Cyclic sighing: what the Stanford study found

How the study was run

The paper appeared in the journal Cell Reports Medicine. Its authors are M. Yilmaz Balban and a Stanford team, with David Spiegel and Andrew Huberman as corresponding authors. Data were collected from June to September 2020, remotely: nobody came into a lab.

There were 108 adult participants with no moderate-to-severe medical or psychiatric conditions. Most were recruited from an undergraduate psychology class at the same university. Each person got an instructional video, a daily text reminder and one assignment: practice five minutes a day for 28 days.

Assignment was random, with one wrinkle. People who shared a household were placed in the same group, which is why the groups aren't the same size.

The four practices

• Mindfulness meditation (24 people): sitting or lying down with eyes closed, watching the breath without changing it and coming back to it when attention drifts.
• Cyclic sighing (30 people): one slow breath in, a second, shorter one to finish filling the lungs, then a slow, complete breath out. Ideally in through the nose and out through the mouth.
• Box breathing (21 people): breathe in, hold, breathe out, hold, each side the same length. The sides weren't always 4 seconds: they ran from 3 to 10, based on a quick test of how long each person took to let a breath out.
• Cyclic hyperventilation with retention (33 people): three rounds of 30 deep breaths, each followed by a hold of about 15 seconds on empty lungs.

What was measured

Before and after every session, each person filled out two short questionnaires: one on anxiety in the moment and one on mood, which separates positive affect (feeling energetic, interested, enthusiastic) from negative affect. At the start and end of the month, the team also measured anxiety as a trait and sleep-related daytime tiredness.

On top of that, everyone wore a wrist strap that tracked breathing rate, resting heart rate, heart rate variability and sleep. So alongside what people said they felt, there were readings from the body.

What it found

In every group, in-the-moment anxiety and negative affect dropped after sessions. On those two measures there were no differences between groups.

The difference showed up in positive affect. On a scale that runs from 10 to 50, it rose by an average of 1.91 points per session with the breathing exercises and 1.22 with meditation. That's a small change. In the statistical analysis, cyclic sighing was the only one of the three exercises whose edge over meditation was significant, and that edge grew with more days of practice.

In the body, resting breathing rate fell more with the breathing exercises than with meditation, and again cyclic sighing was the one that stood out. There were no significant changes in resting heart rate, heart rate variability or sleep. Trait anxiety didn't change in any group either.

The limits the authors point out

• It was exploratory. The team describes it as groundwork for a larger trial. It wasn't registered before it began, only retrospectively.
• The groups were small. The study was sized to compare breathing as a whole with meditation, not the three exercises with one another.
• It lasted four weeks. There was no follow-up, so nobody knows what happens afterward.
• It happened at home. Nobody checked how or how much each person practiced. On average, people in the breathing groups completed about 20 of the 28 days.
• There was no group that did nothing. Everyone did some practice each day. What dropped in every group may be partly due to the pause itself, or to what each person expected.

A few more things worth knowing

Mood and anxiety were measured with questionnaires, and nobody can do an exercise without knowing which one they were assigned. That means expectations can color the answers.

The paper also discloses that the company behind the wrist straps donated them, with no role in the design or analysis, and that people who finished the study kept theirs. It reports as well that Andrew Huberman, who recorded the instructional videos, became an advisor to that company in June 2022, after the data were collected. It's disclosed, which is the right thing.

Finally, the sample came mostly from one university. What happened in that group won't necessarily repeat in people of other ages or with different lives.

What to do with this

The study doesn't show that cyclic sighing is a treatment, or that it's the best breathing exercise. It shows that, in a small group over one month, five minutes a day went along with slightly better mood, a little more than with meditation. It's a single trial.

If you'd like to try it, Cyclic sighing on this site runs five minutes, with a 2-second inhale, 1 more second, and a 6-second exhale through the mouth. The study gave no seconds, only the shape. After a scare uses three sighs and follows them with a minute of slow breathing. If you want something simpler still, start with 4-6 breathing.

If you feel dizzy or tingly, stop and breathe normally. And if anxiety is intense or won't let up, a breathing practice is no substitute for professional help.