If you build a breathwork app, you should be able to say which literature it stands on. This page is exactly that.
We keep updating it. If you miss an important source or spot a mistake, write to us. We are grateful for the pointers.
You find the five anchors themselves in our statement of position. Here we show which studies hold each anchor in place.
Anchor 1Low-stimulus
We design against the outward pull on attention. Instead of pulsing animations, a quiet centre. The effect of low-stimulus design on stress and attentional regulation sits inside an active research area in digital health and human–computer interaction.
Anchor 2Anatomically grounded
We check every breathing technique against the anatomy and physiology we know from clinical practice. A technique that helps a healthy twenty-eight-year-old may harm a sixty-two-year-old with emphysema. So we build contraindications directly into the exercise selection.
Anchor 3Evidence-based
Three fields hold our methodology in place: resonance breathing and cardiac coherence, acoustic breath-phase detection via smartphone microphone, and clinical evidence for HRV biofeedback.
Resonance and coherence
We use resonance breathing at roughly 5 breaths per minute, because at that frequency HRV and respiratory sinus arrhythmia come into phase in most people. The exact frequency varies individually with body size, age, and lung capacity.
A Practical Guide to Resonance Frequency Assessment for Heart Rate Variability Biofeedback. Frontiers in Neuroscience. 2020;14:570400.
Methodological guide for determining an individual resonance frequency, including test-retest reliability and minimum epoch length for HRV indices. Open access, co-authored by one of the authors of the AAPB HRV standards.
- Steffen P, Austin T, DeBarros A, Brown T. The Impact of Resonance Frequency Breathing on Measures of Heart Rate Variability, Blood Pressure, and Mood. Frontiers in Public Health. 2017. DOI
- García C et al. Methods for Heart Rate Variability Biofeedback (HRVB): A Systematic Review and Guidelines. Applied Psychophysiology and Biofeedback. 2023. DOI
- Parrado E, Lalanza JF, Ramos-Castro J, Capdevila L. Resonance frequency is not always stable over time and could be related to the inter-beat interval. Scientific Reports. 2021. DOI
Acoustic breath-phase detection
Our microphone algorithm identifies inspiration and expiration from the smartphone microphone's audio signal. That is not trivial. Ambient noise, microphone quality, and distance to the mouth all change the signal. The study we draw on most validates this exact approach against respiratory inductance plethysmography.
Estimation of respiratory rate and exhale duration using audio signals recorded by smartphone microphones. Biomedical Signal Processing and Control. 2023;81:104318.
Adaptive thresholding method for breath detection from smartphone audio, validated against the RIP gold standard. MAE of 0.2 bpm in the lab, 0.79 bpm across 217 remote recordings from healthy and COVID participants. The paper also includes an audio-quality classifier (AUC 0.81), a pattern we picked up conceptually for our own noisy-environment warnings.
- Salvador-Navarro A et al. Respiratory rate estimation applying non-negative matrix partial co-factorization from breath sounds. IEEE MELECON. 2024. DOI
- Bae S et al. (Google Health). Prospective validation of smartphone-based heart rate and respiratory rate measurement algorithms. Communications Medicine. 2022. DOI
Clinical evidence for HRV biofeedback
HRV biofeedback is the most studied form of breath-based autonomic self-regulation. We lean on the meta-analysis that most follow-up work refers to.
The effect of heart rate variability biofeedback training on stress and anxiety: a meta-analysis. Psychological Medicine. 2017;47(15):2578-2586.
Meta-analysis across 24 studies and roughly 1,700 participants. Clinically meaningful effect sizes for HRV biofeedback on stress and anxiety. Stefan Hofmann (Marburg, previously Boston University) is among the most cited authors in clinical anxiety research.
- Lehrer P, Vaschillo E, Vaschillo B et al. Heart Rate Variability Biofeedback: How and Why Does It Work? Frontiers in Psychology. 2014. The mechanism paper that grounds the resonance approach physiologically.
- Vermetten E, Blase K, Lehrer P, Gevirtz R. Neurophysiological Approach by Self-Control of Your Stress-Related Autonomic Nervous System with Depression, Stress and Anxiety Patients. IJERPH. 2021. DOI
- Gitler A, Bar Yosef Y, Kotzer U, Levine AD. Harnessing non-invasive vagal neuromodulation: HRV biofeedback and SSP for cardiovascular and autonomic regulation. Medicine International. 2025. DOI
Anchor 4We measure. You decide.
We hand you values. Frequency, depth, rhythm, the I:E ratio from the microphone, HRV and RSA from Apple Watch, HealthKit, or a BLE chest strap. What you make of them is up to you. This stance follows from the clinical concept of the informed patient and from research on patient autonomy in digital health applications.
Anchor 5Attention inward
The final anchor is also the test for all the others. With every design decision we ask: does this turn attention inward or outward? Interoception, the perception of one's own body, is its own research field and the foundation for the effect of breath work.
What we deliberately do not claim
We take care to keep the methodology of the app as close to the literature as we can. Still, the picture has to be honest: the app itself is not clinically validated. The effects of resonance breathing, HRV biofeedback, and interoceptive training are established at the level of method. What our specific app does in our specific users, that is something we can only speak to after our own study.
We are sceptical of proprietary scores sold as science. If we ever introduce a number of our own, we will lay the methodology open.
If you know a study from this field we should add here, write to us. This page grows with what we learn.
Low-stimulus. Anatomically grounded. Evidence-based.