4 ms·

> Researchers found the method Apple uses, standard deviation of normal-to-normal, or SDNN, is the most accurate. Other brands implement a system called root me

> Researchers found the method Apple uses, standard deviation of normal-to-normal, or SDNN, is the most accurate. Other brands implement a system called root mean square standard deviation, RMSSD, which Tenan said produces a wider range of error in the measurement This is interesting for a couple of reasons. Firstly, I have one so it’s cool to know Apple designed it well in that respect, and that I shouldn’t lose confidence in the readings I get. The second reason it’s interesting is that this kind of decision, which time-domain measure of heart rate variability metric to use, is the kind of super small design decision that really makes a system special. Having really good developers and designers really shines through for moments like these.

raverbashing 2y ago
That's an interesting assessment I was curious about this and curious if the measurements made by my smart watch were "good" or "bad", and it does calculate RMSSD (allegedly) Though it seems while they have some correlation they're all over the place https://www.researchgate.net/figure/Scatterplot-for-SDNN-A-B-RMSSD-C-D-pNN50-E-F-and-HFnu-G-H-for-men-left_fig1_311620117 https://www.researchgate.net/figure/Scatterplot-for-SDNN-A-B... SDNN also might apply to a different time range than RMSSD
- drilbo 2y ago
  As pointed out by someone further down the thread, the value of HRV really only seems to appear in comparing to your own baseline.
_9omd 2y ago
Just got the latest Apple Watch, so also glad to hear they perform better on HRV. It seems they get a lot of the little details better than competitors in the wearables space. I read previously about how they likely had the most accurate sleep tracking and also had some special sauce in their GPS handling code that made it outperform Garmin.
smallerfish 2y ago
I would love to know from somebody with an Apple watch how you use HRV. I wear a Garmin, which I mostly like. The HRV measurement is predictive enough to be useful. Two occasions from the past 18 months are worth noting: 1) The measure dropped from green through yellow and into red. It was in yellow 7 days. On the first day it was in red, I was hit with covid symptoms. It stayed in red 2 1/2 weeks before starting to climb again; my symptoms had mostly cleared by the end of the first week, but I still felt sluggish for some time after. The kicker is that I hadn't looked at the HRV at all until sometime during the week of symptoms. 2) I went on a trip to a location +8 hours off my usual timezone, and jet lag symptoms were brutal. HRV dropped into yellow on day one of the trip, and stayed there during the 2 weeks of travel. A couple days after I returned it dropped into red, exactly coinciding with the onset of flu/covid symptoms. Again, it stayed in red throughout the illness, and then as I started to get fit again on the other side it ramped up through yellow and into green. So even with 20% accuracy (if that is the case), it's a useful data point! What does 90% accuracy give you?
- tpoacher 2y ago
  Just because it's the most "accurate" doesn'mean it's the most "useful", "correct", or "reliable" in this context. It just means reproducible, if I read it correctly. In the same way that if they mistakenly chose to represent HRV by the heart rate, this measure would probably be very "accurate" (but ultimately false). The issue of which HRV measure is the most useful, and what they show exactly, is still an open issue in HRV research afaiu
- micadep 2y ago
  That sounds useful ! Apple have the best sensors and algorithms, but they are terrible at presenting your data.
  - egwor 2y ago
    Is this true? Do you have a source? I thought that the pros tended to make use of polar’s technology and that their heart rate monitoring (via strap) was considered the most accurate / baseline. (My sources are from what I remember dcrainmaker saying who I consider a well researched reviewer)
SoftTalker 2y ago
Apple had goals to make their these health monitoring features in their watch medically approved, to open up additional markets. For other watch makers, it was a checkbox on a feature list.
dxbydt 2y ago
> root mean square standard deviation, RMSSD This is incorrect. RMSSD means root mean square successive differences. Interestingly, if you actually compute what you wrote ie root mean square standard deviation, you will get the same result as sdnn.

raverbashing 2y ago

That's an interesting assessment I was curious about this and curious if the measurements made by my smart watch were "good" or "bad", and it does calculate RMSSD (allegedly) Though it seems while they have some correlation they're all over the place https://www.researchgate.net/figure/Scatterplot-for-SDNN-A-B-RMSSD-C-D-pNN50-E-F-and-HFnu-G-H-for-men-left_fig1_311620117 https://www.researchgate.net/figure/Scatterplot-for-SDNN-A-B... SDNN also might apply to a different time range than RMSSD

drilbo 2y ago
As pointed out by someone further down the thread, the value of HRV really only seems to appear in comparing to your own baseline.

_9omd 2y ago

Just got the latest Apple Watch, so also glad to hear they perform better on HRV. It seems they get a lot of the little details better than competitors in the wearables space. I read previously about how they likely had the most accurate sleep tracking and also had some special sauce in their GPS handling code that made it outperform Garmin.

smallerfish 2y ago

I would love to know from somebody with an Apple watch how you use HRV. I wear a Garmin, which I mostly like. The HRV measurement is predictive enough to be useful. Two occasions from the past 18 months are worth noting: 1) The measure dropped from green through yellow and into red. It was in yellow 7 days. On the first day it was in red, I was hit with covid symptoms. It stayed in red 2 1/2 weeks before starting to climb again; my symptoms had mostly cleared by the end of the first week, but I still felt sluggish for some time after. The kicker is that I hadn't looked at the HRV at all until sometime during the week of symptoms. 2) I went on a trip to a location +8 hours off my usual timezone, and jet lag symptoms were brutal. HRV dropped into yellow on day one of the trip, and stayed there during the 2 weeks of travel. A couple days after I returned it dropped into red, exactly coinciding with the onset of flu/covid symptoms. Again, it stayed in red throughout the illness, and then as I started to get fit again on the other side it ramped up through yellow and into green. So even with 20% accuracy (if that is the case), it's a useful data point! What does 90% accuracy give you?

tpoacher 2y ago
Just because it's the most "accurate" doesn'mean it's the most "useful", "correct", or "reliable" in this context. It just means reproducible, if I read it correctly. In the same way that if they mistakenly chose to represent HRV by the heart rate, this measure would probably be very "accurate" (but ultimately false). The issue of which HRV measure is the most useful, and what they show exactly, is still an open issue in HRV research afaiu
micadep 2y ago
That sounds useful ! Apple have the best sensors and algorithms, but they are terrible at presenting your data.
- egwor 2y ago
  Is this true? Do you have a source? I thought that the pros tended to make use of polar’s technology and that their heart rate monitoring (via strap) was considered the most accurate / baseline. (My sources are from what I remember dcrainmaker saying who I consider a well researched reviewer)

SoftTalker 2y ago

Apple had goals to make their these health monitoring features in their watch medically approved, to open up additional markets. For other watch makers, it was a checkbox on a feature list.

dxbydt 2y ago

> root mean square standard deviation, RMSSD This is incorrect. RMSSD means root mean square successive differences. Interestingly, if you actually compute what you wrote ie root mean square standard deviation, you will get the same result as sdnn.