In large studies, people with lower heart rate variability (HRV) died earlier and had more heart attacks and strokes, on average, than people with higher HRV. That is a link across groups of people, measured with medical ECG ECG A recording of the heart's electrical activity, taken with electrodes on the skin. ECG-based RR intervals are the reference standard for measuring HRV. More about ECG recordings; it cannot tell you how long you will live, and the HRV number in your app is not a test for heart disease.
What the mortality studies measured
Almost all of this research used an ECG, the electrical recording of the heart, worn for hours. Researchers then calculated HRV the way the 1996 Task Force standards set out. Those standards give separate methods for short recordings of about 5 minutes and long recordings of 24 hours, and they define the two measures you see most often:
- SDNN SDNN The standard deviation of all NN intervals in a recording, in milliseconds. It measures overall variability, and its value depends strongly on how long the recording lasts. More about SDNN , the standard deviation of all normal beat-to-beat intervals. It reflects all the influences on heart rate over the recording, so it grows with recording length.
- RMSSD RMSSD Root mean square of successive differences: an HRV metric built from how much each heartbeat interval differs from the next. It mainly reflects vagal activity. More about RMSSD , the square root of the mean squared differences between successive normal beats. It mainly reflects the vagus nerve Vagus nerve The main nerve of the parasympathetic nervous system, running from the brainstem to the heart, lungs and gut. Its fast signals to the heart drive most beat-to-beat HRV. More about Vagus nerve , the brake that slows your heart at rest.
SDNN: The standard deviation of all NN intervals in a recording, in milliseconds. It measures overall variability, and its value depends strongly on how long the recording lasts. More about SDNN
Most wearables estimate HRV with a light sensor in a ring or watch, during sleep or in a short morning reading (see HRV by device). That is a different recording from a 24-hour ECG, so the numbers and cut-offs in these studies do not carry over to your app (Task Force, 1996).
The key studies
| Study | Who | HRV measure | Follow-up | Main finding |
|---|---|---|---|---|
| Kleiger, 1987 | 808 heart attack survivors | SDNN from a 24-hour ECG, about 11 days after the heart attack | Average 31 months | Death was 5.3 times as likely with SDNN under 50 ms as with SDNN over 100 ms |
| Tsuji, 1994 (Framingham) | 736 older adults, average age 72 | Several measures from 2 hours of ambulatory ECG | 4 years | 74 people died; each 1 standard deviation lower LF power LF power The share of heart rate variability in slow rhythms of 0.04 to 0.15 Hz, about one cycle every 7 to 25 seconds. Once read as sympathetic activity, which is now disputed. More about LF power went with 1.70 times the hazard of death |
| Hillebrand, 2013 | 8 studies, 21,988 people without known heart disease | SDNN, LF and HF HF power The share of heart rate variability in fast rhythms of 0.15 to 0.40 Hz, the range of normal breathing. It mainly reflects vagal activity when breathing is 9 to 24 breaths a minute. More about HF power | Varied | Lowest vs highest SDNN: 1.35 times the risk of a first heart attack, stroke or other cardiovascular event |
| Jarczok, 2022 | 32 studies plus 2 individual data sets, 38,008 people | Eight HRV measures | Varied | Lower HRV predicted higher death rates in healthy and patient groups; lowest quarter of 5-minute RMSSD vs the rest: 1.56 times the hazard |
After a heart attack
The best-known early study came from Kleiger and colleagues (1987). Of everything measured on the 24-hour recordings, HRV had the strongest link with later death. It stayed a significant predictor after the researchers adjusted for other clinical factors, including how well the heart pumped. The authors noted that HRV is also lower in severe coronary disease, heart failure, diabetic nerve damage and ageing. So low HRV can partly reflect illness that is already there.
In the general population
Tsuji and colleagues (1994) re-analysed ECG tapes from original members of the Framingham Heart Study, a long-running US community study. People with irregular rhythms, many extra beats or anti-arrhythmia drugs were excluded. Lower HRV predicted death over four years even after adjusting for other risk factors. Low-frequency power was the strongest single measure.
Pooled evidence
A 2013 meta-analysis by Hillebrand and colleagues pooled eight studies of people without known heart disease. Low HRV was linked with a 32% to 45% higher risk of a first cardiovascular event, depending on the measure. In their model, each 1% higher SDNN went with about 1% lower risk.
