Polar H10 for Clinical Research: The Reference Standard That Is Not a Medical Device

Yes, the Polar H10 records a real single lead surface ECG at 130 Hz and yields millisecond RR intervals, and it hands the raw waveform to the investigator over an openly documented Bluetooth connection with no account, no cloud and no fee. It is also the device that most consumer wearable heart rate and heart rate variability accuracy claims are measured against.

No, it is not a cleared medical device. It has no FDA clearance, no De Novo grant, no US device listing and no unique device identifier record. Its 0.7 to 40 Hz passband forecloses ST segment, QT interval, P-wave and T-wave morphology by design, which means it also forecloses arrhythmia adjudication.

Those two facts sit together uncomfortably, and any protocol using the H10 has to hold both. That combination is what makes Polar H10 clinical research a different proposition from putting a smartwatch on a cohort. What follows is what it senses, why the field treats it as a reference, how to get data out, and where it stops being appropriate.

What Is the Polar H10, and What Is It For?

The H10 is a chest strap with integrated dry electrodes, a single bipolar chest lead, an on board accelerometer and a coin cell, at USD 104.95 (about EUR 99.90) with no subscription. It is best understood as an instrument rather than a tracker.

It has two siblings worth naming once. The Verity Sense is the optical armband version, which trades ECG for PPG but can record without a phone tethered to it. The H9 shares the same electrical front end but has no dual Bluetooth and no internal memory, so it cannot be used for strap only capture. A protocol should confirm which model a participant actually owns.

What the H10 does not have is the entire layer that defines consumer wearables: no sleep staging, no stress score, no recovery index, no vendor defined composite of any kind. That absence is the point of the device.

Why Is the Polar H10 Called the Gold Standard?

Because the field made it one, not because a regulator did.

An empirical reference standard is a device that enough published studies have used as the criterion arm that it becomes the default comparator. A cleared device is one a regulator has authorised for a stated indication. The H10 is the first and not the second.

The field has written this down. The INTERLIVE Network expert statement (Br J Sports Med, 2021) puts it in two parts: the current gold standard reference method for assessing heart rate is 12-lead ECG, and any commercial chest strap providing RR intervals that has been independently validated and shows excellent agreement can serve as an appropriate criterion measure. The H10 clears that second bar on independent data. It does not thereby become a cleared device.

The reason a chest strap rather than a wrist device holds this position is the sensing modality. ECG measures the electrical depolarisation directly. PPG infers pulse timing optically from a blood volume change downstream, which adds jitter that matters at millisecond resolution and collapses under motion. That ECG versus PPG difference, not the brand, is what the criterion role rests on.

The scale of the first is real. As of August 2026, PubMed returns 144 records for the exact phrase "Polar H10", and a separate title and abstract search pairing that phrase with criterion, reference or gold standard language returns 45. Full text search across PubMed Central returns more than 1,800 articles mentioning it, of which 372 also contain the phrase "gold standard". Recent validations of an Apple earbud heart rate feature, a Galaxy Watch cardiopulmonary exercise test and low cost competitor straps all use the H10 as the criterion arm.

This has a consequence that most device comparisons never state. When any article, including ours, says a wrist device shows a given error for heart rate, the reference in the underlying paper is very often a Polar H10 rather than a 12-lead ECG. "Validated against a Polar H10" is a weaker sentence than "validated against 12-lead ECG", and a protocol should say which one it means. Our guide to wearable data quality and validation works through how to read those studies properly.

Polar's own copy calls the strap "widely regarded as the gold standard in wireless heart rate monitoring". That is vendor marketing and should be read as such, not as evidence.

Does the Polar H10 Record Real ECG?

Yes, with precise limits. Polar H10 ECG is a genuine single lead surface recording, not a derived estimate: 130 Hz plus or minus 2%, 24-bit signed values in microvolts, band limited to 0.7 to 40 Hz.

That specification is an R-peak detection front end and nothing more. A 0.7 Hz high pass corner distorts the ST segment. A 40 Hz low pass corner and a 130 Hz sampling rate leave insufficient resolution for QT interval measurement or reliable P-wave and T-wave morphology.

So the honest answer to the question study teams actually ask is: the H10 cannot detect or adjudicate arrhythmia. It gives you beat timing of exceptional quality and it gives you no diagnostic waveform.

This is the mirror image of what a cleared smartwatch ECG feature does. An Apple Watch or WHOOP MG ECG app runs a cleared classification algorithm on a 30 second recording and returns a result. WHOOP withholds the underlying signal entirely. Apple releases the 30 second ECG voltage samples through HealthKit but never the continuous PPG. The H10 runs no classifier at all and hands you everything, continuously.

