review, and the percentages are its summary of other people's series.

50. 50
    Annerino DM, Arshad S, Taylor GM, Adler CH, Beach TG, Greene JG. Parkinson's disease is not associated with gastrointestinal myenteric ganglion neuron loss.Acta Neuropathol · 2012 · 124(5):665-80doi:10.1007/s00401-012-1040-2 · PMID 22941241

    Quantitative comparison of myenteric neuron density along the human gut in Parkinson disease, incidental Lewy body disease and controls. Found no neuron loss, and found that the Lewy pathology follows the distribution of parasympathetic input from the dorsal motor nucleus rather than sympathetic input or the intrinsic enteric neurons. A negative result that redirects attention from the enteric nervous system to the vagal projection itself.

51. 51
    Tysnes OB, Kenborg L, Herlofson K, Steding-Jessen M, Horn A, Olsen JH, Reichmann H. Does vagotomy reduce the risk of Parkinson's disease?Ann Neurol · 2015 · 78(6):1011-2doi:10.1002/ana.24531 · PMID 26418122

    A letter challenging the Danish vagotomy result in the journal that published it. The PubMed record carries no abstract, so it is cited only for the existence and the thrust of the published challenge.

52. 52
    Borghammer P, Hamani C. Preventing Parkinson disease by vagotomy: Fact or fiction?Neurology · 2017 · 88(21):1982-1983doi:10.1212/WNL.0000000000003969 · PMID 28446643

    The editorial published with the Swedish replication, co-authored by an author of the Danish cohort. An editorial, not data; the PubMed record carries no abstract, so it is cited for its existence, its title and its authorship.

53. 53
    Lin SY, Lin CL, Wang IK, et al. Dementia and vagotomy in Taiwan: a population-based cohort study.BMJ Open · 2018 · 8(3):e019582doi:10.1136/bmjopen-2017-019582 · PMID 29602843

    155,944 vagotomized patients and the same number of matched controls in Taiwan. Dementia incidence was 2.43 against 2.84 per 1,000 person-years and the adjusted hazard ratio was 1.09 (0.87 to 1.36), with truncal vagotomy at 1.04 (0.87 to 1.25). Cutting the vagus did not protect against dementia, which limits how broadly the protective claim can be generalized.

54. 54
    Killinger BA, Madaj Z, Sikora JW, et al. The vermiform appendix impacts the risk of developing Parkinson's disease.Sci Transl Med · 2018 · 10(465)doi:10.1126/scitranslmed.aar5280 · PMID 30381408

    Two epidemiological datasets covering more than 1.6 million people and over 91 million person-years found appendectomy decades earlier associated with lower Parkinson risk, especially in rural residents, and with later onset; healthy human appendix tissue contained aggregated and truncated alpha-synuclein. Association in the epidemiology, and the presence of aggregates in healthy tissue cuts both ways: it shows the protein is normally there.

55. 55
    Svensson E, Horváth-Puhó E, Stokholm MG, Sørensen HT, Henderson VW, Borghammer P. Appendectomy and risk of Parkinson's disease: A nationwide cohort study with more than 10 years of follow-up.Mov Disord · 2016 · 31(12):1918-1922doi:10.1002/mds.26761 · PMID 27621223

    Danish registries: 265,758 people appendectomized between 1980 and 2010 against 1,328,790 matched comparison subjects. Beyond ten years, Parkinson incidence was 0.19 against 0.15 per 1,000 person-years, an adjusted hazard ratio of 1.14 (1.03 to 1.27), consistent past twenty years and in both sexes. A slightly raised risk, not protection, from the same group that produced the vagotomy cohort.

