Model-Based Assessment of the Reference Values of CAVI in Healthy Russian Population and Benchmarking With CAVI0

• Published in: American journal of hypertension Vol. 37; no. 1; pp. 77 - 84
• Main Authors: Safronova, Tatiana, Kravtsova, Anna, Vavilov, Sergei, Leon, Cristina, Bragina, Anna, Milyagin, Victor, Makiev, Ruslan, Sumin, Alexei, Peskov, Kirill, Sokolov, Victor, Podzolkov, Valery
• Format: Journal Article
• Language: English
• Published: United States 01.01.2024
• Subjects: Adult; Benchmarking; Blood Pressure - physiology; Cardio Ankle Vascular Index; Female; Humans; Male; Reference Values; Russia; Vascular Stiffness - physiology; Russia; arterial stiffness; reference values of CAVI; CAVI0; blood pressure; multivariable linear regression; model-based analysis; hypertension; CAVI
• ISSN: 0895-7061; 1941-7225; 1941-7225
• DOI: 10.1093/ajh/hpad082

Cardio-ankle vascular index (CAVI) and its modified version (CAVI0) are promising non-invasive markers of arterial stiffness, extensively evaluated primarily in the Japanese population. In this work, we performed a model-based analysis of the association between different population characteristics and CAVI or CAVI0 values in healthy Russian subjects and propose a tool for calculating the range of reference values for both types of indices. The analysis was based on the data from 742 healthy volunteers (mean age 30.4 years; 73.45% men) collected from a multicenter observational study. Basic statistical analysis [analysis of variance, Pearson's correlation (r), significance tests] and multivariable linear regression were performed in R software (version 4.0.2). Tested covariates included age, sex, BMI, blood pressure, and heart rate (HR). No statistically significant difference between healthy men and women were observed for CAVI and CAVI0. In contrast, both indices were positively associated with age (r = 0.49 and r = 0.43, P < 0.001), however, with no clear distinction between subjects of 20-30 and 30-40 years old. Heart rate and blood pressure were also identified as statistically significant predictors following multiple linear regression modeling, but with marginal clinical significance. Finally, the algorithm for the calculation of the expected ranges of CAVI in healthy population was proposed, for a given age category, HR and pulse pressure (PP) values. We have evaluated the quantitative association between various population characteristics, CAVI, and CAVI0 values and established a method for estimating the subject-level reference CAVI and CAVI0 measurements.