HEART-TO-ANKLE PULSE WAVE VELOCITY AS USED IN CARDIO-ANKLE VASCULAR INDEX (CAVI) AND CAVI_0 IS A NEARLY DIASTOLIC WAVE SPEED METRIC

Abstract

Objective: Cardio-ankle vascular index (CAVI) and CAVI_0 are arterial stiffness indices based on the heart-to-ankle pulse wave velocity (haPWV). haPWV is derived from the heart-to-ankle transit time (haTT = tb+tba) as haPWV = L/haTT, with L an estimated transit distance. tb is the time difference between aortic valve closure (causing the aortic dicrotic notch) and the dicrotic notch in the brachial pressure waveform; tba is the time difference between the diastolic feet in the brachial and ankle pressure waveforms. CAVI and CAVI_0 aim to normalise haPWV for blood pressure at the time of measurement. As tb and tba are determined at different pressures — tb at dicrotic notch pressure (P_notch); tba at diastolic blood pressure (DBP; Fig. C) — determining the pressure that governs haPWV (P_haPWV) and hence should ideally be used for normalisation of haPWV is not trivial. CAVI uses P_m = (SBP+DBP)/2 to normalise haPWV, whereas CAVI_0 uses DBP (SBP: brachial systolic blood pressure). We aim to analytically estimate P_haPWV and compare this to the normalisation pressures used in CAVI and CAVI_0. Design and method: Consistent with CAVI/CAVI_0 assumptions, the arterial tree was assumed to consist of tubes with exponential pressure-diameter relationship and uniform stiffness index beta_0, allowing us to calculate PWV at P_notch (PWV_notch) and at DBP (PWV_DBP; Fig. A). P_notch was calculated as 0.55DBP+0.45SBP [Adji et al.,J Hypertens 2012;30:1540–1551]. The resulting PWVs, together with heart-to-ankle (L_ha=170 cm) and heart-to-brachial (L_hb=38 cm) path lengths [Alastruey et al.,11th Int Conf Press Surges 2012;401–442] (Fig. B), allowed analytically estimating tb and tba: tb=L_hb/PWV_notch and tba=(L_ha-L_hb)/PWV_DBP. P_haPWV was finally estimated as the weighted average of P_notch and DBP using tb and tba as weights. Results: Choosing SBP/DBP=120/80 mmHg (and hence P_m=100 mmHg) and beta_0 = 10 yielded tb = 0.048 s and tba = 0.187 s, giving P_haPWV = 84 mmHg. This implies that P_haPWV was much closer to DBP (as used in CAVI_0) than to P_m (as used in CAVI). This result was only marginally influenced by blood pressure: e.g., SBP/DBP = 160/90 mmHg (P_m = 125 mmHg) yielded P_haPWV = 96 mmHg. Conclusions: DBP offers a more accurate estimate of P_haPWV than Pm, supporting the utility of CAVI_0 as an enhancement of CAVI.