James Brickman bought a Fitbit Flex in 2013. Two years later, he was behind a class-action lawsuit filed against the company because the device didn't measure sleep data as promised, and was inaccurate by as much as 67 minutes per night. The company asserts that the lawsuit is meritless; litigation is ongoing.
It wouldn't be the first time fitness trackers would be found to be less than perfect.
Fitness trackers typically log active minutes, calories consumed and burned, duration and intensity of exercise, sleep habits, stairs climbed and steps taken. Some trackers sync with other devices, such as the Fitbit Aria Wi-Fi Smart Scale, or with apps, such as Endomondo. Of all that collected data, health and fitness-related information is what's most important to consumers. Seventy-seven percent of consumers want a wearable that tracks and improves their exercise, and 75 percent want a wearable that stores and tracks their medical information.
Researchers at Iowa State University studied the accuracy of four consumer fitness trackers (the FitBit Flex, Jawbone UP24, Misfit Shine and Nike+ FuelBand SE) and two research activity monitors (the Actigraph GT3X+ and BodyMedia Core), measuring the error rate of each monitor's estimate of calories burned during sedentary, aerobic and resistance activity.
They collected data from 52 participants who completed three sets of activity: 20 minutes of a sedentary activity, followed by 25 minutes of aerobic activity, followed then by 25 minutes of strength training. Participants were allowed five minutes between each activity.
The six activity monitors had overall error rates that were considered reasonable. The FitBit Flex, for example, had a good overall error rate at 16.8 percent, while the 30.4 percent overall error rate of the Misfit Shine rated the worst in the study. But once the tracker's results are examined within specific activities rather than overall, the large amounts of error emerge.
"By looking at the most-commonly performed activities in exercise and daily living settings, we can examine where the errors occur," lead author Yang Bai says in a statement. The FitBit Flex, for instance, resulted in a 29.4 percent error rate during sedentary activities, a 34.7 percent error rate during aerobic activities and a 31.6 percent error rate during resistance activities.
But don't get your hopes up. These disparities aren't significant enough to skew something that may sound silly but is of great importance to many Fitbit users: rankings among buddies who have a bit of friendly competition on the Fitbit dashboard.
“I think the key to a consumer is not so much if the activity monitor is accurate in terms of calories, but whether it's motivational for them and keeps them accountable for activity in a day,” explains Dr. Greg Welk, associate professor kinesiology at Iowa State University in a statement. But there's no case where these trackers could cause a person engaging in a reading a book to appear to be more active than someone who is, for example, walking briskly — these trackers aren't actually to blame for such things as artificially inflating a couch potato's total calories burned. The data, says Welk in an email, doesn't support the idea of “sabotage or some type of conspiracy theory about how the data are tracked."
"The issue with error is that it is largely variable and not systematic," he continues. "Error would tend to wash out and just contribute to noise in these comparisons."