Methods of arterial stiffness calculation and cardiovascular disease events: the multiethnic study of atherosclerosis

Abstract

Background: A wide variety of different formulae have been used to calculate local arterial stiffness with little external validation in relationship to cardiovascular events. We compared the associations of several arterial stiffness calculations in a large, multiethnic cohort. Methods: The multi-ethnic study of atherosclerosis (MESA) is a longitudinal study of 6814 adults without clinical cardiovascular disease (CVD) at enrollment. MESA participants with CVD surveillance through year 2018 and carotid ultrasound (n = 5873) or aorta MRI (n = 3175) at the baseline exam (2000–2002) were included. We analyzed 21 different calculations of local arterial stiffness. Cross-sectional and longitudinal statistical analyses were performed in addition to Cox hazard modeling for associations with CVD events (myocardial infarction, resuscitated cardiac arrest, stroke, adjudicated angina, and cardiovascular death). Results: Carotid artery stiffness calculations had variable correlations with each other (r = 0.56–0.99); aortic stiffness measures were similar (r = 0.66–0.99). Nevertheless, for CVD events, the hazard ratio (HR) per standard deviation change were similar for all carotid stiffness calculations with HRs in the range of 1.00–1.10 (equivalence P < 0.001). For the aorta, aortic distensibility coefficient had a stronger association with CVD events (HR 1.18 [1.02–1.37]) compared to aorta Peterson's elastic modulus (HR 0.98 [0.89–1.07]) and aorta pulse wave velocity (HR 1.00 [0.90–1.11]). HRs between all other aortic stiffness calculations were equivalent (P < 0.01). Conclusion: Different methods of calculating local arterial stiffness largely gave equivalent results, indicating that the variety of different arterial stiffness calculations in use do not cause inconsistent findings.