Research Wearable Comparison.

Compare 17 wearables and home devices used in clinical research by sensors, raw data access, data routes, battery, and regulatory status.

Filter by what your protocol needs, compare up to three devices side by side, and open any device for its full record with sources. Figures are the makers’ published specifications, not our test results.

Data checked on 20 October 2026. 17 devices from 9 makers, grouped by maker and not ranked.

Devices

17 devices

Apple Watch

Series 12, Ultra 4 and SE 3

Apple Inc.

SmartwatchConsumer

Worn on the wrist

  • ECG
  • Heart rate
  • HRV
  • SpO2
  • Sleep stages
  • Steps
Raw sensor data
Yes, from the maker
Battery
18 to 84 hours by model and mode
FDA features (US)
5 cleared or authorised
Phone
iPhone only
Subscription
Not required

WeGuideVia Apple Health

Garmin wrist wearables

vívosmart 5, vívoactive 6, Venu 4 and Forerunner

Garmin

SmartwatchConsumer

Worn on the wrist

  • Heart rate
  • HRV
  • SpO2
  • Sleep stages
  • Respiration
  • Steps
Raw sensor data
Yes, also through WeGuide on supported models
Battery
Up to 21 days by model and mode
FDA features (US)
1 cleared or authorised, on compatible models
Phone
iPhone and Android
Subscription
Not required

WeGuideNative integration

Garmin CIRQA Smart Band

Wrist band with an upper arm band accessory

Garmin

Screenless bandConsumer

Worn on the wrist, upper arm

  • Heart rate
  • HRV
  • SpO2
  • Skin temperature
  • Sleep stages
  • Steps
Raw sensor data
Not confirmed
Battery
Up to 10 days
FDA features (US)
None found
Phone
iPhone and Android
Subscription
Not required

WeGuideVia Apple Health or Health Connect

Google Fitbit Air

Screenless band

Google (Fitbit)

Screenless bandConsumer

Worn on the wrist

  • Heart rate
  • HRV
  • SpO2
  • Skin temperature
  • Sleep stages
  • Steps
Raw sensor data
No
Battery
Up to 7 days
FDA features (US)
1 cleared or authorised
Phone
iPhone and Android
Subscription
Not required

WeGuideSupported

Fitbit Charge 6

Fitness band

Google (Fitbit)

Fitness bandConsumer

Worn on the wrist

  • ECG
  • Heart rate
  • HRV
  • SpO2
  • Sleep stages
  • Steps
Raw sensor data
No
Battery
Up to 7 days
FDA features (US)
2 cleared or authorised
Phone
iPhone and Android
Subscription
Not required

WeGuideVia Apple Health or Health Connect

Google Pixel Watch

Pixel Watch 5 and Pixel Watch 4

Google

SmartwatchConsumer

Worn on the wrist

  • ECG
  • Heart rate
  • HRV
  • SpO2
  • Skin temperature
  • Sleep stages
Raw sensor data
No
Battery
30 to 40 hours, up to 72 with Battery Saver
FDA features (US)
3 cleared or authorised
Phone
Android only
Subscription
Not required

WeGuideVia Health Connect

Samsung Galaxy Watch

Watch9, Watch Ultra2 and Watch8 Classic

Samsung Electronics

SmartwatchConsumer

Worn on the wrist

  • ECG
  • Heart rate
  • HRV
  • SpO2
  • Skin temperature
  • Sleep stages
Raw sensor data
Yes, from the maker
Battery
30 to 60 hours by model
FDA features (US)
3 cleared or authorised
Phone
Android only
Subscription
Not required

WeGuideVia Health Connect

Oura Ring

Ring 5, Ring 4 and Ring 4 Ceramic

Oura Health Oy

Smart ringConsumer

Worn on the finger

  • Heart rate
  • HRV
  • SpO2
  • Skin temperature
  • Sleep stages
  • Steps
Raw sensor data
No
Battery
6 to 9 days (Ring 5)
FDA features (US)
None found
Phone
iPhone and Android
Subscription
Required

WeGuideVia Apple Health or Health Connect

WHOOP

WHOOP 5.0 and WHOOP MG

WHOOP, Inc.

