Heart Rate Variability: What It Is, What It Reflects, What It Isn't

What is HRV?

Heart rate variability is the natural variation in the time between consecutive heartbeats. A healthy heart is not a metronome: the interval between one beat and the next is never exactly the same, and those small differences — the interbeat intervals — are what the term refers to [S1] [S2]. It is a property of the heart's rhythm, not a heart rate: two people can share an average pulse and have completely different variability.

This page is the canonical definition the rest of the science section builds on. The two main metrics have their own pages — RMSSD and SDNN; how closely smartwatches and rings can reproduce the ECG measurement is the question of Can you trust HRV from a smartwatch or ring?; where your numbers sit by age belongs to the normal HRV by age article and the HRV calculator; and the everyday questions — what a low reading means, what to do about it — live in HRV Questions, Answered. None of that is repeated here.

How does HRV work?

Every heartbeat is triggered by the heart's own pacemaker, and the autonomic nervous system continuously adjusts the pace from beat to beat. Two overlapping processes generate the short-term variation: the dynamic relationship between the sympathetic and parasympathetic branches, and regulatory mechanisms that include the breathing-linked rhythm and the baroreflex [S2]. HRV is often described as a window onto autonomic regulation; the honest version is narrower — it is the visible trace of that regulation in the heart's rhythm, and the trace is not the machinery.

The most visible contributor in a short resting recording is respiratory sinus arrhythmia (RSA): the respiration-driven speeding and slowing of the heart, with each breath in briefly speeding the rhythm and each breath out slowing it [S2]. In brief resting recordings RSA is the primary source of the variation [S2] — which is why breathing rate and depth visibly move the number. The breathing-and-HRV mechanism has a dedicated page planned; the practical side of slow breathing lives in how to raise HRV naturally and is not repeated here.

How is HRV measured?

HRV is computed from a series of interbeat intervals, so everything starts with how those intervals are obtained. The reference method is an ECG, which times the heart's electrical activity directly; consumer wearables instead estimate the timing from the optical pulse signal at the skin [S3]. How closely that estimate tracks the ECG — where it agrees, where it degrades — is a measurement question in its own right, and it belongs to the wearable HRV measurement page rather than to this one.

The two metrics you will meet most often are time-domain statistics computed from those intervals. The beat-to-beat one is RMSSD — the root mean square of successive differences between heartbeats [S1]; it isolates the changes between adjacent heartbeats, leans more on the fast parasympathetic pathway [S2], and is the value most sleep and recovery apps report. Its usual companion, SDNN — the standard deviation of the intervals between normal heartbeats [S1], summarizes the overall spread across the recording window. Each has its own concept page — RMSSD and SDNN — with the definitions, the Apple Health specifics and the recording-length story. This page does not repeat them.

The window is part of the meaning. A brief lab recording, an all-day hospital ECG and a watch's overnight samples are different measurement regimes: values grow with recording length, and numbers across regimes are not interchangeable [S1] [S2]. Published norms treat all-day, short-term and ultra-short-term values as separate worlds for the same reason [S2]. For keeping your own readings comparable, the practical guide is how to measure HRV consistently.

What affects HRV?

• Breathing. Breathing rate and depth during the window write themselves into the value through the breathing-linked rhythm [S2], and current guidelines list respiration among the technical factors that shape rigor and reliability [S4].
• Age. On average, 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 [S5]; individuals vary widely, and population averages are not personal targets. Where your numbers sit by age is the normal HRV by age article's question.
• Recording length and time of day. Longer recordings absorb slower rhythms and circadian processes, each adding to the value [S2]; a morning spot check and an overnight average describe different worlds.
• Recording method. An ECG and an optical pulse sensor are different measurement chains [S3]; numbers from the two are not automatically comparable.
• Recording conditions. The setting, body position, breathing and the analysis method all shape the value and the rigor of any comparison [S4].
• Everyday conditions. As with other HRV metrics, a single reading can deviate for ordinary reasons; treat such shifts as observations, not verdicts.

What does the evidence show?

Established. The definition and the physiological origin of the signal come from the field's measurement standards [S1] and methodological reviews [S2]: HRV is generated by the interplay of the two autonomic branches with regulatory mechanisms, and RSA is the dominant source in short resting windows [S2]. The metric properties are equally settled — RMSSD and SDNN summarize different aspects of a recording, and values depend on recording length [S1] [S2]. On average, values decline with age in healthy adults, with wide individual variation [S5]. In clinical cardiology, HRV computed from all-day ECG in patient populations is an established risk-stratification measure — a measurement regime a consumer watch does not reproduce [S2].

Methodological guidance. Current guidelines name the input signal, recording length, setting, breathing and analytical approach as factors that determine rigor and reliability [S4], and call for findings — including those from wearables — to be interpreted and contextualized within these limitations [S4]. That is expert consensus about how to measure and interpret, not direct experimental data about HRV itself.

Context-dependent. Consumer estimates from the pulse signal are a different measurement chain from the ECG reference [S3]; the agreement evidence, its conditions and its limits are covered on the measurement page rather than repeated here.

Unknown. What a specific absolute value means for a specific healthy person; how overnight consumer values map onto the clinical evidence base, which was built on continuous ECG in patients; and how much of a day-to-day shift reflects physiology versus measurement noise. These are open questions — reasons for measurement literacy, not for discarding the data.

What HRV does not tell you

• It is not heart rate. The average pulse and the beat-to-beat variation are different properties of the same rhythm; a slower pulse does not by itself mean more variability [S1] [S2].
• It is not vagal tone. The framing matters. Vagal tone cannot be measured directly; HRV measures such as RMSSD reflect vagally mediated changes in heart rate [S1] [S4].
• It is not a stress or health verdict. a single low HRV reading does not by itself mean you are stressed or unwell [S1] [S4] — and the same caution bounds unusually high values.
• Higher is not automatically better. Abnormal beats and noise can masquerade as variability and inflate the number [S2]; higher values are generally associated with better recovery, with exceptions.
• It is not one universal number. Metrics, windows and methods differ; a value without its method, metric and window is incomplete information [S2] [S3].
• A single value says little. Methodological guidance calls for cautious, contextualized interpretation [S4]; your own recent readings under comparable conditions are the more informative comparison.

In ONDA

ONDA's nightly baseline is built on the HRV values written to Apple Health — from Apple Watch or any tracker whose app syncs heart data there. Apple Health records HRV as SDNN [S6], so that baseline signal is SDNN-based rather than RMSSD-based, and the app's night-by-night comparisons run against your own corridor rather than population targets. The live reading shown during a practice is a surrogate computed from the standard deviation of heart rate, not RMSSD or SDNN, and the phone camera gives pulse, not HRV. ONDA's use of the signal is biofeedback — observing what happens to your physiological signal while you practise — and the HRV calculator puts Apple Watch numbers in context. ONDA describes and compares your own numbers; it does not diagnose anything. See what ONDA measures.