AORTIC STIFFNESS ASSESSED BY HIGH-RESOLUTION PHOTOPLETHYSMOGRAPHY AND SINGLE-LEAD ECG

Abstract

Objective: Aortic stiffness is an independent predictor of cardiovascular risk and plays an important role for the development of cardiovascular disease. However, it is rarely measured in clinical practice as it requires complicated methodology and expensive equipment. We assessed whether high-resolution photoplethysmography (PPG) and single-lead ECG, two methods available in current consumer devices, could be used to measure aortic stiffness. Design and method: Simultaneous recordings of infrared PPG (earlobe and index finger), single lead ECG (PowerLab, ADInstruments) and continuous blood pressure (BP) by the volume-clamp method (Ohmeda 2300, Finapres) at 1000 Hz were made in 33 subjects and repeated after 1–3 wks. Carotid-femoral pulse wave velocity (cfPWV, SphygmoCor, AtCor Medical) and aortic pulse wave velocity (aoPWV, Arteriograph, Tensiomed) were measured as reference. PPG fiducial point identification was refined using pulse wave shapes and the blood pressure signal. Pulse wave analysis extracted 23 features previously associated with arterial stiffness, which were ranked according to univariate strength of correlations with cfPWV and aoPWV. The feature with highest rank was used in a multiple linear regression model with heart rate, body height, mean arterial BP and PAT (ECG R-peak to PPG foot) as covariates. Results: Mean age was 44 (range 21–66) ys, 19 were men, and 5 had hypertension; office BP was 119 ± 12/72 ± 9 mm Hg. PWV was similar at visit 1 and 2 (cfPWV 6.74 ± 1.17 vs 6.80 ± 1.18; aoPWV 7.66 ± 1.80 vs 7.55 ± 1.52). Finger-PPG showed stronger relations with PWV than earlobe-PPG, and 17 of the 23 extracted features showed correlations (P < 0.05) with cfPWV and aoPWV. The ratio of acceleration features (b-c-d-e)/a showed highest correlations with cfPWV and aoPWV (r2 = 0.58 and 0.64, respectively), that improved in multiple regression (r2 = 0.72 and 0.71, respectively; both P < 0.001). Conclusions: High-resolution finger-PPG combined with advanced feature extraction can be used to assess aortic stiffness. Adding body height, BP and single-lead ECG improve the strength of the correlation. Easily available PPG based aortic stiffness may facilitate cardiovascular risk prevention, evaluation, and improve treatment.