Just how accurate are Fitbits?

Many users of activity trackers have always harboured suspicions: How accurate are these things?

A handful of tests by journalists and researchers have tried to bring clarity to the issue. Results, alas, have been mixed.

The latest study, released by the plaintiffs in a class-action lawsuit against Fitbit, found that the pulse-monitoring technology used in the company’s wrist-bound Surge and Charge devices was “highly inaccurate during elevated physical activity.”

Researchers from California State Polytechnic University, Pomona, had 43 subjects wear the devices as they ran, jogged and jumped rope, among other activities, and then compared the readings with those of an electrocardiogram.

During moderate to high-intensity exercise, Fitbit’s sensor was off by an average of about 19 beats a minute, it found.

The study, which was not peer-reviewed and was commissioned by the plaintiffs’ law firm, drew a strong denunciation from Fitbit, which released a statement calling it “biased, baseless and nothing more than an attempt to extract a payout from Fitbit.”

Fitbit, a young company based in San Francisco, is a leader in the wearables industry with a market capitalisation of about $3 billion.

Alex Montoye, an assistant professor of clinical exercise physiology at Ball State University said that the Pomona study needed rigorous vetting before any conclusions could be drawn.

Inaccurate

In another recent study in the journal Medicine & Science in Sports & Exercise, two other functions of Jawbone and Fitbit devices, the step and calories counts, were put to the test by a research team led by Montoye.

The researchers tracked 30 healthy adults as they performed various tasks. When the subjects were engaged in sedentary activities, like using a computer, the devices did relatively well. But both overcounted and undercounted as things intensified, Montoye said.

The researchers were surprised to find, however, that their own laboratory equipment got the measurements wrong, as well.

“The Fitbit and Jawbone monitors worked about as well as the best we can do as far as measuring physical activity,” Montoye said.

The pulse trackers have been a more recent addition to the suite of activity trackers on the market. Most use optical sensors that reflect light onto the skin to detect blood volume.

Independent attempts to assess their accuracy have not exactly settled the question. In 2014, CNET tested several devices and found that they were mostly accurate during rest, but ran into trouble during exercise.

A test by an Indianapolis broadcaster in collaboration with Ball State earlier this year, that included devices from Fitbit and Garmin, also found the heart rate calculations to be off. An assessment by Consumer Reports in January, meanwhile, found that Fitbit’s Charge HR and Surge models accurately recorded heart rates at everything from a leisurely walk up to a fast run.

If that all leaves you unsure about what to trust, there is always a very affordable, perfectly accurate alternative: a stopwatch and two fingers held to your neck.