The short version
- A 2023 meta-analysis of 12 randomised controlled trials in 785 adults found breathwork reduced self-reported stress with a Hedges' g of −0.35 (95% CI −0.55 to −0.14), with heterogeneity of 42%.
- The same analysis found effects of g = −0.32 on anxiety across 20 trials and g = −0.40 on depressive symptoms across 18 trials, and its authors explicitly warned against overselling the result.
- A 2017 meta-analysis of 24 heart rate variability biofeedback studies in 484 participants found a between-group effect of Hedges' g = 0.83, which was not moderated by study year, risk of bias, percentage of female participants, number of sessions or presence of an anxiety disorder.
- In a 2023 Stanford randomised study, five minutes daily of cyclic sighing — breathing with prolonged exhalations — improved mood and lowered respiratory rate more than an equal period of mindfulness meditation over one month.
- The most-cited ashwagandha stress meta-analysis reported effect sizes several times larger than breathwork's — SMD −1.75 for stress and −1.55 for anxiety across 12 trials in 1,002 people — but with heterogeneity of 83.1% and 93.8% and the certainty of evidence rated low for both.
We are a supplement company, so let us be clear about what this article is. It is the one where we tell you that the best-evidenced thing in the stress category is not on our shelf, does not cost anything, and takes five minutes a day.
The number
Fincham and colleagues, in Scientific Reports in 2023, screened 1,325 records and pooled 12 randomised controlled trials covering 785 adults, with self-reported stress as the primary outcome. Breathwork beat control conditions with a Hedges' g of −0.35 (95% CI −0.55 to −0.14, p = 0.0009). Heterogeneity was intermediate at I² = 42%. Secondary outcomes were similar in size: anxiety across 20 trials, g = −0.32; depressive symptoms across 18 trials, g = −0.40. Most included studies were judged at moderate risk of bias, and the authors close by warning against, in their words, a "miscalibration between hype and evidence".
That is a small-to-medium effect. It is not a transformation and the paper does not claim one. What matters is the shape of the estimate: a narrow-ish confidence interval, moderate heterogeneity, and authors actively arguing against overselling their own result.
Compare that to how the big botanical numbers behave. The most-cited ashwagandha meta-analysis pooled 12 randomised trials in 1,002 participants and reported a stress effect of SMD −1.75 and an anxiety effect of SMD −1.55 — several times larger than breathwork's. It also reported heterogeneity of 83.1% and 93.8% respectively, and rated the certainty of the evidence low for both. A large effect size with low certainty and wildly disagreeing studies is a less useful thing to own than a small effect size with a tight interval. Bigger is not more believable.
The version with equipment
Heart rate variability biofeedback is slow breathing with a screen telling you when you are doing it. Goessl, Curtiss and Hofmann pooled 24 studies covering 484 participants in Psychological Medicine in 2017. The pre-post within-group effect was Hedges' g = 0.81 and the between-group comparison against a control condition was g = 0.83 — a large reduction in self-reported stress and anxiety.
The moderator analysis is the interesting part. Efficacy was not moderated by study year, by risk of study bias, by the percentage of female participants, by the number of sessions, or by whether participants had a diagnosed anxiety disorder. Findings that survive that many cuts are usually findings.
Five minutes, and which five minutes
Balban and colleagues at Stanford, in Cell Reports Medicine in 2023, ran a remote randomised controlled study comparing three different daily five-minute breathing exercises against an equal period of mindfulness meditation, over one month. The three were cyclic sighing, which emphasises prolonged exhalations; box breathing, with equal inhale, hold and exhale; and cyclic hyperventilation with retention.
Cyclic sighing — the exhale-focused one — produced greater improvement in mood (p < 0.05) and a greater reduction in respiratory rate (p < 0.05) than meditation. That is an active comparator rather than a sham, which is a stronger test than most of this literature manages, and it gives you something specific to do rather than a general instruction to breathe.
The caveat we'd rather you heard from us
You cannot blind someone to whether they are breathing slowly. Every effect size above contains the participant's expectation, and there is no way to remove it. That is a genuine weakness and it is the same weakness people correctly point out in supplement trials — except a capsule can at least be matched against an identical-looking placebo, and breathing cannot.
The second caveat is about the comparison itself. Nobody has randomised people to slow breathing versus a botanical, so putting these effect sizes next to each other is not a head-to-head. It tells you which body of evidence is better behaved, not which intervention wins.
What we'd actually tell you
Do the five minutes first, for a month, before you spend anything. Exhale longer than you inhale — that is the pattern with the specific trial behind it. Do it daily, at a fixed time, and judge it at four weeks against how the start of the month felt.
If after a month it has done nothing, then a supplement conversation is reasonable, and we would rather have it with someone who has already ruled out the cheap option. Magnesium for anxiety is where most people start and the reviews there are more cautious than the marketing. L-theanine is honest about being an attention effect measured in hours. Ashwagandha has the deepest file in the category and still comes with a low certainty rating. None of them is in a different league from five minutes of breathing, and one of them costs nothing.
Good questions
Is breathwork actually better than taking a supplement for stress?
Nobody has run that head-to-head, so the honest answer is that we do not know. What we can say is that breathwork's pooled estimate is modest, reasonably consistent and comes from trials with control groups, while the largest botanical effect sizes come with low certainty ratings. Better-behaved evidence is not the same as a bigger effect.
How long do I have to do breathing exercises before anything happens?
Judge it at four weeks. The Stanford study ran five minutes a day for one month and found mood improvements against an active comparator. Most trials in the pooled analyses ran weeks rather than single sessions. If a month of daily practice has changed nothing you would have noticed anyway, stop — that result is information and it is free.
Which breathing pattern should I use?
One with longer exhalations than inhalations. In the only randomised comparison of three specific techniques, cyclic sighing — prolonged exhalations — outperformed both box breathing and cyclic hyperventilation, and beat mindfulness meditation on mood and respiratory rate. Five minutes is the tested dose, not twenty.
Doesn't the placebo effect explain all of this?
Some of it certainly does, and there is no way to blind someone to their own breathing. That is a real weakness we are not going to talk around. It is worth noting the effect survived comparison against an active control in one trial and against a range of control conditions in the pooled analysis, but expectation is baked into every number here.
Why would a supplement company tell me to breathe instead of buying something?
Because the alternative is pretending our shelf outperforms an intervention it has never been compared against. If you try the free thing first and it works, you have saved money and we have kept your trust. If it does not work, you come to a supplement conversation having ruled out the cheapest option, which is a better place to start from.
Sources
- Fincham GW, Strauss C, Montero-Marin J, Cavanagh K. Effect of breathwork on stress and mental health: A meta-analysis of randomised-controlled trials. Sci Rep, 2023. View study
- Goessl VC, Curtiss JE, Hofmann SG. The effect of heart rate variability biofeedback training on stress and anxiety: a meta-analysis. Psychol Med, 2017. View study
- Balban MY, Neri E, Kogon MM, et al. Brief structured respiration practices enhance mood and reduce physiological arousal. Cell Rep Med, 2023. View study
- Akhgarjand C, Asoudeh F, Bagheri A, et al. Does Ashwagandha supplementation have a beneficial effect on the management of anxiety and stress? A systematic review and meta-analysis of randomized controlled trials. Phytother Res, 2022. View study
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
