Threshold values of brachial cuff-measured arterial stiffness indices determined by comparisons with the brachial–ankle pulse wave velocity: a cross-sectional study in the Chinese population

BACKGROUND: Arterial Velocity-pulse Index (AVI) and Arterial Pressure-volume Index (API), measured by a brachial cuff, have been demonstrated to be indicative of arterial stiffness and correlated with the risk of cardiovascular events. However, the threshold values of AVI and API for screening increased arterial stiffness in the general population are yet to be established. METHODS: The study involved 860 subjects who underwent general physical examinations (M/F = 422/438, age 53.4 ± 12.7 years) and were considered to represent the general population in China. In addition to the measurements of AVI, API and brachial-ankle pulse wave velocity (baPWV), demographic information, arterial blood pressures, and data from blood and urine tests were collected. The threshold values of AVI and API were determined by receiver operating characteristic (ROC) analyses and covariate-adjusted ROC (AROC) analyses against baPWV, whose threshold for diagnosing high arterial stiffness was set at 18 m/s. Additional statistical analyses were performed to examine the correlations among AVI, API and baPWV and their correlations with other bio-indices. RESULTS: The area under the curve (AUC) values in ROC analysis for the diagnosis with AVI/API were 0.745/0.819, 0.788/0.837, and 0.772/0.825 (95% CI) in males, females, and all subjects, respectively. Setting the threshold values of AVI and API to 21 and 27 resulted in optimal diagnosis performance in the total cohort, whereas the threshold values should be increased to 24 and 29, respectively, in order to improve the accuracy of diagnosis in the female group. The AROC analyses revealed that the threshold values of AVI and API increased markedly with age and pulse pressure (PP), respectively. CONCLUSIONS: With appropriate threshold values, AVI and API can be used to perform preliminary screening for individuals with increased arterial stiffness in the general population. On the other hand, the results of the AROC analyses imply that using threshold values adjusted for confounding factors may facilitate the refinement of diagnosis. Given the fact that the study is a cross-sectional one carried out in a single center, future multi-center or follow-up studies are required to further confirm the findings or examine the value of the threshold values for predicting cardiovascular events.