Impact of Anatomical Placement on the Accuracy of Wearable Heart Rate Monitors During Rest and Various Exercise Intensities.

PURPOSE: This study evaluated the accuracy of arm-based photoplethysmography (PPG) wearable heart rate (HR) monitors in comparison to a validated chest strap reference across various activity levels. METHODS: Twenty-eight healthy adults (14 males, 14 females; aged 23.8 ± 1.1 years) wore six HR monitors: Polar H10 chest strap, Polar Verity Sense on the forearm, Garmin Forerunner on the wrist, and three identical Whoop 4.0 devices placed on the left wrist, forearm, and upper arm. Participants completed rest, warm-up, high-intensity burpees, and graded treadmill exercise. HR data were analyzed using mean absolute percentage error (MAPE), Bland-Altman analysis, concordance correlation coefficient (CCC), and Deming regression. RESULTS: Accuracy was highest at rest and gradually decreased as movement intensity increased. During rest, all devices showed minimal bias and high agreement (CCC > 0.99), with Verity Sense recording the lowest MAPE. In warm-up, Whoop-upper arm and Verity Sense outperformed wrist and Garmin positions, while the Whoop-forearm showed proportional and systematic bias. Burpees resulted in the lowest accuracy across devices (CCC < 0.50), but Whoop-upper arm performed better than other placements. During the Modified Bruce protocol, Verity Sense and Whoop-upper arm had the strongest agreement with the chest strap. Placement of identical Whoop units affected accuracy, with the upper arm outperforming forearm and wrist positions. CONCLUSIONS: PPG wearables provided accurate HR monitoring at rest, during warm-up, and during steady-state graded exercise, particularly when positioned proximally (forearm or upper arm). Accuracy declined during short, high-intensity, full-body activities due to motion artifacts. Both the forearm-mounted Verity Sense and the upper-arm Whoop demonstrated the closest agreement with the chest-strap reference. The intra-device comparison of identical Whoop units confirmed that anatomical placement alone significantly affects accuracy.