Heart-Rate Variability Explained
What heart-rate variability (HRV) and RMSSD measure, how wearables estimate them, why no single number is ideal, and what the research does and does not show.
Content last updated . This page has not been medically reviewed.
Units and conditions
The same person gets different values from a 24-hour recording, a 5-minute recording and a shorter "ultra-short" one, and published norms for these are not interchangeable. [1] Recording length, age and sex all change baseline values. [1] Breathing also matters: in a repeat-test study, paced breathing improved reliability. [2]Normal variability
There is no single healthy HRV.- A review of 44 studies with 21,438 healthy adults found large differences between individuals, and noted that methodological discrepancies underlay disparate values across studies. [5]
- In 39 healthy adults tested on two consecutive days under the same conditions, the second reading of a short recording could be 1.9 times or 0.5 times the first for SDNN, purely from random variation. [2]
- The authors concluded that short-term HRV measures are subject to large day-to-day random variation. [2]
So a single low reading is weak information. We did not find a study showing that following a wearable's nightly HRV trend improves health outcomes (insufficient evidence).
What research shows about health
A meta-analysis of 8 studies and 21,988 people without known heart disease found that low HRV (SDNN) was associated with a 32% to 45% higher risk of a first cardiovascular event. [6] Label: mixed to promising as a population association. This does not show that raising HRV lowers risk, and it used clinical-style recordings, not consumer scores.For breathing-based "HRV biofeedback" practices, see box breathing, resonance breathing and HRV biofeedback. Evidence that training your HRV improves health is covered there, not here.
Appropriate uses, and what HRV cannot do
- Reasonable: noting your own pattern over weeks, in the same position and time of day, as one input beside sleep, mood and training.
- Not reasonable: chasing a "good" number, comparing with other people, or treating a drop as a diagnosis. HRV cannot tell you the cause of a change, and cannot diagnose a heart, nerve or mental health condition.
- Accuracy is not benefit. A watch can estimate HRV well and still not show that tracking it improves health.
When professional assessment may be appropriate
If you have symptoms such as fainting, chest pain, a racing or irregular heartbeat, or new breathlessness, an HRV number is not the guide; contact a clinician. The NHS says palpitations are usually not serious but to get help if they keep coming back or come with other symptoms. [7] In the United States, call 911 for an emergency.Related: resting heart rate, reference ranges versus "optimal" ranges, consumer data versus clinical measurements.
Frequently asked questions
What is a good HRV number?
There is no single good number. Values depend on recording length, age and sex, and norms from different recording lengths are not interchangeable. [1] Healthy adults differ widely. [5]
What does RMSSD mean?
It is the root-mean-square of successive differences between heartbeat intervals, a score of beat-to-beat change. [2]
Can a smartwatch measure HRV accurately?
At rest, studies often find a close match with ECG, but agreement falls with exercise and has not been shown for sleep or free-living use in the meta-analysis we read. [3][4]
Does low HRV mean I am unhealthy?
Not on its own. In groups of people without heart disease, low HRV was linked to more first cardiovascular events, but that is a population association. [6]
More details: evidence, limits and cautions
The details behind this page: cautions, limits of the evidence and where sources disagree. Open any note to read it.
How to read the evidence labels
Labels used on this page: established, mixed, promising but limited, insufficient evidence.
What is measured and what is calculated
- Directly measured: the timing of each heartbeat. The time between consecutive beats is called an interbeat interval. [1]
- Calculated: HRV scores made from those intervals. Time-domain measures describe the amount of variability over a monitoring period that may range from about 2 minutes to 24 hours. Frequency-domain measures split the signal into bands, and non-linear measures describe unpredictability. [1]
- Two common scores: SDNN is the standard deviation of the intervals, and RMSSD is the root-mean-square of successive differences between intervals. [2] Both are given in milliseconds (ms).
- Proprietary scores: many apps combine HRV with other data into a "readiness" or "recovery" score. Those scores have no universal clinical meaning.
Chest strap or watch: pulse rate variability
An ECG or chest strap times beats from the heart's electrical signal. An optical wrist or ring sensor times pulses in the blood flow, so its output is more precisely called pulse rate variability (PRV). See optical sensing and ECG.A systematic review of 18 studies of wearables found the match with ECG-derived HRV ranged from very good to excellent at rest, but fell as exercise level rose. [3] A 2026 meta-analysis of PRV against ECG HRV in healthy people pooled an absolute standardized error of 0.188 for RMSSD, based on only 10 studies, and warned not to generalize to sleep, exercise, stress or free-living settings. [4] Because those pooled results cannot be applied to sleep or free-living use, they do not prove accuracy for a particular watch or ring's overnight number.
References
- Shaffer and Ginsberg, Front Public Health 2017. Read 2026-10-22.
- Bravi et al., Clin Sci 2007. Read 2026-10-22.
- Systematic review, Folia Med 2018. Read 2026-10-22.
- Systematic review and meta-analysis, Sensors 2026. Read 2026-10-22.
- Nunan et al., Pacing Clin Electrophysiol 2010. Read 2026-10-22.
- Hillebrand et al., Europace 2013 (meta-analysis). Read 2026-10-22.
- NHS: Heart palpitations. Read 2026-10-22.