The vagus nerve breathing exercise, and what gets attributed to it

The vagus nerve has no switch you can reach from the outside. What you manipulate is the breathing rhythm, and that is what changes cardiac vagal activity. The nerve runs from the brainstem down into the abdomen, behind the sternum and the lungs. No movement of the hand reaches it.

The detour through the breath, on the other hand, is well described. Gerritsen and Band (2018) laid out an explicit model: a slow exhale increases parasympathetic influence on the heart, because pulmonary stretch receptors and baroreceptors send the brainstem information that slows the heart rate. It is the same circuit that produces respiratory sinus arrhythmia, the familiar rise in pulse on the in-breath and fall on the out-breath that any chest strap will display.

That detour has a consequence the exercise lists pass over in silence. Nobody ever measures the vagus nerve. What gets measured is an index derived from heart rate variability, usually RMSSD or high-frequency spectral power, and cardiac vagal activity is inferred from it. These are not the same thing, and the gap becomes visible as soon as you read the results closely.

One detail about the landscape, for context. In October 2026, the leading English results for this query each offered between five and twelve exercises. Almost none distinguished a measured effect from an anatomical guess.

This question follows the first one everywhere, and the honest answer is uncomfortable. To our knowledge, no diagnostic nomenclature recognises a blocked vagus nerve, and no home test measures one.

The symptom lists that circulate share one flaw: bloating, reflux, unexplained fatigue, a lump in the throat, palpitations at rest, a sense of being permanently on alert. Every one of those is true of almost everyone at some point, and compatible with a dozen unrelated causes. Persistent fatigue can come from anaemia, from sleep apnoea, from a thyroid problem, from exhaustion at work. Diagnosing yourself with a trapped nerve from that list means picking one explanation out of twelve without ruling out the other eleven.

There is a number that can be measured, and it deserves better than the vagueness around it: heart rate variability. Most watches and rings on the market read it with reasonable accuracy at rest, and a persistently low value for your age is real information. It is not a vagus nerve diagnosis, and it says nothing about the cause.

The real marker lies elsewhere. A persistently hoarse voice, difficulty swallowing, repeated fainting or digestive trouble that settles in call for a medical examination, not a breathing exercise. The lump in the throat is almost always something else, and we covered it separately in what loosens a throat tightened by stress.

If only one movement were to survive, it would be this one: lengthen the exhale until you come down to around six cycles per minute. Two recent trials give the measure of it, and they do not say quite the same thing.

You and colleagues (2021) brought 59 participants into the lab five times. Each visit meant one session of slow-paced breathing at six cycles per minute, of a different length, plus a control session with nothing. RMSSD, the most widely used index of cardiac vagal activity, rose in every condition compared with the control. The effect is not subtle, it is immediate, and it requires no skill. Six cycles per minute, that is all.

Magnon, Dutheil and Vallet (2021) tested a single session of deep and slow breathing in 25 younger adults and 22 older ones. High-frequency spectral power, another vagal index, rose in both groups and more so in the older participants, and state anxiety fell on both sides.

And then there is the result nobody quotes. In that same study, RMSSD did not move, and the authors had anticipated it: their protocol used an unequal inhale to exhale ratio, which shifts one index without shifting the other. Two measures supposed to reflect the same activity, two different verdicts on the same session.

This is the strongest argument against reading the subject mechanically. Stimulating the vagus nerve describes a movement whose effects are assessed through imperfect proxies, not a command whose execution you could verify.

The rhythm itself is documented in detail elsewhere: five seconds in and five seconds out (in French) gives exactly six cycles per minute. The extended exhale after a double inhale (in French) is the short version, for the moments when six minutes are not available.

Five minutes does as much as twenty, at least during the session itself. That is the main finding of the trial by You and colleagues, and it was the point of the study: no difference in RMSSD between 5, 10, 15 and 20 minutes, neither during the slow breathing nor in the resting measurement taken immediately afterwards.

What duration does change lies elsewhere, and it is more interesting than the expected number. Spontaneous breathing frequency measured after the session, the rate you settle into without thinking about it once the exercise ends, fell as the session got longer. The first five minutes do the work on the heart. The next fifteen do the work on the habit.

The practical consequence is immediate. If the goal is to settle down now, five minutes is enough and nothing justifies forcing yourself to do more, which is good news for anyone who has abandoned an exercise because it dragged. If the goal is to breathe differently for the rest of the day, duration stops being zeal.

