Recovery Guide

HRV Normal Range by Age and Sex

In a study of 1,906 healthy adults, average SDNN (the HRV measure Apple Watch reports) fell from about 49–50 ms at ages 25–34 to 37 ms at 45–54 and 28–30 ms at 65–74, with similar values for women and men (Voss et al., PLoS One, 2015). The spread at every age is wide, so your own rolling baseline is the number that actually means something.

Last reviewed 12 October 2026 · Sources linked in the text · Not medical advice

This guide gives you the age-by-age reference values, what they can and cannot tell you about a smartwatch reading, why the metric declines with age, and everything else that moves it — so you can stop grading yourself against a chart that was never built for that. Prefer Apple-specific detail? Read Apple Watch HRV explained.

Start with your baseline, not the table

The most useful thing anyone can tell you about HRV norms is how little they apply to you individually. Consider the classic healthy-adult reference for 24-hour SDNN: 141 ± 39 ms (Task Force of the European Society of Cardiology & NASPE, Circulation, 1996). Take that at face value and roughly two in three healthy adults land somewhere between about 102 ms and 180 ms — a near two-fold window, before you split anyone by age at all. Widen it to two standard deviations and healthy values run from the sixties into the low two-hundreds.

That is the whole problem with “normal for my age.” The spread inside a single age band dwarfs the average gap between bands, so a 45-year-old at 130 ms and a 25-year-old at 130 ms are both entirely unremarkable. What is not unremarkable is your own number moving: if you normally sit around 60 ms overnight and you have spent ten days at 40 ms, that is a real signal — regardless of where either figure sits on a population chart.

This is precisely how Reps scores it. Rather than compare you to a table, it learns your personal HRV range from your own history and reads today against that. New to the metric itself? Start with What Is HRV.

HRV by age and sex: the reference table

The most useful published age-by-age figures for short recordings come from Voss and colleagues, who analysed 5-minute ECGs from 1,906 healthy adults aged 25 to 74 in the German KORA S4 population study, recorded lying down after 5–10 minutes of rest (Voss et al., 2015). Values are mean ± standard deviation in milliseconds.

Three things stand out. HRV falls with age on both measures, most steeply between the mid-30s and mid-50s, and the study found no significant differences between the 55–64 and 65–74 groups. Women and men are close on these time-domain measures. And the standard deviations are large: at 25–34, a woman one standard deviation below the mean on RMSSD (about 20 ms) sits close to the average for a 55-year-old. A single age band covers a wide range of perfectly healthy people.

The reference figures that are real

These are the healthy-adult numbers most often quoted in the literature. They pool adults across ages and both sexes — they are not age bands — and they exist here to show you how the measurement window changes the answer, and how wide “normal” is once you account for the standard deviation.

24-hour values are the healthy-adult reference from the 1996 Task Force standards; short-term values are the 5-minute normative figures compiled in Shaffer & Ginsberg, 2017. The final column is simply mean ± 1 SD, written out. Reported norms vary between studies and measurement protocols.

Look at the two SDNN rows. The same metric, in the same healthy population, reads around 141 ms over 24 hours and around 50 ms over five minutes — roughly a three-fold difference driven entirely by how long the recording ran, because a longer window captures more of the slow, circadian swings in heart rate. If you have ever compared a short reading from your watch against a “normal HRV is 100+” figure you found online and felt alarmed, this is why. You were comparing two different measurements.

SDNN and RMSSD are not interchangeable

Two metrics dominate consumer HRV, and swapping one for the other is the most common way people mislead themselves:

• SDNN — the standard deviation of the intervals between normal heartbeats. It reflects total variability from all sources, and it grows with the length of the recording. Apple Watch reports SDNN, which is what lands in Apple Health as Heart Rate Variability, sampled in short windows during the day and around sleep rather than continuously.
• RMSSD — the root mean square of successive differences between beats. It emphasises the fast, beat-to-beat changes driven mainly by the vagus nerve, and it is stable over short recordings, which is why most recovery apps and chest-strap tools favour it.

