Polar has an ANS Recharge metric that is akin to the rHR/HRV spreadsheet that @MedTechCD and @Gerald have created.
@David, is it possible to pull that metric as well for Wellness data?
Note, it’s not the same as their Readiness or Sleep Score metrics.
I started reading that but stopped after less then a third of the page…
HRV from wrist measured values, only taking the first 4 hours of the night into account.
Already 2 major things that have been proved entirely wrong by well-known researchers. I wouldn’t trust it for a cent.
HRV should be measured in the morning or for the whole night and with a quality device able to detect correct RR values.
Measuring only the beginning of the night for example will give you very different values if you regularly work out late in the evening.
There was an interesting paper comparing the various timing and methods of taking HRV. (Unfortunately, I can’t find the reference.) One of the findings was something that I noticed for myself as well; the nocturnal (even partial nocturnal) differed to a degree from morning HRV but the overall trends were consistently aligned, except in one case—sickness. In that case, the nocturnal HRV would be elevated and could give a false sense of recovery for some time before it too returned to baseline. While partial nocturnal may not be ‘optimal’ it does provide information, especially in the case of lacking any other information. I personally stopped taking morning HRV because I found that I had learned how to manipulate the reading. Each morning became a game to try to/not to try to affect the reading. In regards to working out late in the evening, all metrics need to be (re-)evaluated in such cases. IMHO, if the metric is available to be harvested from Polar automatically then make the option available.
I’m not against making the metric available, but there is early research enough that questions these procedures. I have something similar on my Coros watch and Garmin also is doing something in that area. The first problem is that they all do it differently, and they all have different outcomes. It is something that get’s more and more intrest by users and the big players all want a part of the market share. None of those measuring procedures have any scientific evidence.
Here’s a good read:
Yes, comparing between the various players is rather pointless. HRV is a curious metric because it’s both trends over time and acute indicators that are of interest.
@david, FTM, Polar has an avgRR for the duration that it uses with HRV and RR. I’d like to capture that data automatically as well, if possible.
My thought is to come up with a graph that tracks all three (RR, HRV, HR) and does essentially the same thing as the HRV/HR scatterplot-; to provide information for detection of ANS issues and readiness. There has to be a way to create and use ‘z-scores’ for this purpose, tho it’ll take someone with far more maths & stats understanding to find the solution.
The data could be captured however and from whatever source one wishes. (I just happen to use Polar for now and would like to get that data.)
Cheers
Do you really mean avgRR or do you mean the beat-to-beat RR values?
If it is avg RR, that is simply the inverse of avg HR. A RHR of 60bpm translates to an avgRR of 1000msec.
If the Polar values for RHR and HRV (rMSSD) find their way to the wellness metrics, you can use the ithlete chart.
Polar counts avgRR via beat-to-beat interval data.
Yes, off course. That’s what every Heart Rate monitor does. Count the number of R-peaks (the highest peak in the QRS complex of the ECG signal) in a rolling time window and calculate how many that would be in a minute (60 divided by the time window length multiplied by the number of peaks). Or take the average of the time between peaks for a number of peaks (60 divided by the avg interval time). Chest straps, optical sensors and ECG devices all do that same thing to give you instantaneous HR or avgRR which is 60 divided by the HR in msec.
“avgRR” makes me assume that you get one number for a period of time just like bpm is a number of beats per minute. And those are just the inverse of one another.
If you get an avgRR number that is not equal to 60 divided by the RHR, that can only mean one thing:
they are not calculated over the same time window. Some wearables do that, they calculate avgRR for the whole night but isolate a part of the night where your HR is lowest for a short period and consider that as your RHR. But then, you don’t really know “what is what”, unless they specify what and how they measure.
Isolating a part of the night is risky because that part could be plenty of artefacts due to movement or blood flow constriction caused by your sleeping position.
HRV on the other hand, measures every single RR interval and evaluates the magnitude of difference between all those intervals. You need a record of every single RR interval to do that and a proper filtering technique to ignore artefacts. Then you calculate rMSSD or SDNN which gives you a “HRV” number.
You will probably be interested in this. Polar is not in the test but it can give you some ideas on where wearables can fail to provide consistent data.