56. 56
    Palacios N, Hughes KC, Cereda E, Schwarzschild MA, Ascherio A. Appendectomy and risk of Parkinson's disease in two large prospective cohorts of men and women.Mov Disord · 2018 · 33(9):1492-1496doi:10.1002/mds.109 · PMID 30218460

    The Nurses' Health Study and the Health Professionals Follow-up Study. Pooled hazard ratio for self-reported appendectomy 1.08 (0.94 to 1.23); in women, appendectomy for appendicitis 1.23 (1.00 to 1.50) while incidental appendectomy showed nothing. Self-reported exposure, but prospective and large: no association.

57. 57
    Park KW, Woo HT, Hwang YS, Lee SH, Chung SJ. Appendectomy and the Risk of Parkinson's Disease: A Korean Nationwide Study.Mov Disord Clin Pract · 2024 · 11(6):704-707doi:10.1002/mdc3.14031 · PMID 38696328

    16,122 appendectomies among 703,831 subjects in the Korean National Health Insurance Service sample cohort. Unadjusted hazard ratio for Parkinson disease 1.32 (0.97 to 1.80), adjusted 1.42 (0.88 to 2.30); no difference by sex. The first test of the appendectomy claim outside the Western world, and it is null.

58. 58
    Ma L, Wang HB, Hashimoto K. The vagus nerve: An old but new player in brain-body communication.Brain Behav Immun · 2024 · 124:28-39doi:10.1016/j.bbi.2024.11.023 · PMID 39566667

    Review of the vagus as a two-way channel, including the cholinergic anti-inflammatory pathway through alpha-7 nicotinic receptors and transcutaneous auricular stimulation. Cited for framing; it repeats the truncal-vagotomy-protects claim without the weight of the negative replications, which is itself worth noticing.

59. 59
    Kim S, Kwon SH, Kam TI, et al. Transneuronal Propagation of Pathologic α-Synuclein from the Gut to the Brain Models Parkinson's Disease.Neuron · 2019 · 103(4):627-641.e7doi:10.1016/j.neuron.2019.05.035 · PMID 31255487

    The strongest animal support for the vagal route. Preformed alpha-synuclein fibrils injected into the mouse duodenal and pyloric muscle layer produced pathology first in the dorsal motor nucleus, then the caudal hindbrain including the locus coeruleus, and much later the amygdala, dorsal raphe and substantia nigra, with dopamine neuron loss and behavioral deficits; truncal vagotomy and alpha-synuclein deficiency prevented all of it. Mice, and fibrils injected in quantities no stomach ever meets.

60. 60
    Van Den Berge N, Ferreira N, Gram H, et al. Evidence for bidirectional and trans-synaptic parasympathetic and sympathetic propagation of alpha-synuclein in rats.Acta Neuropathol · 2019 · 138(4):535-550doi:10.1007/s00401-019-02040-w · PMID 31254094

    Fibrils into the rat duodenal wall produced pathology along both parasympathetic and sympathetic routes - enteric nervous system, autonomic ganglia, spinal intermediolateral nucleus, heart, dorsal motor nucleus and locus coeruleus - for at least four months, and only in transgenic rats carrying extra human alpha-synuclein. Nothing propagated in wild-type rats. The result that made the sympathetic route a serious competitor, and it depends on protein overexpression.

61. 61
    Wang M, Sun G, Wan P, et al. Duodenal alpha-Synuclein PFF Injection Induces Vagus-Related Gut-to-Brain Pathology in C57BL/6J and A53T Transgenic Mice.Brain Sci · 2026 · 16(8):804doi:10.3390/brainsci16080804 · PMID 42651119

    Preformed alpha-synuclein fibrils into the duodenal muscularis of wild-type C57BL/6J and A53T transgenic mice. At four months the wild-type animals had phosphorylated alpha-synuclein in both the duodenal muscle layer and the striatum, with altered gut bacteria and motor deficits, and truncal vagotomy was associated with less of the protein in the brain. The wild-type result answers the main objection to the 2019 rat study, but the vagotomy finding is an association rather than a controlled test, and the authors state that the transgenic arm had no genetically matched wild-type control.

