The HRV number on a patient’s watch or ring is a pulse-based estimate from one night of sleep or a short reading at rest. It is not the 24-hour Holter SDNN SDNN The standard deviation of all NN intervals in a recording, in milliseconds. It measures overall variability, and its value depends strongly on how long the recording lasts. More about SDNN used in the risk studies, so read it as a personal trend and never as a test.
Take an example. A patient shows you a screenshot in which their ring says their HRV fell from 48 ms to 31 ms this week, with a red “low” label. They want to know if something is wrong with their heart. This post covers what that number is, how far you can trust it, and what to say.
What consumer HRV measures
HRV is the variation in time between normal heartbeats. The 1996 Task Force standards define the two measures patients see most (Task Force, 1996):
- RMSSD RMSSD Root mean square of successive differences: an HRV metric built from how much each heartbeat interval differs from the next, in milliseconds. It mainly reflects vagal activity. More about RMSSD is the root mean square of successive differences between normal beats. It is the main time-domain measure used to estimate fast, vagally driven changes (Shaffer and Ginsberg, 2017).
- SDNN is the standard deviation of all normal-to-normal intervals. It takes in slower swings too, so it depends strongly on recording length.
Most rings and wrist bands report RMSSD averaged over sleep. In one validation study, all five devices tested, from four makers, reported HRV as RMSSD (Dial and colleagues, 2025). Apple Health is the main exception. It stores Apple Watch HRV as SDNN, from samples the watch records automatically (Apple Developer Documentation). So two patients with “an HRV of 40” may be reporting different statistics, recorded at different times.
Chest straps time beats from the ECG ECG A recording of the heart's electrical activity, taken with electrodes on the skin. ECG-based RR intervals are the reference standard for measuring HRV. More about ECG signal. Rings and watches use photoplethysmography PPG An optical method that shines light into the skin and detects the pulse of blood with each heartbeat. Most watches, rings and phone-camera apps use it to estimate HRV. More about PPG (PPG), a light sensor that finds each pulse in the skin, so strictly they give pulse rate variability Pulse rate variability Variation in the time between pulse beats, measured optically by watches, rings and phone cameras. It closely tracks HRV at rest but drifts from it during movement. More about Pulse rate variability , an estimate of HRV.
How accurate is it?
At rest, PPG holds up fairly well: a review of studies comparing pulse-based and ECG-based variability found sufficient accuracy when people were at rest. Pulse data tended to overestimate short-term variability somewhat. Physical activity and some mental stressors made agreement worse, often to an unacceptable degree (Schäfer and Vagedes, 2013).
For overnight readings, Dial and colleagues followed 13 healthy adults over 536 nights, with a chest-strap ECG as the reference. The mean absolute percentage error for nightly RMSSD ranged from 5.96% for the best ring (Oura Gen 4) to 16.32% for the weakest watch (Polar Grit X Pro). At an RMSSD of 50 ms, that is an average error of roughly 3 to 8 ms, and some nights were much further off. The study was small and included only healthy adults, so the figures may not hold in patients with arrhythmias or heart disease.
The HRV by device guide lists what each maker measures and when.
How it differs from clinical HRV
| Consumer wearable | Clinical and research HRV | |
|---|---|---|
| Sensor | Optical pulse (PPG) on finger or wrist; ECG chest strap | ECG, often a 24-hour Holter |
| Measure | Mostly RMSSD; SDNN in Apple Health | SDNN over 24 hours, plus other time and frequency measures |
| Window | One night of sleep, or spot samples | 5-minute or 24-hour recordings under set conditions |
| Beat editing Beat correction Finding the beat intervals that are wrong (missed, extra or ectopic beats) and deleting or replacing them before HRV is calculated. The rule of thumb is to correct under 5% of beats. More about Beat correction | Usually not published | Ectopic beats Ectopic beat A heartbeat that starts outside the heart's normal pacemaker and usually comes early. Ectopic beats are common, but they distort HRV unless removed or corrected. More about Ectopic beat and artefacts edited out by set methods |
| Comparison | Against the person’s own recent values | Against published reference ranges and study cut-offs |
The recording length matters most. Because SDNN depends on the length of the recording, the Task Force standards advise against comparing SDNN values from recordings of different lengths. A later review makes the same point: 24-hour, 5-minute and shorter values are not interchangeable (Shaffer and Ginsberg, 2017). A 24-hour Holter SDNN and a watch’s spot SDNN are different numbers.
