Search this and you will be told, within about four scrolls, to plunge your face into ice water, hum for two minutes, gargle aggressively, tug your left ear, and massage the side of your own neck.
One of those is genuinely unsafe. Two of them have decent evidence. The rest sit somewhere between plausible and decorative.
This article grades all of them. The honest headline first, though, because it changes how you’ll read the rest: you are not resetting a nerve. The vagus is not switched off waiting to be switched on — it is working continuously right now, braking your heart and running your digestion. What these practices do, at best, is shift the moment-to-moment balance toward parasympathetic activity, and, done repeatedly over months, possibly shift your baseline. That’s a worthwhile thing. It’s just a different thing from what the videos are selling.
If you want the map of the nerve itself — where it runs, what it’s wired to, and why the neck is off limits — that’s in where is the vagus nerve and what does it do?
How the grading works
- Strong — replicated human trials, or an accepted clinical use
- Moderate — real human evidence, but limited, short-term, or with methodological problems
- Plausible — a sensible mechanism and some supporting data, not properly tested
- Unsupported — popular, but with nothing solid behind it
- Avoid — a safety problem
Slow breathing — Strong
If you do one thing on this list, do this one. It has the best evidence, it is free, and it works within a minute.
Why it works: your heart rate rises slightly on each inhale and falls on each exhale — respiratory sinus arrhythmia, and it exists because vagal braking of the heart is fast enough to change within a single breath. Breathing slowly, and especially exhaling longer than you inhale, exaggerates that rhythm. At around six breaths per minute — five seconds in, five seconds out — most people’s cardiovascular system hits a kind of resonance where the swings are largest.
A systematic review and meta-analysis of voluntary slow breathing found consistent increases in vagally-mediated heart rate variability indices across studies, and slow-paced breathing at six cycles per minute is one of the most reliably reproduced interventions in the whole field.
One honest caveat: some of that HRV increase is a mechanical consequence of the breathing pattern rather than evidence of a deeper change. The subjective effect on how you feel, though, is not in doubt, and it shows up in anxiety measures too.
How to do it:
Sit or lie down. Breathe in through your nose for a count of four. Breathe out — through the nose or through slightly pursed lips — for a count of six. No straining, no holding. Ten minutes is a real dose; five is fine; two minutes in a car park before a difficult conversation is genuinely useful.
Two refinements worth having:
- The exhale is the active ingredient. Vagal braking is strongest during exhalation. A 4-in / 6-out or 4-in / 8-out pattern will do more than an even 5/5.
- Breathe low. Your hand on your belly should move more than the hand on your chest.
That’s it. There is no supplement, device, or protocol on this list that outperforms this.
Cold on the face — Moderate
The physiology here is solid and specific — and it is facial cold, not whole-body cold, that does it.
Why it works: cold water on the face, particularly around the eyes, forehead, and nose, triggers the diving response through the trigeminal nerve: heart rate drops, peripheral blood vessels constrict, and vagal activity to the heart increases sharply. This is a genuine reflex with a mapped pathway, not a wellness metaphor, and a meta-analysis of diving-response studies confirms a reliable increase in cardiac vagal activity.
Two things are commonly over-claimed. First, the effect is acute — it happens while you’re doing it and shortly after. Evidence that repeated cold exposure durably raises baseline vagal tone is much weaker than the enthusiasm suggests. Second, whole-body cold plunges are a different stimulus — they involve a large sympathetic surge as well, and a systematic review of cold exposure and autonomic control found the picture considerably messier than face immersion alone.
How to do it:
Fill a bowl with cold water — around 10–15°C (50–60°F), cold tap water plus a little ice. Hold your breath and put your face in for 15–30 seconds. Or, more practically: press a cold wet cloth or a gel pack over your forehead, eyes, and the bridge of your nose for 30 seconds. A cold splash at the end of a shower, aimed at the face, does a lighter version.
Safety, and this one matters: the diving response slows the heart, which is the point and also the risk. Skip cold face immersion entirely if you have any heart rhythm disorder, coronary artery disease, uncontrolled high blood pressure, or a history of fainting, and check with your doctor if you’re not sure. Never do it in a bath or anywhere you could pass out and end up face-down in water. Stop if you feel light-headed.
Exercise — Strong
The least exciting entry, and one of the best supported.
