¹Department of Medicine and Dentistry at Royal Perth Hospital, University of Western Australia, Perth, Western Australia, Australia

Correspondence: Dr. Richard J. Woodman, School of Public Health (Building 400), Curtin University of Technology, Kent Street, Bentley, Perth, WA 6102, Australia E-mail: r.woodman@exchange.curtin.edu.au

^*School of Medicine and Pharmacology (RJW, GFW, LJB), Royal Perth Hospital Unit, University of Western Australia and the Cardiovascular Research Centre, Perth, Australia; Alfred Baker Medical Unit Alfred Hospital (BAK, SEH, AMD), and Wynn Domain, Baker Heart Research Institute, Melbourne, Australia

Received 12 January 2004; Revised 25 August 2004; Accepted 25 August 2004.

Abstract

Background: Several new techniques exist for measuring arterial stiffness, but their association with central pulse wave velocity (PWV), an established measure of central arterial stiffness, has not been compared in the same study. The aim of this study was to compare the agreement of the new techniques with central PWV.

Methods: Fifteen men with coronary artery disease (59 2 years of age) and 15 healthy men (28 1 years of age) were recruited. The following measures were performed in a random sequence and repeated after 1 week: central and distal pulse wave velocity (PWV), large and small artery compliance (C1 and C2, respectively), and stroke volume/pulse pressure (SV/PP) (measured by HDI/PulseWave CR-2000), augmentation index (AIx) and central pulse pressure (CPR) (Sphygmocor), stiffness index (SI) (Micro Medical Pulse Trace), systemic arterial compliance (SAC) (area method), and brachial PP (Dinamap). Methods were compared using correlation coefficients and Bland-Altman analysis.

Results: Although all measures of arterial stiffness except PWV correlated significantly with central PWV (P < .01 for each), Bland-Altman analysis showed poor agreement (confidence interval [CI] >3 Z-scores) between central PWV and C1, C2, SV/PP, and SAC). There was good agreement (CI > 2 Z-scores) between central PWV and SI, AIx, CPP, and brachial PP. The coefficient of variation was lowest with central PWV (7.6%), brachial PP (8.0%), and SV/PP (8.6%) and was significantly higher (P < .05) in increasing order with C1 (11.3%), C2 (15.6%), SI (17.8%), SAC (19.3%), AIx (22.4%), and CPP (25.3%).

Conclusions: Based on our study findings, C1, C2, SV/PP, and SAC show poor agreement with central PWV, an established measure of central arterial stiffness. Indices of this type should therefore be useful in providing a more complete understanding of arterial stiffness. In comparison, SI, AIx, and CPP are more closely related to central arterial stiffness.