Breathing exercises for stress: what has actually been tested

There is a moment most people recognise. It is 5.20 in the afternoon, you are standing in the hallway with a shopping bag in one hand, someone in the house is crying for reasons you have not worked out yet, your phone is buzzing in your pocket, and you notice you have been holding your shoulders up around your ears for twenty minutes without knowing it.

That is where breathing exercises belong. Not as a lifestyle, not as a third daily training discipline to squeeze in, but as a tool for exactly that moment. The problem is that this has become one of the noisiest corners of the internet. There are hundreds of named techniques, apps selling subscriptions to timers, and claims about the nervous system that the sources behind them do not support.

This guide does three things. It explains what actually happens in the body, it walks through the three exercises that are properly documented rather than the thirty that circulate, and it reports how big the effect is in numbers. Including the parts that did not work.

What happens in the body when you breathe slowly

There is a physiological basis, it is well described, and it is simpler than most write-ups imply.

Your heart rate is not constant even when you sit still. It rises slightly every time you breathe in and falls slightly every time you breathe out. The phenomenon is called respiratory sinus arrhythmia and the research literature describes it as entirely normal. It has to do with how breathing changes pressure inside the chest, how the diaphragm moves, and how active the vagus nerve is at different points in the breath.

The consequence is the interesting part. If the inhale drives your heart rate up while the exhale pulls it down, then the length of the exhale relative to the inhale becomes a lever you can genuinely take hold of. That is the whole mechanism. Nothing mystical, nothing that needs an app.

An adult at rest usually breathes somewhere between 12 and 20 times a minute. Voluntary slow breathing typically sits around six breaths a minute, roughly a third of that pace. A 2022 systematic review pooled 223 studies on voluntary slow breathing specifically. The conclusion was that vagally mediated heart rate variability rises during the exercise itself, immediately after a single session, and after a period of several sessions too. The authors describe the method as low-tech, low-cost and with few expected adverse effects.

So the rise in heart rate variability during the exercise is well established. Whether it rises in the long run is a separate question we come back to below. What the measure is and is not is covered in the guide to HRV.

Every technique is the same lever in different packaging

If you have searched for this, you have come across 4-7-8, box breathing, coherent breathing, alternate nostril breathing, the physiological sigh and a dozen other names. It looks like a choice between separate methods. It rarely is.

Almost all of them share two features: the pace slows down, and the exhale is made at least as long as the inhale. The differences come down to how the pauses are distributed and how many seconds get counted. These are variations on one principle, not competing schools.

There is one exception, and it matters. Techniques built on fast, forceful breathing in rounds are something else entirely. They belong to their own family with their own safety rule, which appears further down this page.

The three exercises that are properly documented

Of everything in circulation, these three are the ones that either come from a national health service or have been tested in a randomised trial. The others may well work. They are simply not documented in the same way.

1. The NHS calming breath

The NHS describes this as a calming breathing technique for stress, anxiety and panic, and it has no pauses at all. You can do it standing, sitting in a chair that supports your back, or lying down. Sitting or standing, place both feet flat on the ground, roughly hip-width apart.

1. Let your breath flow as deep down into your belly as is comfortable, without forcing it.
2. Try breathing in through your nose and out through your mouth.
3. Breathe in gently and regularly. Some people find it helpful to count steadily from one to five.
4. Let the breath flow out just as gently, counting from one to five again.
5. Keep going for at least five minutes.

The NHS is also clear that you may not reach five at first, that you should not force yourself there, and that you get the most benefit if you do it regularly as part of your daily routine rather than only in a crisis.

2. The short exercise from Sweden's 1177

1177 is Sweden's national health information service. It publishes this exercise under the heading of relaxation through breathing, reviewed by a physiotherapist at Sahlgrenska University Hospital. It is built to be done almost anywhere, a point 1177 makes explicitly: on the bus, in a waiting room, at work.

1. Take a few deep breaths through your nose. Let your shoulders drop. Relax your stomach.
2. Breathe in slowly through both nostrils while counting to four.
3. Hold the breath in while counting to two.
4. Breathe out slowly through your nose while counting to four.
5. Hold the breath out while counting to two.
6. Continue for five to ten rounds.

1177 adds one detail that is the whole point of the page you are reading: if it feels comfortable, you can let the exhale take twice as long as the inhale, counting to eight on the way out. That is the same lever every other technique pulls, stated by a national health service in a single sentence.

3. Cyclic sighing

This one comes from a randomised trial at Stanford, published in Cell Reports Medicine in 2023. It differs from the other two in two ways: a double inhale instead of a single one, and a deliberately long exhale.

