Activity devices have their uses, but don’t rely on them for accurate heart-rate data
People are increasingly relying on fitness devices for a window into their daily heart health, despite mounting evidence these consumer products are imprecise.
A recent study published in JAMA Cardiology showed that wrist-worn heart rate monitors were more inaccurate at higher rates of physical activity.
With cardiac patients increasingly relying on monitors to stay within physician-recommended, safe heart rate thresholds during rehabilitation and exercise, accuracy was important, the authors said.
In the study, four fitness devices – the Apple Watch, Fitbit Charge HR, Basis Peak and Mio Fuse – were compared in 50 health adults against standard ECG measurements.
All four of the devices fell below the gold standard of ECG in terms of accuracy.
The subjects wore both the fitness device and ECG before, during and after exercise on a treadmill. Their heart rate ranged from 49b/m to 200b/m.
Two of the devices, the Fitbit and the Basis Peak, underestimated heart rate during vigorous exercise, and were described as having “suboptimal accuracy during moderate exercise”.
Apple Watch and Mio Fuse performed better, but none of the devices achieved the accuracy of the standard ECG.
“In general, accuracy of wrist-worn monitors was best at rest and diminished with exercise,” the authors said.
Associate Professor Andre La Gerche, head of sports cardiology at Baker IDI, told TMR the wrist-worn fitness devices were inappropriate for monitoring heart rates in cardiac patients because they were not consistently accurate.
“This study shows they are 80% accurate in some cases and my anecdotal experience is that they might be 80% accurate on one day and 0% accurate on another,” he said.
However, not all experts agreed. Professor Robyn Gallagher, the chair of the Cardiovascular Nursing Council Cardiac Society of Australia and New Zealand, told TMR that her research had shown that Fibits were “accurate, sensitive and specific”.
“No doubt there can be an overcount of 10% compared to the gold standard measure worn on the waist, however, this overcount is primarily due to the wrist location.
“For us, few patients will tolerate waist worn monitors in their normal lives, so it’s essential that acceptable trackers are found to not only monitor activity but to motivate it.”
But Professor La Gerche said, in his experience, wrist-worn heart-rate monitors were “virtually useless” for individual general fitness tracking.
“If you really want to know your heart rate, and how it’s going up and down, then use a heart rate monitor,” he said.
However, Professor La Gerche said wrist worn devices had the potential to deliver clinically useful long-term data on individuals.
“If we are using it as a weight loss or activity tracker, or to tell people that we want them to do more exercise, then it is very good,” he said.
Wrist-worn heart monitors have received media attention in Australian as life-saving devices, with an Apple Watch helping to pick up previously undetected atrial fibrillation and a Fitbit alerting a man to a life-threatening arrhythmia.
But their accuracy has also been called into question, notably by plaintiffs in a US-based class action lawsuit filed against Fitbit in January.
Professor La Gerche said the inaccuracies of the devices did not mean the data was useless in population-level studies.
“There is someone sitting in Silicon Valley right now with access to hundreds of thousands of people’s data – anonymised hopefully – that will be describing population activity levels,” he said.
“Knowing what we are doing over one decade compared with the next could be really important discoveries for public health.”