Regular aerobic exercise raises vagally-mediated heart rate variability over weeks and months, across multiple systematic reviews and meta-analyses. Unlike almost everything else here, this is a change in baseline rather than a moment’s shift, and the mechanism — improved cardiac vagal control with training — is well established.
It’s also the entry with the largest effect on everything else the vagus is involved in: inflammation, digestion, sleep, mood.
How to do it: the usual, unromantic answer. Moderate aerobic work most days, something that raises your breathing but lets you still talk. Add strength work twice a week. Consistency matters far more than intensity here; very hard training temporarily lowers HRV while you recover.
Humming, chanting, and singing — Plausible
Widely recommended, mechanistically reasonable, thinly evidenced.
The case for: the vagus supplies the muscles of the larynx and pharynx, so vocalising involves vagal motor activity directly. Humming and chanting also force a slow, long exhale — which is the mechanism in the slow-breathing section, doing the real work. Small studies of Bhramari pranayama (humming bee breath) have reported increases in HRV indices, and a pilot comparison of humming breathing against slow-paced breathing found effects on HRV and affect.
The much-quoted nitric oxide finding, handled honestly: humming increases nasal nitric oxide roughly fifteen-fold compared with quiet exhalation, by flushing gas out of the paranasal sinuses. That’s a real, well-replicated finding from 2002. It is a sinus ventilation effect, though — it says nothing about vagal tone, and the leap from “humming raises nasal NO” to “humming tones your vagus nerve” is one that the study does not make.
The honest verdict: humming is pleasant, it’s free, it enforces a long exhale, and the studies that exist are small and mostly unblinded. Do it because you like it and because the breathing pattern is good. Don’t do it instead of the slow breathing above — it is the slow breathing above, with a sound attached.
How to do it: inhale through the nose for four, then hum on the exhale for as long as is comfortable, six to ten seconds, feeling the buzz in your face and throat. Five to ten rounds. Singing along to something for ten minutes does much the same job with better company.
Gargling and the gag reflex — Plausible
The pharyngeal branches of the vagus supply the muscles you use to gargle, and the gag reflex has a vagal efferent limb. So the mechanism is sound on paper.
What’s missing is any meaningful human evidence that gargling until your eyes water changes vagal tone, mood, digestion, or anything else people gargle for. It is a popular recommendation built almost entirely on anatomy.
Verdict: harmless, possibly useful, entirely untested. If you enjoy it, gargle vigorously with warm salt water for 30 seconds. Don’t induce actual gagging; there’s no evidence it helps and it can trigger a vasovagal faint in susceptible people.
Ear stimulation and taVNS devices — Moderate, and improving
This is the one non-invasive method with a genuine anatomical claim, because the auricular branch of the vagus supplies the skin of the concha and part of the ear canal — the only patch of body surface with vagal sensory supply.
Transcutaneous auricular vagus nerve stimulation (taVNS) puts a mild electrical current through an ear clip or in-ear electrode on that spot. It’s been studied for depression, epilepsy, insomnia, stroke rehabilitation, tinnitus, and inflammatory conditions, and meta-analyses have found positive pooled effects in several of these areas — insomnia and post-stroke rehabilitation among the better supported, with a safety review finding adverse events mild and uncommon (mostly local tingling, itching, and occasional headache).
Three honest problems:
- Sham control is genuinely hard. You can feel the stimulation, so blinding is imperfect, and different trials use different “sham” locations on the ear — some of which may not be inert.
- Protocols are all over the place. Frequency, intensity, duration, and electrode position vary wildly between studies, which makes pooled results shakier than the effect sizes suggest.
- Study quality is often low, with small samples and short follow-up.
Verdict: legitimate, actively researched, not yet something to spend a lot on. If you’re curious, the cheap consumer ear clips are a lower-stakes experiment than the price of the premium devices implies. Don’t use one if you have a pacemaker or other implanted electrical device, a history of seizures without medical guidance, or heart rhythm problems, and don’t stimulate both ears at once.
The free version: simply massaging the concha — the bowl of the ear just outside the canal opening — with a finger for a minute is often recommended on the same anatomical logic. There’s no trial behind it. It’s also completely safe, which is not nothing.
Vagal maneuvers for a racing heart — Strong (medical, specific)
Worth including because it’s the clearest proof that deliberately stimulating the vagus does something powerful in humans.
