Does your wrist affect optical heart rate accuracy?

Most athletes trust the number on their wrist. A heart rate reading during a tempo run, a zone calculation between intervals, a recovery figure from the previous day’s workout: all feed decisions about training volume, fitness levels, and race preparation. The question is whether all wrists produce an equally accurate reading.

Optical heart rate monitors embedded in every modern smartwatch, including those from Garmin, Apple, Fitbit, and WHOOP, are based on reflected light technology. Skin tone and wrist size are both plausible sources of interference. The evidence on each is less settled than the headlines suggest.

How optical heart rate sensors work

All wrist-worn heart rate monitors use a method known as photoplethysmography (PPG). A light-emitting diode, typically green, shines into the skin. A photodetector measures how much light is reflected. Blood absorbs light more strongly than surrounding tissue, so each heartbeat produces a detectable change in reflection. The sensor counts those changes to calculate beats per minute.

Green light is used because it provides a strong signal-to-noise ratio even in the presence of motion. Its limitation is absorption by melanin, the pigment that determines skin colour. Higher melanin content absorbs more green light before it reaches the photodetector, thereby reducing signal quality. Modern devices attempt to compensate through hardware and software, but the extent of that compensation varies by device and condition.

What the research actually shows about skin tone

The most extensive analysis is Koerber et al. (2023), published in the Journal of Racial and Ethnic Health Disparities. Across 10 studies, 469 participants, and 26 devices, four studies showed no skin-tone effect on accuracy, four found decreased accuracy for participants with darker skin, and two produced mixed results. That split does not support confident universal conclusions in either direction.

A widely cited 2020 paper by Bent et al. found no significant differences in heart rate accuracy across skin tones using the Apple Watch 4, Fitbit Charge 2, and Garmin Vivosmart 3. The protocol covered seated rest, paced breathing, and walking to 50 per cent of maximum heart rate. The sample at the darkest end of the Fitzpatrick scale was small, and exercise intensity was low throughout, which limits the extent to which the findings generalise.

More recent research points in a different direction during harder efforts. Hung et al. (2025) found no skin-tone differences at rest using the Fitbit Charge 5, but as exercise intensity rose, accuracy diverged. Mean error for darker skin tones at greater than 60 per cent of heart rate reserve reached 11.7 bpm, compared with a baseline error for lighter skin tones at less than 40 per cent heart rate reserve. A separate 2025 study by Icenhower et al. at Wake Forest University, testing the Garmin Forerunner 45, found no significant skin-tone effect overall, but did find that accuracy dropped during periods of rapidly changing heart rate, regardless of skin tone.

Taken together, the current evidence suggests skin-tone-related differences are most likely to emerge during moderate-to-vigorous exercise rather than at rest. The literature remains sufficiently inconsistent that broad conclusions across all devices and users are difficult to sustain.

The measurement tool is also the problem.

Running through all the studies above is a methodological limitation that makes the literature hard to interpret.

The Fitzpatrick Skin Type scale, used in virtually all of these studies, was developed in 1975 to determine PUVA phototherapy doses, initially in lighter-skinned populations. Types V and VI were added in 1988. The scale measures reported sun sensitivity, not melanin concentration, and within each of its six categories, actual melanin content varies substantially. That within-category variance weakens any finding that relies on Fitzpatrick groupings.

Mulholland, MacDonald, and Aguiar (2025), published in the European Journal of Applied Physiology, is among the first studies to assess accuracy using objective measurement of melanin content via colourimetry rather than Fitzpatrick self-report. It represents the direction the field needs to move.

What manufacturers have said and not said

Manufacturer transparency on this is inconsistent.

• Garmin’s support documentation acknowledges that optical HR tracking can affect battery life under some conditions. However, the company has not published a detailed methodology for melanin compensation or the internal validation data underlying it.
• Fitbit has referenced thousands of hours of internal testing without providing any demographic breakdown of participants or results by skin tone.
• Apple has stated that skin perfusion and tattoos can affect optical sensor performance and that the watch is designed to increase LED brightness and sampling rate when signal quality is poor. Skin tone is not mentioned directly.

The industry’s response to hardware has been to add wavelengths. Newer devices from Apple, Garmin, and Fitbit combine red and infrared with green. Red and infrared wavelengths penetrate deeper into tissue than green light and are less strongly absorbed by melanin, improving signal quality across different skin tones. Whether those hardware changes have closed the accuracy gap observed in validation studies has not yet been established by independent data. None of the companies has disclosed how many participants with darker skin tones were included in their internal validation, what results those participants produced, or how algorithms were adjusted as a result.

