Put a Garmin on one wrist and a clinical ECG on the other, and the heart rate number comes back almost perfect. The HRV number does not.
That is the short version of a new peer-reviewed study out of the Sports Research Centre at Universidad Miguel Hernandez in Elche, Spain, working with the Institute for Health and Biomedical Research of Alicante. The team tested the Garmin Forerunner 955, specifically its Elevate v4 optical sensor, against a Biopac ECG reference in 23 healthy adults. The paper landed in the journal Sensors on October 14, 2026, under DOI 10.3390/s26206549.
For resting heart rate, the watch basically nailed it. For heart rate variability, the sensor measured something real, but not the same thing the ECG measured. And that gap matters if you are using Garmin's recovery features to decide whether to train hard tomorrow.
How the test worked
The setup was deliberately simple and controlled. Participants wore the Forerunner 955 and the ECG at the same time, recording short windows while lying down and again while seated, breathing to a metronome so the data were comparable. Then the researchers compared two things:
- Resting heart rate from the watch against the ECG
- rMSSD, the heart rate variability metric, from the watch's optical pulse data against rMSSD derived from the ECG
rMSSD is the number Garmin reports as HRV. It is the same basic metric Oura, Whoop, Polar and Apple report in some form, and it is the input behind a lot of the recovery advice you see on the wrist.
Where the watch shines
Resting heart rate is the good news, and it is not close. The optical sensor showed what the authors describe as a near-perfect linear association with the ECG, with correlation coefficients of at least 0.94, plus substantial concordance. Translation: when you look down at your resting heart rate on a Garmin, that number is trustworthy.
That lines up with what most long-term Garmin owners already suspect. Averaged resting heart rate is easy for optical sensors because the signal is slow, steady and strong, and the watch has all night to sample it.
Where it falls apart
HRV is the harder problem, and the study is blunt about it. The watch's pulse-derived rMSSD had good to excellent test-retest reliability within a session, which means it was consistent with itself. But agreement with the ECG was limited. The paper reports a significant proportional bias and wide limits of agreement, with relative measurement errors ranging from -43% to +52% in the supine position and -41% to +118% in the seated position in isolated individual cases.
Read that second range again. In the worst seated cases, the watch reported roughly double the HRV the ECG measured, or less than two thirds of it. Direction was not consistent either, which is the part that ruins comparisons between people or between devices.
The authors do not hedge in their conclusion. The Garmin Elevate v4 is, in their words, "highly reliable for basic HR tracking and within-subject PRV trends, but it lacks the absolute accuracy required to replace reference ECG-derived HRV for individual resting autonomic evaluation."
Why this matters for your Garmin
Garmin does not present HRV Status as a lab-grade measurement, and this study is a decent argument for why. The feature works off a personal baseline: your nightly reading gets compared with your own rolling range, and you get told whether you are balanced, unbalanced or low. That kind of relative tracking is where the sensor performs reasonably well, per this paper.
The failure mode is the one users fall into anyway. Comparing your rMSSD against a friend's, against a Whoop, against an Oura, or against a research paper's number, is not a comparison the hardware can currently support. If your HRV Status drops for a week, that is a trend worth acting on. If your absolute number is 62 and someone on a forum says theirs is 84, that difference tells you almost nothing.
Garmin's own documentation leans the same way, describing HRV Status as a personal trend tool that needs a baseline period of consistent overnight wear before it says anything useful.
The caveats you should hold onto
This is one study, with real limits. Twenty-three participants is a small sample. The recordings were short and lab-based, in two fixed postures with paced breathing, not a night of normal sleep on your own arm. Conditions like that favor clean signal capture, so the night-long picture in the real world could be better or worse.
The sensor tested is also a generation behind. The Forerunner 955 runs Elevate v4, while newer Garmin watches such as the Fenix 9 and Forerunner 970 ship with a newer optical generation. Nothing in this paper tells you how those perform, and Garmin has not published an equivalent validation either way.
One more practical note: the paper is an early access version on Sensors, so the complete HTML and PDF are still pending while the abstract and metadata are live. The numbers above come straight from that published abstract and the article record, not from a press summary.
If absolute HRV values are the point of your training, a chest strap remains the honest answer. A Garmin HRM chest strap measures the electrical signal directly rather than inferring it from blood flow, which is the same reason the lab used an ECG and not a watch. For everything else, the watch is doing more than most people give it credit for.
What to do with it
Nothing on your wrist needs to change today. Wear the watch, keep your overnight HRV logging consistent, and treat the number as a direction, not a measurement. The pattern over weeks is the signal. The absolute value is not.
That is also the honest read of the wider picture. Independent teams keep publishing validation work on wrist sensors, and the results keep landing in the same place: heart rate has quietly become a solved problem on a good watch, while HRV is still a noisy estimate dressed up as a clinical metric. We saw the same split in the independent testing roundups that followed the Apple Watch's heart rate claims in our look at Apple's heart rate accuracy study, and Garmin's own firmware history shows how fragile the sensor pipeline can be, as in the Fenix 8 optical heart rate failure that 23.40 finally fixed.
If you are in the market and this study is what pushes you toward a newer sensor generation, the Forerunner 955 on Amazon is still the exact model that was tested here, which makes it the useful reference point even though newer hardware has moved on.
Bottom Line
Garmin's optical heart rate sensor is genuinely excellent. The study puts resting heart rate at a near-perfect match against a clinical ECG, which is a strong result for a device you wear on your wrist all day. Its HRV reading is the weak link, with relative errors that reached beyond 100% in some seated cases and no consistent direction of error.
Use HRV Status the way Garmin describes it, as your own trend line, and it will serve you fine. Use it as a lab measurement and the hardware is not there yet.
Sources: Sensors (MDPI), 2026, 26(20), 6549, DOI 10.3390/s26206549, published October 14, 2026, Sports Research Centre, Universidad Miguel Hernandez and ISABIAL; Garmin HRV Status technology and support documentation. The study's full text is still in early access, so figures here are drawn from the published abstract and article record. No second independent write-up of the study was available to cross-check against at the time of publication.
