Heart rate variability (HRV) needs to be calculated across an extended time window. It is not a suitable measure for event-related tasks, such as valsalva, hand grip and so on. ONLY EXQUISITE-QUALITY DATA CAN BE USED FOR HRV ANALYSIS!

It is computed from the time between each set of R peaks. Below an experiment from our data.

Open

The first 7 rows are from one participant and the second set of 7 rows from a different participant.

The column “None missing” is computed from a 5-minute good quality EKG dataset. All markers were checked for correctness and were accurately placed for this analysis.

We are looking at the sympathetic-to-vagal balance on the bottom in red. Participant one had a balance of 0.93. To understand how fragile HRV measures are, I removed a random R peak marker within the 5-minute period and the balance value changed to 0.78. I then removed a second and even third marker and you can see how the values changed quite drastically.

Open

In summary, removing a single marker for subject 1 changed the sympathetic/vagal balance by 15% and for subject 2 by 27%!