The Science of Slow Breathing

Slow breathing is one of the few self-help practices that has both a plausible physical mechanism and a real research literature behind it. It is also surrounded by claims that literature does not support. This page is the reference desk for the rest of the site: every study we lean on, summarised in plain language, with the limits stated as clearly as the findings.

Everything below comes from a paper we opened and read. Where a result is small, or the sample was 40 people, or the trial ran for four weeks and then stopped, you will see that written down next to the result. That is not hedging for its own sake. Breathing research is a young field with a lot of enthusiastic coverage and a fairly thin evidence base underneath it, and you deserve to know which is which before you decide how much weight to give it.

How slow breathing affects the autonomic nervous system

Your autonomic nervous system runs the machinery you do not think about: heart rate, digestion, blood vessel tone, sweat. Breathing is the unusual one. It runs automatically, but you can also take the wheel whenever you want to. That single fact is why breathing practices exist at all — it is the most accessible voluntary input into an otherwise involuntary system.

The connection is not mystical. Your heart rate already rises slightly as you breathe in and falls as you breathe out; this is called respiratory sinus arrhythmia, and it is visible on any decent heart rate trace. Slowing and lengthening your breath exaggerates that swing. The vagus nerve, which carries most parasympathetic traffic to the heart, is the main pathway involved.

The most careful summary of what happens next is a 2018 systematic review in Frontiers in Human Neuroscience. Its authors write: Slow breathing techniques promote autonomic changes increasing Heart Rate Variability and Respiratory Sinus Arrhythmia paralleled by Central Nervous System (CNS) activity modifications. They add: EEG studies show an increase in alpha and a decrease in theta power. That is a brainwave pattern usually associated with relaxed wakefulness.

That is the mechanism, and it is worth being careful about how far it stretches. The same review is candid about the state of the field: the authors found interesting albeit limited evidence of a relationship between physiological parameters and psychological/behavioral outcomes in healthy subjects undergoing slow breathing techniques, and note that the paucity of collected evidence is mostly ascribable to the heterogeneity of the investigated techniques. In other words, we can measure the physiology reliably. Connecting that physiology to how you actually feel is the part that still has gaps in it.

Resonance frequency and heart rate variability

Heart rate variability is the beat-to-beat variation in the gaps between heartbeats. A healthy heart is not a metronome; the intervals flex constantly in response to breathing, blood pressure and posture. Higher variability generally reflects a more responsive autonomic system, and it is the outcome most breathing studies measure, because it is cheap, non-invasive and sensitive to what you are doing right now.

There is a breathing rate at which this effect is largest. Paul Lehrer and Richard Gevirtz call it the resonance frequency, and describe a resonant system as one that, when stimulated, produces high-amplitude oscillations at a single frequency, recruiting or overshadowing other frequencies, to produce a sine wave oscillation with very high-amplitude. For breathing and the heart, they report that the maximum heart rate oscillations at respiratory frequency occurred at approximately 0.1 Hz (six breaths per minute), the one frequency at which heart rate oscillates with breathing at a 0° phase relationship, i.e., exactly in phase.

Six breaths a minute is five seconds in and five seconds out. That is the entire reason our heart coherence timer uses a 5-5 pattern, and it is a good default for anyone who wants one number to remember.

The proposed reason the effect is so large involves the baroreflex, the loop that keeps blood pressure steady by adjusting heart rate. Lehrer and Gevirtz describe what happens when the system is stimulated at the specific frequency causing maximum heart rate oscillations and a 180° phase relationship between heart rate and blood pressure — breathing at that rate lines up with a control loop the body already runs, and the two reinforce each other. The same authors are open that this is not the whole story: they conclude that it is probable that a number of mechanisms are involved in various effects, in addition to baroreflex stimulation.