62. 62
    Manfredsson FP, Luk KC, Benskey MJ, et al. Induction of alpha-synuclein pathology in the enteric nervous system of the rat and non-human primate results in gastrointestinal dysmotility and transient CNS pathology.Neurobiol Dis · 2018 · 112:106-118doi:10.1016/j.nbd.2018.01.008 · PMID 29341898

    The same experiment with the opposite result. Rats showed minor dorsal motor nucleus and locus coeruleus pathology at one month that was gone later, and monkeys had no brain pathology at twelve months despite robust and persistent enteric pathology. The authors conclude that sustained spread from gut to brain is a rare event and enteric pathology may be an epiphenomenon.

63. 63
    Arotcarena ML, Dovero S, Prigent A, et al. Bidirectional gut-to-brain and brain-to-gut propagation of synucleinopathy in non-human primates.Brain · 2020 · 143(5):1462-1475doi:10.1093/brain/awaa096 · PMID 32380543

    Baboons given patient-derived Lewy body extracts into either the striatum or the gut developed nigrostriatal and enteric pathology either way - but no alpha-synuclein lesions were found in the vagus nerve. The authors state the study does not support transmission through the vagus and the dorsal motor nucleus, and propose the general circulation instead. A primate study, which makes the negative finding hard to dismiss.

64. 64
    Yu X, Persillet M, Zhang L, et al. Evaluation of blood flow as a route for propagation in experimental synucleinopathy.Neurobiol Dis · 2021 · 150:105255doi:10.1016/j.nbd.2021.105255 · PMID 33421564

    Mice were surgically joined so they shared a circulation; one received patient-derived alpha-synuclein aggregates into the substantia nigra. The injected animal lost dopamine neurons, the partner showed no lesion and no change in phosphorylated alpha-synuclein. Tests and fails to support the bloodstream route proposed by the baboon study, within the time frame used.

65. 65
    Ahn EH, Kang SS, Liu X, et al. Initiation of Parkinson's disease from gut to brain by δ-secretase.Cell Res · 2020 · 30(1):70-87doi:10.1038/s41422-019-0241-9 · PMID 31649329

    Proposes a molecular mechanism for the gut-to-brain route: the enzyme delta-secretase cuts alpha-synuclein at residue 103 and tau at 368, the fragments aggregate and travel along the vagus faster than full-length protein, and oral rotenone triggers the process. Cell, mouse and human tissue work; it supplies a mechanism for a route whose existence in people is still disputed.

66. 66
    Horsager J, Andersen KB, Knudsen K, et al. Brain-first versus body-first Parkinson's disease: a multimodal imaging case-control study.Brain · 2020 · 143(10):3077-3088doi:10.1093/brain/awaa238 · PMID 32830221

    Thirty-seven newly diagnosed Parkinson patients split by whether they had REM sleep behavior disorder, plus 22 with the sleep disorder alone, imaged four ways. Both Parkinson groups had equally reduced putaminal dopamine uptake while two-thirds of the sleep-disorder group were normal; the sleep-disorder-positive patients had lower cardiac MIBG signal and lower colonic donepezil uptake. The study that split the disease into two trajectories. Case-control, small groups, cross-sectional.

67. 67
    Skjærbæk C, Munk OL, Andersen KB, et al. Cardiac sympathetic degeneration informs the duration of the prodromal stage of body-first Lewy body disease.NPJ Parkinsons Dis · 2026 · online 2 Jul 2026doi:10.1038/s41531-026-01455-z · PMID 42386728

    Longitudinal imaging from two cohorts fitted to progression curves. Cardiac sympathetic signal reached 50 percent of normal at about 5 years and 25 percent at about 10; dopaminergic signal reached those marks at about 8 and about 15 years. Patients with isolated REM sleep behavior disorder had a median cardiac signal of 21.3 percent of normal against a median dopaminergic signal of 81.5 percent, implying that cardiac sympathetic loss precedes nigrostriatal involvement by 8.6 years and that the whole prodromal phase exceeds 20 years. Modeled curves, not observed individual histories.

