I’ve been using both Elite HRV and KubiosHRV. The data and recs are often at odds. After a moderately challenging ride (TR Hunter) yesterday, Elite gave me a 6 (out of 10) and suggested recovery. RMSSD and SDNN were 28 and 40, respectively. KubiosHRV rated me at 94% (an all time high) with the same two readings 91 and 51. Readings were taken back-to-back.
2 Likes
Not surprised. I see this between EliteHRV and HRV4Training at my wake-up HRV measurement. I see them align more often than not. At this point I tend to align with HRV4Training but will admit it only influences my training decisions with the help of how I feel, how I think I slept and how my Garmin tells me I rested.
1 Like
Runalyze.com can do the same LT1 estimation if your watches, head unit can record RR intervals (typically all Fenix series, FR9x5, etc.)
Cold outside + Amstel Gold → spent 9€ on HRV logger to play around
2min averaging steps + Polar H10 via Bluetooth
did more testing later in the ride, quite a scatter of DFA alpha1 around my LT1 HR of 136bpm.
not really convinced yet.
1 Like
If I am reading this correctly you had a DFAa1 reading of .29?
Seems like you’ve got something weird going on.
So more on this, first of all the entire ride:
Distinct cardiac drift. I measured La along, the final 16min at 253W are interesting. Lactate corresponded to the heart rate (e.g. 141bpm, ~0.4 mmol above baseline), whereas the power is clearly below LT1 (when fresh). I’ve highlighted a 15min interval after an hour, baseline Lactate: 270W.
Take home message: heart rate seems to correlate well with La, power changes throughout a ride. However, I’ve eaten a few bananas during the ride. Impact?
Now onto DFA a1, plot HR vs a1 for the enire ride (well, not the entire ride as I only recorded the first 1:20h and the highlighed 30min and 16min intervals.
Entire ride:
LT1 ~ 138bpm
Cherry picking the cluster > 125bpm
1 Like
What HR threshold estimate does the Python script spits out for you?
Never got to this point when playing with the notebook, data was always too messy.
I would just use historical rides and trust the 5% threshold for artifact removal and see where it lands for you. One thing I did notice is that it’s that you get a more accurate reading if you go above in terms of intensity…As recommended by Bruce Rogers.
1 Like
Have you got a record of how many artefacts you got for the different points because that apparently can cause errors in the DFA value.
I had a large number of artifacts with my tickr, which made the numbers useless.
I moved to a Polar H10, and the artifacts went away and only happen occasionally. The numbers are in a more reasonable range now, but they are so variable that I find it hard to do anything with them. Here is from a session at 160 watts, in ERG mode. The spike early on I stepped of the bike, the rest of the session was steady at 160 watts. (4 artifacts total in the whole workout.)
This type of variability is typical for me. If I were to do a step test, I’d need steps of 30-60 minutes I think. Alpha 1 decreases with intensity for me, but with this much noise/variability in the data it doesn’t really help me find my aerobic threshold with much precision.
2 Likes
A couple of points.
-
It decreases with intensity for everybody, meaning the series go form uncorrelated to anti correlated.
-
It seems that you were going too easy in this effort.
-
For a protocol, forget about watts, try to get your HR between 80% to 90% of LTHR in steady steps…and for the final step, go beyond, like 92%-95%.
-
I don’t use HRV logger up, only the Python script on historical data. I get very consistent estimation of my AeT HR. For me, nothing revelatory, but a nice confirmation of RPE.
1 Like
I have done harder efforts, but I still see the highly variable data, where I’ll get data points varying between (for example) 0.80 and 1.00 or so - a very wide range when I’m doing a consistent effort for 30-60 minutes. I’ll try to give the python script a go, as ‘eyeballing’ the data doesn’t give me confidence in it producing a consistent/accurate number.
Wow, I feel like I just took a wrong turn and entered a lab 
Science sure has progressed a lot since I last used an HRV measuring. Or maybe we make science of everything now?
To get back to the original question, I have been using Elite HRV myself 4 years back, but quit after I no longer could trust the data.
My two cents on the methodology:
- Too new science if you ask me and requires more refined hardware and software, as well as proper research, though I may be out of date. 4 years ago, I had little use of my measurements each morning over many months and quit because it did not add anything to my training schedule but unnecessary routine and doubts. At least, not more than how I felt in my body.
- What it clearly indicated was illness, days ahead. Poor HRV showed a flu maybe two days ahead, for that it is perfect, but usually you feel illness immediately at training start.
- There is a simple method to TEST the consistency of the readings (for EliteHRV at least): repeat the test 3 times in a row right after each other. It takes a few minutes only, it should not make much of a difference. My findings: shocking differences while still lying still, fluctuations without pattern. This made me decide in the end that it did not work properly. You cannot have a value of 8 and 4 in only minutes, clearly wrong data.
- Avoid falling into the trap of wishful thinking, by reversed logic: we start explaining figures when we see them, even if we do not agree with them. From then on the data does not help us. Better to estimate it yourself and write it down first each day, then verify what the measurement gives. That is a much better learning.
My advice overall: listen to your body. I am a tech geek myself since childhood, but HRV gadgets are just not proven science for me, yet. I did not have much gain from what the app told me and I never adjusted training, only when I fell ill. Usually a fit person can already see during the first 10 minutes into a training how it will go and adjust accordingly. Definitely on a trainer.
2 Likes
Interesting read of your experience. I have been using it a year. It hasnt told me anything I didnt already realize.
On the subject generally, assuming you’ve had a really bad reading (not the daily readiness nonsense, but an unusually low HRV figure compared to normal), and you’re feeling it too… how exactly would you tweak your workout?
I have a particularly tough session planned in, and I know I should change it, but I’m not sure what’s best. More rest intervals, lower intensity, lower TSS?
Delay it a day and do a recovery or endurance ride instead.
2 Likes
I might give this a go. The only time I’ve done something similar in the morning is before, say a long group ride that day that I’m more or less committed to. If my first reading happens to give a low reading which just puts it into the “take it easy” range I’ll take it a again a few times until it doesn’t!! Then I’m good to go
1 Like
Yeah with Whoop if I change the sleep I get a different score.
One day I altered it and went from 40% to 1% !!!
I tend to ignore the score, the HRV number on its own shouldn’t really change much at all (assuming you’ve done a long enough test / have a decent sensor).