Beat editing matters too. HRV formulas assume normal sinus beats. Ectopic beats and atrial fibrillation make the intervals uneven and can make reliable analysis hard or impossible (Peltola, 2012). Clinical analysis removes or corrects these beats first. Most consumer devices do not say how they handle them.
What the outcome research shows
The research is real, but it describes groups measured with clinical methods. Treat it as background only, not as a reason to act on a consumer reading.
- After a heart attack. Kleiger and colleagues analysed 24-hour Holter recordings from 808 survivors, taken about 11 days after the event. Over a mean of 31 months, mortality was 5.3 times higher with SDNN below 50 ms than above 100 ms. SDNN stayed a significant predictor after adjusting for ejection fraction and other factors (Kleiger and colleagues, 1987).
- Across healthy and patient groups. A 2022 meta-analysis covered 32 studies and two individual data sets, with 38,008 people. Lower HRV predicted higher mortality across ages, sexes and recording lengths. In the studies that compared the lowest quarter of 5-minute RMSSD with the other three quarters, the pooled hazard ratio was 1.56 (95% CI 1.32 to 1.85) (Jarczok and colleagues, 2022).
Age weakens what any cut-off means. In 24-hour recordings from 260 healthy people aged 10 to 99, RMSSD in the sixth decade had fallen to 47% of its value in the second decade. Among healthy people over 65, 12% had RMSSD below published mortality cut-offs (Umetani and colleagues, 1998).
None of these studies used wearables, and none tested whether raising HRV changes outcomes. The HRV and longevity page explains the relative and absolute risk for patients.
What patients misread
Comparing with other people. Patients look up charts by age and worry when they sit below the middle. The spread between healthy people is wide, and the charts often used other measures and recording lengths.
Comparing devices. An Apple Watch SDNN and a ring’s overnight RMSSD will not match, even on the same night, and switching device means building a new baseline Baseline Your own usual HRV level and range, built from weeks of readings taken the same way. Comparing a reading with your baseline says more than comparing it with other people. More about Baseline .
Reading one night as a verdict. Overnight HRV moves with alcohol, poor sleep, hard training and infection, as the what affects HRV guide sets out. A low reading after a heavy weekend or a hard session is expected.
Treating a high number as reassurance. Irregular rhythms such as atrial fibrillation can raise HRV (Shaffer and Ginsberg, 2017). A value that is unusually high or jumps from night to night is not proof of good health.
Reading a study cut-off onto their own number. “Under 50 is dangerous” comes from Holter SDNN after a heart attack. It does not apply to a ring.
How to talk about it
A short set of questions settles most of these conversations:
- Which device and which measure? RMSSD or SDNN, overnight or spot, ring or watch.
- How long has it lasted? A few days or several weeks, and is it a weekly average or a single night?
- What else changed? Sleep, alcohol, training load, illness, new medicines, shift work, stress.
- Any symptoms? Palpitations, chest pain, breathlessness, dizziness or fainting carry the same weight they would without a wearable, whatever the HRV says.
Then reframe the number. It is a personal trend, useful for noticing patterns in sleep, alcohol and training. Low HRV is linked to higher risk in research groups, but it is not a diagnosis, and a normal value does not rule anything out. Showing a patient their own weekly average against their own range puts a single red label in context.
Where to send patients on this site
- What is HRV? for a plain introduction.
- What does a low HRV mean? for the common, harmless reasons.
- Why your devices disagree for SDNN versus RMSSD and device differences.
- Building your HRV baseline for reading weekly trends.
- Can HRV detect heart problems? for limits and red-flag symptoms.
The free HRV course covers all of this step by step, with a short quiz after each lesson.
Where to go next on hrv.expert
- Standards and references: the papers and guidelines the site relies on.
- HRV norms dataset: the reference values behind the checker, with their sources.
- HRV and longevity: the mortality studies in more detail.
- Is wrist HRV as accurate as a chest strap?: the accuracy evidence by setting.
- HRV limits, myths and red flags: the lesson to share with patients who worry.
- HRV for health coaches: how coaches are taught to use HRV and when to refer.