68. 68
    Scheperjans F, Borghammer P, Pfeiffer RF, Appel-Cresswell S, Keshavarzian A. The Gut-Brain Axis in Parkinson's Disease: From Clinical Observation toward Mechanism-Based Precision Medicine.Mov Disord · 2026 · online 8 Oct 2026doi:10.1002/mds.70566 · PMID 42845110

    A perspective published days before this page was written. Treats gastrointestinal dysfunction as a core feature of Parkinson disease, notes that microbiome associations are reproducible but confounded by geography, constipation, medication, diet, disease stage and reverse causation, and states that trials of probiotics, fecal transplantation and diet have not established disease modification. Opinion, not data.

69. 69
    Dorsey ER, De Miranda BR, Horsager J, Borghammer P. The Body, the Brain, the Environment, and Parkinson's Disease.J Parkinsons Dis · 2024 · 14(3):363-381doi:10.3233/JPD-240019 · PMID 38607765

    Proposes that inhaled toxicants seed the olfactory route and ingested ones the gut route, generating testable predictions about prevalence and imaging patterns. The authors list the unanswered questions themselves. A hypothesis paper; no exposure has been shown to produce either subtype in people.

70. 70
    Yu M, Zhang L, Zhang X, et al. Impaired brain-gut circuit for gut motility dysfunction in mouse models of Parkinson's disease.Cell Rep · 2026 · 45(9):117877doi:10.1016/j.celrep.2026.117877 · PMID 42611706

    Several mouse models mimicking genetic and sporadic Parkinson disease. Alpha-synuclein pathology significantly impaired the excitability of the cholinergic neurons of this nucleus, and chemogenetically activating them restored gut motility, which makes the hypoexcitability causal in those animals rather than incidental; electroacupuncture at ST36 raised the same neurons' excitability and relieved the dysmotility through the cholinergic brain-gut axis. The first causal link between what this nucleus does and the gut symptoms that precede diagnosis. Mice, artificial models, one laboratory.

71. 71
    Ye K. Parkinson's gut dysmotility: A role for DMV cholinergic neurons.Cell Rep · 2026 · 45(9):117909doi:10.1016/j.celrep.2026.117909 · PMID 42715097

    The invited commentary published with Yu and colleagues, summarizing the finding that alpha-synuclein pathology in this nucleus compromises neuronal excitability and drives Parkinson-related gut dysmotility through a DMV-gut circuit that electroacupuncture reactivates. A commentary, not data.

72. 72
    Adler CH, Janarthanam C, Serrano GE, et al. Alpha-Synuclein Seed Amplification Assay in CSF, Skin, and Submandibular Gland From Incidental Lewy Body Disease and Parkinson Disease.Neurology · 2026 · 107(2):e218121doi:10.1212/WNL.0000000000218121 · PMID 42348807

    Autopsy-confirmed cases: 19 controls, 40 with incidental Lewy body disease, 15 with Parkinson disease. In the incidental cases the seeding assay was positive in 75.8 percent of cerebrospinal fluid samples but only 35.1 percent of submandibular gland and 37.5 percent of skin; in established Parkinson disease all three tissues performed well. Small numbers, and the authors call the results preliminary.

73. 73
    Zheng T, Camilleri M. Management of Gastroparesis.Gastroenterol Hepatol (N Y) · 2021 · 17(11):515-525PMID 35466306

    Sets out the several separate mechanisms that produce the same symptom picture: weak antral and duodenal contraction, failure of the pylorus to relax, failure of the stomach to accommodate a meal, and heightened visceral sensation. Useful for the point that delayed emptying is a finding, not a cause. A review; the record carries no DOI.

