Beat-to-beat variability of aortic pulse wave velocity: implications for aortic stiffness measurements

Aortic pulse wave velocity (aPWV) predicts cardiovascular risk. Being the reference method for aortic stiffness evaluation, invasive aPWV is also recommended for validation of noninvasive devices. Because of intrinsic haemodynamic variability and processing issues, aPWV shows beat-to-beat variability. We aimed to quantify this variability and evaluate its implications for the reliability and use of aPWV as reference in validation and clinical application studies.

The study included n  = 84 patients, in whom two datasets of invasive data were recorded: 1) simultaneous ascending aorta and iliac pressure acquisitions using a dual-tip intra-aortic catheter, and 2) an additional ascending aorta pressure acquisition. By combining the iliac and ascending aorta pressure recordings from the first and second acquisitions, respectively, we evaluated how a sequential acquisition protocol affects variability. We compared three pressure waveform foot identification methods to investigate the effect of data processing on variability. Furthermore, we estimated how averaging over nbeats consecutive heartbeats affects the standard deviation (SD) of such nbeats -averaged estimate of aPWV.

The simultaneously acquired invasive aPWV showed a 5% beat-to-beat SD (variability), with small but significant differences between foot identification methods. The sequential acquisition protocol doubled aPWV variability compared to simultaneous acquisition. However, because averaging had a much stronger effect on sequentially measured aPWV, the two acquisition protocols yielded comparable variabilities at nbeats  = 10 (2% vs. 3%).

Our study suggests that, independently from the acquisition protocol and data processing, the intrinsic beat-to-beat variability of aPWV becomes manageable when aPWV values of at least ten heartbeats are averaged.