Trackers can wildly overcount calories
A high‑profile social summary of a Stanford finding warns many fitness trackers can overestimate calorie burn — in some cases by as much as 93% — so “eating back” exercise calories can undo weight goals. (Coaches are using that study to tell people not to automatically reward workouts with extra food without checking the math.) (x.com)
A fitness watch can be pretty good at one job and terrible at another. Stanford researchers found six of seven wrist trackers kept heart-rate error under 5%, but none measured calories burned well. (med.stanford.edu) Heart rate is the easy part because the watch is reading a direct body signal from your wrist. Calories are the hard part because the device has to guess how much fuel your whole body used from motion, pulse, and your profile data. (mdpi.com) Stanford tested seven devices on 60 volunteers while the volunteers sat, walked, ran, and cycled. The researchers compared each watch to lab methods including continuous heart monitoring and indirect calorimetry, which estimates energy burn from breathing gases. (mdpi.com) The calorie numbers were not just a little off. In the Stanford study, the best device still missed energy expenditure by 27% on average, and the worst missed by 93%. (med.stanford.edu) The errors also were not evenly spread across people or activities. The paper says calorie estimates got worse for walking, for men, for higher body mass index, and for darker skin tone, which means the same watch can behave differently on different wrists. (mdpi.com) That is why “I burned 600 calories, so I can eat 600 extra calories” can break fast. If your watch overstates the workout by even 200 to 300 calories several times a week, the math can wipe out the calorie deficit many people are trying to create. (med.stanford.edu) Stanford’s team said a consumer device used outside medicine should stay under 10% error to be truly useful. Their study found no device got energy-expenditure error below 20%, which is why the authors said calorie numbers should be used with caution in health programs. (med.stanford.edu) (mdpi.com) Researchers have since shown why wrist devices struggle here. A 2021 Nature Communications paper found a smartwatch had 42% cumulative error on common activities, while a system that measured leg movement from the shank and thigh cut that to 13%, because legs reveal the work of walking and running more directly than a wrist does. (nature.com) So the useful way to read a tracker is to trust trends more than rewards. Steps, workout time, and heart-rate patterns can show whether you are moving more than last month, but the calorie number is closer to a rough guess than a receipt. (med.stanford.edu)
