Optical heart rate: where it struggles and why.

Almost every wrist device measures heart rate the same way: it shines light into the skin and watches what comes back. It is a clever trick, and in the right conditions it works well. This piece is about the conditions where it works less well, what the published research says about each, and where the research itself is thin.

What the sensor is actually doing

The method is called photoplethysmography, or PPG: an optical way of measuring changes in blood volume under the skin.¹ A light source sits against the skin, a photodetector measures how much light is reflected back, and the changes in blood volume are worked out from the amount detected.⁴ The resulting signal has two parts: a pulsing component that follows each heartbeat, riding on a slower baseline shaped by breathing, the nervous system and temperature regulation.⁴

Notice what that means. The device never sees a heartbeat. It sees a faint pattern in reflected light and infers the beat. Anything else that changes the reflected light can be mistaken for, or can bury, the pulse.

Movement is the big one

Motion artifacts are typically caused by the sensor moving over the skin, skin deformation, and changing blood flow, and they can show up as missing or false beats.¹ There is a particular trap with repetitive movement: sensors can lock on to the rhythm of walking or jogging and mistake it for the heart, an effect the authors call “signal crossover.”¹

The numbers bear this out. In a study of 53 people wearing six devices against an ECG patch, error during activity was on average 30% higher than at rest, and the authors concluded that the devices were all reasonably accurate at rest and during prolonged elevated heart rate, but differed in how they responded to changes in activity.¹ A separate 2026 study of 45 adults and ten commercial wearables found the same shape: stop-start walking increased errors for several devices, and accuracy was strongly device-specific, from a mean absolute error of 4.5 bpm for the best performer to 9–14 bpm for several others.⁵

Fit and position

Because the light path matters, so does where the sensor sits. Physical activity can displace the sensor from its original position, which changes the path of the light and so the signal, and the pressure the device applies to the skin controls how strong the returned signal is.⁴ At least one maker’s published guidance lists wearing the device too tightly and flexing the wrist among the things that hurt accuracy.¹ The practical reading: snug but not tight, in the same place each time, and expect readings during activities that flex the wrist to be the less dependable ones.

Skin tone: a real question, not a settled one

Darker skin contains more melanin, which absorbs more green light, and earlier research suggested wrist heart-rate readings were inaccurate up to 15% more often in dark skin.¹ The 53-person study above was designed to test this directly, with participants spread evenly across the Fitzpatrick scale, and found no statistically significant difference in accuracy between skin tones.¹

That is reassuring, and it is not the last word. A published response pointed out that the sample was small, that the Fitzpatrick scale is a subjective visual scale with known limits, and that errors in subsets of the darkest group could be missed.² The original authors replied that visual scales are imperfect and that a study can never prove the null hypothesis true, and they encouraged more work.³ So the honest position is: one carefully designed study found no skin-tone effect on heart rate, a reasonable critique says it cannot rule one out, and larger studies with objective colour measurement would settle it.

Tattoos

Ink sits right where the light needs to go. A 2026 systematic review explains the proposed mechanism: tattoo pigment absorbs and scatters light in the skin, which can shrink the pulsing part of the signal below the level the device will accept, and how bad it gets depends on how well the ink overlaps the colours of the sensor’s light.⁶ The same review is candid about the evidence: of 17 studies, only one directly tested tattoos on a wearable under controlled conditions, reporting complete signal dropout 36% of the time at rest. Its conclusion is that the typical failure is dropout — no reading — rather than a confidently wrong one.⁶ One study is a thin base, so treat the 36% as a warning, not a rate.

Cold, sweat and perfusion

These are the commonly repeated explanations, and we could not verify most of them. Device makers do list skin perfusion among the factors that affect performance.¹ But in the 2026 chamber study, which tested the same devices at 23 °C, 36 °C and 10 °C, the climate conditions did not significantly affect accuracy; movement did.⁵ The 2020 authors, too, reported that their literature search found no publications showing an effect of arm hair, sweat or ambient temperature on PPG.³ Cold may matter for some people and some sensors. The published evidence we read does not say it is a major driver.

What this means in practice

• Trust the resting number most. That is where the devices in these studies did best, and it is where trends over weeks live.
• Be sceptical of exercise spikes, especially in intervals and stop-start effort, where errors were largest.
• Treat a gap as a gap. Dropout, not wrong values, is the typical failure for at least one cause (tattoos), and a flat or missing stretch is more honest than a made-up one.
• Choose the device, not just the technology. Accuracy varied far more between devices than with climate in the 2026 study, which is one reason to test a device against your own week before you rely on it.

Sources

The studies and standards behind the claims above, so you can read them yourself. Listing a source is not an endorsement of any device it tested.

1. Bent B, Goldstein BA, Kibbe WA, Dunn JP (2020). NPJ Digital Medicine 3, 18, doi 10.1038/s41746-020-0226-6 — six wearables vs ECG patch, 53 participants across the full range of skin tones.
2. Colvonen PJ (2021). NPJ Digital Medicine 4, 38, doi 10.1038/s41746-021-00408-5 — Matters Arising response on sample size and skin-tone scales.
3. Bent B, Enache OM, Goldstein B, Kibbe W, Dunn JP (2021). NPJ Digital Medicine 4, 39, doi 10.1038/s41746-021-00409-4 — the original authors’ reply.
4. Castaneda D, Esparza A, Ghamari M, Soltanpur C, Nazeran H (2018). International Journal of Biosensors & Bioelectronics 4(4), 195–202, doi 10.15406/ijbsbe.2018.04.00125 — review of wearable PPG sensors.
5. Gielen J, Van Oost CN, Debard G, et al. (2026). JMIR Formative Research, e85186, doi 10.2196/85186 — ten commercial wearables, 45 adults, three climate conditions.
6. Rojas-Valverde D, Gamboa-Salas J (2026). Revista de Ciencias del Ejercicio FOD 21(2), doi 10.29105/rce-fod.v21i2.184 — systematic review of tattoo pigment and wearable PPG, 17 studies.

← All articles

This article is general wellness education, not medical advice. SoluneFit is a wellness tool, not a medical device. It does not diagnose or treat any condition. Talk to a clinician about health concerns.