Screenless bandConsumer

Worn on the wrist, upper arm, chest, waist or hip

  • Heart rate
  • HRV
  • SpO2
  • Skin temperature
  • Sleep stages
  • ECG on WHOOP MG
Raw sensor data
No
Battery
14+ days
FDA features (US)
1 cleared or authorised, on WHOOP MG
Phone
iPhone and Android
Subscription
Required

WeGuideVia Apple Health or Health Connect

Polar H10

Chest strap heart rate sensor

Polar Electro Oy

Chest strapConsumer

Worn on the chest

  • Heart rate
  • RR intervals
  • ECG waveform
  • Acceleration
Raw sensor data
Yes, from the maker
Battery
400 hours on a replaceable coin cell
FDA features (US)
None found
Phone
iPhone and Android
Subscription
Not required

WeGuideVia Apple Health or Health Connect

Polar Verity Sense

Optical heart rate sensor

Polar Electro Oy

Optical armbandConsumer

Worn on the upper arm, forearm

  • Heart rate
  • PP intervals
  • PPG waveform
  • Motion sensors
Raw sensor data
Yes, from the maker
Battery
Up to 30 hours of continuous use
FDA features (US)
None found
Phone
iPhone and Android
Subscription
Not required

WeGuideVia Apple Health or Health Connect

Withings ScanWatch

ScanWatch 2, ScanWatch Nova and ScanWatch Nova Brilliant

Withings (France)

SmartwatchConsumer

Worn on the wrist

  • ECG
  • Heart rate
  • HRV
  • SpO2
  • Sleep stages
  • Steps
Raw sensor data
Yes, from the maker
Battery
Up to 35 days
FDA features (US)
1 cleared or authorised
Phone
iPhone and Android
Subscription
Conditional (Withings+)

WeGuideVia Apple Health or Health Connect

Withings BPM blood pressure monitors

BPM Vision, BPM Connect, BPM Pro 2 and BPM Connect Pro

Withings (France)

Blood pressure cuffMedical

Worn on the upper arm

  • Blood pressure
  • Pulse rate
Raw sensor data
No
Battery
6 to 12 months by model
FDA features (US)
1 cleared or authorised, held by Guangdong Transtek
Phone
iPhone and Android
Subscription
Conditional (Withings+)

WeGuideVia Apple Health or Health Connect

Withings under mattress sleep mat

Sleep Rx and Sleep Analyzer

Withings (France)

Under mattress sensor matMedical

Not worn, placed under the mattress

  • Breathing events
  • Heart rate
  • Sleep stages
  • Snoring
Raw sensor data
No
Battery
Mains powered
FDA features (US)
1 cleared or authorised, Sleep Rx, prescription
Phone
iPhone and Android
Subscription
Conditional (Withings+)

WeGuideVia Apple Health or Health Connect

ActiGraph LEAP

Includes LEAP Core

Ametris (formerly ActiGraph), a Signant Health company

Research actigraphResearch grade

Worn on the wrist, upper arm, waist or hip, ankle

  • Raw acceleration
  • Activity counts
  • Steps
  • Sleep measures
  • Heart rate, investigational
Raw sensor data
Yes, from the maker
Battery
Up to 32 days at default settings
FDA features (US)
1 cleared or authorised, prescription
Phone
Optional app, iPhone or Android
Subscription
Research contract

WeGuideNot integrated

ActiGraph CentrePoint Insight Watch

Model CPW01

Ametris (formerly ActiGraph), a Signant Health company

Research actigraphResearch grade

Worn on the wrist

  • Raw acceleration
  • Activity counts
  • Steps
  • Sleep measures
Raw sensor data
Yes, from the maker
Battery
Up to 30 days at default settings
FDA features (US)
1 cleared or authorised, prescription
Phone
Optional app, iPhone or Android
Subscription
Research contract

WeGuideNot integrated

ActiGraph wGT3X-BT

Legacy model

Ametris (formerly ActiGraph), a Signant Health company

Research actigraphResearch grade

Worn on the wrist, waist or hip, ankle, thigh

  • Raw acceleration
  • Activity counts
  • Steps
  • Sleep measures
  • Ambient light
Raw sensor data
Yes, from the maker
Battery
Up to 25 days at 30 Hz
FDA features (US)
1 cleared or authorised, listed under K080545
Phone
Not needed, USB to a Windows PC
Subscription
Research contract

WeGuideNot integrated

  • Specifications are the makers’ published figures, not test results. Battery life is shorter when raw data are captured continuously.
  • Regulatory status is recorded per feature and per region. No device is FDA cleared as a whole, and a cleared feature is not a validated trial endpoint.
  • Validation applies to one metric, population, wear location, and firmware version. A listed metric is not evidence that it suits your endpoint.
  • WeGuide labels: Native integration (Garmin wrist wearables), Via Apple Health, Via Health Connect, or both (summary data that the maker’s app writes to the phone), Supported (Fitbit Air), and Not integrated.

