Fitbit's optical sensors don't accurately record heart rates during exercise, according to a new study published last week by researchers at Cal Poly. The average difference between Fitbit devices and an electrocardiogram (ECG) is approximately 20 beats per minute.

That's well beyond any reasonable margin of error, according to the Lief Cabrasser law firm, which commissioned the study as part of a consumer fraud class-action suit it filed earlier this year. The plaintiffs claim that several different Fitbit models using the company's PurePulse technology don't accurately measure heart rates.

PurePulse, introduced in 2014 with the Fitbit Charge HR, is an optical monitor that reads the wearer's heart rate once every second. During exercise, you can set a target heart rate zone, or rely on what Fitbit calls "Simplified Zones" that have ranges for fat burning, cardio, and peak heart rate.

The Cal Poly researchers tested PurePulse using the Charge HR and Fitbit Surge on 43 separate subjects. The tests lasted for 65 minutes each during a variety of activities such as jogging, stair climbing, and jumping rope. The subjects were also connected to an ECG, which doctors routinely use to assess the heart's electrical and muscular functions.

The researchers concluded that during moderate to high-intensity exercise, the PurePulse trackers recorded a heart rate that differed from the ECG by an average of 19.2 beats per minute. They were also inconsistent, with the Charge HR measuring different heart rates compared to the Surge.

A Fitbit spokesperson downplayed the accuracy of the study, and said the company is "vigorously defending" itself against the class action lawsuit. 

"What the plaintiffs' attorneys call a 'study' is biased, baseless, and nothing more than an attempt to extract a payout from Fitbit," the spokesperson said in a statement. "It lacks scientific rigor and is the product of flawed methodology. It was paid for by plaintiffs' lawyers who are suing Fitbit, and was conducted with a consumer-grade electrocardiogram – not a true clinical device, as implied by the plaintiffs' lawyers. Furthermore, there is no evidence the device used in the purported 'study' was tested for accuracy."

In her review of the Charge HR last year, PCMag's Jill Duffy said found that "the heart rate monitor was reasonably accurate when I was at rest, based on what I know about my body, other heart rate monitors that I used for comparison, and the two-fingers-to-the-jugular test. I consider plus or minus 5bpm to be roughly accurate.

"When I was in motion with an elevated heart rate, either due to exercise or running to catch a subway, the Charge HR was less reliable," she said. "A few times, I noticed an oddly high reading, and I'd first readjust the band to make sure it was snug and a few inches up my wrist. When I did that, the heart rate reading very quickly returned to a measurement that was more in line with my expectations."