How Accurate Is the Polar H10?

Polar H10 accuracy is outstanding in the population it has been tested in, and that population is narrow. The same caveat applies to Polar H10 HRV, where the headline agreement figures cover time domain metrics only.

Three things follow, and the third is the one most often missed.

First, the headline figures are real. Sub millisecond RR accuracy against a criterion ECG is exceptional and justifies the strap's status. The study that started the reputation, Gilgen-Ammann and colleagues in 2019, reported 99.4% RR signal quality against a Holter's 89.8% during high intensity activity, in ten participants. Polar's own engineering document still cites that single ten person study as its accuracy reference.

Second, that accuracy degrades in patients. Error roughly quintuples moving from healthy adults to a cardiac rehabilitation cohort. Every anchor validation used healthy, largely young participants, and the largest had 43 people in it. If your cohort is older, arrhythmic or clinically unwell, you are extrapolating.

Third, accuracy does not transfer across derived metrics. Sub millisecond beat timing does not mean sub millisecond quality nonlinear HRV. Detrended fluctuation analysis alpha-1 shows limits of agreement spanning roughly plus 58% to minus 41% at high exercise intensity, from the same RR series that produced near perfect timing. Validate the metric you will actually report, not the one upstream of it.

How Do You Get Data Out of the Polar H10?

Three paths exist and they yield very different things.

Four operational consequences worth designing around:

• There is no offline raw capture. If your protocol needs continuous raw ECG without a tethered phone, the H10 is the wrong device. That is a device selection decision, not a configuration one.
• AccessLink is a different pipe with a different consent model. It runs over a consumer Polar Flow account, which means the participant's personal account becomes part of your data chain. Its completeness and processing are undocumented.
• Polar Sensor Logger is what most academic teams actually use, and it is unofficial. Android only, single maintainer, no validation dossier, no audit trail, no 21 CFR Part 11 posture. Fine for feasibility. Not suitable as unqualified regulatory grade source data.
• SDK Mode is the reason to build the direct path. With the on device algorithms disabled and your own version locked R-peak detector in the analysis chain, there is no vendor black box between the electrode and the endpoint. That is a rare and genuinely valuable property, and it is worth the engineering cost when the endpoint depends on it.

Whichever path you choose, the study still has to carry consent, symptom context and adherence around it. WeGuide's integration engine ingests the resulting files alongside the ePRO the same participant is completing. What WeGuide does not do is the Bluetooth side raw capture app, or the beat detection and HRV derivation that turns a waveform into an endpoint. Those stay with you or your engineering partner. We carry the participant facing layer around them.

Is the Polar H10 FDA Cleared?

No, and the answer is unusually clean. Searches of the FDA 510(k), De Novo, registration and listing, and unique device identifier databases return nothing for Polar Electro: no clearance, no De Novo grant, no US device listing, no identifier entry. Polar holds FCC and ISED radio authorisations for the strap, and a radio authorisation is not a device authorisation.

Polar's own H10 manual states that the device is not for medical purposes. The separate Polar BLE SDK licence goes further, disclaiming life critical and medical use of the licensed materials and capping Polar's liability at 50 euros. That pairing is what a sponsor's quality function will stop on. In Europe the picture matches: as of 13 August 2026 Polar Electro has no EUDAMED registration as a medical device manufacturer and no device registered under the Medical Device Regulation. The H10 is CE marked as radio equipment, which is a different conformity route entirely.

This means "Is the Polar H10 FDA approved?" has a stranger answer than usual. Neither "approved" nor "cleared" applies, because neither process was ever entered. It is a consumer product used as a scientific instrument, and a protocol relying on it should say exactly that rather than borrowing regulatory language it has not earned. Cleared, granted and approved are three different FDA outcomes, and none of the three applies here.

Polar H10 Versus the Consumer Wearables

The pattern is the single most useful thing to know when choosing between these devices: every device in this category with regulatory clearances withholds its continuous raw signal, and the one device that hands over the raw signal has no clearances. Apple is the partial exception, and only for the 30 second on demand ECG, never for continuous PPG. No product gives you unrestricted raw signal and a clearance together.

