A hardcore user of the Oura ring told me that a measure of how restful one’s sleep has been is reflected in one’s HRV during sleep (or hours not awake, whichever you prefer).
They said that whatever one’s RHR or waking heart rate is - say 65 - is a baseline. And that until one’s heart rate descends and reaches this baseline one is not getting restful sleep. So the total duration of hours of sleep prior to this could be taken as a measure of how high the quality of sleep is.
Can anyone else with insight into sleep tracking confirm if this is the case?
We’ve run an experiment with PSG, Oura, Apple Watch, and Phillips Respironics devices to understand the sensitivity of all these devices. We validated that the Oura and the Apple Watch have very accurate sensors for measuring HRV compared to a clinical grade ECG (R2 is about .80 when comparing the HRV sensors of Apple Watch and Oura to clinical sensors), but that the algorithms for measuring sleep quality have some room for improvement.
When people fall asleep their HRV decreases and in deep sleep the HRV decreases as well. So a decrease in HRV during sleep is indicative of sleep quality since deep sleep is more regenerative
So generally, yes, I would say that this is the case.
I think there’s some ambiguity in the original question about resting heart rate (RHR) and heart rate variability (HRV). Just to clear this up: Is it the case that we expect both RHR and HRV to fall in deep sleep? What’s the expected relationship in general between RHR and HRV, if known?
Thank you for feedback @dgartenberg. I would like to echo @Agaricus and it’s my fault for mixing up terms. Is it accurate that pulse rate measurement should decline in deep sleep?
Ya its an interesting question. From what I’ve seen in the data, HRV is distinctly lower in N3 deep sleep, but not necessarily overall heart rate. I’ve actually seen that heart rate is higher in N3 than N2, which surprised me. I haven’t revisited this empirical finding in awhile so I’m just confirming with one of our scientists as well and I’ll get back to you.
Interesting that you found Apple Watch and Oura are both roughly comparable to clinical grade HRV sensors. Have you published the data? I’m unable to find a good description for how to get high-quality HRV data out of the Apple Watch (the built-in HealthKit-derived XML seems to only show SDDN and then only through the Breathe App).
What I was wondering was if you go to sleep with a high pulse rate and fall asleep that any sleep until you’re at your RHR isn’t high quality sleep. I imagined a graph where the pulse rate descended until at the RHR level at anything above that baseline could be discarded
My sense is that heart rate drops during deep sleep, but that it could then rise again in other phases of sleep. Is that right?
The MiTools app also registered it as bad quality sleep, so perhaps that HRV decline is already factored into the evaluation of sleep that these apps perform?
If you have an iPhone you may use Knomee to collect HRV from you Apple Watch and other related metric, and then either look at the data using Knomee analytics or send the CSV files back to yourself. Knomee gets both average and rest rates from HealthKit APIs.
you may combine heart data with sleep data from AutoSleep (for instance) and find interesting correlations …
