Impact of brachial-ankle pulse wave velocity and ankle-brachial blood pressure index on mortality in hemodialysis patients

Abstract

Background: Pulse wave velocity (PWV) and ankle-brachial blood pressure index (ABPI) are markers for atherosclerosis, and each predicts mortality in patients undergoing hemodialysis. However, there have been no studies in the past that compared head-to-head the clinical validity of these 2 parameters. Compared with conventional aortic PWV, brachial-ankle PWV (baPWV) is considered simple and thereby easily applicable to clinical use.

Methods: To clarify the relationship between baPWV and ABPI and assess their prognostic values, we analyzed 785 hemodialysis patients with a mean age of 60.2 +/- 12.5 (SD) years for whom ABPI and baPWV at baseline had been measured simultaneously and who were followed up for 33.8 +/- 10.8 months.

Results: Of 785 patients, 131 deaths were recorded. In Kaplan-Meier analysis, all-cause mortality was progressively and significantly greater from the lowest quartile of baPWV onward (log-rank test, 41.8; P < 0.001). However, in Cox proportional hazards analysis, the impact of baPWV was insignificant when ABPI was included as a covariate. ABPI maintained strong predictive power in this model. When patients who had advanced peripheral arterial occlusive disease (ABPI < 0.9) were excluded from analysis, patients with the highest quartile of baPWV had significantly increased hazard ratios of all-cause (hazard ratio, 4.08; 95% confidence interval, 1.46 to 11.43; P < 0.007) and cardiovascular (hazard ratio, 7.03; 95% confidence interval, 1.49 to 33.08; P < 0.014) mortality. The predictive power of baPWV in this population was independent from other covariates associated with atherosclerotic disorders.

Conclusion: In a head-to-head comparison, ABPI, but not baPWV, showed strong power in predicting the mortality of hemodialysis patients. However, baPWV was useful to pick a high-risk population in patients with ABPI greater than 0.9. Thus, screening hemodialysis patients by means of baPWV and ABPI provides complementary information in identifying a high-risk population.