Smartwatch Calorie Estimates Often Miss the Mark
New research indicates that popular fitness trackers frequently over- or under-estimate energy expenditure by up to a quarter, with accuracy varying significantly based on user body composition.
Many adults rely on wrist-worn devices to monitor their daily health, assuming the numbers displayed are precise scientific measurements. However, a recent study published in PLOS One suggests that these calorie counts are often significantly off. Researchers analyzed data from four major brands and found that estimates deviated from actual energy expenditure by an average of 15 to 25 percent. This discrepancy means that a workout logged as 500 calories might actually have burned considerably more or less, depending on the individual.
The investigation, highlighted by GN auto tech/wearables: fitness tracker, involved 58 adults cycling on stationary bikes while wearing smartwatches from Apple, Garmin, Samsung, and Fitbit. Participants were simultaneously connected to laboratory equipment that measured energy expenditure directly by analyzing breath gases. This method provided a ground truth against which the wearable devices could be tested. The results showed that no single brand was consistently accurate for every person, challenging the assumption that these devices offer a universal standard for health tracking.
Body composition affects accuracy
A key finding from the study is that the error rate increased as participants' body fat percentage rose. Lead author Jason Kostrna, a professor at Florida International University, emphasized that for a technology to be broadly useful, it must perform similarly across different bodies. While skin tone did not show a significant effect in this specific test, the researchers noted that their sample was limited. The primary variable driving inaccuracy was the physical difference in how optical sensors interact with different tissue densities, leading to skewed data for some users compared to others.
Among the brands tested, Apple devices were found to be the most accurate overall. Garmin and Samsung tended to overestimate calorie burn, while Fitbit devices showed the highest variability, including occasional missing data points. This lack of consistency suggests that users should not expect their devices to behave like laboratory instruments. Instead, the numbers should be viewed as rough estimates rather than definitive records of metabolic activity.
Algorithms rely on indirect signals
The inaccuracy stems from how these devices operate. They do not measure calories directly. Instead, they track movement, optical pulse waveforms, and location, then use proprietary algorithms to estimate metabolic output. Adam Lepley, a kinesiology professor at the University of Michigan, explains that errors can creep in at multiple stages. Sensor noise, the way the watch is positioned on the wrist, and motion artifacts all introduce uncertainty. The final number is a calculated guess based on assumptions that do not always hold true for every user.
Despite these limitations, experts advise against discarding smartwatches entirely. The data remains useful for tracking relative changes in activity levels over time. If a user sees a consistent trend in their daily movement or heart rate, that pattern is valuable. The catch is that the absolute numbers, such as total calories burned, should be treated with caution. Users are better off focusing on the consistency of their habits rather than the precise numerical value of each session.
Interpreting wearable data responsibly
Health professionals suggest that consumers interpret wearable data as a directional guide rather than a medical diagnosis. Amanda Velazquez, an obesity medicine director at Cedars-Sinai, notes that while the tools have flaws, they encourage engagement with health. The trade-off is clear: convenience comes at the cost of precision. Users must understand that their device is estimating a complex biological process using limited input signals. Recognizing this limitation helps prevent frustration when the numbers do not align with expectations or other health metrics.
As wearable technology continues to evolve, manufacturers face pressure to improve the inclusivity of their algorithms. Until then, the most prudent approach is to use these devices for broad trends. The study underscores that while smartwatches are convenient, they are not infallible. For those using them to manage weight or training loads, cross-referencing with other methods or consulting with healthcare providers can provide a more balanced view of one's true energy expenditure.