The largest pooled analysis so far, by Jarczok and colleagues (2022), covered 38,008 people. Lower HRV predicted higher death rates across ages, sexes, continents, health status and recording lengths, and most HRV measures gave similar hazard ratios.
Relative risk vs absolute risk
Every number above is a relative risk: how much more often something happened in one group than in another. It does not tell you how likely it is to happen to you. That is the absolute risk, and it depends on how common the event is to begin with.
Here is a made-up example to show the difference. Say that in a group of 100 people like you, 4 would have a heart attack or stroke over the next ten years. A 35% higher relative risk would raise that to about 5 in 100. The headline sounds large, but the change is about 1 person in 100, and 95 of the 100 would still have no event.
Real numbers make the same point. Even in the older Framingham group, 74 of 736 people died over the four years, so about 9 in 10 were still alive at the end, whatever their HRV.
To estimate absolute risk, doctors use tools built from large groups of people. The European SCORE2 model, for example, estimates 10-year risk of heart attack and stroke from age, sex, smoking, blood pressure and cholesterol (SCORE2 working group, 2021). HRV is not part of it.
Why your number is not a forecast
HRV falls with age. In 24-hour recordings from 260 healthy people aged 10 to 99, RMSSD in people in their 50s was about 47% of the value in people aged 10 to 19, and it levelled off after that. Among healthy people over 65, 12% had RMSSD below cut-offs that earlier studies had linked to higher death risk (Umetani and colleagues, 1998). A low number can be normal for your age.
HRV moves for everyday reasons. Sleep, alcohol, hard training, stress and a coming cold can all pull it down for a while. The Kleiger and Framingham studies used one recording per person, which cannot tell a short dip from a lasting level. A low week in your app is not a signal about your lifespan. See why HRV changes so much.
A link is not a cause. These studies show that lower HRV and higher risk go together. None tested whether raising HRV lowers risk. Low HRV may partly be a sign of disease already present, as the 1987 study noted.
Irregular beats can mislead. HRV methods are meant to use only normal beats, and the Framingham analysis left out people with many extra beats. Irregular beats distort the result, so a wearable number can be off in either direction. Can HRV detect heart problems? explains why.
What health age scores do and do not mean
Many apps show a “heart age”, “fitness age” or “cardiovascular age”. These scores take one estimate from your device and compare it with typical values for people of your age.
They are not all built on HRV. Oura says its Cardiovascular Age is based on an estimate of pulse wave velocity, a measure of artery stiffness, taken from the shape of the pulse signal, and that the ring is not a medical device. Garmin says its fitness age is based on your VO2 max estimate, and newer watches also use activity intensity, resting heart rate Resting heart rate How many times your heart beats per minute while you are fully at rest. For adults it is usually 60 to 100 beats per minute, and lower in well-trained athletes. More about Resting heart rate and body fat or BMI.
These scores can give you a simple trend to follow and a nudge to stay active. They cannot diagnose disease, rule it out, or predict your lifespan. The mortality studies above did not test any of them. Treat a change over months as a prompt to look at your habits, not as a verdict.
What to do with this
- Follow your own trend, measured the same way, and compare it with your own baseline Baseline Your own usual HRV level and range, built from weeks of readings taken the same way. Comparing a reading with your baseline says more than comparing it with other people. More about Baseline rather than with study cut-offs. The normal HRV by age page shows typical ranges.
- If you want to lift your HRV, how to increase HRV lists what the evidence supports. Regular exercise is the best-supported habit, and it is good for your heart whatever it does to your number.
- If you want to know your heart risk, ask a doctor. Blood pressure and cholesterol are part of the risk tools doctors use.
When to see a doctor
Your symptoms matter more than your HRV. Call emergency services for chest pain, fainting, shortness of breath, unusual sweating or dizziness. Book a prompt appointment if you often feel extra beats, notice new or different palpitations, or have a resting pulse over 100 that exercise, anxiety or fever does not explain. Do the same if you have heart disease or risk factors such as high blood pressure, diabetes or high cholesterol (MedlinePlus). Bring your HRV trend if you like, but let the doctor decide what tests you need.