The stiffness index (SI) is used to estimate cardiovascular risk in humans. In this study, we developed a refined SI for determining arterial stiffness based on the decomposed radial pulse and digital volume pulse (DVP) waveforms. In total, 40 mature asymptomatic subjects (20 male and 20 female, 42 to 76 years of age) and 40 subjects with type 2 diabetes mellitus (T2DM) (23 male and 17 female, 35 to 78 years of age) were enrolled in this study. We measured subjects' radial pulse at the wrist and their DVP at the fingertip, and then implemented ensemble empirical mode decomposition (EEMD) to derive the orthogonal intrinsic mode functions (IMFs). An improved SI (SInew) was calculated by dividing the body height by the mean transit time between the first IMF5 peak and the IMF6 trough. Another traditional index, pulse wave velocity (PWVfinger), was also included for comparison. For the PWVfinger index, the subjects with T2DM presented significantly higher SInew values measured according to the radial pulse (SInew-RP) and DVP signals (SInew-DVP). Using a one-way analysis of variance, we found no statistically significant difference between SInew-RP and PWVfinger when applied to the same test subjects. Binary logistic regression analysis showed that a high SInew-RP value was the most significant risk factor for developing T2DM (SInew-RP odds ratio 3.17, 95% CI 1.53-6.57; SInew-DVP odds ratio 2.85, 95% CI 1.27-6.40). Our refined stiffness index could provide significant information regarding the decomposed radial pulse and digital volume pulse signals in assessments of arterial stiffness.
Keywords: digital volume pulse; ensemble empirical mode decomposition (EEMD); pulse wave velocity; radial pulse; stiffness index (SI); type 2 diabetes mellitus (T2DM).