What a normal HRV looks like by age and gender: the baseline ranges nobody publishes

Sam Health is a wellness app, not a medical device. The data and ranges discussed below are wellness signals for general awareness. Consult a healthcare professional for medical advice.

TL;DR

In the largest healthy-adult HRV study with proper age and sex stratification — 1,906 people from the KORA S4 cohort, recorded with 5-minute ECG — mean SDNN drops from about 49–50 ms at age 25–34 to about 28–30 ms at age 65–74. Sex differences in SDNN and RMSSD are smaller than most blogs claim and largely non-significant at any age decade. Your Apple Watch number will typically read lower than this table — recent validation work shows Apple Watch underestimating SDNN by about 8 ms compared with a chest-strap reference. The number that matters is not the population mean; it is your own seven-to-fourteen-day baseline trend, measured the same way every day.

What HRV actually measures

Heart rate variability is the variation in the time interval between successive heartbeats. A healthy autonomic nervous system produces a heart rhythm that varies — the gaps between beats are not uniform. More variation usually means more parasympathetic (rest-and-recover) influence; less variation usually means more sympathetic (stress, exertion) influence.

There are several ways to summarize that variation into a number. The two most-cited time-domain metrics are:

• SDNN — standard deviation of normal-to-normal beat intervals across the recording window. Apple Watch reports this metric (Apple Developer, heartRateVariabilitySDNN, accessed 15 May 2026). It is the metric most strongly associated with overall cardiac risk in clinical 24-hour recordings.
• RMSSD — root mean square of successive differences between adjacent beat intervals. Oura, WHOOP, Polar, and Garmin tend to expose RMSSD or a derivative. It is more sensitive to short-term, beat-to-beat parasympathetic activity.

There are also frequency-domain metrics (LF, HF, LF/HF ratio) and nonlinear metrics that some research devices report. For everyday use on a consumer wearable, SDNN and RMSSD are the two numbers you'll actually encounter, and the rest of this article focuses on them.

What "normal" HRV looks like in healthy adults

The most defensible reference table for non-athlete adults comes from Voss et al. 2015, published in PLOS ONE. They analyzed 5-minute ECG recordings from 1,906 healthy participants in the KORA S4 cohort (a German population study), stratified by sex and by 10-year age band from 25 to 74. Subjects with cardiovascular disease, diabetes, or relevant medication use were excluded (Voss et al., 2015, PLOS ONE, accessed 15 May 2026).

The values below are means and standard deviations from that paper's Tables 5 and 7. They are 5-minute clinical ECG values, not wrist-watch values. We explain the gap further down.

Female participants, 5-minute clinical ECG

Source: Voss et al., 2015, Table 5 (PLOS ONE).

Male participants, 5-minute clinical ECG

Source: Voss et al., 2015, Table 7 (PLOS ONE).

A few patterns worth noticing:

• The age decline is real and steep. Mean 5-minute SDNN drops from roughly 50 ms in the 25–34 band to under 30 ms in the 65–74 band. Mean RMSSD halves over the same range. This is normal physiology, observed across many independent cohorts.
• The standard deviations are huge. Look at the SD column. A female aged 25–34 with a 5-minute SDNN of 30 ms is roughly one standard deviation below the mean, but is still well within the healthy distribution. HRV varies enormously between healthy individuals.
• Sex differences in time-domain HRV are smaller than most internet sources claim. Voss et al. tested gender differences directly in each age decade and found no statistically significant difference between men and women for SDNN or RMSSD in any age band from 25–34 through 65–74 (Voss et al., 2015, Table 9). Sex differences do appear in some frequency-domain (LF/HF) and nonlinear metrics, but for the simple time-domain numbers most consumer wearables report, age matters far more than sex.

The Nunan et al. 2010 systematic review of 44 short-term HRV studies (21,438 participants) reached a similar headline conclusion: there is reasonable cross-study agreement on means, but inter-individual variability is enormous, and short-term values from the literature tend to run lower than the older ESC Task Force long-term norms (Nunan et al., 2010, Pacing Clin Electrophysiol, accessed 15 May 2026).

What your Apple Watch shows, and why it doesn't match the table

Apple Watch calculates SDNN over short windows using its optical heart-rate sensor — green LEDs flashing through your wrist's blood vessels, captured at high frequency, then converted to beat intervals (Apple Support, Monitor your heart rate with Apple Watch, accessed 15 May 2026). The HealthKit type is heartRateVariabilitySDNN and is captured most reliably during Breathe sessions and quiet, still moments throughout the day (Apple Developer, heartRateVariabilitySDNN, accessed 15 May 2026).

There are three reasons your Apple Watch number probably reads lower than the Voss table:

1. The sensor and measurement window are different. Voss used a 5-minute resting ECG with chest electrodes. Apple Watch uses wrist photoplethysmography over short, often non-fixed windows. These are not the same measurement, even if they share an SDNN label.
2. Apple Watch underestimates SDNN compared with a chest-strap reference. A 2024 prospective validation study from University College Dublin compared Apple Watch Series 9 and Ultra 2 against a Polar H10 chest strap paired with the Kubios HRV reference software, in 39 healthy adults across 316 paired measurements over 14 days. Apple Watch underestimated SDNN by an average of 8.31 ms (95% CI −11.04 to −5.59 ms), with a mean absolute error of about 20 ms and a mean absolute percentage error of about 29%. Apple Watch resting heart rate, by contrast, was within ~0.1 bpm of the reference on average (O'Grady et al., 2024, Sensors, accessed 15 May 2026).
3. Daily context shifts the number around. Sleep the night before, alcohol, training load, position (sitting vs. lying), time of day, hydration, stress, and even ambient temperature all move HRV. The Voss table is one number per age band; your watch produces dozens per week, captured under different conditions.