That forces a real choice. If your endpoint needs a defensible, version locked derivation you control, the H10 is the instrument. If it needs a cleared classification you can point a regulator at, it is not, and the Samsung Galaxy Watch or Apple Watch are the honest answers. If it needs multi day passive wear, a chest strap will not survive contact with adherence and the Oura Ring or a wrist device is the better bet. The framework for choosing between them is in our guide to wearables in clinical trials, and the full head to head sits in consumer wearables in clinical research.

When to Choose the Polar H10, and When Not To

That last row deserves emphasis. The commercial wrapper is thin in both directions. There is no subscription, no API fee and no per seat cost. Equally, Polar publishes no data processing agreement, no business associate agreement or HIPAA posture, no service level agreement, no sandbox and no research contract vehicle. None of that is a refusal, it's an absence of published terms, so confirm in writing with Polar before assuming there is anything procurement can sign.

Running an H10 Study in Practice

1. Provision the straps yourself. This is not a bring your own device candidate: almost nobody already owns an H10, and those who do may own an H9 or an older firmware.
2. Design around the Bluetooth tether. A dropped link is a lost recording, not a degraded one. Decide in advance what a failed session means for the analysis.
3. Plan the refurbishment loop. Straps are consumable, electrodes degrade, and reverse logistics is the part everyone budgets last.
4. Write consent for what is actually collected. Continuous single lead ECG is a richer signal than "heart rate", and it can surface findings nobody planned for. The consent should name the signal, the collection window, what happens to incidental observations, and what deletion means. This is exactly the passive sensing problem our eConsent layer is built for.
5. Capture the context the waveform cannot supply. Symptoms, caffeine, medication timing and sleep at each session boundary are ePRO questions, not sensor outputs, and without them the signal is much harder to interpret.
6. Prespecify and version lock the processing chain. R-peak detector, artefact correction rules, and the exact HRV metric definitions, before enrolment opens.

Polar H10 Clinical Research FAQs

Can Polar H10 measure ECG?

Yes. It records a genuine single lead surface ECG at 130 Hz, 24-bit, in microvolts, and streams it over Bluetooth. The signal is band limited to 0.7 to 40 Hz, which is sufficient for R-peak detection and insufficient for waveform morphology.

Can Polar H10 detect arrhythmia?

No. It runs no arrhythmia classifier, and its passband and sampling rate rule out the morphology analysis that adjudication requires. It can tell you precisely when beats occurred. It cannot tell you what kind of beats they were.

Is Polar H10 FDA approved?

Neither approved nor cleared. The FDA 510(k), De Novo, device listing and unique device identifier databases hold no Polar Electro record for the H10. Polar holds FCC and ISED radio authorisations only, and its own manual states the device is not intended for medical purposes.

Is Polar H10 more accurate than Apple Watch?

For beat timing, yes, which is why the H10 is routinely used as the criterion arm in studies that measure wrist devices rather than the other way round. For cleared arrhythmia classification the comparison does not apply, because the H10 offers none.

What is the most accurate device to measure HRV?

For time domain HRV in healthy adults during bounded sessions, the H10 is the practical answer among accessible devices. Be careful with nonlinear metrics: the same RR series that produces near perfect timing can produce very wide limits of agreement on detrended fluctuation analysis at high intensity.

Does the Polar AccessLink API give raw ECG?

No. AccessLink exposes an RR interval series inside exercise samples and continuous heart rate at 5 minute resolution. Raw ECG is available only over the direct Bluetooth path using the Polar BLE SDK.

What is Polar Sensor Logger?

An unofficial Android application, widely used by academic teams, that records raw streams from Polar sensors. It is maintained by a single developer, has no validation dossier and no audit trail, and should be treated as a feasibility tool rather than as regulatory grade source software.

How do you export Polar H10 data?

Three routes: direct Bluetooth capture with the Polar BLE SDK for raw ECG and RR intervals, the strap's internal memory for 1 Hz heart rate from a single session, or the AccessLink cloud API for RR series inside exercise samples. Only the first gives you the waveform.

The Polar H10 Verdict for Study Teams

The H10 is an academic instrument, not a registrational one, and its own trial record says so. Of the 213 registered studies naming it, 181 are phase not applicable, 208 have an academic lead sponsor, and the median enrolment is 44.

Read that as a fit statement rather than a criticism. For method development, device validation, physiology work and any endpoint where you need to own the entire derivation chain, nothing else in this price range comes close. For anything that needs a cleared classification, multi day passive adherence, or a vendor you can sign a contract with, it is the wrong instrument and the fact that it is the field's yardstick will not change that.

The yardstick isn't certified, and both halves of that sentence belong in your protocol.