Fitness trackers exaggerate burn

A fitness tracker is good at counting motion and pulse, but calories are a different problem because your body burns energy like a hybrid car uses fuel: some for movement, some for idling, and some for systems you never notice, like temperature control and digestion. (med.stanford.edu) That is why a Stanford team tested seven wrist devices against laboratory measurements in 60 volunteers walking, running, and cycling, instead of just comparing one gadget to another. The devices were Apple Watch, Basis Peak, Fitbit Surge, Microsoft Band, Mio Alpha 2, PulseOn, and Samsung Gear S2. (med.stanford.edu) Heart rate came out looking solid: six of the seven devices stayed within 5 percent error for pulse. Energy expenditure, which is the technical term for calories burned, was the weak spot on every device they tested. (med.stanford.edu) The worst device missed calorie burn by 92.6 percent, and even the best device was off by about 27.4 percent. In plain English, a screen that says you burned 500 calories could be closer to 365, or much less, depending on the device and the workout. (med.stanford.edu) The reason is simple: wrist wearables mostly see signals like arm motion and heart rate, then use proprietary formulas to guess oxygen use and total energy burn. Those formulas struggle when two people have the same pulse but different body size, fitness, efficiency, or exercise style. (med.stanford.edu) Stanford’s researchers said activity type, skin tone, body mass index, and fitness level all affected accuracy, which means the same watch can be decent for one person and sloppy for another. The team’s senior author, Euan Ashley, said there was “no device” that hit acceptable error rates for energy expenditure. (med.stanford.edu) This is not the same as saying wearables are useless. A 2021 systematic review and meta-analysis in JAMA Network Open found that wearable physical activity trackers were associated with increases in daily steps of about 1,800 and moderate-to-vigorous physical activity of about 37 minutes per week in adults with cardiometabolic conditions. (jamanetwork.com) So the better use case is behavior, not bookkeeping: use the watch to keep a walking streak, catch a rising resting heart rate, or compare today’s run with last week’s run under the same conditions. Using the calorie number to decide you “earned” a 700-calorie dinner is where the math starts to drift. (med.stanford.edu; jamanetwork.com) If someone is trying to lose weight, food logging usually has tighter numbers than exercise logging because a nutrition label on a yogurt cup is more stable than an algorithm guessing your metabolism from your wrist. The tracker can still help by showing whether you moved for 20 minutes or 60 minutes, which is a cleaner signal than whether it says 240 calories or 410. (med.stanford.edu) The Stanford paper that sparked this conversation was published in 2017, but it keeps resurfacing because the basic warning has held up: heart rate is measurable, calorie burn is inferred. If you want one rule of thumb, trust the watch more for trends than totals. (med.stanford.edu; nature.com)