If you rely on a smartwatch to track your health data, you’re not alone: In 2019, 21% of American adults said they used a fitness tracker, according to Pew Research Center—and the number has likely only grown since then. It’s certainly convenient to track your step count, calories, and heart rate with a small accessory.
But how credible are those statistics, really?
A new study adds to a growing body of research that suggests fitness trackers are likely not as accurate as you’d hope. Published in PLOS One, the study analyzed the accuracy of caloric data generated by smartwatches from four brands (Apple, Garmin, Samsung, and Fitbit). The researchers found that the devices had a significant error rate, with estimates deviating an average of 15-25% from participants’ actual calorie expenditures. Across brands, the error rate increased with participants’ body fat percentage.
“If a technology is going to be used broadly for health and exercise, we need to know not only how accurate it is overall, but whether it performs similarly across different people,” says Jason Kostrna, Ph.D., C.M.P.C., lead study author and associate professor of sport and exercise psychology and kinesiology and exercise science at Florida International University.
But don’t toss your Fitbit just yet: Experts say smartwatch data can still be useful—as long as you interpret it in the right way. Ahead, scientists break down the results and what they mean for you.
Meet the experts: Jason Kostrna, Ph.D., C.M.P.C., lead study author and associate professor of sport and exercise psychology and kinesiology and exercise science at Florida International University; Adam S. Lepley, Ph.D., A.T.C., director of the PRIME Lab and assistant professor at the University of Michigan’s School of Kinesiology; Amanda Velazquez, M.D., director of obesity medicine at Cedars-Sinai.
What did the study find?
The researchers had 58 adults cycle on a recumbent bike while wearing different smartwatches. While they exercised, participants were hooked up to lab equipment that measures exact energy expenditure by analyzing gases in the breath. Then, the researchers checked the smartwatches’ caloric estimates against their lab data, and measured the difference—and assessed whether the accuracy varied based on participants’ body composition or skin tone. The project was a collaboration between Dr. Jessica Ramella-Roman at FIU’s Medical Photonics Laboratory and Dr. Cristina Palacios at FIU’s Robert Stempel College of Public Health and Social Work; Dr. Ramella-Roman in particular “has examined how factors such as body composition and skin tone can influence the optical signals used by wearable devices,” Kostrna explains.
Ultimately, they found that all of the devices gave inaccurate estimates, but this varied across brands. Apple was the most accurate, Garmin and Samsung tended to overestimate calorie burn, and Fitbit “exhibited the largest variability, with extreme outliers and occasional missing data,” the researchers write.
While the percentage of error increased with participants’ percentage of body fat, skin tone did not have an effect. “That does not mean skin tone can never affect wearable accuracy,” Kostrna notes. “We examined a specific range of skin tones under a specific exercise condition.”
Why might smartwatch caloric data be inaccurate?
Smartwatches give us lots of metrics, many of which aren’t measured directly. “Instead, watches measure signals such as movement, optical pulse waveforms, or location, and then use proprietary algorithms to estimate things like calories burned, VO2max, sleep stages, or recovery scores,” says Adam S. Lepley, Ph.D., A.T.C., director of the PRIME Lab and assistant professor at the University of Michigan’s School of Kinesiology. “Error can be introduced at several points along the way, from sensor noise and how the device is positioned on the wrist, to motion artifact, differences between individuals, and the assumptions built into the algorithm.”
Add Prevention as a preferred source in your Google search results.
Use your Google account to add this to your settings.
All in all? “A smartwatch can give you a very precise-looking number, but that does not necessarily mean the underlying estimate is equally precise,” he says. In fact, this was something he researched himself, too—Lepley was the lead author of a recent narrative review on smartwatch data, published in Sensors, which concluded that the accuracy of smartwatch data varies significantly between metrics.
Kostrna agrees: “Even if the watch measures heart rate or movement reasonably well, that does not necessarily mean it can translate those measurements into an equally accurate estimate of calories burned.”
Should you use smartwatch data to help with weight loss?
Many of the smartwatch users most interested in caloric data are thinking about weight loss. However, people should be wary of relying too heavily on this data—especially those with a higher percentage of body fat, for whom the numbers may be even more inaccurate.
“Where I become concerned is when someone treats the calorie estimate as an accounting system, for example, ‘my watch says I burned 600 calories, so I can eat an additional 600 calories.’ Based on our findings, I don’t think the technology currently supports that level of precision,” Kostrna says.
What’s more, even if the data were perfect, it wouldn’t be a good idea to fixate on it: “Exercise has tremendous benefits beyond the number of calories burned,” he says. “I wouldn’t want an inaccurate calorie estimate on a watch to become the measure of whether someone had a ‘good’ workout.”
Amanda Velazquez, M.D., director of obesity medicine at Cedars-Sinai, is wary of overrelying on calorie-counting in any way—digital or analog. “I caution patients about becoming too focused on calorie tracking and advise them to take a step back, and view nutrition more holistically and focus more so on nutrition quality,” she says. “The main takeaways are to follow a balanced, diversified nutrition plan, such as the Harvard Healthy Eating Plate, eat at relatively consistent times each day, and maintain adequate hydration with water.”
This is particularly true for those living with an eating disorder. “Calorie tracking can sometimes worsen anxiety, restriction, and lead to undernourishment,” Dr. Velazquez says.
What’s the takeaway?
Experts say smartwatch data can still be valuable when used correctly. Rather than getting caught up in individual numbers and daily variations, try looking at broader changes over time.
“I use one for my own health tracking. So, I think there are definitely benefits to using smartwatches, as long as people have the right expectations,” Lepley says. “They can be very useful for tracking behavior and monitoring trends over time, especially for things like daily activity, step count, and heart rate under many common conditions.”
Kostrna agrees. “Pay attention to things like heart rate, exercise duration, distance, and pace when appropriate, as well as your patterns over time,” he says. “I also think there is value in combining the data with what you know about yourself. How hard did the workout feel? What does your heart rate normally look like at that intensity? How are your workouts progressing over time? If weight management is your goal, what is actually happening to your weight and body composition over weeks and months?”
Your smartwatch data can complement that self-reflection: “A smartwatch can add useful information to that picture, but it shouldn’t replace it,” Kostrna says.
Read the full article here
