FIU Study: Smartwatches Overestimate Calories Burned

An FIU study of four popular smartwatches found all overstated calories burned, with errors rising sharply as wearers' body fat percentage increased.

A Florida International University team led by Jason Kostrna, associate professor of kinesiology and exercise science, tested the Apple Watch Series 8, Fitbit Sense 2, Samsung Galaxy Watch 5 and Garmin Forerunner 955 against a COSMED K5 metabolic analyzer, the gold-standard device clinical labs use to measure true energy use. Fifty-eight adults rode a recumbent bike alternating between moderate and vigorous intensities for ten minutes while wearing the devices. Every watch overstated the calories people burned, according to the Tampa Bay Times.

The researchers report a median error of roughly 15% to 25% across devices, but the averages for the worst-performing watches climbed much higher. Apple had the smallest systematic error, while Garmin and Samsung produced the largest overestimates, according to the Tampa Bay Times. Fitbit failed in a different way: Tampa Bay Times reports it occasionally logged implausibly low counts, including just one calorie for an entire workout, and sometimes failed to generate a reading at all, leading FIU researchers to exclude much of its data. The Conversation adds that seven Fitbit readings exceeded 450% of the measured value.

The unexpected finding, stated plainly by lead author Jason Kostrna in The Conversation: as wearers' body fat percentage rose, every device tested became less accurate. For participants with higher body fat, errors for some watches ranged from about 50% to 100% or more — a 100-calorie workout could display as 150 or even 200. The Conversation notes the team recruited 58 Hispanic adults in Miami across a wide range of body fat percentages and skin tones, and found skin tone had no significant effect under the tested conditions. People most likely to use watches for weight loss may get the least reliable numbers.

Small errors compound quickly. According to The Conversation, someone with 35% or more body fat exercising four hours a week could see their watch stray by 1,200 to 2,400 calories per week — nearly an entire day's worth of food consumed in error, assumed to have been burned in workouts. Kostrna put it bluntly to FIU News, as quoted by the Tampa Bay Times: you could end up in a calorie surplus while thinking you are in a deficit, because 'you can't treat the calories burned number it gives you as an accurate number, because it's not.'

The study does not make wearables useless. In a companion study using the same protocol, the FIU team found these devices measure heart rate with reasonable accuracy, and they reliably track workout duration, distance and pace depending on the activity, according to The Conversation. Wrist devices combine accelerometer movement data and light-based heart rate readings with user profile details, then run proprietary algorithms to estimate energy expenditure — a prediction problem that body composition, fitness and movement efficiency all complicate. The researchers recommend treating the calorie figure as a rough guide, useful for patterns over time, not as a precise measurement.

FIU's lab is now extending its smartwatch experiments to strength training, where whole-body movement combined with wrist motion and varying intensity may make accurate estimation even harder, according to The Conversation. Because each manufacturer guards its proprietary algorithm, the team cannot pinpoint why higher body fat percentage degrades accuracy so sharply, though calibration choices likely play a role. The takeaway for the roughly 100 million U.S. smartwatch owners is pragmatic: be conservative using the calorie number to decide how much to eat, watch trends alongside heart rate and body weight changes, and remember that consistency is not accuracy — a watch can overestimate systematically and still look perfectly steady workout to workout.