In emergency departments, a modified Valsalva maneuver is a first-line treatment for supraventricular tachycardia. In the REVERT trial, 433 patients were randomised to standard Valsalva versus a modified version — a 60-second strain while semi-recumbent, then immediately lying flat with the legs raised to 45° for 15 seconds. Conversion to normal rhythm: 17% with standard Valsalva, 43% with the modified version (odds ratio 3.7), with no significant difference in adverse events.
That is a vagal intervention stopping an abnormal heart rhythm in nearly half of patients, with no drugs.
Important: this is a treatment for a diagnosed arrhythmia, performed by people who can tell an SVT from something that must not be treated this way. It is not a wellness practice, and straining hard against a closed airway is not something to practise for general “vagal tone.” It’s here because it shows what the nerve can do, not as an instruction.
Related and more clearly off-limits: carotid sinus massage is a genuine medical maneuver, and a genuine cause of stroke when done by the wrong hands on the wrong neck. Doctors do it with monitoring. Nobody should do it to themselves.
Feeding the gut — Plausible in humans, strong in animals
The gut is the vagus nerve’s largest sensory territory — on the order of 80% of its sensory neurons — so what happens in there is, quite literally, vagal input.
The animal work is striking. A Lactobacillus rhamnosus strain changed anxiety-like behaviour and brain GABA receptor expression in mice, and the effect vanished when the vagus was cut — one of the cleanest demonstrations that gut bacteria talk to the brain along this nerve.
In humans, it is much less settled. Probiotic trials for mood are mixed, strain-specific, and often small, and none of them demonstrate a vagal mechanism in people.
What to do with that, honestly: eat in a way that supports a decent gut environment — fibre, fermented foods, a varied diet — for its own well-established reasons, and treat any vagal benefit as a bonus rather than the point. The almanac’s digestive standbys sit here: CCF tea with cumin, coriander, and fennel seed after meals; cardamom for a heavy stomach; ginger for nausea, which is the one with the strongest trial evidence of the three.
Warm, slow touch and aromatics — Plausible
No oil reaches the vagus nerve. It runs inside a fascial sheath beside your carotid artery, and topical aromatics work in the skin’s outer layers and local circulation. That’s the mechanism honestly stated.
What warm oil massage does do is reliably lower arousal — slow rhythmic touch, warmth, and a predictable ritual all reduce the body’s guarding, and that’s usually what people are chasing when they say “vagal tone.” Small studies of inhaled lavender and bergamot report shifts in HRV and subjective calm; they’re short, small, and nearly impossible to blind, since you can tell what you’re smelling.
Verdict: real but modest, working through smell and touch rather than through the vagus directly. Which is a perfectly good reason to do it.
In practice: vata abhyanga oil for a warm full-body self-massage, ghee foot oiling as a nightly wind-down, or a magnesium foot soak before bed. For inhalation, a drop of lavender, roman chamomile, frankincense, or vetiver on a tissue is enough — check what dilution percent actually means before putting any of them on skin.
The calming herbs — Moderate for calm, not vagal
Worth stating clearly because they’re constantly folded into vagus content: lemon balm, chamomile, passionflower, holy basil, and ashwagandha have their own evidence for anxiety, sleep, and stress — GABAergic effects, cortisol modulation, adaptogenic action. None of them is a vagal stimulant, and none needs to be to be worth drinking.
A tulsi-rose tea when you’re wound up, ashwagandha milk in the evening, or a magnesium foot soak before bed are good practices. See adaptogens 101 for what that class does and doesn’t do.
Implanted and cervical stimulation — Strong, but medical
For completeness, because it’s the source of most of the excitement:
- Implanted VNS is FDA-approved for drug-resistant epilepsy and treatment-resistant depression. The large RECOVER trial implanted devices in nearly 500 people with markedly treatment-resistant depression (averaging thirteen prior failed treatments). Its primary endpoint at 12 months was not met — no significant difference between active and sham on the main depression scale — but multiple secondary measures, including quality of life and functional capacity, did improve significantly, with most benefit emerging in the final months.
- Implanted VNS for rheumatoid arthritis produced a modest but significant benefit in 242 biologic-refractory patients, as covered in the companion article.
- Non-invasive cervical stimulation (gammaCore) is FDA-cleared for cluster headache and migraine, held against the side of the neck under medical direction. This is a regulated device with a known safe placement — it is not a licence to press on your own neck.