Device record

Device Details

How to Use the Research Wearable Comparison

  1. Write down what you need to measure first: the outcome, the population, how long participants wear the device, and the data resolution your analysis needs. The device follows from that list.
  2. Filter the devices by wear location, device class, raw sensor data, an FDA cleared or authorised feature, how data leaves the device, phone platform, and whether the device works with WeGuide. Chips in one group widen the list, and each extra group narrows it.
  3. Search for a device, a maker, or a metric such as ECG, HRV, or SpO2.
  4. Tick Compare on up to three devices and open the side by side view. On a phone it reads one criterion at a time, with each device underneath.
  5. Open Full details for sensors, sampling rates, data routes, the regulatory record of each feature, validation studies with PubMed links, and the sources behind every field.
  6. Share your shortlist. Filters, compare selections, and an open device record are saved in the page address, so a copied link opens the same view.

Every record was checked against the maker’s current pages, developer documentation, the FDA database, and PubMed on 20 October 2026. Products, firmware, and access terms change often, so check them again at protocol lock. For a wider introduction, read our guide to wearables in clinical trials.

Choose the Measure First, Then the Device

A wearable is fit for purpose only for a defined measure in a defined population. The V3 framework splits that question into verification of the sensor, analytical validation of the algorithm, and clinical validation of the measure in the people you study[1]. FDA guidance on digital health technologies asks sponsors to show that a device is fit for purpose in the proposed population. Its example is a step algorithm that works in healthy participants but may not work for people with Parkinson’s disease who walk with a shuffling gait[2].

Wear location changes the number

In a network meta-analysis of free living studies, wrist and ankle research devices recorded 15% to 30% more daily steps than hip devices, and consumer wrist devices recorded 3% to 15% more[3]. Do not pool step counts across placements without calibration, and write the wear location into the protocol.

Name the endpoint before the device

Our digital biomarkers library lists measures with their evidence, which helps you name the measure before you shortlist devices. The first regulatory qualification of a primary endpoint derived from a wearable, stride velocity 95th centile in Duchenne muscular dystrophy, came from EMA in 2023. It used an ankle worn sensor that is not in this comparison[4].

Getting Data Out of a Wearable

Summary data are values the maker’s algorithm has already computed, such as daily steps, resting heart rate, or sleep stages. Raw data are the sensor signals underneath, such as accelerometer samples or a PPG waveform. Raw data let you rerun one fixed algorithm across devices and firmware versions, but access is usually gated and the files need processing.

Cloud API

The maker’s server sends data after each participant links an account. Google switches off the legacy Fitbit Web API on 30 October 2026, and new builds must use the Google Health API, with every participant giving consent again[5]. Oura’s API caps apps at 10 users until Oura approves them, and it exposes no raw PPG or accelerometer data[6].

Device SDK

Code in your study app reads the device directly. Samsung’s Health Sensor SDK exposes raw accelerometer data at 25 Hz, PPG at 25 Hz continuously or 100 Hz on demand, and ECG at 500 Hz on demand. An app that uses it needs Samsung partner approval[7]. The Polar H10 streams single lead ECG at 130 Hz through the free Polar BLE SDK, over a live Bluetooth link only[8].

Apple Watch

Apple has no server or OAuth API for Apple Watch data. A study app reads HealthKit on the participant’s iPhone with permission for each data type, and it cannot tell whether read access was denied[9]. Raw signals need SensorKit, which requires an Apple approved study and entitlement, holds data for 24 hours before release, and gives access to 7 days of prior data[10].

Phone health stores

Apple Health and Health Connect hold the summaries that each maker’s app chooses to write. Oura does not write HRV to Apple Health[11], and WHOOP writes no HRV to either store[12]. WeGuide reads these stores, which is why most devices on this page are listed as indirect.

