The Physiological Sigh

Someone is upset, and you hear yourself saying it before you have decided to. Take a deep breath. It may be the most-repeated piece of advice in the English language, handed out by parents, coaches, midwives, referees and strangers at bus stops, with such confidence that hardly anyone stops to ask whether it does anything at all.

One particular version of it has been tested properly, which makes it worth a closer look. The pattern goes: a full breath in through the nose, then a second short sip of air on top of it before you have let any out, then a long slow release through the mouth. Researchers call it cyclic sighing. Doing it for a minute or two is the whole practice.

The belief

The folk version is deep breaths calm you down, and it is offered as though it were obvious. It rarely comes with instructions. How deep, how many, how long, in what shape, and for how long the effect is meant to last are all left to the person who is already too rattled to work it out. It is also, quite often, delivered at exactly the moment nobody wants advice.

There is a real kernel in it. What the last few years of research have added is a shape and a rough size, along with a fairly awkward finding about why it works.

The verdict

Solid. Cyclic sighing has been through a randomised controlled trial against three comparison practices, a preregistered field study that caught people in genuine anxious moments as they happened, and a pilot trial in a hospital waiting room. The felt effect shows up every time. What has not survived is the explanation attached to it in almost every article written about it, and that gap is most of what follows.

What the studies actually did

The trial that started the interest randomised 108 adults to twenty-eight days of a five-minute daily practice: cyclic sighing, box breathing, cyclic hyperventilation with breath retention, or mindfulness meditation. Cyclic sighing produced a significantly greater daily rise in positive mood than meditation, and a significantly greater fall in breathing rate, with the benefit growing the longer people stayed with it.¹

That is a genuinely interesting result, and it is worth reading slowly, because three things sit underneath the headline. State anxiety and negative mood dropped significantly in every arm, meditation included, with no significant difference between them; the win was on positive affect alone. There was no significant change in heart rate variability or resting heart rate in any group, which leaves respiratory rate as the only objective marker of arousal that moved — and respiratory rate falling during a breathing practice is close to circular. The arms were small, between 19 and 32 people, the whole thing was delivered remotely and unsupervised with adherence self-reported, and nobody was followed up past the twenty-eight days.¹

Does it survive contact with real life?

Laboratory calm is easy. The harder question is whether a minute of breathing does anything when you are actually anxious, at a bus stop, on a Tuesday. A preregistered field study went after exactly that. Forty-seven students flagged real anxious episodes as they occurred and were randomly assigned in the moment to seventy seconds of cyclic sighing, sixty-nine seconds of box breathing, or carrying on as usual, producing 782 usable measurements. Five minutes later, the cyclic sighing episodes showed a significantly larger drop in state anxiety than the control ones, and the within-person effect was very large.²

Several things temper that. Box breathing performed just as well, with no significant difference between the two, so whatever is happening does not appear to belong to the sigh specifically. It was an unblinded pilot with self-report and no physiological measurement of any kind. Forty-three of the forty-seven participants were women aged between 18 and 30. People with higher trait anxiety saw state anxiety fall less over those five minutes whatever they had just done, control episodes included; trait anxiety did not moderate either breathing exercise, so that looks like a fact about how five minutes passes for an anxious person rather than a sign the practice serves them worse. And there was no cognitive benefit at all: cyclic sighing did not reduce commission errors, and it was associated with slower reaction times and lower attention-efficiency scores.² It made people feel better without making them sharper, and possibly a touch duller for a few minutes.

A third study took it somewhere genuinely unpleasant. Patients in an orthopaedic walk-in clinic, waiting for an X-ray, were given either four minutes of pre-recorded cyclic sighing or a time- and attention-matched control. The sighing group reported significantly less pain intensity and less pain unpleasantness.³ Anxiety and depression scores did not differ by condition, so what this shows is an acute shift in how an unpleasant state is experienced. It is also explicitly pilot-scale, single-site, unblinded and self-reported. The detail worth keeping is that it was a recording. There was no instructor in the room, and it still moved.

The mechanism, kept separate from the proof

Here is where the story people tell starts to come apart from the evidence.

