What a Single HRV Value Can and Can't Tell You

What can a single HRV value tell you?

One heart rate variability value answers a narrow question: how much did the time between heartbeats vary during that recording? Most of that short-term variation comes from vagal influence on the heart. The framing matters. Vagal tone cannot be measured directly; HRV measures such as RMSSD reflect vagally mediated changes in heart rate [S3].

So a reading describes your heart rhythm at that moment, in those conditions, as captured by that device and summarized by that metric. Change any of these and the number changes: the signal source (chest ECG or optical pulse), the length of the recording, where it was taken, how you were breathing and how it was analysed all shift the result [S4]. A short daytime reading and a night-long average are not the same quantity, and values from recordings of different length cannot be compared directly [S2]. RMSSD and SDNN are different metrics, and how a watch or ring estimates HRV adds its own error.

That is all one number can say. It is easy to read more into it than it holds [S3].

Why is HRV so personal?

"You can compare your HRV with a friend's." Hardly at all. Healthy adults differ enormously in HRV [S6], and a lower value than a peer's does not necessarily mean poorer physiological status [S5]. On average HRV falls with age — HRV tends to fall with age — in our night-time RMSSD table the median drops by about 4–8 ms from one age band to the next [S7] — and the metric, the device, the time of day and the recording length differ between any two people's numbers [S4]. A friend's higher reading tells you nothing about your health.

The useful comparison is with yourself: your own readings, taken the same way, over days and weeks. How that works is the subject of the HRV baseline. Age tables are in normal HRV by age; read them as group context, not as a target.

When is a high value a reason to check the data?

"Higher is always better." In sports science, HRV at or above your own baseline is used as one of the signs of recovery: in one small randomized trial, hard sessions were scheduled only on mornings when HRV had not dropped [S18]. But a higher number is not always a better one. HRV is computed from the intervals between beats, so anything that makes those intervals irregular for reasons other than normal vagal rhythm raises the number. Abnormal beats and noise can masquerade as HRV [S2].

• Extra beats and arrhythmia. Ectopic (premature) beats and atrial fibrillation count as physiological artefacts in the beat-interval series [S8]. Ectopic beats add mathematical artefact to both time- and frequency-domain measures, and leaving them uncorrected significantly altered HRV in simulations [S9].
• Recording artefacts. Loose contact and movement create false beats; such beats are common and can make a reliable analysis difficult or impossible [S8].
• Erratic sinus rhythm. In one cohort of older adults, abnormal heart-rhythm patterns that raised many HRV indices were common and were associated with higher, not lower, mortality [S10] (single observational study). Researchers describe this as high short-term HRV that reflects a breakdown of heart-rhythm control rather than healthy vagal influence, and are developing measures such as heart rate fragmentation to separate the two [S11] (emerging research).

A sudden, unusually high value is therefore a reason to check the data — was the device loose, were you moving, does the pulse feel irregular — before treating it as good news.

What can make a reading low?

Many factors move HRV [S4]. The ones most commonly reported are:

• a short or poor night's sleep;
• alcohol;
• an illness starting, such as a cold or fever;
• a hard training session or a heavy training block;
• measurement conditions: a different time of day, posture, movement, talking or a loose sensor.

How each of these works, and how large the effects are, belongs to the planned page on why HRV changes from day to day. Practical next steps after a low reading are in what to do after a low HRV reading, and the Apple Watch specifics in why your Apple Watch HRV is low.

A single low HRV reading does not by itself mean you are stressed or unwell [S1]. A single low value is not a diagnosis and not a measurement of stress; it is one point that only gains meaning next to your own recent readings.

What does the evidence show?

Established — population risk. In patient groups and large populations, low HRV is associated with worse outcomes. The classic study by Kleiger and colleagues measured SDNN from all-day ECG recordings in people after a heart attack: lower HRV was associated with higher mortality and remained a significant predictor after adjustment for other clinical factors [S14]. Its thresholds cannot be compared with the SDNN an Apple Watch shows, because the watch calculates SDNN from short readings, not from a whole day of ECG. In population studies of people without known heart disease, mostly with ECG recordings, a meta-analysis found about 32–45% higher risk of a first cardiovascular event in those with low HRV [S15]. In patients with cardiovascular disease, a meta-analysis of cohort studies found lower HRV associated with a higher risk of death and of cardiovascular events [S16].

These are population associations from clinical-grade ECG recordings, mostly in patients or older adults. They do not transfer to a single value from a watch: one reading does not tell a healthy person their risk, and these studies do not show that raising your HRV lowers it.

Context-dependent — readiness and recovery scores. Many apps and wearables turn HRV, resting heart rate, sleep and other inputs into a single readiness or recovery score. These are each manufacturer's own formulas: an analysis of HRV apps found that many such scores rely on proprietary algorithms that are not transparently described, so it is hard to assess independently how they were derived and validated [S12]. A review of smartwatch metrics describes readiness scores as algorithmic estimates rather than direct physiological measurements, and finds wearable data most useful for within-person trends rather than diagnosis or cross-device comparison [S13]. A score can be a convenient summary, but it is not a standard scientific metric, and two brands' scores are not comparable.

Context-dependent — single readings. HRV findings are easy to misread without their measurement context [S3]. What holds up best is the direction of your own readings over time, taken under similar conditions.

What it does not tell you

• Not your stress level or a diagnosis. One value says nothing specific about your mental state or about disease [S1].
• Not your rank. A higher or lower number than someone else's says little about either of you [S5] [S6].
• Not your fitness or "readiness". Neither the raw value nor a proprietary score is a validated measure of these on its own [S12] [S13].
• Not a personal risk forecast. Population risk findings do not apply to one healthy person's reading [S15] [S16].
• When HRV is worth attention. A shift away from your own baseline that lasts many days, together with symptoms such as unusual tiredness, poor sleep, fever or a racing or irregular pulse, is worth noticing — the number is a prompt to look at how you feel, not the finding itself.
• When to see a doctor. As a general precaution, see a doctor if you notice an irregular pulse, palpitations, or a lasting change in your readings that comes with symptoms. Seek urgent care for chest pain, fainting, or severe breathlessness. If you take your data to an appointment, doctors and your data explains what helps, with guides for the GP and the cardiologist.

More everyday questions are answered in HRV questions answered, and you can work through your own numbers in the HRV calculator.

In ONDA

ONDA reads HRV (SDNN) from Apple Health, written there by an Apple Watch or another device that syncs heart data [S17]; the iPhone camera gives pulse, not HRV. ONDA never judges a single value against other people. ONDA compares each night with your own corridor — the average of your recent nights plus or minus one standard deviation — and flags a night only when it is at least 1.5 standard deviations outside and has changed by a minimum amount. These signals are descriptive comparisons with your own baseline, not measurements of stress [S17]. ONDA has no single readiness or recovery score and does not diagnose anything. See what ONDA measures.