A cycling study linked higher body fat to larger smartwatch calorie errors. Learn why workout calories should not set your food budget.
In 2026 PLOS One published “Body fat, skin tone, and the accuracy of smartwatch caloric expenditure estimates,” which tested four devices during controlled cycling. NPR reported on the research on September 22, 2026.
The researchers compared smartwatch estimates of physical-activity energy expenditure with indirect calorimetry. Their findings give the calorie number on a workout screen some necessary context:
The PLOS One authors concluded that the tested watches substantially misestimated energy expenditure under these conditions. They said current consumer devices do not yet provide reliable individual caloric monitoring.
The study reported both error and average bias. These describe different questions: how far an estimate sits from the comparison measurement and whether a device tends to read higher or lower on average.
A small average bias is not the same thing as consistently accurate readings for individuals. An average can conceal substantial variation, so it needs to be read alongside the rest of the results.
The reported average biases were:
The Apple, Garmin and Samsung mean biases were statistically different from zero. Fitbit’s mean bias was not statistically significant after the outliers were removed.
That last detail matters. The paper reports seven Fitbit estimates above 450% of the metabolic-system values. Including those outliers raised Fitbit’s mean bias to 128.6 kcal.
The Fitbit result therefore needs its full context. Selecting only its small average bias would leave out the extreme readings that changed that average substantially. These results do not provide a straightforward shopping verdict for everyone who wants a more accurate watch.
The evidence has a defined boundary. Participants were young Hispanic adults with Fitzpatrick skin types III to V. This was not a study specifically of gay men over 35, and its results should not be transferred unchanged to older populations or other groups.
The controlled cycling test also does not establish accuracy during strength training or other exercise. Nor does it establish accuracy during ordinary daily movement. The four named models should not stand in for every smartwatch on the market.
The researchers found no robust main effect of Fitzpatrick skin type on energy-expenditure error. They cautioned that interpretation was limited by the small number of participants in the darkest skin-type subgroup. This is not proof that skin tone never affects calorie-estimate accuracy.
The body-fat finding describes an association within the study. Proprietary device algorithms prevented the researchers from explaining exactly why the errors occurred. The paper also did not test whether using these calorie estimates changes food choices or weight outcomes, and it did not measure long-term health effects.
Finally, the authors left open the possibility that wearables could help track relative changes within a person over time. They said that possibility warrants further study in free-living conditions. It should not be treated as an established benefit of the calorie readout.
The authors called for greater transparency from manufacturers. Without access to proprietary algorithms, researchers cannot fully account for the differences between devices or the relationship between error and body fat.
They also suggested that better models may need datasets separated by characteristics such as adiposity, sex and age. Better activity recognition and additional sensing approaches were among the possibilities they identified.
The editorial implication is straightforward: accuracy across body types is a design question worth asking. A polished interface does not answer it. Neither does a calorie total displayed with apparent precision.
For the person wearing the watch, that shifts the focus from how convincing the screen looks to how well its estimate supports the decision being made.
The practical adjustment is to give the calorie display less authority over eating decisions. It can remain an estimate on a screen without becoming a precise food allowance.
For a gay man over 35, that distinction leaves room for several legitimate priorities. Looking good may matter. So may staying strong, feeling energetic and protecting long-term health. None of those goals requires treating a workout’s displayed calorie total as an exact nutritional instruction.
A watch showing calories burned does not establish an equally precise amount to “eat back.” Treating the two numbers as interchangeable gives the exercise estimate more authority than this study supports.
The researchers warned that large energy-expenditure errors may mislead people managing weight or energy balance. The practical interpretation is not to skip food after exercise. It is to stop treating the display as an automatic calculation of what a meal should contain.
There is also no dependable correction factor here. The study’s percentage errors cannot be turned into a universal discount applied to every workout reading. Device differences and individual variation make that shortcut unsupported.
If a personal eating target is needed, that decision belongs in a conversation with a qualified clinician or registered dietitian. This paper supplies no dietary prescription. Its contribution is a reason to question one input into that decision rather than a replacement target.
The next step is to distinguish a workout number from the goals behind the workout. A calorie total is not itself a verdict on whether training served your priorities.
GayFitLab’s broader approach to body composition keeps appearance and long-term health in the same conversation. That makes looking beyond a single fat-loss number a more useful editorial frame than letting one device estimate dominate the discussion.
For physique goals, the question can remain whether your approach fits the strength and daily function you value. Our coverage of keeping muscle and everyday capability in view provides another place to continue that discussion without making the watch’s calorie figure the central goal.
This is a change in how information is used, not a demand to abandon technology. A device does not need unquestioned authority over food decisions to have a place in someone’s routine.
A practical boundary is to describe the number accurately: it is an estimate of workout energy expenditure. That wording keeps it from quietly becoming a measurement of what you should eat next.
Separate the question of whether a watch fits your routine from the question of whether its calorie value deserves that nutritional role. Those are different decisions. The calorie estimate should carry only the weight that its supporting evidence warrants.
The same editorial discipline applies when interpreting recovery scores without letting them run the routine. A useful discussion starts with the specific metric and the decision attached to it, rather than treating every number on a wearable as equally authoritative.
It should change the authority assigned to the estimate, not the respect assigned to your body.
A larger device error is not a judgment about fitness or attractiveness. It is a problem for interpreting the device’s output. The sensible response is to question the number’s suitability for a decision rather than turn it into another reason to scrutinize your appearance.
Body composition can still be a valid personal goal. This finding does not require giving up that goal or pretending appearance is irrelevant. It asks for a cleaner separation between what you want for your body and what a particular algorithm claims to know about it.
That separation is useful whether your priority is a leaner look or maintaining strength. It also fits a routine where health matters but does not occupy every conversation or meal. The watch can stay on your wrist without becoming the final authority at the table.
Treat your watch’s workout calories as an estimate rather than a precise food budget, and base personal nutrition decisions on a broader assessment with qualified support when needed.
After separating workout calorie estimates from food decisions, the next step is building a broader approach to nutrition and body composition that fits life after 35. GayFitLab is a research-led publication for gay men 35+ that connects nutrition and fitness with healthy aging while separating supported findings from marketing claims.
Stay connected for research-led guidance on fitness, body composition, skin, grooming, vitality, sexual wellbeing, relationships and healthy aging. Clear ideas for gay men 35+ who want to look good, feel strong and age with confidence.
Build habits and systems that support clear thinking, steady energy and long term capacity throughout a demanding career.
explore the Blog
