I compared my Apple Watch and Amazon Halo to a clinical grade heart monitor while running a treadmill echo cardiogram stress test. The results raise for me some questions about the usefulness of consumer wearables for serious health tracking.
The blue shaded area is the time I was on the treadmill, running an echo stress test (Bruce Protocol). Strapped with ECG leads and a blood pressure cuff, I was instructed to run as long as possible, after which they attached me to an ultrasound device to see how my heart reacts under stress.
The Apple Watch, apparently to save power, didn’t sample my heart rate often enough to record useful data during this short session. Next time I’ll need to remember to manually set the Watch to “Running” mode, which hopefully would increase the sampling interval. Halo, on the other hand, sampled adequately, but didn’t register this session as particularly rigorous. The clinical test showed a max heart rate of 183, but Halo barely topped 140.
This isn’t necessarily a surprise. A more formal study published in Nature npj Digital Medicine (2020) found:
Our conclusions indicate that different wearables are all reasonably accurate at resting and prolonged elevated heart rate, but that differences exist between devices in responding to changes in activity.
This figure, from their paper, shows Apple Watch most closely resembles the clinical-grade ECG for mean absolute error (MAE). But as I learned, that’s only true if the sampling rate is high enough.
So what would work better? Obviously I could get one of the clinical grade sensors, perhaps one worn around the chest. But who wants to wear such an unwieldy device all the time?
For now my conclusion is that everyday wearables, like Apple Watch, are good for showing overall patterns – are you walking more now than a year ago? – but they’re not as accurate as I’d like if you need serious understanding of what’s happening when your body is under stress.
See Bent, B., Goldstein, B.A., Kibbe, W.A. et al. Investigating sources of inaccuracy in wearable optical heart rate sensors. npj Digit. Med. 3, 18 (2020). https://doi.org/10.1038/s41746-020-0226-6