NATION NOW

Your fitness tracker could be way way off

Sean Rossman
USA TODAY
Your fitness tracker may not be that accurate.

That smart device dangling from your wrist may not be the best way to track your calories. In fact, it could be leading you astray.

The Stanford University Medical Center tested seven popular wearable fitness trackers and found most failed to accurately track how much energy people were burning during workouts. Actually, according to the study, the wearables weren't even close compared to medical-grade measurements.

However, wearable makers stand by their numbers, despite the massive inaccuracies found in the report.

The study took 60 people and had them perform a range of exercises — walking, running, cycling — at different intensities while wearing the devices, which included the Fitbit Surge, Apple Watch, Basis Peak, Microsoft Band, Mio Alpha 2, PulseOn and the Samsung Gear S2. The results for heart rate and energy expended were compared to oxygen and carbon dioxide measurements and electrocardiograph readings, which researchers described as "gold standards" of the medical field.

Graduate student Anna Shcherbina, a lead co-author on the study, said a 10% range of error would have been acceptable for home use. But the most accurate devices were off by an average of about 27%. The least accurate, the PulseOn, was off by 93%. 

The devices were accurate when it came to measuring heart rate, with all coming in under 10% error.

"The heart rate measurements performed far better than we expected," said senior author Euan Ashley, a Stanford professor, "but the energy expenditure measures were way off the mark. The magnitude of just how bad they were surprised me."

Read more:

Here's what happens to your body when you walk those recommended 10,000 steps

How precise is your fitness band or Wi-Fi scale?

Shcherbina said the devices tended to underestimate calories burned during the less stressful exercises, and overestimated during the more strenuous workouts.

The best performing devices were the Apple Watch and the Fitbit, which both posted median errors of a little more than 20%. They were followed by the Microsoft Band and the Basis Peak. And finally, the PulseOn. Shcherbin said the Mio Alpha 2 and Samsung Gear S2 weren't included in the energy expenditure study because they didn't provide measurements every minute.

The takeaway, Shcherbina said, is the devices can be useful, but shouldn't be used as an exact measurement of your daily activity.

"You can't really trust them at this point," she said. "They can be a useful guideline, but still, you can't rely on the device 100%."

The authors didn't provide a reason for way the energy expenditure numbers were so off. Shcherbina said it's difficult to create a device that's accurate in the face of different weights, heights and fitness levels.

Fitbit issued a statement saying its confident in the performance of its device.

"Overall, the success of Fitbit products comes from enabling people to see their overall health and fitness trends over time," the statement said, "it's these trends that matter most in achieving goals."

PulseOn CEO Jaakko Hattula suggested the researchers may not have properly adjusted the user settings - such as height, weight and fitness level - for each participant. Not doing so, he said, could alter the outcome.

"We think the excess error reported in (energy expenditure) is not representative in this study, due to this methodological error," Hattula said in an email.

The study was published in the latest edition of the Journal of Personalized Medicine.

Follow Sean Rossman on Twitter: @SeanRossman