The size of the heart rate variability effect is one of the better-replicated findings in this whole area. A 2022 systematic review and meta-analysis in Neuroscience and Biobehavioral Reviews screened 1,842 abstracts and included 223 studies. Its result is stated flatly: Results indicate increases in vmHRV with VSB, DURING, IM-AFTER1, and AFTER-INT — that is, vagally-mediated heart rate variability rose while people breathed slowly, straight after a single session, and after longer multi-session programmes. The authors' conclusion is measured rather than triumphant: slow breathing could be advised as a low-tech and low-cost technique to use in prevention and adjunct treatment purposes, with few adverse effects expected.

Two honest caveats. First, a rise in heart rate variability is a physiological marker, not a symptom you can feel, and the leap from marker to wellbeing is exactly the leap the 2018 review said the field has not fully made. Second, resonance frequency varies from person to person; six breaths a minute is a good average, not a personal prescription.

Cyclic sighing: the clearest single trial we have

If you want one well-designed, recent, randomised study to point at, it is Balban and colleagues, published in Cell Reports Medicine in 2023. It is the paper behind our cyclic sighing timer.

The design, in the authors' words: Here, we report a remote, randomized, controlled study (NCT05304000) of three different daily 5-min breathwork exercises compared with an equivalent period of mindfulness meditation over 1 month. One hundred and eight people enrolled, and the arms were uneven: From the 108 subjects enrolled, 24 were randomized into the Mindfulness Meditation control condition and 84 were randomized to the treatment conditions (30 Cyclic Sighing, 21 Box Breathing, 33 Cyclic Hyperventilation with Retention). Everyone was asked for five minutes a day for four weeks, and adherence was partial rather than perfect: On average, mindfulness meditation participants completed 17.71 ± 9.25 of the 28 days, and breathwork participants completed 19.61 ± 7.73 of the 28 days.

The headline finding: breathwork, especially the exhale-focused cyclic sighing, produces greater improvement in mood (p < 0.05) and reduction in respiratory rate (p < 0.05) compared with mindfulness meditation. Cyclic sighing is a double inhale through the nose — a full breath followed by a short top-up — and then a long, slow exhale through the mouth. It is the pattern our timer runs at roughly two seconds in, one second in again, six seconds out.

Now the limits, which the authors state plainly. The study ran remotely: The remote nature of the study limited the monitoring of how closely participants followed the instructions on a daily basis. Nobody watched people breathe. And the findings of this study are limited to 4 weeks with no additional follow up, so we do not know whether the mood benefit persisted, faded, or grew. Mood was self-reported. The comparison group was mindfulness meditation rather than an inert control, which is a strength for interpreting the difference but means the trial does not tell you what would have happened with nothing at all.

Our reading: this is a genuinely encouraging result from a well-run trial, and it is one trial. That is why cyclic sighing is the technique we recommend when someone wants the option with the most direct evidence behind it, and why we do not describe it as proven.

Slow breathing and blood pressure

Blood pressure is where breathing research gets the most attention and where the evidence needs the most caution, because the stakes of getting it wrong are real.

The relevant pooled analysis is a 2023 systematic review and meta-analysis in the International Journal of Cardiology: Cardiovascular Risk and Prevention. Fifteen studies with roughly 1,097 participants were combined. The result: Breathing exercises have a modest but significant effect on decreasing systolic blood pressure (−7.06 [−10.20, −3.92], P = <0.01) and diastolic blood pressure (−3.43 [−4.89, −1.97], P = <0.01) mm Hg. Heart rate fell too: Breathing exercises were also observed to cause a significant decrease in the heart rate (−2.41 [−4.53, −0.30], P = 0.03) beats/minute.

A seven-point drop in systolic pressure would be clinically meaningful if it held up. Here is why we are not going to tell you it does. The review's own quality assessment is blunt: All 15 studies which were included in the review have an unclear or high risk of bias. Heterogeneity between the trials was very high — around 90% for the systolic result — which means the individual studies disagreed with each other a great deal, and the pooled number papers over that disagreement rather than resolving it.