The long exhale does seem to be doing the felt work, and that part has support. Thirty participants breathed at six or twelve breaths a minute, crossed with an inhale-to-exhale ratio that either favoured the exhale or favoured the inhale. The long-exhale pattern produced higher self-reported relaxation, stress reduction, mindfulness and positive energy, along with greater high-frequency heart rate variability — though the heart rate variability advantage appeared only at the slow six-a-minute rate. Slowing the breath on its own, without lengthening the exhale, improved only the ‘positive energy’ measure.⁴ It is a small, acute, single-session laboratory study in healthy volunteers, unavoidably unblinded, with subjective primary outcomes, and it says nothing whatsoever about the double inhale.

The explanation everybody repeats is the next step along: the long exhale raises vagal tone, and you can see it in your heart rate variability. That step is where the evidence gives way. An original study of 26 undergraduates and an internal replication of 16 more, all of them practised rather than naive breathers with one to two semesters of paced-breathing training in the resonance-frequency range behind them, compared a one-to-one against a one-to-two inhale-to-exhale ratio at six breaths a minute. Neither found any difference in time-domain, frequency-domain or non-linear heart rate variability metrics; the authors concluded that the longer exhale may simply not alter the measures people prize.⁵ That contradicts the physiological half of the earlier finding, and it sits neatly alongside the null heart rate variability and resting heart rate results from the twenty-eight-day trial.¹

Being fair to the vagal story: both of those samples are small groups of trained undergraduates, and they tested a modest one-to-two ratio rather than the much longer exhale that cyclic sighing actually uses.⁵ So this undercuts the confident version of the mechanism rather than disproving any exhale effect at all. Where that leaves us is a practice whose effect on how people feel has replicated across a randomised trial, a field study and a clinic, and whose supposed mechanism has quietly failed to. It works, and the reason given for why it works is running well ahead of the data.

The part that is quietly remarkable

Underneath all of this is something that has nothing to do with technique.

You sigh spontaneously every few minutes, without ever deciding to.⁶ In mice, a small population of neurons in the brainstem — in the retrotrapezoid nucleus and parafacial respiratory group, expressing one of two neuropeptides — projects to about two hundred receptor-carrying neurons in the preBötzinger Complex, the region that sets the rhythm of breathing. Deliver either peptide there and the animal sighs. Block or remove either receptor and basal sighing drops; block both and it stops. Take out the receptor-carrying neurons and spontaneous sighs disappear, along with the ones triggered by low oxygen, while ordinary breathing carries on untouched.⁶

What those two hundred cells appear to exist for is reinflating the small air sacs that quietly collapse between one sigh and the next. A dedicated circuit, listening for two specific molecules, converting an ordinary breath into a sigh several times an hour for your entire life, and you have never once noticed it happening.

That is mouse anatomy, and the honest limit is worth stating plainly: it does not show that a voluntary double inhale engages that circuit, and it is not evidence that the practice calms a human being.⁶ But it does establish that a sigh is a discrete, hard-wired event with machinery of its own. The double inhale was not invented by anyone. It is a reflex your body has been running all along, borrowed on purpose.

Two sips of air in through the nose, one long sigh out. Three rounds, about a minute. Tested head-to-head against meditation it came out ahead on positive mood, with anxiety falling about as much in both, and it holds up in the middle of real life as well as in a lab.

What this might mean for your evening

Mostly it means there is a free, unremarkable thing you can try, and the evidence for it is better than for most things in this category, which is a low bar honestly cleared.

Expect what the studies actually found. The best-supported effect is on how you feel over the following few minutes, and the strongest single result was on positive mood rather than on anxiety, where the meditation arm did just as well.¹ One minute is a real dose; the trial used five minutes a day and the field study got its effect from seventy seconds.^1, 2 Nothing here suggests it does anything to a diagnosis, and nobody has followed anyone past a month.

Two practical notes fall out of the detail. It does not need a teacher — the waiting-room study used a recording and still found an effect,³ which makes it unusually portable. And it is probably not the thing to do immediately before something demanding your full concentration, since reaction times came out slightly slower afterwards.² Late evening, when nothing much is being asked of your attention, is a more forgiving time to find out whether it suits you.

If you would like a quiet stretch of evening to try it in, our interactive soundscape is free to sit with, and the rest of the Lab is where we keep working through this sort of thing — including a longer look at the longer exhale on its own.

← Back to The Lab