Validation of photoplethysmography-derived short-term heart rate variability using a wearable device
摘要
Heart rate variability (HRV) reflects cardiac autonomic function and cardiovascular health. While electrocardiography (ECG) remains the gold standard for HRV assessment, photoplethysmography (PPG)-based wearables offer an alternative for monitoring. This study aimed to validate the wrist-worn PPG device Bora band® for short-term HRV assessment against standard ECG and to characterize domain-specific agreement patterns. In this prospective, single-center observational study, 66 participants in sinus rhythm underwent simultaneous high-resolution 12-lead ECG and wrist-based PPG recording for 5 min and 30 s. PPG-derived interbeat intervals were extracted and compared to ECG-derived R-R intervals. Sixteen HRV metrics from time-, frequency-, and non-linear domains were computed for both modalities. Agreement was assessed using biweight mid-correlation (ρ), Cliff’s delta (δ), and equivalence testing (TOST). Strong agreement was observed for mean heart rate (ρ = 1.0, δ = 0.104, TOST p < 0.001), standard deviation of normal-to-normal intervals (ρ = 0.98, δ = 0.037, TOST p = 0.003), deceleration capacity of heart rate (ρ = 0.95, δ = 0.071, TOST p = 0.014), coefficient of variation of normal-to-normal intervals (ρ = 0.98, δ = 0.035, TOST p = 0.002) and Poincaré Plot standard deviation 2 (ρ = 0.99, δ = 0.074, TOST p < 0.001). Moderate agreement was noted for long-term fractal scaling exponent and very low-, low- and high-frequency power. Weaker agreement appeared for short-term variability and entropy metrics. Bland-Altman analysis indicated minimal bias without systematic error. Under controlled resting conditions, wrist-based PPG provides reliable HRV indices compared with ECG-derived HRV. These findings support the use of selected PPG-derived HRV parameters for short-term assessment in clinical settings. Further validation in real-world settings will be necessary.