8. Usefulness of Brachial-Ankle and Carotid-Femoral Pulse Wave Velocity as Predictive Values of Cardiovascular Events

Abstract

Background

The measurement of carotid-femoral pulse wave velocity (cfPWV) is useful to predict storke and cardiovascular events in hypetensive patients as our previous cohort study Non-invasive Atheroslerotic evaluation in Hypertension (NOAH) sutdy. As the cfPWV is useful, but not easy to measure, it couldn’t be suitable for a practical use. Therefore we evaluated the usefulness of brachial-ankle PWV (baPWV) as a predicting factor of cardiovcascular events, which was measured by widely-used equipment formPWV/ABI (Colin Co. Ltd.), compared with cfPWV.

Methods

We designed this study as a part of NOAH study. We selected 414 outpatients (male/female = 242/172, mean age = 61.2 ± 12.0 y.o.) from anticipants of NOAH study (n = 813 with essential hypertension), who were simultaneously undergone baPWV and cfPWV measured by AT-form, and their prognoses were followed by questionnaire or medical records. Mean follow-up period was 43.0 ± 17.2 months. We set stroke and/or cardiovascular disease (CVD) as primary endpoint, and stroke, CVD, coronary artery disease (CAD) and mortality as secondary endpoint. During this follow-up period, 36 primary endpoints, 18 brain attacks, 19 heart diseases, 15 CADs and 10 deaths were recorded.

Results

The baPWV were strongly correlated with cfPWV (regression analysis; r = 0.580). Patients were equally divided into 4 groups by either baPWV or cfPWV and evaluated the prognostic impact by Kaplan-Meier analysis (Log-rank test). For primary endpoint, baPWV and cfPWV showed significant distributions (p = 0.0268 and 0.0002, respectively). The baPWV did not show significant distribution for CVD, CAD, and mortality, but only for stroke (p = 0.0015). On the other hand, cfPWV showed significant distribution for CVD, stroke (p = 0.0094 and 0.0338, respectively), but neither for CAD nor mortality. By Cox proportional hazard model adjusted with confounders; age, sex, blood pressure, serum creatinine, diabetes and dyslipidemia, only cfPWV was adopted as predictive factor, but not baPWV.

Conclusion

Although a further large scale multicenter trial is necessary, measurement of cfPWV may be better to predict cardiovascular event, but baPWV also can be a useful screening marker and predictor of future cardiovascular event.