Does BMI or body fat affect wearable heart rate and blood oxygen accuracy?

This is a distinct question from whether skin tone affects wearable accuracy, even though the two often get lumped together or tested in the same studies. Melanin absorbs light at the sensor's wavelengths; body fat adds physical distance and scattering tissue the light has to travel through. They're separate variables that happen to share one sensor technology, and — as the evidence below shows — they don't always move together.

How body fat interferes with an optical sensor, mechanically

Nearly every consumer wearable — Apple Watch, WHOOP, Garmin, Fitbit, Oura — reads heart rate and blood oxygen with photoplethysmography (PPG): an LED shines light into the skin, and a photodiode measures how much bounces back as blood volume shifts with each heartbeat. That signal has to travel through skin, fat, and tissue in both directions before the photodiode ever sees it.

More subcutaneous fat under the sensor means a longer optical path and more light scattering before it reaches a blood vessel and scatters back out. Researchers using Monte Carlo light-transport simulations — modeling how photons actually scatter through layered tissue with the optical properties measured in people with obesity — have found the PPG signal is measurably altered, with one simulation reporting up to a 40% change in the AC-to-DC signal ratio (the part of the waveform the algorithm actually uses to find your pulse) tied to body composition. That's a modeling result, not a direct accuracy study, but it gives the correlational findings below a plausible physical cause rather than leaving them as an unexplained statistical pattern.

A second, less obvious mechanism compounds it: wrist circumference. A looser-fitting strap on a larger wrist allows more relative motion between the sensor and the skin during activity, adding motion artifact on top of the weaker underlying signal — a problem that's really about fit, not fat, but correlates with BMI in practice.

What the heart-rate studies actually found

The two studies, nine years apart and testing different device generations, land on the same structural finding: BMI is an independent predictor of heart-rate error, not something that only matters because it correlates with skin tone or anything else. Both variables load onto the same optical weak point, but they do it separately — the 2026 study found each one predicted error on its own, after accounting for the other.

The size of that effect, in absolute terms, is smaller than you might expect from how often this question comes up online. Across both studies, resting and steady-activity heart rate error for most devices stayed within a few percent even for higher-BMI participants — the gap is real and statistically independent, but it's a second-order effect on top of the much bigger activity-type and device-model differences visible in the same data.

Calorie-burn estimates are where body fat shows up hardest

This is the metric where the body-composition effect is least ambiguous, and it isolates the variable in a useful way. A 2026 FIU study tested four current devices (Apple Watch Series 8, Fitbit Sense 2, Samsung Galaxy Watch 5, Garmin Forerunner 955) against a clinical metabolic cart during a cycling protocol in 58 adults, and found calorie-burn error rose with body fat percentage across every single brand — while skin tone, measured in the same people, showed no significant effect on calorie error at all.

That contrast is the clearest evidence in this whole topic that body fat and skin tone are genuinely separate variables rather than two labels for the same underlying bias: in this dataset, one metric (heart rate) was sensitive to both, and another (calories) was sensitive to only one of them. The mechanism for calories is also different from the optical-signal story above — these algorithms estimate energy expenditure from heart rate plus a proprietary model of body size and composition, so an error in the body-composition input compounds with any heart-rate error already present. It's worth remembering, too, that calorie-burn accuracy is weak even in average-BMI users: a 2017 validation found no device under roughly 20% error for anyone, with some running past 90%. A device reading too high or low by 20-30% at a healthy weight and 35-45% at a higher one is the same underlying weakness, amplified.

Blood oxygen: the physics is the same, but the evidence basically isn't there yet

Blood oxygen (SpO2) uses the identical optical principle as heart rate — the same LED, the same tissue path, the same scattering problem — so there's no reason to expect BMI matters less here. But a direct, well-powered validation study that measures consumer wrist SpO2 error against a clinical reference and stratifies cleanly by BMI essentially doesn't exist yet. A 2023 scoping review of consumer wearables used in people with obesity flagged this gap directly: most validity studies for these devices don't report participants' BMI at all, so there's no way to know from the existing literature how large the effect actually is for the population most likely to experience it. For the fuller picture of what SpO2 can and can't tell you day to day, the dedicated accuracy guide covers the parts of this that are better established.

The one place higher BMI makes a reading more trustworthy, not less

Every mechanism above points toward wearable numbers getting noisier at higher BMI. Overnight SpO2 is the exception, and it runs the opposite direction for a reason that has nothing to do with sensor physics: obstructive sleep apnea is strongly linked to body weight. Across studies of people with obesity, roughly half show OSA on formal testing, and that share climbs further — into the large majority — among people with BMI 40+ who are screened for sleep symptoms. A repeating pattern of nighttime SpO2 dips in a higher-BMI wearer is therefore more likely to reflect a genuine physiological event than it would be in someone at lower risk, not sensor noise to explain away — the opposite of how to read the same dip pattern on the heart-rate side of this guide.

A practical self-check if you suspect your numbers are off

1. Cross-check heart rate against a chest strap during a real workout, not just at rest — the gap, where it exists, is largest during activity, and a chest strap sidesteps the optical problem entirely because it reads electrical signal rather than light.
2. Treat your device's calorie-burn number as a rough trend, not a precise input for a calorie deficit — it's the weakest-validated metric any wearable reports, for everyone, and the gap widens somewhat with body fat percentage on top of that.
3. Make sure your strap fits snugly, one to two finger-widths above the wrist bone, especially on a larger wrist — a looser fit adds motion artifact on top of whatever optical attenuation is already happening.
4. Don't dismiss a recurring overnight SpO2 dip pattern as "probably just my weight throwing off the sensor." If it repeats most nights and comes with loud snoring, witnessed breathing pauses, or daytime sleepiness, that combination is worth raising with a doctor regardless of what any wearable shows — only a sleep study, not a wrist sensor, can diagnose sleep apnea.
5. Weight toward trends over weeks, not single readings, for every metric in this guide — the device-level and activity-type differences in the research above are consistently larger than the BMI effect on its own.

The bottom line

BMI and body fat are linked to measurably more wearable error, through the same physical mechanism as skin tone but as a genuinely separate variable — strongest for calorie-burn estimates, real but modest for heart rate, and still too thinly studied to size up for blood oxygen. The one exception runs the other way: a recurring overnight SpO2 dip is more likely to be real, not sensor noise, in someone at higher risk for sleep apnea. Vita pulls heart rate, HRV, and SpO2 from whichever source you connect — WHOOP, Apple Health, or both — into one recovery picture built to weight your own trend over time rather than any single reading, and Vita's AI coach can help you reason through whether an unusual number is worth a second look or just normal noise for your situation.