Two studies show the use of consumer wearables is increasing in clinical research, but limitations in accuracy still exist

By Heather Mack
Share

The proliferation and ever-increasing sophistication of wearable activity trackers makes clear that we have come a long way from analog pedometers and clunky heart rate monitors, but new research suggests those metrics may be the only ones modern wearables can accurately track.

Stanford researchers took a hard look at seven different consumer wearables with a group of 60 volunteers, who wore the devices while walking or running on treadmills or riding a stationary bike. At the same time, their heart was also measured with the gold standard of electrocardiograph. Additionally, metabolic rate was estimated with an instrument for measuring the oxygen and carbon dioxide in breath.

According to the results of the study, published in the Journal of Personalized Medicine, all the devices were pretty spot-on with heart rate. Of the Apple Watch, Fitbit Surge, Basis Peak, Microsoft Band, Mio Alpha 2, PulseOn and the Samsung Gear S2, six had an error of less than 5 percent with heart rate. Some were more accurate than others, and were affected by things like the user’s weight or skin tone.

Once it got to caloric burn, however, none of the devices fared well. The most accurate device was off by an average of 27 percent (Fitbit Surge), and the worst was off by almost 93 percent (PulseOn).

“The heart rate measurements performed far better than we expected,” lead study author Euan Ashley told Stanford News. “But the energy expenditure measures were way off the mark. The magnitude of just how bad they were surprised me.”

Even though consumer devices aren't held to the same standards as medical grade devices, Ashley noted, people are still using them to base life decisions on the data provided. But, to the devices' credit, the researchers conceded it's tough to create algorithms strong enough to predict caloric expenditure for individuals. 

"My take on this is it's very hard to train an algorithm that would be accurate across a wide variety of people because energy expenditure is variable based on someone's fitness level, height and weight, etc," gradute researcher Anna Shceherbina told Stanford News.

Measuring caloric burn is a tough one, and no one device on the market seems to have cracked the code yet. That hasn't stopped everyday people and researchers from using them, however. Just a few weeks ago, researchers at the University of California Davis tapped Healbe, makers of the GoBe 2 Smart Life Band fitness tracker, to validate its caloric intake and expenditure capabilities. 

But even with the energy expenditure limitations, researchers are increasingly turning to consumer wearables in clinical studies. This is especially true for Fitbit, which has been part of more than 350 publications. The Fitabase Research Librarycompiled by Fitbit’s data-crunching and research facilitation partner Fitabase, puts every single research study or trial that used Fitbit in one place. A recent analysis in the Journal of the Federation of American Societies for Experimental Biology (FASEB) found Fitbit devices were the most common consumer physical activity monitors in published work, accounting for 83 percent of clinical trials using consumer devices, and 95 percent of NIH-funded research. The most common primary outcome measure is physical activity, and most conditions studied were mental and brain related. Oddly, body weight-related conditions were not the first.