Is SDNN of any use? (HRV readiness readings)

Hi all,

I’m reading HRV from two sources every morning: from Elite HRV and from Kubios. Both are 2:30 min readings in the morning, and I get both rMSSD and SDNN (or can calculate them).

Is the SDNN parameter of any use and worth following, considering that it’s read from just a 150 second sample?

Great question! It’s fantastic that you’re tracking HRV from multiple sources to gather insights.

Regarding SDNN:

  • While SDNN is a valuable measure of overall HRV (it reflects both short-term and long-term variability), it’s typically more reliable when measured over longer time periods (e.g., 5 minutes or 24 hours). This is because SDNN captures both high-frequency (short-term) and low-frequency (long-term) HRV components, and a shorter 150-second sample may not provide enough data for an accurate assessment of these longer-term trends.
  • For shorter recordings like 2:30 minutes, rMSSD is generally preferred because it focuses specifically on short-term parasympathetic activity (vagal tone) and is less sensitive to duration. rMSSD is also less influenced by factors like breathing patterns, making it more consistent for morning readings.

Yeah, I know. Still, Apple uses it. I have also read about some quite harsh critisicm about this. Are any one of you aware of any occasion where apple has defended this usage and countered this criticism.

I mean, technically it’s not hard to switch from one calculation to another. SD is just the standard deviation, whereas rMSSD is the sum of the differences, and then you square them and take the root, because it doesn’t matter if the differences are positive or negative. Both are measures of the variance of the samples.

The main criticism is that everyone uses RMSSD as the premier HRV parameter except Apple…
Lot’s of other apps, studies use both RMSSD & SDNN because SDNN can give a slightly different view on HRV and help in diagnosing parasympathetic saturation amongst other.
But the widespread use of RMSSD is something Apple keeps ignoring. If they don’t want to switch to RMSSD, they could at least record it as a second parameter.