Research platforms and exports

ActiGraph LEAP samples acceleration at 32, 64, 128, or 256 Hz and delivers raw files for each participant and day through the CentrePoint V3 API, after upload from a phone app, a cellular hub, or a site PC[13]. Withings offers raw accelerometer and PPG data for ScanWatch only under a research contract, and switching it on disables other watch features and cuts battery life to 3 to 4 days[14].

Garmin

Garmin offers a cloud Health API for business use and two Health SDKs. The Standard SDK keeps data off Garmin servers, the Companion SDK works alongside Garmin Connect, and Garmin notes that some device models may not be supported[15]. WeGuide’s Garmin integration collects raw accelerometer and gyroscope data at 25 Hz on supported models.

Regulatory Status, Feature by Feature

The FDA clears or authorises software features for a named indication, not whole devices. On Apple Watch, the ECG app (De Novo DEN180044, 2018), irregular rhythm notifications (De Novo DEN180042, 2018), AFib History (K213971, 2022), the Sleep Apnea Notification Feature (K240929, 2024), and the Hypertension Notification Feature (K250507, 2025) each have their own record[16]. Apple labels its Blood Oxygen app as wellness only[17].

What a clearance does and does not mean

  • It covers one feature and one indication. Apple’s hypertension notifications were cleared with 41.2% sensitivity and 92.3% specificity in the maker’s study, and they give no blood pressure value[18].
  • It is not a trial endpoint. The only FDA qualification for trial use we found among these devices is Apple’s AFib History, qualified in 2024 as a Medical Device Development Tool for weekly AFib burden as a secondary endpoint in cardiac ablation device studies[19].
  • The holder may not be the brand. Current Withings blood pressure cuffs are marketed under 510(k)s held by the contract manufacturer Guangdong Transtek, and the 510(k) linked to Oura temperature data, K202897, belongs to Natural Cycles[16].
  • It is a US status. Apple says hypertension notifications may not be available on Series 12 or Ultra 4 in some regions, including Australia, while further clearances are in process[20].

Not cleared usually means wellness

FDA’s January 2026 general wellness guidance says FDA may treat noninvasive optical estimates of blood pressure, oxygen saturation, or HRV as general wellness products when they are intended only for wellness, but not when they are intended for “screening, diagnosis, monitoring, alerting, or management of a disease”[21]. WHOOP’s ECG feature is cleared (K243236, 2025), while its Blood Pressure Insights feature is not FDA cleared[16][22]. Wellness outputs can still serve as exploratory measures, but they carry no FDA performance claim.

Validation Evidence and Its Limits

Validation is specific to a metric, a population, a wear location, and a firmware version. The studies in each device record favour independent work and clinical populations, and studies funded by or written with the maker are flagged.

  • Movement raises heart rate error. Across devices, optical heart rate error during activity was on average 30% higher than at rest[23].
  • Lab results overstate accuracy in daily life. In lab studies, 45.2% of consumer step count comparisons were within ±3%, but only 42% of free living comparisons were within ±10%[24].
  • Sleep staging is the weakest common output. Six wrist devices compared with polysomnography all detected more than 90% of sleep epochs, but specificity for wake ranged from 29% to 52% and Cohen’s kappa from 0.21 to 0.53[25]. Seven consumer devices showed the same pattern and did worse on disrupted nights[26].
  • One task, different results. In 201 cardiology patients, single lead ECG detection of atrial fibrillation reached 85% sensitivity and 75% specificity for both Apple Watch 6 and Galaxy Watch 3, and 58% sensitivity and 75% specificity for Withings ScanWatch, with 17% to 24% of tracings inconclusive[27].
  • Skin tone evidence is mixed. One study found no statistically significant difference in heart rate accuracy across skin tones[23]. A systematic review of 10 studies found 4 reporting lower accuracy in darker skin, 4 no effect, and 2 mixed results[28]. A meta-analysis by authors affiliated with Verily found no statistically significant pulse rate bias by pigmentation but SpO2 accuracy problems in all groups[29].
  • Firmware updates change results. FDA asks sponsors to keep a record of the timing and nature of updates for each device, because changes in the measurement after an update may invalidate the results[2]. A review of activity and heart rate monitors notes that updates “can, and do, change device performance”[30].
  • Large studies are often maker funded. In the Apple Heart Study, 0.52% of 419,297 participants received an irregular pulse notification, and Apple coauthored the study[31]. The Fitbit Heart Study included 455,699 participants and was sponsored by Fitbit[32].