74. 74
    Pitra S, Smith BN. Musings on the wanderer: What's new in our understanding of vago-vagal reflexes? VI. Central vagal circuits that control glucose metabolism.Am J Physiol Gastrointest Liver Physiol · 2021 · 320(2):G175-G182doi:10.1152/ajpgi.00368.2020 · PMID 33205998

    Review arguing that chronic dysregulation of blood glucose remodels dorsal vagal complex circuits, and that it is unsettled whether the remodeling drives the metabolic problem or compensates for it. States the open question rather than resolving it.

75. 75
    Hjerpsted JB, Flint A, Brooks A, Axelsen MB, Kvist T, Blundell J. Semaglutide improves postprandial glucose and lipid metabolism, and delays first-hour gastric emptying in subjects with obesity.Diabetes Obes Metab · 2018 · 20(3):610-619doi:10.1111/dom.13120 · PMID 28941314

    Thirty people with obesity, semaglutide 1.0 mg weekly against placebo in a crossover design. First-hour gastric emptying was slowed (area-under-curve ratio 0.73, 95 percent confidence interval 0.61 to 0.87) but overall gastric emptying was not statistically different from placebo. Commercial conflict of interest: four of the six authors (Hjerpsted, Flint, Axelsen, Kvist) were Novo Nordisk employees, the other two being at Covance Clinical Research Unit and the University of Leeds, and the company makes the drug.

76. 76
    Dahl K, Brooks A, Almazedi F, Hoff ST, Boschini C, Baekdal TA. Oral semaglutide improves postprandial glucose and lipid metabolism, and delays gastric emptying, in subjects with type 2 diabetes.Diabetes Obes Metab · 2021 · 23(7):1594-1603doi:10.1111/dom.14373 · PMID 33710717

    Fifteen people with type 2 diabetes (mean age 58.2 years, glycated hemoglobin (HbA1c) 6.9 percent, weight 93.9 kg) given oral semaglutide escalated to 14 mg. Gastric emptying in the first postprandial hour was delayed, a 31 percent fall in the acetaminophen (paracetamol) area-under-curve. Fifteen subjects, 13 of them male. Commercial conflict of interest: four of the six authors were Novo Nordisk employees.

77. 77
    Rayner CK, Watson LE, Phillips LK, et al. Effects of Sustained Treatment With Lixisenatide on Gastric Emptying and Postprandial Glucose Metabolism in Type 2 Diabetes: A Randomized Controlled Trial.Diabetes Care · 2020 · 43(8):1813-1821doi:10.2337/dc20-0190 · PMID 32471908

    Thirty patients, eight weeks of the short-acting agonist lixisenatide against placebo, gastric emptying measured by scintigraphy. Retention of the test drink rose 2.19-fold (1.82 to 2.64, p less than 0.001) and postprandial glucose fell in proportion to how much emptying slowed. The paper states that tachyphylaxis for slowing of emptying is seen with continuous exposure to GLP-1, which is why the short-acting drug was tested; it is a 30-patient parallel-group trial.

78. 78
    Nersessian RSF, da Silva LM, Carvalho MAS, et al. Relationship between residual gastric content and peri-operative semaglutide use assessed by gastric ultrasound: a prospective observational study.Anaesthesia · 2024 · 79(12):1317-1324doi:10.1111/anae.16454 · PMID 39435967

    Two hundred and twenty fasted patients before elective surgery, assessed by point-of-care gastric ultrasound. Increased residual gastric content in 43 of 107 who had taken weekly semaglutide within ten days (40 percent) against 3 of 113 unexposed (3 percent), propensity-weighted odds ratio 36.97 (16.54 to 99.32), with no case of pulmonary aspiration. Observational, single network, and it measures stomach content rather than this nucleus.

