Breathing Exercises is a research and writing project about one narrow question: what actually happens in the body when you change how you breathe, and which of the techniques built on that are worth your time. We publish free timers for the techniques that hold up, written guides that show the evidence behind them, and an honest label on the ones that do not hold up as well as the internet says they do.
- Every technique guide is built from published research, and every claim in it is traceable to a specific paper you can open yourself.
- We rate the strength of that research openly, including when it is weak.
- Where a breathing pattern has been tested against an active control and failed, the page says so.
- The timers are free, require no account, and are not a funnel to anything.
- Every page names the writer and the person who checked its citations.
- Nothing here replaces advice from a professional who knows your history.
Who Writes This
Breathing Exercises is written by a small team of researchers and writers. Jay Newmark is the lead writer and editor: he sets the evidence standard described below, and nothing publishes until it clears that standard. Contributing writers take on individual guides, and a separate fact-checking pass runs on each one before it goes live.
Every page names the person who wrote it and the person who checked its citations, along with the date each of those happened. That is deliberate. You should be able to see who stands behind a specific claim instead of trusting an anonymous house byline.
What the team has in common is method rather than background. We read the primary literature on respiratory physiology and behavioral breathing interventions, trace popular claims back to the studies they came from, and put what survives into plain language. Where the honest answer is “the studies are small and nobody has replicated this,” that is what the page says — which is the part of the job most breathing content skips.
How We Research a Technique
Every technique guide on this site follows the same process, in this order.
- Find the origin of the claimMost breathing advice online cites another article, which cites another article, which cites nothing. We follow the chain back until it ends at a study or it ends at no study at all. Both outcomes are publishable, and the second one is often the more useful page.
- Read the paper, not the abstractSample size, control condition, duration, and what was actually measured. A protocol that “reduces stress” in twelve undergraduates over one session is a different fact from one tested in three hundred people over eight weeks, and the guide says which one it is.
- Check the control arm before citing anythingA trial reported as positive can turn out to have had a control group that improved more, and a reference can resolve to a different paper than the one being cited. Neither is visible unless someone opens the source and reads it against the sentence it is attached to.
- Prefer the strongest available evidenceSystematic reviews and meta-analyses first, then randomized controlled trials, then smaller controlled work, then physiological mechanism studies. Anecdote and tradition are described as anecdote and tradition.
- Separate the mechanism from the outcomeThat slow exhalation raises heart rate variability is well measured. That a given change in heart rate variability makes you sleep better is a separate claim needing separate evidence. Guides keep these apart, because collapsing them is how a plausible mechanism gets sold as a proven result.
- Write the protocol so it is reproducibleExact counts, exact durations, position, and how long before anything is expected to happen — matching what the study protocol actually used, not a rounded version of it.
- Cite it, inlineEvery technique guide carries a numbered reference list, and each substantive claim in the body links to its entry. If a sentence cannot be tied to a source, it either gets marked as our own reasoning or it comes out.
How We Rate Evidence
Two Scales, Because a Technique and a Single Claim Are Not the Same Thing
Every technique carries one badge summarizing the state of the research behind it. The badge is visible on the guide, on the timer, and on every hub card that lists it, so the rating travels with the technique instead of being buried in the text.
| Rating | What it means |
|---|---|
| Well Supported | A physiological effect measured repeatedly and replicated by groups with no stake in the result. That paced breathing near six breaths a minute shifts heart rate variability is this kind of finding. This tier is not available for outcome claims — see the ceiling below. |
| Qualified | Real controlled evidence exists, but it beats doing nothing rather than beating an active alternative, or the samples are small, the follow-up short, or the result unreplicated. Worth trying, and worth knowing the limits of. |
| Thin Evidence | Mechanistically plausible and widely taught, but the specific claim the page makes has not been tested directly. Harmless to practice, and we say plainly that the popular claims outrun the data. |
| Unrated | We have not assessed it yet. The tier exists so that an unassessed page is visibly unassessed, rather than quietly implied to be fine. |
The Ceiling on Outcome Claims
This is the part most breathing sites leave out, and it is why the top badge is defined the way it is above.
