Counterintuitive HRV measurements

Quantified Self QS Open Forum
1

Hello all, this is my first post on this forum.

I am trying to learn how to manage my stress using the free Azumio StressChecker app on my android phone. It’s pretty neat. You put your finger over the phone lens for 2 mins, and the app measures your pulse by noting slight changes in the colour of your skin. It seems pretty reliable. I used the Heart Rate app (also from Azumio) and compared its readings to that of a chest band, and they are always within 1 beat/min of each other.

That said, I am very puzzled by the Stress Meter readings. As I understand it, it measures Heart Rate Variability (HRV), which has been thoroughly studied scientifically and correlates well with your level of relaxation. However, the readings I get from the app are very counter-intuitive… it often says I am stressed when I feel relaxed and vice versa. Not sure how to interpret this. Is the device’s reading wrong, or is it me that has a poor sense of when I am stressed vs relaxed?

For example, yesterday, it said my stress level maxed out (97%) after a 30mins spinning session. Yet, from my point of view, that’s probably when I felt most relaxed during my day. Granted, when I waited for 30 mins, it went down somewhat, but was still high (73%). After 90 mins, it came back down to 40%, but that’s still higher than what I measured the rest of the day. I also measured my stress level before and after a 60 mins nap, and it registered about the same (29% versus 23%) even though I felt MUCH, MUCH more relaxed after the nap. I also notice that whenever I take a measurement before and after a 2 mins deep breathing “relaxation” session, it’s almost always higher AFTER the deep breathing (that said, I always have trouble with meditation and it does feel like it doesn’t actually relaxes me).

Conversely, the moment when I registered the lowest reading (1%!) was after I spent 30 mins “fighting” with Microsoft Excel, at which point I felt somewhat frustrated and stressed.

Do others have similar experience with those HRV devices? How should I interpret these readings? As a flaw in the app, or as meaning that I have trouble assessing my own level of stress?

Thx.

Alain

2

i suggest you take a look at some of the material on HRV at SweetBeatLife. I track my HRV and galvanic skin response (gsr). While HRV is good for measuring stress (defined as physical stress from over exertion or exhaustion), GSR is a better reflection of stress in the sense of “stress caused by psychological fear or worry”. Hence the high readings after a workout.

I can get in arguments with my wife and HRV looks very good – calm, if you will. But GSR spikes like crazy during this time. I also track HRV and GSR during sleep. Comparing the EEG, GSR clearly spikes when I am having a troubled dream. hRV seems only to correlate with overtraining in a general sense. Bad dreams don’t reflect in the running HRV.

If you want to truly explore this, wear a chest strap and run the SweetBeat app. gSR is most accurately measured with the Neuma watch. But the basis peak also measures it. The neuma combines GsR with Google,PA’s so you can see where and when you had high readings and this is immensely useful in true-ING up the results

3

Don’t know how Azumio calculates “stress”, but I’ve noticed that simple measures like the HRV SDNN reflect my heart rate and breathing more than anything else…

4

Thx. I’ll look into GSR devices.

What you wrote is consistent with my own experience, but it surprises me somewhat, given that a lot of relaxation/biofeedback devices use HRV (most notably, the emWave, which lots of people are raving about).

Alain

5

I think there is still debate on the issue. I know the founder of Neumitra. He is a research neuroscientist from MIT and has a good grasp of these metrics. One thing I would add is that HRV is more than one metric. If you use Kubios, you can get the LF, HF rMSSD, and a bunch of other analytics. Many use the LF/HF ratio, but the literature on this is very mixed. I tend to like the rMSSD measure, but use HRV for overtraining monitoring.

I had a challenging year end review on Friday and could see my GSR go from low to high as we discussed money. And I did not perceive myself as stressed. I really have to see more side by side data to have a stronger opinion. I would like to see a study on it, but I am highly suspicious of data collected,in the lab. White coat syndrome is such a confounding variable. The great thing about the Neumann is that it collects,research quality data unobtrusively in real life settings.

If anyone comes across any academic papers, please PM me.

6

Very interesting topic. I can see that some hrv apps market their product as a means able to tell when you are ready to work out again, so that you do not overtrain. Hi stress levels on the azumio app after the work out, as you experienced, seems to be consistent with this, i.e. it’s nothing wrong with the measurements.

When you are fighting with your wife you are actually taking action to reduce stress - it’s a coping mechanism to get to a solution, and i guess you are the dominant party and usually win these arguments which give you some degree of satisfaction? So maybe the advantageous hrv measurements aren’t that suprising.

What I get from this is that what one feel isn’t necessarily a good indicator of how stressed one’s body is.

Scientific papers on the topic would be highly appreciated.