The vagus nerve and trauma: what the science actually shows

The first thing the vagus nerve actually is.
A real anatomical bridge.

The vagus is the tenth cranial nerve, the longest of the cranial nerves, and it is bidirectional. Roughly 80% of its fibres are afferent, carrying signals from the viscera up to the brain. The remaining 20% are efferent, carrying signals from the brain down to the viscera. This is unusual for a peripheral nerve and is the anatomical basis for the now-familiar observation that the gut and the heart and the lungs do not just receive instructions from the brain, they also continuously inform it about their state.

Within the vagal system, two functional branches are distinguished. The ventral vagal complex, anchored in the nucleus ambiguus, is myelinated, fast-conducting, and provides the parasympathetic input that allows the heart to slow gracefully under safe social conditions. The dorsal vagal complex, anchored in the dorsal motor nucleus, is unmyelinated, slower, and is implicated in the deep autonomic shutdown patterns that show up in dissociation and freeze. This two-branch structure is the architectural backbone of Porges' Polyvagal Theory, formally introduced in 1995 and developed across subsequent decades (Porges, 2011).

Functionally, the vagus modulates heart rate variability (HRV), gut motility, vocal modulation through the larynx, and aspects of facial expression coordination through its connections to the facial nerve. Heart rate variability in particular has a substantial empirical literature behind it as a marker of vagal tone and a correlate of emotion-regulation capacity (Thayer & Lane, 2007). When the wellness conversation talks about “vagal tone,” there is a real underlying physiology being referred to. The question is what that physiology can and cannot do.

The second.
Where Porges' actual claims end.

Polyvagal Theory is a substantial, integrative framework, and treating it as a strawman is intellectually unfair. Porges proposed a series of claims about how mammalian autonomic regulation evolved, how the ventral vagal circuit supports social engagement, and how the dorsal vagal circuit deploys under conditions of overwhelming threat. Some of these claims have empirical support, particularly the connection between vagal tone (as indexed by HRV) and emotion-regulation capacity. The theory has been clinically generative, providing a vocabulary that many practitioners and patients have found useful for describing autonomic states.

Other claims in the theory are theoretical inferences that the available evidence supports less directly. Grossman and Taylor (2007) argued that the specific evolutionary narrative Porges proposed for ventral vagal myelination is not consistent with comparative neuroanatomy. Grossman (2023) extended the critique, arguing that several of the theory's foundational premises face fundamental challenges and likely refutations. The point of citing this is not to settle the scientific debate. The point is that the theory is a model, with strong areas and weak areas, and the model has been treated outside academic circles as if it were settled physiology.

The patient who reads a wellness blog claiming that “the polyvagal nervous system” has been definitively mapped is reading marketing copy that has compressed a contested theoretical framework into a settled product specification. The compression is the problem, not the framework.

The third.
How the wellness extension oversteps.

The wellness market that has formed around the vagus nerve is large, sincere, and not uniformly wrong. Humming has small effects on HRV. Cold water immersion does activate vagal afferents. Slow breathing with extended exhales does shift the sympathetic-parasympathetic balance toward parasympathetic dominance, briefly. These effects are measurable in research conditions and are not invented.

It is also worth saying that many of the practitioners who teach these tools are doing careful work in good faith, often with patients whose previous treatment offered them nothing in this register at all. The objection here is not to the tools or the people teaching them. It is to a specific compression of language in which a tool that produces a small autonomic shift gets sold as a tool that resolves the structural pattern that makes the autonomic shift necessary in the first place.

The overreach is what gets claimed about them. “Tone your vagus nerve to heal trauma” is a structurally different claim from “this exercise produces a small, transient increase in HRV.” The first claim implies that engaging the vagal system at the level the exercises can reach will retire the prediction templates that drive the trauma response itself. The second claim is about a measurable physiological signal in the autonomic system. The two claims live in different parts of the architecture.

The patient who tries the exercises and notices that they do something but do not reach the trauma pattern is not failing at the exercises. The exercises were not designed to do what the marketing said they would do. The signal they produce is real and the signal is at the cortical-autonomic interface, not at the layer where the archived prediction lives (the architecture of that layer is described here in detail). The wellness extension closes that distinction in its language and opens it in the patient's life.

This is where the structural cost shows up. A patient who spends two years on vagal exercises while their actual trauma pattern continues to fire is not lazy and is not failing the protocol. They are using a real tool on a layer the tool does not reach. Naming this accurately is the first step toward the work that does reach the layer.

The fourth.
What the bridge can actually do in trauma work.

The vagal system is genuinely useful in the right place in a trauma protocol. It is an input and output channel for autonomic state. Working with breath, voice, posture, and the felt sense of being met by another nervous system in the room does engage the vagal pathways, and the engagement does produce real, measurable effects on the regulatory baseline.

What this work does well, in clinical practice, is build capacity. The patient who can spend more time in the window of tolerance, who can come back from activation more reliably, who can tolerate sensation in the body without immediately deploying the freeze or dissociation protocol (the dorsal vagal protocol is described here in detail), is a patient who has more usable bandwidth for the work that retires the prediction template. Vagal-system work is the floor that makes the rest of the work possible.

What it does not do, on its own, is retire the prediction template itself. The template was set by an event the cortex is no longer having a vote on. Updating it requires the input the prediction circuit can read as evidence: physiological outcome that contradicts the archived prediction, in the moment the prediction is active. Vagal exercises in calm conditions do not produce that input. They produce a different, also useful, input. Conflating the two is the error the wellness market has standardised.

In a small clinical-phenomenological study (Laugman, 2026), the regulatory shifts that patients reported alongside the trauma work included markers consistent with improved vagal tone: deeper breath, easier transition into sleep, more available range of vocal expression. These were companion changes, not the engine of change. The engine was the work done at the prediction layer. The vagal markers reorganised because the load that had been keeping them suppressed came down.

The bridge is real. It carries traffic. It is not the destination, and treating it as one keeps the patient busy on the bridge, with the destination still in front of them, year after year.