1. Sit or lie down. Set a timer for five minutes.
2. Breathe in slowly through your nose.
3. Once your lungs feel full, breathe in once more on top of that first inhale, to fill them completely. The second inhale is shorter and smaller. That is how it should be.
4. Then exhale slowly and fully through your mouth, all the way out.
5. Repeat the pattern for five minutes, then return to normal breathing.

In the trial, participants were told that ideally both inhales should go through the nose and the exhale through the mouth, but that anyone who preferred could do the whole exercise entirely through the nose.

How big is the effect, in numbers

This is where this guide parts company with most writing on the subject. The effect is there. It is not large.

The most complete picture comes from a meta-analysis in Scientific Reports in 2023. The researchers searched five databases and two clinical trial registers, got 1,325 hits and settled on twelve randomised controlled trials with 785 adult participants in total for the primary outcome of self-reported stress. The result was an effect size of 0.35 in reduced stress compared with controls, with a confidence interval running from 0.14 to 0.55. That counts as small to moderate. For anxiety the figure was 0.32 across twenty trials, and for depressive symptoms 0.40 across eighteen.

Most of the included trials were judged to be at moderate risk of bias. The authors' own closing line is unusually blunt for a scientific paper: they urge caution and call for study designs at lower risk of bias, explicitly in order to avoid a mismatch between hype and evidence. The first author declares that he has trained as a breath teacher, which is stated in the paper.

The second important study is the Stanford trial. It compared three breathing exercises against five minutes of mindfulness meditation, five minutes a day for 28 days. 108 participants were randomised: 24 to meditation, 30 to cyclic sighing, 21 to box breathing and 33 to cyclic hyperventilation.

Read the result carefully, because it is usually reported wrong:

• All four groups lowered their negative mood and lowered their acute anxiety from before the exercise to after it, the meditation group included. Positive mood rose in all four as well, but in the meditation group that rise did not reach statistical significance.
• The difference between groups appeared in only one of those three measures: positive mood. On acute anxiety and negative mood there was no difference at all between breathing and meditation.
• Cyclic sighing was the only individual breathing group that differed significantly from meditation. The other two pointed the same way without getting there.
• Breathing rate fell further in the breathing groups over the month, most of all in cyclic sighing.

What did not change

The last three rows of that table are the ones almost nobody quotes, and they are at least as useful as the rest.

Across 28 days of daily breathing practice, no change could be detected in heart rate variability or resting heart rate, in the breathing groups or the meditation group. Participants wore a heart rate strap around the clock, so it was not for lack of data. It does not mean heart rate variability is untouched by breathing. It rises reliably during slow breathing, as the review of 223 studies shows. But the idea that it rises lastingly on five minutes a day for a month gets no support from this trial.

Sleep is the same story. Neither sleep duration, sleep efficiency, sleep score nor daytime sleepiness measured with a standardised questionnaire changed in any group. That is worth knowing if you took up breathing exercises in order to sleep better. Breathing may well help you wind down in bed, but the mechanisms that actually govern sleep after 40 are covered in the guide to sleeping better after 40. And if you wake regularly in the middle of the night, that is its own thread, handled in the guide to waking at three in the morning.

The trial has limitations the researchers report themselves: it was originally intended as an exploratory study and was registered as a clinical trial retrospectively, the groups were small, it ran for four weeks with no follow-up, and participants practised at home with no way for anyone to check how closely they followed the instructions. The senior author became an adviser to the maker of the heart rate strap in June 2022, after the data had been collected, which is also disclosed.

The one safety warning that is real

Slow, calm breathing is harmless for the vast majority of people. The review of voluntary slow breathing describes it as a method with few expected adverse effects. There is a boundary, and it is sharp.

Techniques built on fast, forceful breathing in rounds must never be done in or near water. The reason is physiological. Forceful breathing lowers the carbon dioxide level in your blood, and it is carbon dioxide that triggers your urge to breathe. If you then hold your breath underwater, your oxygen can fall to a level where you lose consciousness before that urge arrives. This is called hypoxic blackout, and it comes without warning.

This is not a theoretical risk. A review published by the US Centers for Disease Control and Prevention in 2015 mapped sixteen cases in New York State between 1988 and 2011 in which drownings or near-drownings could be linked to deliberate underwater breath-holding behaviours. The review was prompted by the deaths of two healthy young men in 2011 who had deliberately hyperventilated before going under at a swimming facility. Combining more than one such behaviour raised the risk of a fatal outcome.

How to choose

Four situations cover most of everyday life.

The fifth situation is your stomach. When it locks up before a meeting, the exhale is the only tool that reaches it within a minute, but it does not solve what sits underneath. What stress does in the gastrointestinal tract, and what is actually proven to help over time, is in our guide to stress and stomach problems.

