Impact of Anatomical Placement on the Accuracy of Wearable Heart Rate Monitors During Rest and Various Exercise Intensities
Photoplethysmography (PPG) wearables are widely used for heart rate (HR) monitoring, although accuracy varies by body placement and exercise intensity.
PURPOSE: This study evaluated five PPG-based wearable trackers (WT) against a validated chest strap across rest, warm-up, high-intensity, and graded exercise. A secondary aim was to test whether placement alone alters accuracy when identical devices are worn at different arm sites.
METHODS: Twenty-eight healthy adults (23.8 ± 1.1 yrs; 14 males, 14 females) wore a chest strap reference, a forearm optical armband (WT1), a smartwatch on the wrist (WT2), and three identical wristband-style trackers on the left wrist (WT3), forearm (WT4), and upper arm (WT5). Participants completed 5-min supine rest, 5-min unloaded cycling warm-up, 30-s maximal burpees with 1-min recovery, and five stages of the Modified Bruce treadmill test. HR data were analyzed using mean absolute percentage error (MAPE), Bland–Altman plots, concordance correlation coefficient (CCC), and Deming regression.
RESULTS: Accuracy was highest at rest, with all devices showing minimal bias and strong concordance (CCC > 0.99); WT1 recorded the lowest MAPE (1.6 ± 0.8%). During warm-up, WT5 (upper arm) had the lowest error (2.9 ± 2.2%) and strong agreement (CCC = 0.99), whereas WT2 and WT3 (wrist) showed higher errors (≥5.6%). Burpees yielded the lowest accuracy across devices (CCC < 0.50), though WT5 still outperformed other placements. In the Modified Bruce protocol, WT1 (MAPE: 1.9 ± 1.2%) and WT5 (MAPE: 1.9 ± 1.2%) showed the strongest agreement with the chest strap (CCC > 0.99), while wrist placements displayed proportional and systematic bias.
CONCLUSION: Proximal placements (upper arm, forearm) outperformed distal sites, with accuracy declining most during short, high-intensity exercise. Wrist-based sensors were more susceptible to motion artifacts, whereas proximal sites provided more stable signals and stronger agreement with the chest strap.
SIGNIFICANCE/NOVELTY: Placement alone significantly influenced accuracy in identical trackers, with the upper arm outperforming the forearm and wrist. These findings highlight the value of proximal PPG placements as reliable alternatives to chest straps in clinical monitoring, rehabilitation, and steady-state training, while emphasizing caution with wrist devices during dynamic activities.