So when you look at the Voss numbers and then look at your Apple Watch and see, say, a 28 ms SDNN at age 32, you are not "low" in any clinical sense. You are looking at the result of a different measurement method and a different measurement context. The right reference point is your own baseline, not a published mean.

SDNN vs. RMSSD: a quick translation

If you're switching from another wearable or you're comparing your Apple Watch HRV with a friend's Oura number, two things to know:

• Apple Watch shows SDNN. Oura and WHOOP show RMSSD or an RMSSD-derived score. Polar can show either. Garmin's "stress" metric is computed from RMSSD.
• RMSSD numbers are typically smaller than SDNN numbers from the same recording. In the Voss data above, mean RMSSD is about 80–90% of mean SDNN in the younger age bands and falls faster with age than SDNN does. They are different metrics that capture different aspects of variability, not different health states. Higher SDNN and higher RMSSD both generally reflect better recovery and parasympathetic balance for a given person, but their absolute values are not interchangeable.

There is one technical note worth flagging for completeness: in the Voss 2015 paper, the Poincaré plot index SD1 carries the exact same information as RMSSD with a fixed mathematical conversion (SD1 × √2 = RMSSD). If you ever see "SD1" in a research paper, treat it as RMSSD wearing different clothes.

What this means for reading your own data

A few practical rules, given everything above:

• Use a rolling personal baseline. A seven- to fourteen-day rolling average of your Apple Watch SDNN, measured under similar conditions, is the most useful single reference. Sam Health is built around this principle — it surfaces a deviation when your daily reading drifts meaningfully from your own baseline, not when you fail to match a population chart.
• Watch the direction, not the absolute number. A working-age non-athlete whose SDNN trends from 42 ms down to 30 ms across three weeks is showing a deviation worth attending to. The same person sitting flat at 30 ms for months is just sitting at their baseline.
• Be patient with the baseline. Apple's overnight Vitals app needs about seven nights to establish a typical range for its overnight metrics (Apple Support, Track your overnight vitals with Apple Watch, accessed 15 May 2026). HRV's day-to-day variability is high enough that you should give yourself 14–30 days before reading too much into a trend.
• Cycle phase matters for menstruating users. HRV typically drops in the luteal phase relative to the follicular phase, which means a "normal HRV" framing without cycle awareness can mislead. We unpack this in cycle phase and HRV, sleep, and resting heart rate.
• What about the ESC Task Force 50 ms 24-hour SDNN threshold? It comes from full 24-hour Holter ECG recordings (ESC/NASPE HRV Task Force, 1996, accessed 15 May 2026). Apple Watch does not produce that recording. The clinical threshold is not directly comparable to your wrist number — using it as a personal cutoff is a category error.

Where Sam Health fits in

Sam Health reads your Apple Watch SDNN through HealthKit and presents it as a trend against your own personal baseline rather than against a population chart. It surfaces insights when there's new context worth knowing about your HRV — for example, when a sustained downward shift is co-occurring with other deviations in your overnight metrics — and lets you build healthier habits with daily insights. It is a wellness companion, not a medical device.

Try Sam Health

Sources

• Voss A. et al. — Short-Term Heart Rate Variability — Influence of Gender and Age in Healthy Subjects, PLOS ONE 10(3):e0118308 (2015) — https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0118308 — accessed 15 May 2026.
• Nunan D. et al. — A Quantitative Systematic Review of Normal Values for Short-Term Heart Rate Variability in Healthy Adults, Pacing and Clinical Electrophysiology 33(11):1407–1417 (2010) — https://pubmed.ncbi.nlm.nih.gov/20663071/ — accessed 15 May 2026.
• Shaffer F. & Ginsberg J.P. — An Overview of Heart Rate Variability Metrics and Norms, Frontiers in Public Health 5:258 (2017) — https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2017.00258/full — accessed 15 May 2026.
• ESC / NASPE HRV Task Force — Heart Rate Variability: Standards of Measurement, Physiological Interpretation, and Clinical Use (1996) — https://www.escardio.org/static-file/Escardio/Guidelines/Scientific-Statements/guidelines-Heart-Rate-Variability-FT-1996.pdf — accessed 15 May 2026.
• O'Grady B., Lambe R., Baldwin M., Acheson T., Doherty C. — The Validity of Apple Watch Series 9 and Ultra 2 for Serial Measurements of Heart Rate Variability and Resting Heart Rate, Sensors 24(19):6220 (2024) — https://www.mdpi.com/1424-8220/24/19/6220 — accessed 15 May 2026.
• Apple Developer Documentation — heartRateVariabilitySDNN — https://developer.apple.com/documentation/healthkit/hkquantitytypeidentifier/heartratevariabilitysdnn — accessed 15 May 2026.
• Apple Support — Monitor your heart rate with Apple Watch — https://support.apple.com/en-us/120277 — accessed 15 May 2026.
• Apple Support — Track your overnight vitals with Apple Watch — https://support.apple.com/en-us/120142 — accessed 15 May 2026.