How to Measure HRV: ECG vs PPG and a Practical Protocol

Learn how to measure HRV consistently, compare ECG-derived HRV with PPG-derived pulse variability, and interpret RMSSD and SDNN without universal cutoffs.

Heart rate variability, or HRV, describes variation in the time between consecutive heartbeats. It is not the variation in heart rate from minute to minute, and it is not a score that should be as high as possible.

Measure HRV from reliable beat-to-beat intervals, under repeatable conditions, and compare the same metric with your own baseline. A five-minute resting ECG is a widely used reference protocol. PPG wearables can be convenient for trends at rest, but pulse-rate variability derived from PPG is not automatically interchangeable with ECG-derived HRV.

How to measure HRV consistently

1. Use the same device and the same metric each time.
2. Measure at a similar time of day, before exercise, caffeine or other factors that change from day to day.
3. Use the same posture, usually seated or lying down, and settle quietly before recording.
4. Record for five minutes when practical. Measurements of different lengths are not equivalent.
5. Stay still and breathe normally unless your protocol tells you to control breathing.
6. Check signal quality and correct artifacts or abnormal beats with a defined method.
7. Compare the result with your own rolling baseline, not a universal table from the internet.

Consistency doesn't remove every source of variation, but it makes changes easier to interpret. Posture, breathing, time of day, recent exercise, sleep, alcohol, illness, medication and recording artifacts can all affect the result.

ECG vs PPG for HRV

PPG can agree well with ECG in some resting protocols and devices, but that doesn't make every optical reading equivalent to HRV. The pulse reaches the wrist or finger after the heart's electrical activation, and that transit time can vary. Recent studies therefore distinguish pulse-rate variability, or PRV, from ECG-derived HRV.

PPG is still useful. The mistake is to mix methods, devices, body positions, durations or metrics, and then read the difference as a physiological change.

What ECG vs PPG studies actually found

The size of the difference depends on the device and the protocol. Three comparisons show the range.

At rest, posture changed the agreement

In 31 healthy adults, the Polar OH1 optical sensor was compared with the Polar H10 chest ECG during five-minute measurements lying down and sitting. Lying down, agreement was strong: ICC 0.955 for RMSSD and 0.980 for SDNN. Sitting, it fell to 0.834 and 0.921. Mean biases ranged from −2.1 to −8.1 ms.

ICC describes agreement across people, not the error of one reading, so a high ICC can coexist with a bias of several milliseconds. For tracking over time, the strongest setup is the one you can repeat with the same sensor, posture, duration and processing method.

PPG read lower than ECG

In 931 adults across healthy and disease groups, ECG and finger PPG were recorded at the same time. PPG systematically underestimated RMSSD and SDNN. The mean SDNN difference was 8.50 ms in the cardiovascular group and 11.84 ms in the neurological group, so a PRV value shouldn't be dropped into an ECG-HRV reference range as if the methods were identical.

Sampling rate matters for beat timing

In 54 adults standing at rest, PPG was sampled at different rates and compared with ECG-derived intervals. About 100 to 200 Hz was needed to keep the average PRV bias below 2%. At 40 to 50 Hz, the signal met only a much wider 20% limit. When precise beat-to-beat timing matters, sampling rate is not a minor specification.

What are RR and NN intervals?

R-R intervals are measured between successive R peaks. HRV analysis commonly uses corrected normal-to-normal intervals.

An ECG shows a prominent QRS complex for each ventricular depolarization. The time in milliseconds between successive R peaks is the RR interval.

Raw RR intervals can include missed detections, movement artifacts, premature beats or other rhythm changes. Many HRV analyses therefore use NN intervals: intervals between normal beats after a documented correction process. A metric calculated perfectly from uncorrected artifacts can still mislead.

For research, keep the raw ECG when possible. It lets you check whether an unusual HRV result came from physiology, poor contact, movement or beat-detection errors. A stream of intervals without the waveform behind it is harder to audit.

RMSSD vs SDNN

RMSSD and SDNN answer different questions and depend on the recording protocol.

RMSSD is the root mean square of successive differences between normal intervals. It emphasizes short-term, beat-to-beat changes and is commonly used for short resting measurements.

SDNN is the standard deviation of normal intervals across the recording. It reflects all the variability present in that recording and depends strongly on its duration, so a five-minute SDNN and a 24-hour SDNN are not comparable reference values.

Take four normal NN intervals: 1000, 980, 1020 and 1000 ms. Their mean is 1000 ms, and the successive differences are −20, +40 and −20 ms.

Applying the time-domain formulas:

RMSSD is built from the changes in the bottom row, SDNN from each interval's distance to the mean, so the same four intervals give two different numbers.

Use the same metric for repeated comparisons. Don't convert one into the other with a generic threshold, and check whether an app reports a raw value, a logarithmic transformation, a score or a personalized deviation.

What is a good HRV?

There is no single good HRV value for every person. Age, physiology, health, training background, measurement method, posture, duration and data processing all influence the number.

A result such as 40 ms can't be called good or poor without knowing whether it is RMSSD or SDNN, how long the recording lasted, how artifacts were handled and how it compares with that person's usual values.

Build a personal baseline from measurements taken under similar conditions, and look for a repeated change rather than reacting to one morning. HRV can add context to recovery or stress tracking, but it doesn't identify the cause of a change and doesn't diagnose overtraining, illness or a heart condition.

When is HRV useful?

HRV is most useful when the measurement feeds a clearly defined comparison:

• following a personal recovery trend under consistent conditions;
• comparing phases of a research protocol;
• studying how beat timing relates to breathing, sleep or activity;
• checking whether a change stays visible across several measurements.

It is less useful when devices and protocols are mixed, signal quality is hidden, or every short-term fluctuation triggers a conclusion.

Measuring HRV with Aidlab 1

Aidlab 1 records a 1-lead ECG from the chest and derives heart rate, RR intervals and HRV. It also records breathing, skin temperature, movement and body position, so a researcher or advanced user can review several synchronized signals from the same session.

Aidlab 1 is a wellness and research device, not a medical device, and its ECG and HRV results are not a diagnosis. Aidlab 2 is a separate next-generation wellness and research system. Aidmed One is a separate medical product for professional workflows.

For measurement during exercise, read chest strap vs wrist heart-rate monitor. To compare available signals and workflows, see Aidlab vs Polar H10.

Common questions

How long should an HRV measurement last?

Five minutes is a widely used duration for comparable short-term HRV recordings. Shorter protocols exist, but their validity depends on the metric, device, population and purpose.

Should HRV be measured while sitting or lying down?

Either posture works in a defined protocol, but posture changes HRV. Choose one position and keep it the same across measurements.

Is higher HRV always better?

No. HRV should be read relative to the person, metric, recording length and context. Very high or abrupt values can also come from artifacts, abnormal beats or analysis errors.

Can a smartwatch measure HRV?

Many watches estimate variability from optical pulse data, and some record short single-lead ECGs. Check which signal, metric, duration and artifact handling the device uses. Optical PRV may track ECG-derived HRV in some resting conditions, but the methods are not universally interchangeable.

Sources

• Task Force standards for HRV measurement and interpretation
• Recommendations for HRV experiment planning and reporting
• Pulse-rate variability is not the same as ECG-derived HRV
• Review of PPG pulse-rate variability against ECG
• Polar OH1 PPG compared with Polar H10 ECG in 31 adults
• PPG sampling rate and pulse-rate variability bias
• Aidlab intended use and safety