The Vayu Pilot Study: What 199 Participants Showed Over 4 to 6 Weeks
In a naturalistic pilot study of 199 Vayu users, heart rate variability (HRV) increased 28.6% (Cohen's d = 0.68), resting heart rate decreased 7.4 bpm (d = 0.86), and perceived stress measured on the PSS-10 fell 2.5 points, a 16% reduction (d = 0.41), over 4 to 6 weeks of app-guided breathing practice. No adverse events were reported. The manuscript is in revision at Human Factors (Sage). The research was conducted with the SFU Metacreation Lab and supported by NRC IRAP funding.
What did the study measure?
Three outcomes: HRV (the beat-to-beat variation in heart rate, a standard marker of autonomic nervous system function), resting heart rate, and perceived stress using the PSS-10, a widely used 10-item self-report scale. Physiological measures came from participants' own wearables during normal app use.
What were the results?
| Outcome | Change | Effect size |
|---|---|---|
| Heart rate variability (HRV) | +28.6% | d = 0.68 (medium-to-large effect) |
| Resting heart rate | −7.4 bpm | d = 0.86 (large effect) |
| Perceived stress (PSS-10) | −2.5 points (−16%) | d = 0.41 (small-to-medium effect) |
| Adverse events | None reported | — |
For context: Cohen's d is a standard measure of how large an effect is. Values around 0.2 are considered small, 0.5 medium, and 0.8 large. The resting heart rate change (d = 0.86) was the strongest effect observed; the perceived stress change (d = 0.41) the most modest.
How was the study designed?
This was a naturalistic pilot: 199 participants used Vayu in their daily lives for 4 to 6 weeks, without laboratory visits or supervised sessions. Outcomes were compared before and after the practice period. Naturalistic designs trade experimental control for ecological validity: the results reflect how people actually use a breathing app, not how they perform in a lab.
What are the limitations?
Several, and they matter for interpreting the numbers. There was no control group, so the design cannot separate the effect of the practice from regression to the mean, seasonal factors, or expectation effects. Participants were self-selected app users, who may be more motivated than the general population. Perceived stress was self-reported. And the study was conducted by Prana Labs researchers on the company's own product (see the disclosure below). These are normal constraints for a pilot; a pilot's job is to justify more rigorous follow-up work, not to settle the question.
Who conducted and funded it?
The study was run by Prana Labs in an ongoing research partnership with the Metacreation Labat Simon Fraser University, with funding support from Canada's NRC IRAP program. The manuscript (Adhia et al.) is in revision at Human Factors (Sage); this page will link the published article when it is available.
Where these numbers appear
The figures on this page are the source for the statistics quoted on our clinical evidence, science, and HRV breathingpages. Methodology notes for the app's own metrics live in docs.
