Wearables

Don't look to your wrist for accurate heart rate monitoring, study says

View 2 Images
Cleveland Clinic study casts doubt on wrist-mounted heart rate monitors
VectorStory, remixed by Emily Ferron
Cleveland Clinic study casts doubt on wrist-mounted heart rate monitors
VectorStory, remixed by Emily Ferron
The Fitbit Charge HR, one of the fitness trackers used in the study
Simon Crisp/New Atlas

One recently published study indicates that wrist-based heart rate monitoring may not be as reliable as hoped. Researchers at the Cleveland Clinic's Heart and Vascular Institute found that four leading devices were of inconsistent accuracy, and that accuracy drops during workouts.

The study tested the Fitbit Charge HR, Apple Watch, Mio Fuse and Basis Peak wrist-worn devices. 50 healthy adults were assigned two different wrist monitors at random, along with a proven Polar H7 chest strap monitor as a control. The participants' heart rates were measured at rest, and on a treadmill at 2, 3, 4, 5 and 6 miles per hour.

Researchers found that "[i]n general, accuracy of wrist-worn monitors was best at rest and diminished with exercise." Of the devices tested, the Apple Watch and Mio Fuse were most accurate, followed by the Fitbit Charge HR and lastly, Basis Peak.

But even the top-rated devices display a wide and difficult to predict range of results. For some perspective, 95-percent of the Apple Watch and Mio Fuse readings were between -27 BPM and +29 BPM of the accurate reading. For the Fitbit Charge, 95-percent of readings were between -34 and +39 BPM. The corresponding range for the Basis Peak was between -39 and +33 BPM.

Device performance also skewed differently depending on the level of exertion. For instance, Basis Peak overestimated HR during moderate exercise, with median differences of -8.9 and -7.3 beats per minute (BPM) at 2-3 miles per hour. On the other hand, Fitbit Charge HR skewed low during more vigorous exercise, with median differences of 7.2 and 6.4 BPM at 4 miles per hour.

The Fitbit Charge HR, one of the fitness trackers used in the study
Simon Crisp/New Atlas

Although this study is limited in scope, its results are not encouraging for fitness-tracker fans, many of whom paid a premium for heart rate tracking capabilities. Findings like these could also bolster the class action lawsuit filed against Fitbit earlier this year.

More importantly, they indicate the danger of mistaking a wrist-based heart rate monitor for a medical-grade device, particularly for individuals with cardiovascular conditions. The authors of the study write that "[e]lectrode-containing chest monitors should be used when accurate HR measurement is imperative," especially for cardiac patients that rely on monitors to stay within physician-recommended thresholds.

Given the explosion of the health and fitness wearable market, studies like these could prompt greater investigation and consumer awareness. It could also spur a trend toward non-wrist technology, such as the Moov HR headband-worn heart rate sensor, or other types of heart rate monitoring altogether.

Source: JAMA Cardiology

  • Facebook
  • Twitter
  • Flipboard
  • LinkedIn
4 comments
christopher
Sounds dubious - measuring heartrate is a very simple software problem - for everything to be so wrong, it suggests an experimental messup, not an "everything is bad" result.
wle
the problem is, the method.
a chest strap picks up very accurately, the electrical signals that run the heart.
these things work a different way, shining a laser into skin hoping that some random capillary will be more or less opaque in sync with the pulsing blood,then looking for the returned light.
it;s very iffy, it;s been tried since the 70s at least. it works when you lie in a hospital bed still, in the dark more or less
but add sunlight, movement, knocking about, etc and it gets unreliable. plus not everyone has good circulation in the fingertips or wrist area.
wle
Jeff Goldstein
The chest straps are inconvenient and not very comfortable. I find I am using mine less and less when I exercise. I wonder what will eventually be accurate enough to replace them - wrist with improved sensors, headband or earphone.
Jethro
I realize I'm a little late to the game here, but here's my two cents worth. When I went from my Garmin Chest HRM which ultimately failed to transmit, to a Mio Go wrist HRM there was no difference in my chart from one day to the next. Also, if the Mio Go was plus or minus 28 bpm then the chart would have a very pronounced variance from one moment to the next, yet it is a relatively smooth line. So two devices show very similar results, both very consistent with what I know my heart rate to actually be at any given moment. Ipso facto, these devices are both very accurate.