Stiffness index beta (β) and cardio-ankle vascular index (CAVI) are proposed pressure-independent measures of arterial stiffness that quantify the intrinsic exponent (βo) of the exponential relationship between arterial pressure and diameter. We aim (i) to demonstrate that β and CAVI do not strictly approximate the pressure-independent parameter βo, and (ii) to provide a suitable correction for the inherent pressure dependency of beta and CAVI.Design and method:
The exponential relationship between pressure (P) and diameter (d) (Figure A) is valid when expressed using a reference pressure (Pref) and corresponding diameter (dref): P = Pref*exp[βo(d/dref-1)]. In practice, diastolic blood pressure (BP) and diastolic diameter are substituted for Pref and dref. The resulting β (Figure B) is not equal to βo, and can be analytically shown to depend on BP. This dependence affects both β and CAVI. CAVI calculation also relies on an approximation of incremental stiffness dP/dd (Figure D), which further increases pressure dependence.Design and method:
Both effects can be analytically and fundamentally corrected for to obtain clinically usable equations to estimate pressure-independent βo and CAVIo (Figures C,F). The pressure dependence of CAVI and the proposed corrected parameters were quantified in a numerical simulation of a hypertensive patient group. Simulation was done by randomly assigning to each subject a single pressure-independent βo value (no change in actual P-d relationship), and baseline and treated (lowered) diastolic and pulse pressures (DBP and PP). Values were drawn from normal distributions (mean ± SD: βo 10 ± 2, baseline DBP 110 ± 10 and PP 50 ± 10 mmHg; treated DBP 80 ± 10 and PP 40 ± 10 mmHg).Results:
Results of pressure dependence of β and CAVI are shown in Figures B,E. Corrected measures (Figures C,F) showed no change with BP.Results:
After power calculation, simulation of n = 154, with systolic/diastolic BP (SBP/DBP) lowered from 160 ± 15/110 ± 11 to 120 ± 14/79 ± 11 mmHg, resulted in pulse wave velocity reduction from 8.2 ± 1.1 to 6.9 ± 1.0 m/s (p < 0.001). CAVI changed significantly from 8.1 ± 2.0 to 7.7 ± 2.1 (p = 0.023), while CAVIo did not change (9.7 ± 2.3 vs. 9.9 ± 2.6, p = 0.393).Conclusions:
We conclude that β and CAVI as regularly used inherently exhibit pressure dependence. We derived clinically applicable equations to evaluate intrinsic pressure-independent changes in arterial stiffness.