The wrist size question: less evidence, no signal

Wrist circumference as a variable affecting PPG accuracy is under-researched.

The physics are plausible. A narrower wrist places the sensor over different subcutaneous anatomy, including greater proximity to muscle and bone, and varying vascular density relative to the radial artery. Device fit relative to wrist circumference affects contact pressure and movement artefact. Both influence signal quality.

Published literature references wrist size as a variable but not as a primary research focus. Some studies have found a larger wrist circumference associated with reduced accuracy; others have not. The directionality is unsettled even for larger wrists, and the literature says nothing specific about smaller wrists.

A 2025 validation study in JMIR Cardio compared the Polar Verity Sense upper-arm sensor and Polar Vantage V2 wrist watch against a Polar H10 ECG chest strap across nine activities in 16 participants. The Verity Sense produced a mean absolute error of 1.43 bpm. The Vantage V2 produced a mean absolute error of 6.41 bpm with considerably higher variability across activities. The study did not examine wrist circumference or other anatomical factors, so its results cannot be applied directly to the wrist size question. What it does confirm is that sensor placement alone produces meaningful differences in PPG accuracy.

The claim that smaller wrists affect optical HR accuracy is biologically reasonable, practically intuitive, and supported by anecdote. Controlled research to quantify it does not yet exist.

What this means for training decisions

At rest, the available evidence does not support a meaningful difference in optical HR accuracy by skin tone. Available evidence suggests that differences in resting heart rate accuracy across skin tones are generally small and unlikely to meaningfully affect most recovery metrics. However, this has not been directly tested in long-term studies.

During moderate-to-vigorous exercise, the picture changes on some devices. Hung et al. (2025) found substantially larger heart rate errors in participants with darker skin tones at higher exercise intensities on the Fitbit Charge 5. For some runners, consistent underestimation could lead to training harder than the displayed zone suggests, affecting training load calculations, VO2 max estimates, Recovery Time, and Training Readiness, all of which are derived from heart rate data. For others, using a different device, the effect may not appear at all.

Device selection matters more than general claims about wearable sensors. The published research covers specific models: Fitbit Charge 5 and Garmin Forerunner 45 feature more skin-tone data than WHOOP, Oura, or Apple Watch. Knowing how a specific sensor behaves is more useful than any category-level statement.

For serious training, the chest strap remains the benchmark. The Polar H10, used as the reference device in every study cited here, measures heart electrical activity. It is not affected by melanin or wrist anatomy.

The regulatory gap nobody is closing

Consumer wearables are not classified as medical devices and are not subject to the validation requirements that apply to clinical pulse oximeters or heart rate monitors. No regulatory body currently requires manufacturers to validate optical HR accuracy across the full Fitzpatrick spectrum before a product reaches market.

The precedent is relevant. Pulse oximeters used in hospitals were shown to overestimate blood oxygen saturation in patients with darker skin, contributing to delayed care during the COVID-19 pandemic. The FDA issued guidance, not a mandate, recommending manufacturers validate across skin tones. Consumer wearable companies have not faced equivalent scrutiny.

Until broader validation requirements are introduced, it will remain difficult to determine how consistently wearable manufacturers address potential differences in accuracy across skin tones.

The bottom line

The evidence base has grown meaningfully in the past few years. It is concentrated on specific devices and conditions: exercise intensity between 40 and 60 per cent of heart rate reserve, with the Fitbit Charge 5 and Garmin Forerunner 45 producing most of the data. It is also consistently limited by small sample sizes at the darker end of the skin tone range and by reliance on the Fitzpatrick scale, which introduces substantial within-category variance.

The physics is established, and several controlled studies have found that skin tone can influence the accuracy of optical HR during exercise. Whether it affects any individual athlete depends on the device, the exercise intensity, and their own skin tone.

Where the research disagrees: a single error value that applies across all devices and skin types, a precise wrist size at which accuracy declines, and any claim that an existing device has solved the problem.

The practical test for any runner is a direct comparison between the chest strap and wrist during a hard workout. That comparison is the only way to know whether the effect is present in your own data.

This article is part of the site’s coverage of female athlete tech, which covers wearables, physiology, and performance for female endurance athletes. For more on heart rate monitoring, optical sensor comparisons, and how to set training zones, see the heart rate guide. For the underlying physiology and peer-reviewed research on wearable accuracy, see the sports science section. For how this site tests heart rate accuracy against reference devices, see the testing methodology.

Last Updated on 19 June 2026 by the5krunner