79. 79
    Pai SL, Nimma SR, Beam WB, et al. Assessment of Gastric Content Using Gastric Ultrasound in Patients on Glucagon-Like Peptide-1 Receptor Agonists Before Anesthesia.Anesth Analg · 2026 · 142(4):640-649doi:10.1213/ANE.0000000000007764 · PMID 41032460

    Three hundred and sixteen fasted patients on these drugs across three hospitals; 113 (35.8 percent) had high residual gastric content. Receiver operating characteristic analysis gave cutoffs of 7.5 days or less of withholding a weekly injection and 21.3 hours or less of fasting from solid food. Observational and without an unexposed comparison group, so it gives thresholds rather than a risk ratio.

80. 80
    Vlaeminck N, Van de Putte P, Dekeyser M, et al. Gastric ultrasound in patients receiving semaglutide: a prospective, multicentre, matched control study.Anaesthesia · 2026 · 81(6):801-809doi:10.1111/anae.70129 · PMID 41631344

    Forty-four patients on semaglutide matched one to one with controls on age, body mass index and diabetes status. A full stomach in 21 of 43 (49 percent) against 8 of 44 (18 percent), odds ratio 4.29 (1.63 to 11.29); solid content in 18 of 43 (42 percent) against 3 of 44 (7 percent), odds ratio 9.85 (2.57 to 37.76); calculated gastric volume did not differ. Matching for diabetes and obesity is what distinguishes it; the numbers are small, and the authors recommend using the ultrasound to guide management.

81. 81
    Boudreau C, Couture M, Rousseau-Saine N, et al. Residual gastric content after holding of glucagon-like peptide-1 receptor agonists before elective surgery: a cross-sectional study - the RESIDUAL study.BMC Anesthesiol · 2026 · 26(1)doi:10.1186/s12871-026-03999-2 · PMID 42265610

    Ninety-three fasted patients at two Montreal hospitals: increased residual gastric content in 22 of 52 (42 percent) who had held a weekly agonist for at least seven days against 10 of 41 controls (24 percent), prevalence ratio 1.73 (0.96 to 3.75) unadjusted and 1.65 (0.7 to 3.7) adjusted. The authors state plainly that the study was underpowered and the signal clinically relevant but statistically inconclusive.

82. 82
    Grossman P. Fundamental challenges and likely refutations of the five basic premises of the polyvagal theory.Biol Psychol · 2023 · 180:108589doi:10.1016/j.biopsycho.2023.108589 · PMID 37230290

    The single-author forerunner of the 2026 consensus paper. Argues that the theory's premise of distinct dorsal-vagal and ventral-vagal control of heart rate, and its use of respiratory sinus arrhythmia as an index of vagal tone, are untenable or highly implausible. A critical review.

83. 83
    Wang YB, Saunders SE, Campbell JN, Boychuk CR. Cardiac vagal motor neurons.Curr Opin Neurobiol · 2025 · 93:103068doi:10.1016/j.conb.2025.103068 · PMID 40570503

    Review from the same laboratory, describing the role of the cardiac vagal neurons of this nucleus as a controversy now being resolved rather than a settled absence, and surveying their circuitry and their place in cardiometabolic and mental health disease. A review, not new data.

84. 84
    Wang YB, Dow KE, Boychuk CR. GABA-A receptor delta-subunit mediates increased GABAergic inhibition in cardiac DMV neurons after high-fat diet.iScience · 2025 · 28(4):112268doi:10.1016/j.isci.2025.112268 · PMID 40264791

    Fifteen days of a high-fat diet increased tonic, but not phasic, GABA-A inhibition selectively onto the cardiac-projecting neurons of this nucleus in mice, an effect abolished by deleting the receptor delta-subunit from cholinergic motor neurons. Mouse slice electrophysiology with single-cell transcript measurement; diet-induced plasticity specific to this population.

