Does Breathing Out Longer Actually Calm You Down?
Breathe in for four, out for six. Every app says some version of it. Ours does too: the default Breath Pattern in 5 Minutes is a four second inhale and a six second exhale. That is a design choice we made, so we went and checked whether the research actually backs it. The answer is more interesting than a yes.
Has anyone tested five minutes a day?
We wrote last time that nobody has properly tested five minutes of daily meditation. The shortest doses in good trials were ten and thirteen minutes. That is still true.
Breathing is the exception. In a 2023 Stanford trial, 108 people did five minutes a day for 28 days. They were randomized into four groups: mindfulness meditation as the control, and three breathing exercises. Box breathing, equal in and out and hold. Cyclic hyperventilation, fast in and short out. And cyclic sighing, which is two inhales through the nose followed by one long slow exhale.
So the one practice anyone has tested at exactly five minutes a day is breathing. That alone made it worth reading closely.
Did breathing beat meditation?
On one measure out of three.
The headline result is real: the breathwork groups gained more positive mood per session than the meditation group. The average daily change was 1.91 points for breathwork against 1.22 for meditation, on a scale that runs from 10 to 50. Cyclic sighing, the exhale-focused one, did best of the three.
Here is the part that did not travel. On the other two measures, the groups were the same. In the authors' words, "there were no differences between the two groups in state anxiety and negative affect changes." And if you look at the raw daily numbers rather than the model, meditation was slightly ahead on both. Negative affect fell 1.62 points for meditation against 0.98 for breathwork. State anxiety fell 3.95 points against 3.03, on a scale of 20 to 80.
So breathing did not out-calm meditation. It out-cheered it. That is a narrower claim than "breathwork beats meditation", which is how the study got summarized almost everywhere.
One finding does deserve the attention it got. The advantage grew with practice. People who had been on the protocol longer got a bigger lift from each session, and that pattern was significant for cyclic sighing and not for the others.
What happened in the body?
Less than you would expect.
Breathing slowed people's resting respiratory rate more than meditation did, and the size of that slowdown tracked the mood gain (r = -0.24, meaning the people whose breathing rate dropped most gained the most positive affect). Modest, but in the direction you would predict.
Then the part that rarely gets quoted: there were no significant changes in heart rate variability or resting heart rate in any group. HRV is the mechanism every breathing app invokes. In the trial that everyone cites as proof, it did not move.
Why would slowing down help at all?
Because the slowing itself has a real evidence base, separate from this trial.
A 2018 systematic review of slow breathing, defined as fewer than ten breaths a minute, found consistent increases in heart rate variability and respiratory sinus arrhythmia, shifts in EEG toward more alpha and less theta, and subjective reports of more comfort and relaxation alongside less arousal, anxiety and anger. A 2022 meta-analysis pulled in 223 studies of voluntary slow breathing and found that the vagally-mediated part of heart rate variability rose during the breathing, right after a single session, and after a multi-session course.
That is a lot of evidence, and it is all about rate. None of it is about the ratio.
Does the exhale specifically matter?
This is where our four-in, six-out default stops being supported.
A 2024 paper went looking for exactly this answer and found the literature split. Four studies favored longer exhales than inhales. Three found the ratio made no difference. One favored equal inhales and exhales. One favored longer inhales. Not a consensus, an argument.
So the authors ran it themselves, twice, at six breaths a minute. In the original study of 26 people, a one-to-two ratio produced no different heart rate variability than a one-to-one ratio, on time-domain or frequency-domain measures. The replication with 16 people confirmed it and found no difference on nonlinear measures either.
Small studies, and absence of a difference is not proof of no difference. But it is the most direct test of the exhale rule anyone has run, and it came back flat, twice.
There is a second problem with reading the Stanford trial as proof of the exhale. Cyclic sighing differed from the other conditions in two ways at once: the longer exhale and the double inhale. The design cannot tell you which one did the work, and the authors say so.
How strong is this evidence overall?
Honest answer: promising and thin.
A 2023 meta-analysis of 12 randomized trials covering 785 adults found breathwork lowered self-reported stress with an effect size of 0.35. Anxiety and depression came out at 0.32 and 0.40 across a larger set of trials. Small to medium, and real. The authors, one of whom is a trained breath teacher, still closed by urging caution and warning against "a miscalibration between hype and evidence".
The same lead author then did the harder experiment. A 2024 placebo-controlled trial gave 200 adults three weeks of either intense breathwork or a credible placebo breathing practice, with both groups blinded to which they had. Both improved on stress. Neither improved more than the other. When the control group gets something that feels like a real practice, the gap can vanish.
The Stanford trial has its own limits, and the authors list them plainly. It was exploratory, not pre-registered as a clinical trial and registered only afterwards. Most participants came from one Stanford undergraduate psychology class. It ran four weeks with no follow-up. Each group was small. Participants were given a wearable worth around 350 dollars and a free subscription, and one author became an advisor to that company.
None of that makes the findings wrong. It makes them a starting point rather than a verdict.
What should you actually do?
Three things.
- Slow down. That is the lever, and it is the one thing every strand of this research agrees on. Under ten breaths a minute is the range the reviews cover, and the biofeedback work centers on about four and a half to six and a half.
- Stop optimizing the ratio. Four in and six out, or five and five, or three and seven. The best available test says it does not change your physiology. Pick whichever you can hold without straining, because straining is its own kind of arousal.
- Judge it over weeks. The only effect that clearly grew in the Stanford trial grew with days practiced, not with technique.
We are keeping the Calm pattern at four in and six out. Not because a trial proved that ratio wins, because no trial has. We keep it because it lands you near six breaths a minute, and because a long exhale is easy to follow when your mind is somewhere else. That is a usability reason, not a physiological one, and it seems fair to say so out loud.
The research does not show that a longer exhale is a secret switch. It shows that breathing slowly, most days, for a few minutes, does a small and genuine amount of good. That is a smaller promise. It also happens to be one you can keep.
Sources
This article shares general information about meditation research. It is not medical advice. If you have concerns about your mental health, talk to a qualified professional.
Five minutes. Every day. That is the whole app.