Slow breathing has real, measurable short-term physiological effects. What has not held up is the idea that any particular pattern is special. Every breathing parameter that has been isolated and tested against a matched active control has come back null:
- Coherent breathing at 5.5 breaths a minute. Four hundred people randomized against a matched breathwork placebo at twelve breaths a minute. Null on anxiety, depression, wellbeing and sleep.
- A longer exhale than inhale. Tested directly at a matched breathing rate. Null on every heart rate variability metric.
- Tuning the rate to an individual’s resonance frequency rather than using a fixed six breaths a minute. Null over four weeks, in a trial where resting heart rate variability did not improve in any arm.
- The abdominal emphasis in diaphragmatic breathing. Never isolated and tested at all. No trial compares abdominal-led against chest-led breathing at matched rate, volume and ratio. It is a teaching device with no demonstrated independent effect.
The honest reading is that slow breathing is doing the work, and the named protocols are largely delivery mechanisms for it — ways of getting a person to breathe slowly without having to think about it. That is not nothing. A pattern you will actually follow beats an optimal one you will not, and the counting is what makes it followable.
No technique on this site is rated Well Supported on an outcome claim, and none will be until a pattern beats an active breathing control on something a person would notice. Where you see Well Supported here, it is attached to a physiological effect, and the outcome claims built on top of that effect are rated separately and lower in the evidence table.
We would rather show you a ceiling than pretend there is not one.
Claim Level
Inside a guide, individual claims get their own verdict, because a single technique often mixes strong findings with weak ones. Box breathing has reasonable evidence for short-term calming and essentially none for the focus benefits it is usually sold on. One badge cannot carry that, so each claim is rated separately.
| Verdict | What it means |
|---|---|
| Supported | Controlled evidence directly tests this claim and finds the effect. |
| Qualified | The effect appears, but under narrower conditions than the popular version of the claim implies. |
| Not Shown | The claim is widely repeated and has not been demonstrated. Absence of evidence, stated as such. |
A claim rated Not Shown also keeps the technique out of that use-case archive. If a guide rates the focus claim Not Shown, that technique does not appear under Focus, however much traffic the pairing would attract.
When new work lands, the badge changes and the guide’s last-reviewed date changes with it. Several ratings on this site are lower than they were at first publication. A rating that only ever improves is a marketing device, not an assessment.
Fact-Checking, Dates, and Corrections
Before a technique guide publishes, someone other than its writer opens every citation on it and checks the source against the sentence it supports. That pass is what the fact-check credit on a page refers to, and it is the reason some pages sit unpublished for a while. It is a fact-check, not a medical review, and we do not describe it as one.
Every page shows two dates, kept separate on purpose. Published is when the page first went live. Last reviewed is the most recent date somebody re-opened the sources, checked for newer research, and confirmed the page still says the right thing. A lot of health sites quietly change one date to look current. Showing both is the only version of this that means anything.
If something here is wrong, we want to fix it and we want to fix it visibly. Email corrections@breathingexercises.com with the page and what is wrong with it. Substantive corrections are noted on the page itself with the date, not silently edited away. A page whose central claim no longer holds is rewritten rather than patched, and its evidence rating is lowered to match. Pointing at a study we missed is the most useful message you can send us.
Where AI Is and Is Not Used
Automation is part of how this site is built, so here is where it sits.
- Used for: drafting and restructuring prose, checking spelling and internal consistency, generating the code behind the timers, and producing first-pass summaries that a person then verifies against the source.
- Not used for: deciding an evidence rating, deciding what to publish, or writing anything that reaches you without a person having read the underlying papers.
- Never: generating citations. Every reference on this site was located by a person, opened, and checked to resolve to the paper it claims to be. This is not a hypothetical safeguard. Fabricated and misdirected citations are the most common failure mode in health content, and they are invisible to a reader who does not click through.