Record the model, firmware, and app version for each participant, report skin tone, and check subgroup performance before you rely on a metric.

Battery, Phones, and Subscriptions

Charging gaps are a common source of missing data in continuous wear studies. Each card shows the maker’s published battery figure, measured under the maker’s own assumptions, and continuous raw capture shortens it. Withings states 3 to 4 days for ScanWatch with raw capture switched on[14].

Phones decide who can take part

Current Apple Watch models need an iPhone 11 or later with iOS 27[33]. Galaxy Watch pairs only with Android phones, and its regulated features also need a Samsung Galaxy phone[34]. A device that works on one platform turns a cohort into iPhone owners or Android owners unless you provide phones.

Subscriptions can stop data

Oura Ring Gen3 and later need an active membership for API access and for sync to Apple Health or Health Connect[35]. WHOOP stops collecting data when a membership ends[36]. For Withings accounts created on or after 13 October 2026, Withings+ is needed to share data with other apps, including Apple Health and Health Connect[37].

Bring your own or provide the device

Letting participants use their own device widens reach but mixes models and firmware versions. Providing one model keeps data consistent but adds logistics and cost. Our guide to BYOD and provisioned devices compares the two, and our article on choosing consumer wearables goes deeper on selection. WeGuide supports provisioned devices in wearable device studies.

How We Built This Comparison

We list current models as shown on each maker’s website on the date in the record. Specifications such as sensors and battery life are the maker’s published figures, not our test results. Data access routes come from the maker’s developer and support documentation.

Regulatory status is recorded per feature, not per device, and every US clearance or De Novo number links to its FDA database record. “No FDA record found” means we found no 510(k), De Novo, or PMA for that feature on the FDA’s openFDA database. Validation highlights are a small selection of peer reviewed studies, chosen to favour independent studies and clinical populations, and each cites its PubMed record. We flag studies funded by, or written with, the device maker.

We do not score or rank devices, because fitness for purpose depends on your endpoint, population, and protocol. WeGuide’s own integration status is shown for each device, and for devices without a native WeGuide integration we describe what data can reach WeGuide and how. Some devices on this page are made by companies that compete or partner with WeGuide, and the same rules apply to all of them. Products, firmware, and APIs change often. If you spot an error, tell us through the contact page and we will correct it.

Planning a wearable study?

WeGuide collects wearable data alongside ePRO, eConsent, and automated reminders in one participant app. Garmin is a native integration, with raw accelerometer and gyroscope data at 25 Hz on supported models, and Apple Health, Health Connect, and Fitbit Air are supported. Provisioned devices are supported too.

Book a demo

Frequently Asked Questions

Answers about raw data, regulatory status, phone health stores, and validation when you choose a wearable for a study.

Have another question?

Which wearables give researchers raw sensor data?

Several do, but almost always behind a gate such as an approval or a contract. Garmin, Samsung Galaxy Watch, Apple Watch, Polar H10, Polar Verity Sense, Withings ScanWatch, and the ActiGraph devices document raw or near raw routes. Oura, WHOOP, and Fitbit Air do not. Through WeGuide, raw accelerometer and gyroscope data are available from supported Garmin models only.

Is an FDA cleared wearable cleared for everything it measures?

No. The FDA clears or authorises specific software features for specific indications, such as an ECG app that classifies atrial fibrillation in adults. Heart rate, HRV, sleep stages, and SpO2 on consumer wearables are usually wellness features. Check the feature and its 510(k) or De Novo number, which this tool links to the FDA database.

Can we pull Apple Watch data from Apple’s cloud?

No. Apple has no server or OAuth API for Apple Watch data. A study app reads HealthKit on each participant’s iPhone, with permission for each data type, and then uploads the data. Raw signals such as PPG and accelerometer data need SensorKit, which is only available to research studies that Apple has approved.

What does via Apple Health or Health Connect mean?

It means the maker’s phone app writes summary data into the phone’s health store, and WeGuide reads it from there. Only the data types that the maker writes can arrive, and they differ by maker. For example, WHOOP writes no HRV to either store, and Oura writes no HRV to Apple Health. It is not a direct integration with the maker.

Do participants need a paid subscription?