85. 85
    Radziwon J, Sławek J. Ultrasonographically measured atrophy of vagus nerve in Parkinson's Disease: clinical and pathogenetic insights plus systematic review and meta-analysis.Neurol Neurochir Pol · 2024 · 58(5):471-483doi:10.5603/pjnns.99592 · PMID 39140586

    Twelve case-control studies pooled: the vagus nerve is modestly smaller on neck ultrasound in Parkinson disease, effect size minus 0.79 standard deviations (minus 1.34 to minus 0.25, p equal to 0.004), more pronounced on the right and smaller in women, correlating with cardiac parasympathetic decline and motor ratings. The authors state that the reduction is not readily apparent, is individually variable and is of unclear clinical significance. It measures the nerve, not the nucleus.

86. 86
    Parkinson Study Group STEADY-PD III Investigators. Isradipine Versus Placebo in Early Parkinson Disease: A Randomized Trial.Ann Intern Med · 2020 · 172(9):591-598doi:10.7326/M19-2534 · PMID 32227247

    Three hundred and thirty-six patients with Parkinson disease of less than three years' duration, randomized to immediate-release isradipine 5 mg twice daily or placebo for 36 months at 57 sites. Adjusted least-squares mean change in Unified Parkinson's Disease Rating Scale parts I to III was 2.99 (0.95 to 5.03) against 3.26 (1.25 to 5.26) points, treatment effect minus 0.27 (minus 3.02 to 2.48), p equal to 0.85; secondary outcomes showed no effect. The investigators note the dose may have been insufficient to engage the target calcium channels. The one clinical test of the calcium-fragility mechanism, and it is negative — for a drug aimed mainly at the midbrain, not at this nucleus.

87. 87
    Sousa BP, Santos KMD, Antunes VR. Chemogenetic Stimulation of the Dorsal Motor Nucleus of the Vagus Mitigates Autonomic Dysfunction and Memory Decline in Rat Model of Alzheimer-Like Disease.Acta Physiol (Oxf) · 2026 · 242(7):e70261doi:10.1111/apha.70261 · PMID 42178898

    Transgenic Long Evans rats given intracerebroventricular streptozotocin. Male rats developed raised cardiac sympathetic tone, poorer object recognition and more hippocampal amyloid; females showed none of this at the same time point. Fifteen days of chemogenetic activation of the cholinergic neurons of this nucleus in males returned sympathetic tone to control levels, improved recognition memory and lowered amyloid density. The first attempt to use this nucleus as a target rather than study it as a casualty. Rats, an artificial model, males only, one laboratory.

88. 88
    Yin L, Liu F, Li M, et al. An mPFC-DMV cortical-autonomic-immune axis governs stress-induced immune dysregulation in sepsis.Mol Psychiatry · 2026 · online 19 Sep 2026doi:10.1038/s41380-026-03913-9 · PMID 42763340

    Mice given chronic restraint stress before endotoxemia had worse mortality, more systemic inflammation and impaired function of a circuit running from the medial prefrontal cortex to this nucleus, with reduced splenic nerve activity and disturbed splenic norepinephrine and acetylcholine signaling. Chemogenetically activating the circuit restored splenic autonomic output, lowered inflammatory cytokines, improved survival and rescued regulatory T cell function, and chemical destruction of the splenic sympathetic nerves largely abolished those benefits. Mice, one laboratory, an endotoxin model rather than clinical sepsis.

89. 89
    Mathis KW, Das-Earl P, Nguyen-Alley K, Bruce A. Autonomic-Immune Crosstalk in Lupus: Impaired Cholinergic Signaling and Consequences for Cardiovascular-Renal Physiology.Am J Physiol Cell Physiol · 2026 · online 25 Sep 2026doi:10.1152/ajpcell.00111.2026 · PMID 42788284

    A review of one group's own program in murine lupus. Systemic galantamine, which enhances vagal cholinergic anti-inflammatory signaling, lowered autoantibodies, blood pressure and inflammatory mediators; but direct localized chemogenetic stimulation of the parasympathetic neurons of this nucleus changed neither blood pressure nor autoantibody levels and improved only the renal pathology. Cited chiefly for that dissociation, which argues against reading every benefit of vagal stimulation back onto this nucleus. A review of the authors' own mouse work.