A Novel Way to Establish Your Optimal Breathing Rate

Introducing Journey Mode to the Yudemon HRV Biofeedback App

Max Frenzel, PhD
Yudemon
Published in
5 min readFeb 16, 2023

--

Photo by Maxime Horlaville on Unsplash

Heart rate variability (HRV) biofeedback training — also known as resonance breathing — uses breathing as a powerful method for training your autonomic nervous system, boosting your stress-resilience, adaptability, and overall mental and physical performance.

The optimal breathing rate that triggers the biggest training effect, the so called resonance frequency, is highly individual and depends on your unique physiology. While the average falls at around 6 breaths per minute, it can be as low as 5 or as high as 7 breaths per minute for different people. If you are curious about the physiological mechanisms underlying this, I recommend checking out this article:

While practicing breathing at any of these rates will still be better than not doing any breath practice at all, if you want to get serious about HRV biofeedback training and achieve optimal results, the first course of action is usually to establish your own resonance frequency.

The Limitations of Current Methods

To do this, essentially every single HRV biofeedback tool available so far offers the same protocol that roughly looks like this: Over a single short session (often 6 minutes) you breathe at several different rates, transitioning from 7.5 breaths per minute to 5 breaths per minute in steps of 0.5. Then the rate with the highest HRV is identified as your resonance frequency. And that’s it.

When I first started my own HRV biofeedback practice, that is the method I used, and which told me that my resonance frequency was 5.5 breaths per minute.

However, I was pretty dissatisfied with this. It felt like it lacked accuracy and personalization. Not to mention that a single faulty measurement or some noisy data — a common issue with heart rate monitors — could give you a completely wrong value.

So I decided to dive in deeper and collect much more extensive and fine grained data to narrow down my resonance frequency. I shared my experience and results around this in the following article:

Rather than using one minute of data at different breathing rates separated by large steps of 0.5 breaths per minute, I collected a total of 8 hours of data, 30 minutes per breathing rate with a much finer step size of 0.1 breaths/min. The key result that came out of this self-study was this graph:

HRV (RMSSD in ms) across breathing rates for 8 hours of total data, 30 minutes per rate.

It allowed me to refine the estimate of my resonance frequency from 5.5 breaths per minute to somewhere between 5.3 and 5.4 breaths/min (see the article for more details).

Maybe even more interestingly, it also showed just how quickly my HRV decreases when not breathing as this optimal rate. Even a small change from 5.4 to 5.6 breaths/min led to an HRV decrease of over 10% on average, indicating just how much potential might be missed if the resonance frequency is determined with a resolution of only 0.5 breaths/min.

If your resonance frequency falls right between the coarse intervals other tools use, say at 5.25 or 5.75 breaths/min, you might loose a lot of training potential as they lump you in at 5, 5.5, or 6 breaths/min.

Journey Mode

While this study helped me get a much deeper and more statistically relevant insight into my own resonance frequency, it had a few issues.

For one, it requires a lot of data, and a commitment that most people are probably not willing to make. Second, it still has limited accuracy with its fixed bins separated by 0.1 breaths/min.

So I set out to find a solution to both.

The result is Journey Mode, which was just added to the Yudemon HRV app for iOS.

It’s an entirely new method for establishing your resonance frequency with arbitrary precision. Starting from an estimate of 6 breaths/min, every additional session your practice in journey mode allows Yudemon to update and refine this estimate until it eventually converges to your true resonance frequency.

Every session in journey mode lasts ten minutes and covers only a small range of breathing rates. For example, in your first session you will be breathing at rates between 5.8 and 6.2 breaths per minute. The data from this is then used to update Yudemon’s estimate of your resonance frequency and to determine the rates for the next session.

The graph below shows my own progression across several experiments as I was developing and testing different algorithms for updating the resonance frequency and the breathing rates across journey sessions.

Estimate of my resonance frequency (breaths/min) over the course of 15 journey sessions (comparing different algorithms for the update procedure).

Consistently, after around six sessions into my journey, I started to converge onto a rate that I knew to be correct from my previous data, somewhere between 5.3 and 5.4 breaths/min.

My progress with the final, further refined algorithm that made it into the app can be seen in the screenshot below.

Screenshot of the Yudemon HRV iOS app showing my own journey progress after ten sessions.

Again, after around five or six journey sessions, the estimate started to get close to the range I had known to be correct from my previous experiments. After fifteen sessions, Yudemon’s estimate of my resonance frequency was at 5.36 breaths/min. And with every further session I commit to the journey, this will be further strengthened and refined.

No single session is perfect, and data outliers or genuine differences in practice naturally mean that the result of each session fluctuates a little bit. However, combing data across many sessions it’s possible to get an accurate estimate with arbitrary precision and increasing confidence.

And I’m really excited that this is now available to every user of the Yudemon HRV iOS app (assuming you have a heart rate monitor available — I strongly recommend the Polar H10, which was used in all my experiments).

I do have to note that this algorithm was developed with myself as a test subject, and has further been confirmed on my girlfriend (who has a very different resonance frequency from me). In both cases it worked extremely well. However, while I do strongly believe in the robustness and universality of this new algorithm, I have not (yet) tested this on a large population.

I would love to hear your feedback, especially if you are already an experienced HRV biofeedback practitioner and have previously established your resonance frequency with one of the existing tools, or if you are a researcher and would like to rigorously test this method with a larger cohort.

With Journey Mode now being available, I am fully convinced that Yudemon is one of the best HRV biofeedback tools out there — and certainly the most unique — and I’m excited to be sharing this with all of you.

I wish you a pleasant journey!

--

--

Max Frenzel, PhD
Yudemon

AI Researcher, Writer, Digital Creative. Passionate about helping you build your rest ethic. Author of the international bestseller Time Off. www.maxfrenzel.com