Heart rate variability: a 5-minute ECG, not a wrist reading

Lower HRV was associated with higher death on a 5-minute ECG. That is not overnight wrist RMSSD.

Where you stand

No all-cause figure for overnight consumer-wrist RMSSD has been found. This page is the mismatch between what was measured and what a person can enter.

Heart Rate Variability — what each study measured, and what your watch measures

What heart rate variability studies measured, versus what your own watch measures, by instrument, measurement window, index, endpoint and result.
Instrument	Study, endpoint and result
ECG 5 min · RMSSD	Jarczok et al. 2022 all-cause mortality HR 1.56 (1.32-1.85), unadjusted
ECG rhythm strip (ARIC) 2 min · HRV	Dekker et al. 2000 all-cause and CHD mortality no ratio in our record
fingertip PPG SDNN	Saito et al. 2025 all-cause mortality HR 1.51 (1.15-1.98)
ECG (Rotterdam) 10 sec · SDNN	de Bruyne et al. 1999 cardiac mortality HR 1.8 (1.0-3.2)
— device not in our record · LF power	Tsuji et al. 1994 all-cause mortality HR 1.70 (1.37-2.09)
chest-worn ECG (ActiHeart) multiday · SDNN proxy	Schaarup et al. 2025 all-cause mortality incidence rate ratio 0.79 (0.66-0.94)

wrist PPG (consumer wearable) overnight · RMSSD	your watch — no study on record

What it is

Beat-to-beat variation in heart rhythm. The evidence on this page is a 5-minute ECG. That is not overnight wrist RMSSD.

How it is modeled (not scored)

Technical notes

Across 32 studies covering over 38,000 people, lower HRV predicted higher mortality. The headline HR 1.56 is the UNADJUSTED comparison of the lowest quartile of 5-minute ECG RMSSD versus the other three quartiles — not wearable nightly HRV. As of 2026-08-29 this calculator's curve uses the paper's smaller covariate-adjusted estimate (~1.23) instead. The measurement mismatch remains: users enter overnight wearable RMSSD. Saito 2025 (fingertip PPG, 5-min sitting rest) and Schaarup 2025 (ActiHeart chest ECG, multi-day SDNN) are nearby but not that construct — no all-cause HR for overnight consumer-wrist RMSSD has been found.

How to improve it

Improve sleep quantity and regularity
Train aerobically but respect recovery
Practice slow breathing; limit late alcohol

What the studies found

Endpoint	Grade	Finding
All-cause mortality (RMSSD lowest quartile)	C	Bottom Quartile (Jarczok 2022) Technical notes In unadjusted analyses, the lowest quartile of 5-minute ECG RMSSD was associated with all-cause mortality HR 1.56 (95% CI 1.32–1.85) versus the other three quartiles.
All-cause mortality (overnight wrist RMSSD)	X	Not overnight wrist RMSSD Technical notes Jarczok 2022 is 5-minute ECG. Saito 2025 is 5-minute fingertip PPG; Schaarup 2025 is chest-ECG multi-day SDNN. No all-cause HR for overnight consumer-wrist RMSSD has been found.
All-cause mortality (per 1-SD lower LF power)	B	1.70x per SD of LF Power (Tsuji 1994) Technical notes Framingham: in 736 elderly adults (mean age 72) followed 4 years, each 1-SD lower low-frequency HRV power carried all-cause mortality HR 1.70 (1.37–2.09), independent of traditional risk factors — a different metric again (LF power), and different from the overnight wearable RMSSD users actually enter.
Cardiac mortality (SDNN quartile, U-shaped)	C	U-Shaped, Not Just Low HRV (de Bruyne 1999) Technical notes Rotterdam Study measured heart-rate variability (SDNN) via a standard 10-second ECG in over 5,000 adults 55+: the lowest SDNN quartile carried 80% higher cardiac mortality than the middle quartile (HR 1.8, 95% CI 1.0–3.2) — but so, unexpectedly, did the highest quartile (HR 2.3, 95% CI 1.3–4.0), a U-shaped pattern the authors said needed further study.

What argues against this

No source in this set reverses the direction of the low-HRV-worse association outright, but de Bruyne 1999 (Rotterdam Study, >5,000 adults 55+) complicates the curve's implicit monotonic framing: the lowest SDNN quartile carried 80% higher cardiac mortality (HR 1.8, 1.0–3.2) versus the middle quartile, but so did the highest quartile (HR 2.3, 1.3–4.0) — a U-shaped pattern suggesting very high HRV is not straightforwardly protective either. It measures cardiac mortality via 10-second ECG SDNN, not the all-cause mortality / 5-minute RMSSD construct Jarczok 2022 and this calculator use, so it disagrees on shape and construct, not sign.

Last reviewed 12 October 2026

Sources

6 cited

Jarczok et al. (2022) — Neuroscience & Biobehavioral Reviews
Heart rate variability in the prediction of mortality: a systematic review and meta-analysis
View source
Dekker et al. (2000) — Circulation
Low heart rate variability in a 2-minute rhythm strip predicts risk of coronary heart disease and mortality from several causes: the ARIC Study. Atherosclerosis Risk In Communities
View source
Saito et al. (2025) — Circulation Reports
Pulse Rate Variability and All-Cause Mortality Risk in the General Japanese Population
View source
de Bruyne et al. (1999) — American Journal of Epidemiology
Both decreased and increased heart rate variability on the standard 10-second electrocardiogram predict cardiac mortality in the elderly: the Rotterdam Study
View source
Tsuji et al. (1994) — Circulation
Reduced heart rate variability and mortality risk in an elderly cohort. The Framingham Heart Study
View source
Schaarup et al. (2025) — Diabetes, Obesity and Metabolism
Cardiovascular autonomic dysfunction precedes cardiovascular disease and all-cause mortality: 11-year follow-up in the ADDITION-PRO study
View source

Not part of your score

Heart Rate Variability is tracked here for its research. It does not count toward the 8-marker MarkerScore. See what does in the calculator.

Open calculator

Educational content only — not medical advice, diagnosis, or treatment. Not a medical device; not FDA evaluated. Consult a qualified healthcare professional about your own health, and call emergency services for urgent symptoms. Full medical disclaimer →