For Oura Ring Gen3 and later and for WHOOP, yes, because data access or collection stops without one. New Withings accounts created on or after 13 October 2026 need Withings+ to share data with other apps, including Apple Health and Health Connect. Garmin, Apple, Samsung, Polar, and Fitbit devices work without a subscription for core data.

If a device is validated, is my endpoint valid?

Not automatically. Validation applies to one metric, population, wear location, and firmware version. FDA guidance asks sponsors to show that a device is fit for purpose in the proposed population, and the V3 framework separates verification, analytical validation, and clinical validation. Use published studies to shortlist devices, then plan the evidence your endpoint needs.

Should we choose a ring, a wrist device, or a chest strap?

It depends on the measure and the wear schedule. Rings and screenless bands suit continuous overnight wear, wrist devices offer the broadest set of metrics, and chest straps such as the Polar H10 give beat to beat intervals but suit supervised sessions with a set start and end. Wear location also changes step counts, so do not mix placements without calibration.

What happens if firmware updates during my study?

Measurements can shift after an update. FDA guidance asks sponsors to keep a record of the timing and nature of updates for each device, so record the model, firmware, and app version for each participant and log the update dates. Where raw data are available, you can reprocess every participant with one fixed algorithm.

Can one device serve both iPhone and Android participants?

Yes, several can. Garmin, Fitbit Air, Fitbit Charge 6, Oura, WHOOP, Polar, and Withings devices work with both platforms. Apple Watch works only with iPhone, while Galaxy Watch and Pixel Watch work only with Android phones. Use the phone platform filter and choose Both to see the devices that suit a mixed cohort.

Can a new study still use the Fitbit Web API?

No. Google switches off the legacy Fitbit Web API on 30 October 2026, and new builds must use the Google Health API. Its health scopes are restricted and need a Google privacy and security review, and research use also needs an attestation with ethics approval. Every participant must also give consent again under the new API.