The lesson to take from all three: the effects that survive proper testing are modest, slow, and delivered by devices. That’s a reason to be measured about what two minutes of humming will do — not a reason to skip the free practices, which have their own justification.
What doesn’t hold up
- Massaging the “vagus nerve point” on your neck. Covered above and worth repeating: the nerve is inside the carotid sheath with your carotid artery and jugular vein. Vigorous neck massage has caused arterial dissection and stroke. Avoid. If your neck is genuinely tight, the safe routine is in how to release deep cervical fascia, which deliberately stays on the back and sides.
- “Vagus nerve reset” eye exercises. The popular version — lying down and holding your gaze to one side — has no supporting evidence. Harmless, but it isn’t doing what the video says.
- Supplements sold for vagal tone. No supplement has been shown to raise vagal activity. Magnesium, omega-3s, and B vitamins have their own uses; this isn’t one of the demonstrated ones.
- HRV score-chasing. Your ring’s number moves with your breathing, your sleep position, your alcohol intake, and the time you took the reading. Tracked against your own baseline over weeks it’s a useful load signal; read day to day, it mostly generates anxiety, which is precisely backwards.
A ten-minute daily practice
Built only from the things that hold up, in the order that makes sense:
- Slow breathing, 5 minutes. Four in, six out. Low and easy. This is the bulk of the benefit.
- Humming, 2 minutes. Ten rounds — inhale four, hum out for eight. It’s the long exhale doing the work, and it’s more pleasant than counting.
- Cold on the face, 30 seconds. A cold wet cloth over forehead, eyes, and nose. Skip entirely if any of the heart or fainting cautions above apply to you.
- Warm oil, 2 minutes. Feet, neck, and shoulders with vata abhyanga oil or plain warm sesame oil. Slow strokes. Back and sides of the neck only.
- Somewhere in your week: the exercise. Which does more for your baseline than the other four combined.
And then leave it alone. This works on the timescale of weeks, and checking a number every morning to see whether it’s working is itself a sympathetic activity.
Safety, gathered in one place
- Never press firmly on the front or front-sides of your neck, and never on both sides at once. Never press on a pulse.
- Skip cold face immersion if you have a heart rhythm disorder, coronary artery disease, uncontrolled hypertension, or a fainting history — and never do it where you could pass out into water.
- Don’t practise hard Valsalva straining for general wellness; it’s a medical maneuver for a diagnosed rhythm problem.
- Don’t use taVNS devices with a pacemaker or other implanted electrical device, and get medical guidance if you have epilepsy or a cardiac condition.
- Stop anything on this list that produces dizziness, light-headedness, chest discomfort, visual changes, or nausea.
- See a clinician for hoarseness lasting more than three weeks, new difficulty swallowing, fainting on exertion or without warning, or persistent vomiting of undigested food. These are vagal symptoms that need assessment, not exercises.
The honest summary
- Slow breathing at around six breaths a minute, with a longer exhale, is the best-supported method — and the free one.
- Regular aerobic exercise is the only entry that reliably shifts your baseline rather than the moment.
- Cold on the face triggers a genuine, mapped reflex, acutely — with real cardiac cautions attached.
- Humming and gargling are plausible and pleasant; the humming benefit is mostly the long exhale, and the famous nitric oxide finding is about your sinuses, not your vagus.
- Ear-clip taVNS is the one non-invasive method with a real anatomical claim and a growing evidence base, held back by weak blinding and inconsistent protocols.
- The impressive clinical results come from implanted devices, and even those are modest.
- Neck massage over the vagus is the one thing on the list to actively avoid.
Where to go next
- Where is the vagus nerve and what does it do? — the anatomy, the five jobs, and an honest look at “vagal tone”
- How to release deep cervical fascia — the safe way to work on a tight neck
- Vata abhyanga oil — warm, grounding self-massage
- Ghee foot oiling — the nightly wind-down
- CCF tea — the daily digestive
- Adaptogens 101 — what the calming herbs actually do
- What dilution percent actually means — before you mix any massage oil
If you’re working on your nervous system and want help sorting the real from the decorative, send a note to the almanac. Every submission is read and answered.
This article is educational and not a substitute for medical advice. Anyone with a heart condition, a history of fainting, epilepsy, or an implanted device should speak with their doctor before trying cold exposure, breath-holding, or electrical stimulation.