What decides whether it sticks

The most useful number in the entire Stanford trial is not an effect size. It is adherence. Participants in the breathing groups did their five-minute exercise on just over nineteen of the twenty-eight days on average. The meditation group came in at just under eighteen. Roughly two days in three, in other words, in a study where they got daily text reminders and had volunteered to take part.

That says something important about what to expect of yourself. If study participants with reminders miss every third day, a missed day at home is not a failure. It is the normal case.

At the same time there was a finding pointing the other way: the benefit of cyclic sighing grew with the number of days a participant had kept it up. The further into the month, the larger the daily effect. That makes this a habit rather than a button, and habits are built in a way described in the guide to breaking a bad habit.

One last note on where this belongs in a day. A breathing exercise is a short break, not recovery in the real sense. What actually counts as recovery, and which levers work in which order, is covered in the guide to managing stress after 40. Breathing is a good first step precisely because it is free and takes five minutes. It does not replace sleep, movement and people.

When to seek care

Breathing exercises are a tool for a moment. They are not a treatment for a condition, and there are situations where they should not be your answer.

• Sweden's 1177 advises contacting a health centre or occupational health service if high stress is causing sleep problems, anxiety or heart palpitations. In the UK, that conversation starts with your GP.
• Breathlessness, chest tightness or palpitations that appear at rest or on light exertion are not a breathing-technique question. They should be assessed by a clinician.
• In exhaustion disorder, 1177 lists breathing training among the things that can help once you are starting to feel better, meaning inside a treatment set by clinicians. The full picture of that condition is in the guide to burnout and exhaustion disorder.

This guide is general health information. It does not diagnose anything and does not replace an assessment by a clinician.

Frequently asked questions

Which breathing exercise is best for stress?

The one you actually do. That is not a dodge, it is what the evidence shows. In the Stanford trial, acute anxiety and negative mood fell in all four groups, including the group that simply sat still and observed its own breathing. Positive mood rose in all four as well, but in the meditation group that rise did not reach statistical significance. The difference between groups was small and showed up in a single measure. If you still want to pick one, cyclic sighing has the strongest individual support: it was the only group that beat the control on positive mood, and the only one where the benefit grew with the number of days practised. If you are new to this, the NHS version is the gentlest way in, because it involves no breath holding at all.

How long do I have to keep it up before it makes a difference?

Five minutes a day was enough in the trial that measured the dose. The Stanford group had participants do their exercise for five minutes daily over 28 days, and the effects on mood and breathing rate were measurable at that dose. Two things are worth knowing about that number. Participants did the exercise on roughly 20 of the 28 days on average, so not every day. And the benefit of cyclic sighing grew the more days a participant had kept it up, which suggests this is a habit rather than a one-off button.

Does 4-7-8 breathing work?

It pulls the right lever. 4-7-8 means four counts in, seven held, eight out, which makes the exhale twice as long as the inhale. That is the same principle Sweden's 1177 describes when it says you can let the exhale take twice as long as the inhale. What you should know is that this particular set of numbers is not the one tested in the studies on this page. Those tested cyclic sighing, box breathing and cyclic hyperventilation. So choose 4-7-8 if it suits you, but do not assume the number combination itself is the active ingredient. The long exhale is.

Can breathing exercises be dangerous?

Slow, calm breathing is considered a technique with few expected adverse effects. There is one clear boundary, and it sits at techniques built on fast, forceful breathing. Deliberately hyperventilating before holding your breath underwater can cause hypoxic blackout, meaning you lose consciousness without warning. There are documented deaths. So the rule is absolute: that kind of breathing is done sitting or lying down on dry land, never in or near water, never in the bath, never in the pool. If an exercise leaves you dizzy or tingling in the hands, you have breathed too much or too hard. Stop and let your breathing return to normal.

Can I breathe my way out of burnout or anxiety?

No. A breathing exercise is a tool for a moment, not a treatment for a condition. The researchers behind the meta-analysis on this page say so themselves: they urge caution and call for better studies, specifically to avoid a mismatch between hype and evidence. Sweden's 1177 advises contacting a health centre or occupational health service if high stress is causing sleep problems, anxiety or heart palpitations. Five minutes of breathing does not replace that advice. In exhaustion disorder, 1177 lists breathing training among the things that can help once you are starting to feel better, meaning inside a treatment set by clinicians rather than instead of one.

Why do I feel nothing when I try it?

That is a common and expected first impression. 1177 states plainly that the first few times you do the exercise you may feel restless, and that most people get used to it after a few goes, at which point the relaxation arrives faster. In the Stanford trial, 90 per cent of participants reported a positive experience of the exercises while 10 per cent reported a negative one. The effect in the research is also small to moderate in size, so it is not something that flips a mood like a light switch. If it still feels wrong after a week, switch exercises rather than pushing harder at the same one.