References

  1. Goldsack JC, Coravos A, Bakker JP, et al. Verification, analytical validation, and clinical validation (V3): the foundation of determining fit-for-purpose for Biometric Monitoring Technologies (BioMeTs). NPJ Digit Med. 2020;3:55. PubMed 32337371
  2. US Food and Drug Administration. Digital Health Technologies for Remote Data Acquisition in Clinical Investigations. Final guidance, December 2023, sections IV.C and IV.H.4. fda.gov
  3. Paluch AE, Evenson KR, Eberl EM, et al. Comparison of daily step counts among wearable devices in free-living adults: A systematic review and network meta-analysis. J Sport Health Sci. 2026;101162. PubMed 42492741
  4. Servais L, Strijbos P, Poleur M, et al. Evidentiary basis of the first regulatory qualification of a digital primary efficacy endpoint. Sci Rep. 2024;14(1):29681. PubMed 39613806
  5. Google. Google Health API: about. Google for Developers. Accessed 20 October 2026. developers.google.com
  6. Oura. Oura API v2 documentation. Accessed 20 October 2026. cloud.ouraring.com
  7. Samsung. Samsung Health Sensor SDK: data specifications and app verification. Samsung Developer. Accessed 20 October 2026. Data specifications, app verification
  8. Polar Electro. Polar BLE SDK, version 8.4.0. GitHub. Accessed 20 October 2026. github.com/polarofficial
  9. Apple. HealthKit: protecting user privacy. Apple Developer Documentation. Accessed 20 October 2026. developer.apple.com
  10. Apple. SensorKit: SRFetchRequest and SRSensorReader. Apple Developer Documentation. Accessed 20 October 2026. SRFetchRequest, SRSensorReader
  11. Oura. Apple Health integration. Oura Help. Accessed 20 October 2026. support.ouraring.com
  12. WHOOP. Apple Health integration and Health Connect integration for Android. WHOOP Support. Accessed 20 October 2026. Apple Health, Health Connect
  13. US Food and Drug Administration. 510(k) K231532, ActiGraph LEAP activity monitor. 2023. FDA record. Ametris. CentrePoint V3 API: data access files. GitHub. Accessed 20 October 2026. github.com/actigraph
  14. Withings. Advanced Research API and raw data documentation. Withings Developer. Accessed 20 October 2026. Research APIs, raw data
  15. Garmin. Health SDKs overview. Garmin Developers. Accessed 20 October 2026. developer.garmin.com
  16. US Food and Drug Administration. 510(k) premarket notification and De Novo databases: DEN180044, DEN180042, K213971, K240929, K250507, K240102, K223374, K202891, K202897, and K243236. Checked on openFDA, data dated 5 October 2026. 510(k) database, De Novo database
  17. Apple. Blood Oxygen app on Apple Watch. Apple Support. Accessed 20 October 2026. support.apple.com
  18. US Food and Drug Administration. 510(k) summary K250507: Hypertension Notification Feature. 2025. accessdata.fda.gov
  19. US Food and Drug Administration. Medical Device Development Tool: summary of evidence and basis of qualification, Apple AFib History Feature. 2024. fda.gov
  20. Apple. watchOS 27 feature availability. Accessed 20 October 2026. apple.com
  21. US Food and Drug Administration. General Wellness: Policy for Low Risk Devices. Final guidance, January 2026. fda.gov
  22. US Food and Drug Administration. Close-out letter, WHOOP Inc., MARCS-CMS 709755. 17 June 2026. fda.gov
  23. Bent B, Goldstein BA, Kibbe WA, et al. Investigating sources of inaccuracy in wearable optical heart rate sensors. NPJ Digit Med. 2020;3:18. PubMed 32047863
  24. Fuller D, Colwell E, Low J, et al. Reliability and Validity of Commercially Available Wearable Devices for Measuring Steps, Energy Expenditure, and Heart Rate: Systematic Review. JMIR Mhealth Uhealth. 2020;8(9):e18694. PubMed 32897239
  25. Schyvens AM, Peters B, Van Oost NC, et al. A performance validation of six commercial wrist-worn wearable sleep-tracking devices for sleep stage scoring compared to polysomnography. Sleep Adv. 2025;6(2):zpaf021. PubMed 40303381
  26. Chinoy ED, Cuellar JA, Huwa KE, et al. Performance of seven consumer sleep-tracking devices compared with polysomnography. Sleep. 2021;44(5). doi:10.1093/sleep/zsaa291. PubMed 33378539
  27. Mannhart D, Lischer M, Knecht S, et al. Clinical Validation of 5 Direct-to-Consumer Wearable Smart Devices to Detect Atrial Fibrillation: BASEL Wearable Study. JACC Clin Electrophysiol. 2023;9(2):232-242. PubMed 36858690
  28. Koerber D, Khan S, Shamsheri T, et al. Accuracy of Heart Rate Measurement with Wrist-Worn Wearable Devices in Various Skin Tones: a Systematic Review. J Racial Ethn Health Disparities. 2023;10(6):2676-2684. PubMed 36376641
  29. Singh S, Bennett MR, Chen C, et al. Impact of Skin Pigmentation on Pulse Oximetry Blood Oxygenation and Wearable Pulse Rate Accuracy: Systematic Review and Meta-Analysis. J Med Internet Res. 2024;26:e62769. PubMed 39388258
  30. Collins T, Woolley SI, Oniani S, et al. Version Reporting and Assessment Approaches for New and Updated Activity and Heart Rate Monitors. Sensors (Basel). 2019;19(7). doi:10.3390/s19071705. PubMed 30974755
  31. Perez MV, Mahaffey KW, Hedlin H, et al. Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation. N Engl J Med. 2019;381(20):1909-1917. PubMed 31722151
  32. Lubitz SA, Faranesh AZ, Selvaggi C, et al. Detection of Atrial Fibrillation in a Large Population Using Wearable Devices: The Fitbit Heart Study. Circulation. 2022;146(19):1415-1424. PubMed 36148649
  33. Apple. Apple Watch Series 12 tech specs. Accessed 20 October 2026. apple.com
  34. Samsung. Watch and phone compatibility, support answer ANS10004668. Samsung US Support. Accessed 20 October 2026. samsung.com
  35. Oura. Oura Membership. Oura Help. Accessed 20 October 2026. support.ouraring.com
  36. WHOOP. Canceling your membership. WHOOP Support. Accessed 20 October 2026. support.whoop.com
  37. Withings. API plans, and Partner apps: what is Apple Health. Accessed 20 October 2026. API plans, support.withings.com

Device records checked against the makers’ pages, developer documentation, openFDA, and PubMed on 20 October 2026 by the WeGuide editorial team. Some devices are made by companies that partner or compete with WeGuide, and the same rules apply to all of them. This tool is for research planning and education. It is not a medical device and does not give medical advice.