RMSSD, lnRMSSD and measurement windows
A protocol-level guide for research and product teams choosing how to record and report heart-rate variability (HRV). Sources checked 5 November 2026.
What RMSSD captures
RMSSD is the root mean square of successive differences between normal heartbeats. It reflects short-term, beat-to-beat variation, which is why it is widely used when recordings are short or taken during sleep. It describes variation; on its own it does not explain why that variation changed.
Why log-transform it
RMSSD values are usually skewed: most readings sit in a narrow range, with a long tail of high values. Taking the natural logarithm (lnRMSSD) makes the distribution more symmetric, so a change of the same size means roughly the same thing at different levels. That matters for within-person analysis, which is how ATS-1 uses HRV.
Window length
The 1996 Task Force standards on HRV measurement treat five-minute recordings as the standard short-term window (Task Force of the ESC and NASPE, 1996). Many devices now report ultra-short readings of under five minutes; a published review examined whether ultra-short features can stand in for short-term ones and found that the answer depends on the feature and the recording conditions (State-of-the-art review of ultra-short HRV features (PMC5998753)). Report the window you used, every time.
Context to record
| Context | Why it matters | ATS-1 rule |
|---|---|---|
| Sleep period or spot reading | Posture, movement and breathing change daytime readings | Nightly lnRMSSD only; daytime spot HRV never enters indices |
| Device model, firmware and algorithm version | Devices compute HRV differently | Required on every record |
| Timezone | Night boundaries depend on local time | Required on every record |
| Valid-night count | Short histories make baselines unstable | At least 14 valid days in a 28-day baseline |
What to report
- The measure (RMSSD or lnRMSSD) and its units
- The window and when it was taken
- The device, firmware and algorithm version
- How many valid nights the baseline used
- Comparisons only against the person's own baseline, never against a population threshold. See the State Index.
What this guide does not claim
No HRV value on this site is presented as normal or abnormal. ATS-1 is for research and general wellness only. It does not diagnose, treat, prevent or predict any disease, including any mental health condition. For questions about your own health, speak to a healthcare professional.
Questions
Is lnRMSSD better than RMSSD?
They carry the same information. lnRMSSD is easier to compare across days and people because the log transform reduces skew.
Can a one-minute reading replace a five-minute one?
Sometimes, depending on the measure and conditions, according to published reviews. ATS-1 avoids the question for its indices by using nightly readings.
Sources
Keep reading
Take the next step.
RMSSD and lnRMSSD for HRV research: why the log transform, how recording windows differ, what context to record, and what to report.