They answer different questions and produce different numbers from the exact same heartbeats. Never compare an SDNN value to an RMSSD chart, never compare a five-minute reading to a 24-hour norm, and never compare across two devices that sample at different times of night. If you also wear a Whoop or Oura, see how to read your Whoop or Oura HRV for why those numbers never quite agree either.

Why HRV declines with age

The downward drift is real and reasonably well explained. The dominant factor is reduced parasympathetic (vagal) influence on the heart: the rest-and-digest brake that produces most beat-to-beat variability becomes less pronounced with age. Alongside it, baroreflex sensitivity — the reflex that adjusts heart rate to blood-pressure changes — tends to blunt, arteries stiffen, and the sinoatrial node becomes less responsive to autonomic input. Fewer, smaller corrections from moment to moment means less variability to measure. Age dependency of HRV is well documented in the standards literature (Task Force, 1996) and in normative reviews (Shaffer & Ginsberg, 2017).

Two caveats keep this in proportion. First, it is a population average — plenty of older adults have HRV above plenty of younger ones. Second, aging is entangled with everything that travels with it: fitness, body composition, sleep quality, alcohol intake, medication and stress load. Some of what looks like an age effect is really a lifestyle effect that happens to accumulate over decades.

What else moves your HRV

On any given morning, these will shift your number far more than another year of age:

• Sleep. Short or fragmented nights suppress HRV. Overnight readings also depend on how much deep sleep the window captured.
• Alcohol. One of the most reliable overnight HRV suppressors, often visible even after a modest evening drink.
• Illness. HRV frequently drops before you feel obviously unwell, and recovers as you do.
• Training load. A hard session or a heavy block can depress HRV for a day or more — that is expected adaptation, not a warning, provided it rebounds.
• Psychological stress. Deadlines, conflict and anxiety pull the sympathetic branch up and HRV down.
• Measurement timing and position. HRV is higher lying down than seated or standing, higher during slow-wave sleep than on waking, and higher when you breathe slowly. A supine morning reading and a seated afternoon reading are not comparable.
• The everyday stuff. Hydration, a late heavy meal, caffeine timing, room temperature, travel and time-zone shifts all leave fingerprints.

The practical consequence: measure the same way every time, then judge the 7-day direction rather than any individual morning.

Why the cut-offs you see quoted may not apply to you

You will find SDNN thresholds circulating online — under 50 ms, 50 to 100 ms, above 100 ms — presented as health grades. Those bands come from clinical research using full 24-hour Holter recordings, largely in the context of cardiac risk stratification in patient populations. They were never designed to grade a one-minute reading from a smartwatch, and applying them to one will make a completely ordinary value look frightening. If a threshold does not state the recording length, the metric and the population it came from, it is not usable.

How Reps makes HRV easier to read

Reps is an AI recovery & training coach for iPhone and Apple Watch. It builds a rolling HRV baseline from your own history, and then asks the only question that generalises: where does today sit relative toyour normal? Reps shows that trend beside a daily readiness score built from sleep, resting heart rate and recent training load, so one noisy night after a late dinner does not get mistaken for the whole recovery picture.

Because the comparison is personal rather than population-based, it works the same whether you are 22 or 62, and it keeps working as your baseline itself shifts over the years. Rex, the AI coach built into Reps, can then explain what your trend looks like in plain language and adjust the plan — easing off when your recovery signals are down, giving you the green light when they are back. No subscription hardware required; if you already wear an Apple Watch, you already have the sensor.

Frequently asked questions

The bottom line

HRV really does decline with age, but the decline is slow and the healthy range at every age is wide enough to swallow it. Chasing a number from an age chart is the least useful thing you can do with the metric. Measure consistently, learn your own range, watch the direction over a week or two, and treat any sharp or sustained change — not any particular value — as the thing worth acting on.