In the past year, a few people have brought up Dr. Levine’s interview, in which HRV is dismissed as an unreliable metric.
I will borrow from a (similar) question on Reddit, to highlight the main points brought up in the podcast. The user spumoni620 writes:
Essentially, as I understand it from his discussion, breathing and baseline sympathetic neural tone operate at two different frequencies and are two variables that influence the HRV. By altering breathing, one can artificially inflate or deflate the HRV. I know that point has been discussed already and there are apps that allow you to measure with a set breathing pattern to measure deviations from a set baseline.
But the point that surprised me is that he mentioned after controlling for all the potentially modifiable factors in a lab setting among athletes, despite all the controls, there was still a random 25% variation in HRV (as I understand what he described).
What is going on here? Is it possible that the 25% variation that Dr Levine talked about was actually the variable of interest (ie variability of heart rate reflecting actual recovery status rather than random error)?
The question continues with more considerations, you can find it here.
So, what do we make of this?
Personally, I think the user has captured it perfectly with this "What is going on here? Is it possible that the 25% variation that Dr Levine talked about was actually the variable of interest (ie variability of heart rate reflecting actual recovery status rather than random error)?"
In my opinion, the view discussed in the podcast is quite far from modern use, especially in terms of repeatability, and how the data changes on e.g. consecutive days. This comes from ‘thinking like a doctor’, i.e. you do a ‘test/check’ every few months or years, and learn something about your body. You want the parameter you measure to be stable at least across days, as otherwise what you measure is not really indicative of your chronic state. This is also how HRV was used in research many years ago, e.g. to see if you could distinguish people with a certain chronic health condition, with respect to healthy controls.
However, with HRV, we have learned that’s a meaningless approach, which is why I would never tell you to measure your HRV every 6 months or to do anything with the absolute value (see: Absolute values of Heart Rate Variability (HRV)).
The data must be different on a daily basis as your stress response is different: the whole point of measuring it is to look at relative changes over time - on a daily basis - with respect to your normal range. I think this aspect of the application, more recently developed, is missed in the interview.
In the podcast, HRV is thought of as an absolute number with a meaning, which is quite the opposite of what I’d consider an effective and meaningful use of this marker (see: Getting Started with Heart Rate Variability (HRV)).
The breathing problem seems linked to how HRV was analyzed in the old days, in the ‘frequency domain’, and I also discuss in my blog why that way of looking at the data is indeed flawed and too tightly coupled to breathing (see: Heart rate variability (HRV) numbers: what do they mean?). Here you can also learn why certain methods are not currently used by us or others (e.g. looking at low frequencies as a marker of sympathetic activity, something that was done in the past).
This being said, I do agree with many points that Dr Levine brings up, most importantly that HRV is not a reliable marker of stress in the way it is used today, for many of the issues he mentions, and many more, which I discuss here in the context of continuous HRV measurement: Issues with continuous heart rate variability (HRV) measurements.
I also agree that to use HRV effectively, we need good protocols, and measuring at the right time, so that we can use HRV as a marker of our stress response, which is why I highly recommend a morning protocol and measuring intentionally (as opposed to wearables).
I hope that provides some useful context around HRV and how to use it.
Thank you for reading!
Marco holds a PhD cum laude in applied machine learning, a M.Sc. cum laude in computer science engineering, and a M.Sc. cum laude in human movement sciences and high-performance coaching.
He has published more than 50 papers and patents at the intersection between physiology, health, technology, and human performance.
He is co-founder of HRV4Training, advisor at Oura, guest lecturer at VU Amsterdam, and editor for IEEE Pervasive Computing Magazine. He loves running.
Social:
Twitter: @altini_marco (currently inactive)
Personal Substack