Validation of the Polar 360 Upper Arm Strap for Heart Rate Measurement Across a Variety of Activities and Intensities

Abstract

Introduction & Purpose: Wearable devices using photoplethysmography (PPG) are increasingly used for non-invasive monitoring of physiological responses during daily life and physical activity. However, accuracy can vary with movement and sensor placement, and continuous validation remains essential. Compared with wrist-worn devices, upper-arm placements showed higher-quality PPG signals and are less affected by motion artefacts (Schweizer & Gilgen-Ammann, 2025). This study aimed to validate heart rate (HR) measurements obtained from the Polar 360 upper-arm strap across a wide range of activities and intensities.

Methods: Fourteen healthy adults (7 females; 173.4 ± 9.0cm, 71.9 ± 10.2kg, 33.4 ± 9.2years) with fair skin completed six laboratory-based activities, each separated by a two minutes of transition/rest: (1) 2min lying, (2) 5min sitting, (3) treadmill walking–jogging–running–jogging–walking sequence (5×3min), (4) ergometer cycling at low–moderate–high–very high–low intensities (5×3min), (5) 8min strength training circuits, and (6) 8min high-intensity interval training (HIIT) circuits. HR was measured concurrently using the Polar 360 PPG (1Hz) sensor worn on the non-dominant upper arm and the Polar H10 electrocardiogram (ECG, 1Hz) chest strap as criterion. HR data were averaged in 10-s intervals and analyzed for systematic bias (mean difference), mean absolute error (MAE), mean absolute percentage error (MAPE), Lin’s concordance correlation coefficient (CCC) (McBride, 2005), and the percentage of data within ±5bpm of the criterion.

Results: Criterion HR values ranged from 69.4 ± 11.9bpm (lying) to 161.1 ± 15.9bpm (HIIT). Overall, the Polar 360 showed substantial agreement with the criterion (bias = −0.65bpm, MAE = 1.93bpm, MAPE = 1.62%, CCC = 0.987), with 92.3% of 10-s HR intervals within ±5bpm of the H10 reference. Activity-specific accuracy varied markedly: cycling showed excellent agreement (bias = −0.02bpm, MAE = 0.62bpm, MAPE = 0.44%, CCC = 0.999, 99.5% within ±5bpm), whereas HIIT produced poorer results (bias = −1.93bpm, MAE = 4.18bpm, MAPE = 2.71%, CCC = 0.828, 81.3% within ±5bpm). Walking demonstrated near-excellent agreement, lying yielded similarly low accuracy as HIIT, and sitting and strength training showed moderate agreement (CCC = 0.904 and 0.910, respectively).

Discussion & Conclusion: The Polar 360 demonstrated strong overall validity for HR measurement compared with ECG, though accuracy varied by activity. Consistent with previous findings, PPG-based HR accuracy decreased with greater motion and reduced peripheral blood flow (Schweizer & Gilgen-Ammann, 2025). Nevertheless, the Polar 360 outperformed most commercial wearables across comparable activity conditions. Its combination of accuracy, comfortable upper-arm placement, and long battery life make it a promising alternative to ECG-based chest straps for continuous HR monitoring in both research and applied settings.

References

McBride, G. B. (2005). A proposal for strength-of-agreement criteria for Lin’s concordance correlation coefficient (NIWA Client Report: HAM2005-062). National Institute of Water & Atmospheric Research. https://www.medcalc.org/download/pdf/McBride2005.pdf

Schweizer, T., & Gilgen-Ammann, R. (2025). Wrist-worn and arm-worn wearables for monitoring heart rate during sedentary and light-to-vigorous physical activities: Device validation study. JMIR Cardio, 9, Article e67110. https://doi.org/10.2196/67110