What We Will Not Publish
- Treatment content for diagnosed conditions. No COPD, asthma, sleep apnea or hypertension guides, and nothing on breathing during pregnancy. Those belong with the clinician treating you or with a national health service, not with an independent site.
- A health claim with no traceable source behind it.
- The words cure, heal or detox applied to a breathing pattern.
- Supplements, and any protocol presented as a substitute for treatment.
- Fear framing — the “you have been breathing wrong your whole life” opening that most of this genre runs on.
- A timer, calculator or protocol locked behind an email address.
- A recommendation we would not give someone we know.
Several of those are searchable, high-traffic subjects. We would rather send you away than rank for them.
How the Site Makes Money
Two ways, both worth stating plainly. The site carries advertising, and some links to products earn a commission if you buy through them. Neither one costs you anything and neither one buys influence over what a page says.
| Source | What it may influence |
|---|---|
| Advertising | Nothing about what we publish or how we rate it. Advertisers are placed by a network and we do not select them. |
| Affiliate commissions | Which products we spend time testing, since testing costs money. Never the verdict, the score, or whether a fault gets mentioned. |
The practical rules behind that: no company pays for coverage or sees a page before it publishes. Hardware is normally bought with our own money rather than sent to us, and on the occasions a manufacturer does supply something, the review says so at the top of the page rather than in a footnote. Where a product does not justify its price, the review says so and the link stays, because the honest verdict is the only thing worth linking to. And the timers stay free and account-free regardless — they are the point of the site, not the bait.
Common Questions
Are the techniques on this site safe?
For most healthy adults, slow breathing at a comfortable depth is very low risk. The exceptions matter though, and they are printed on the individual pages: breath holds and intense hyperventilation-style methods carry real risk in and around water, during pregnancy, and for anyone with a cardiovascular or seizure history. Any technique that makes you dizzy, tingly or panicked is telling you to stop.
Why does a technique I like have a low evidence rating?
A low rating is a statement about the published research, not about your experience. Plenty of practices help people while sitting on thin evidence, and that is a normal state for a field this young. What the rating protects against is a specific thing: paying for a program, or delaying real treatment, because a claim sounded settled when it was not.
If the trials keep coming back null, why publish any of this?
Because the nulls are about the choice between patterns, not about slow breathing itself. Deliberately slowing your breathing does measurable things, reliably, within a minute or two. What has not been shown is that four counts in and four counts out beats some other slow pattern. That makes the counting a delivery mechanism rather than an active ingredient — which is genuinely useful to know, and is the opposite of a reason to stop.
Can I use these instead of medication or therapy?
No. Breathing techniques are worth using alongside treatment, not in place of it, and any decision to change a prescribed treatment belongs with the clinician who prescribed it.
Where do the studies come from?
Peer-reviewed journals, located mainly through PubMed and the Cochrane Library. Each guide lists its sources in full at the bottom, with links, so you can read the original rather than take our summary of it.
Do you accept guest posts or sponsored articles?
Sponsored articles, no — nobody pays for placement or for a favorable rating, ever. Writing for us is a different question: everything published here is commissioned and edited by our own team rather than accepted as a submission, but if you have relevant research experience and something you want to cover, write to us and say so.
Get in Touch
Corrections and missed studies go to corrections@breathingexercises.com. Anything else — a technique you want covered, a question about a protocol, a research collaboration — goes to hello@breathingexercises.com. We read everything and reply to most of it.
Medical Disclaimer
This site publishes research summaries and practice tools for general education. It is not medical advice, and using it does not create a clinical relationship of any kind. Nothing here is a diagnosis, a treatment plan, or a substitute for care from a qualified professional who knows your history.
Talk to a doctor before starting any breathing practice if you are pregnant, or if you have a respiratory condition such as asthma or COPD, a cardiovascular condition, epilepsy or a seizure history, uncontrolled high or low blood pressure, glaucoma, panic disorder, or any condition affected by changes in blood pressure or blood gases. Never practice breath holds in or near water. If you feel faint, chest pain or shortness of breath, stop and seek medical attention.