So: a promising signal, from a body of work that is not yet good enough to act on alone. If you have high blood pressure, slow breathing is a reasonable thing to add on top of what your clinician has prescribed. It is not a reason to change or stop anything. Please read the medical disclaimer before you treat anything on this site as advice.

Breathing and anxiety

The pooled analysis we lean on for mental health outcomes is Fincham and colleagues, Scientific Reports, 2023. It pooled randomised controlled trials only, which is the right bar for this question.

For the primary outcome, the authors report: The primary outcome self-reported/subjective stress included 12 randomised-controlled trials (k = 12) with a total of 785 adult participants, and The random-effects analysis yielded a significant small-to-medium mean effect size, g = − 0.35 [95% CI − 0.55, − 0.14]. Anxiety and depressive symptoms were secondary outcomes: Meta-analyses for secondary outcomes of self-reported/subjective anxiety (k = 20) and depressive symptoms (k = 18) showed similar significant effect sizes: g = − 0.32, p < 0.0001, and g = − 0.40, p < 0.0001, respectively.

A small-to-medium effect is a real effect. It is also, in plain terms, a nudge rather than a transformation, and it was measured with questionnaires rather than anything objective. The authors' own closing note is the most quotable sentence in the paper and we think it belongs on every page like this one: However, we urge caution and advocate for nuanced research approaches with low risk-of-bias study designs to avoid a miscalibration between hype and evidence.

If you are here because anxiety brought you here, our page for anxious moments is written for that, and it carries crisis lines at the top. Breathing is a tool for turning down the volume on an anxious moment. It is not treatment for an anxiety disorder, and it does not replace a therapist or a prescription.

A note on single, small studies

Most breathing research is small. The 2022 study on 4-7-8 breathing in Physiological Reports is a fair example of both the value and the ceiling of that kind of work. Forty-three healthy adults aged 19 to 25 were split into a sleep-deprived group and a control group, and both did 4-7-8 breathing while lying down.

In the rested control group, the authors report that Low- and very-low-frequency powers significantly decreased…whereas high-frequency power significantly increased…in the control group — a shift toward parasympathetic dominance — and that The control group had significantly decreased HR, SBP, DBP, MAP, and RPP levels, meaning heart rate and blood pressure both came down during the practice.

That is a genuine within-session physiological effect, measured properly. It is also 43 young, healthy university-age adults in a single lab, doing one session lying on their backs. The authors flag the limits themselves — The small sample size of males in this study appears to be insufficient to accurately reflect the general population's condition, and Future research should investigate the long-term effects of this breathing practice on cardiovascular and cardiac autonomic functioning over a period of months or years. It tells you something happens while you breathe that way. It tells you very little about what a month of practice does for a 55-year-old with insomnia.

How to do it

If you have read this far and want the practical version, here it is. This is a general slow-breathing session at roughly six breaths a minute — the resonance-frequency default described above, with no breath holds.

1. Sit or lie somewhere you can stay still for five minutes. Upright in a chair with your feet on the floor is fine. So is lying on your back. Comfort matters more than posture.
2. Let your shoulders drop and unclench your jaw. Rest one hand on your belly if it helps you feel the movement. You are aiming for the belly to rise more than the chest.
3. Breathe in through your nose for a slow count of five. Not a deep, dramatic inhale. A smooth, unhurried one that fills you comfortably and stops before it strains.
4. Breathe out through your nose or mouth for a slow count of five. Let it fall out rather than pushing it. The exhale is the part that does most of the work.
5. Keep going for five minutes. That is about thirty breaths. Your attention will wander; when you notice, come back to the count without making anything of it.
6. Stop if you feel lightheaded, and breathe normally. Feeling dizzy means you are breathing more than your body wants. Make the breaths smaller, not bigger.

If five seconds each way feels like a strain, start at four in and four out and lengthen it over a week. If it feels too easy, try four in and six out — a longer exhale is the pattern behind our extended exhale timer.