A word on regularity, because the question always follows. Both trials cited here studied a single isolated session: they say nothing about an effect accumulated over weeks. The available data do not allow us to claim that daily practice durably raises resting heart rate variability. That is not a reason not to practise. It is a reason not to promise.

Humming appears in every list, and it deserves to. Trivedi and colleagues (2023) fitted 23 people with Holter monitors and compared four activities: humming, physical activity, emotional stress, and sleep. Humming produced the lowest stress index of the four, sleep included.

The result is striking, and it has to be taken for what it is: a pilot study on 23 participants, without randomisation, and without replication as far as we know. A signal, not proof.

The plausible mechanism, though, is anatomical and solid. The vagus nerve supplies the larynx through its laryngeal branches, so a sustained vibration literally occurs along its path, which is true of no other voluntary movement. And a held hum mechanically lengthens the exhale, so it inherits the effect described above. Two reasons that stack, neither of which asks you to take anything on faith.

Then there is the study cited wrongly, and it is cited everywhere. Weitzberg and Lundberg (2002) showed that humming multiplies nasal nitric oxide fifteenfold in ten healthy subjects. That work circulates as the demonstration of humming's vagal effect. It measures nothing of the kind: it concerns sinus ventilation, and the authors propose it as a test of sinus ostial patency. Nasal nitric oxide is not an index of vagal activity, and nobody in that paper claims otherwise.

If making the sound feels ridiculous, that is the most common obstacle, and it has a workaround. A closed-mouth hum (in French) settles the question: nobody hears it two metres away.

The other popular techniques rest on very uneven levels of evidence, and the sorting is worth doing.

The ear is the most instructive case, because there genuinely is a literature behind it. Frangos, Ellrich and Komisaruk (2015) stimulated the cymba conchae of twelve adults inside a functional MRI scanner and observed activation of the nucleus tractus solitarius, the central relay of the vagus, absent when the stimulation was applied to the earlobe instead. So the auricular branch of the vagus nerve does reach the brainstem in humans, which until then was known only in cats and rats. But the stimulation in that study consisted of electrical pulses of 0.25 milliseconds at 25 hertz delivered by an electrode to a precise area of the outer ear. Massaging your ear with your fingers is not a gentle version of that protocol. It is a different act, and its effect has not been measured.

Facial immersion in cold water rests on a different and older foundation: the dive reflex, well described, slows the heart through a vagal route. The effect exists and it is fast. It is also short, and the first contact produces a sharp sympathetic surge, which makes it a poor tool for anyone who already dreads strong bodily sensations.

Gargling, finally, is an openly acknowledged extrapolation: the vagus supplies the pharyngeal muscles, so gargling recruits them. That a measurable rise in cardiac vagal activity follows is a reasonable inference, not a published result, and the available data do not allow us to assert it.

So there is what has been measured in humans with a reproducible protocol, what has been measured partially, and what rests on anatomy alone. Exercise lists present all three on the same line.

Here is the assembly that concentrates what holds up in the sections above. Six minutes, no equipment, and a single parameter to watch.

Sitting, back supported, one hand resting on the belly. First minute: nothing to do but let the breath settle, without counting it. The next four minutes: in through the nose for a count of four, out through the mouth for a count of six, without reaching for volume. That rhythm gives six cycles per minute, the only setting that mattered in the trials above.

Last minute: keep the same rhythm, but replace the blown exhale with a closed-mouth hum on a low, comfortable note. The vibration should be felt in the throat and the upper sternum. If it tickles the nose, the note is too high.

Two markers for knowing whether it is right. The exhale should stay silent and even: if it ends in a forced sigh, the count is too long for you, drop to five. And your head should not swim, which happens when people breathe bigger instead of breathing slower. The point is the rhythm, not the volume.

One note for anyone who counts badly. Holding the count occupies exactly the attention you are trying to release, and plenty of people give up there rather than on the exercise itself: a voice guiding you does the counting instead.

A session of slow breathing acts on the state of the moment, and that is already a lot. It does not treat an established anxiety disorder, and none of the trials cited claims it does: they measure cardiac indices and state anxiety, over one session, in healthy volunteers.

This practice does not suit everyone. People who find attention to internal sensations difficult can feel discomfort rise as they focus on their breath, and that is a known reaction rather than a failure of method: an external anchor, a sound or a point of contact, works better for them than the belly and the throat. Where there is a cardiac or respiratory condition, medical advice comes before the breathing exercise.

If coherent breathing has already left you more irritated than calm, the technique is rarely the problem, and the usual causes are identifiable enough that we listed them separately: what gets in the way when it does not work.