When to use it, and when not to

Good moments. Before something you are dreading. After something that rattled you. In the gap between work and home. Lying in bed with a mind that will not settle. Any moment where you have five minutes and a reason to want your nervous system a notch quieter.

Less good moments. While driving, cycling or operating anything with a motor. In water. During a task that needs your full attention. Immediately after a large meal, when lying flat and breathing deeply is simply uncomfortable.

Not a substitute. If you are in crisis, or if a symptom is new, severe or getting worse, breathing is not the right tool and this is not the right website. Contact your doctor or your local emergency service. If chest pain, severe breathlessness or fainting is involved, treat it as urgent.

Who should be careful

Gentle slow breathing with no holds — the pattern above — is low risk for most people. Two other families of technique are not, and both appear elsewhere on this site: patterns with long breath holds (such as the seven-count hold in 4-7-8 or the two holds in box breathing), and patterns with deliberately fast breathing (our energizing breath).

None of the research summarised on this page studied people with those conditions. The trials above recruited healthy adults, and a finding in healthy adults does not transfer automatically to anyone else.

What the evidence does not show

This is the section most pages like this one leave out, so here it is in full.

• It does not show that breathing treats or cures any condition. Not anxiety disorders, not hypertension, not insomnia, not asthma. The trials measure symptoms and markers over weeks, not disease outcomes over years.
• It does not show that one named pattern beats another. The Balban trial found cyclic sighing came out ahead of the other two breathwork arms and of meditation on mood, in one study of 108 people. That is a lead, not a ranking. Nobody has run the head-to-head trials that would settle whether 4-7-8 beats box breathing, and we will not pretend otherwise on our comparison page.
• It does not show long-term benefit. The Balban trial ran four weeks with no follow-up. Most others are shorter. There is essentially no good evidence about what a year of daily practice does.
• It does not rest on objective outcomes. Mood, stress and anxiety in these studies are self-reported questionnaires. That is standard practice in the field and it is also a real limitation, particularly where participants know which group they are in.
• It does not come from large samples. Individual trials in this area typically enrol tens of people, not thousands. The 2023 blood pressure meta-analysis pooled around 1,097 participants across fifteen studies — and rated every one of those studies as unclear or high risk of bias.
• It does not show that a timer app improves outcomes. No trial has tested this website, or any breathing app, against practising without one. A timer removes the counting; that is the honest extent of the claim.

What is left after all that is still worth having: a free, portable, side-effect-light practice with a measurable physiological signature and a small-to-medium effect on how stressed people say they feel. We think that is a good thing. We just do not think it is a miracle, and we would rather you heard the caveats from us than found them out later.

How we choose sources

A short, checkable policy, because a citation hub is only worth as much as the rules behind it.

1. We open every source. No citation appears on this site unless someone here loaded the paper and read the sentence being relied on. Every reference below links to the full text or the record we actually opened.
2. We prefer primary literature. Randomised controlled trials, systematic reviews and meta-analyses in peer-reviewed journals. Not press releases, not podcast episodes, not other websites summarising a study we have not read.
3. We quote the supporting sentence. If we say a study found something, the study's own wording is on the page in quotation marks, so you can check that our paraphrase is fair.
4. We publish the limitations alongside the findings. Sample size, follow-up length, risk of bias and heterogeneity get stated in the same breath as the result, using the authors' own assessments where they made one.
5. We never write "studies show". If we cannot name the study, we do not make the claim.
6. We do not claim clinical outcomes. Nothing here treats, cures, prevents or diagnoses anything, and we will not imply that it does.
7. We correct things. If you find a misread result or a broken link, write to contact@helpmebreath.com and we will fix it and say what changed.

Georges Rayess writes and maintains this site. He is not a clinician, and the about page says so plainly. This page is a careful summary of published research by a non-specialist, not a clinical review.

Common questions

Related

A note on what this is

This page summarises published research. It is not medical care, it is not a diagnosis, and it does not replace treatment or advice from a qualified clinician. If a symptom is new, severe or worsening, contact a doctor. Read the full medical disclaimer.