General applicability of a recently developed arterial path length formula for carotid-femoral pulse wave velocity measurements

Background: Carotid-femoral pulse wave velocity (PWV), an index of large artery stiffening, is a significant prognostic marker for hypertension. Despite its additive value beyond traditional cardiovascular risk factors, the measurement of PWV in clinical practice is relatively limited due, in part, to practical aspects of measurement. In particular, accuracy of PWV is dependent on the arterial path length measurement, which is most commonly calculated from body surface measurements and subject to inter- and intra-observer variability. A regression formula for calculating carotid-femoral path length was developed recently (Weir-McCall et al., Hypertension 2018;71:937–45) and validated in two European cohorts, but its general applicability in other populations has yet to be investigated. Aim: To investigate the general applicability of the carotid-femoral path length formula in a heterogenous Australian population. Methods: The formula was used to calculate carotid-femoral PWV distance (distform) in 70 participants (aged 65±12 years (range 29–86), 24 female, 20 Asian, 50 Caucasian) who previously had PWV measured using the 3-point subtraction method for distance measurement. PWV was recalculated using the transit time (TT) previously recorded (PWVform = distform/TT). The formula-calculated values were then compared to the actual measured distance (distmeas) and PWV (PWVmeas) values, and their differences were compared to the differences observed in the validation study (distdiff_val, PWVdiff_val). Results: The mean difference in measured and calculated distances (distmeas – distform) was –4.5±40.4 mm (P = 0.35), and the mean difference between measured and calculated PWV (PWVmeas – PWVform) was –0.06±0.90 m/s (P = 0.59). These differences were comparable to the validation study results for distance (distdiff_val = –7.8±24.9 mm; P = 0.53) but not for PWV (PWVdiff_val = 0.87±0.93 m/s; P < 0 .001). Using Bland-Altman analysis, a proportional bias was observed in the formula calculated distances, whereby the formula overestimated short distances but underestimated long distances. Conclusions: Although the formula for calculating carotid-femoral path length overestimates short distances and underestimates long distance, the resulting calculated PWV values had a reasonable agreement with actual measured values, indicating the formula was generally applicable in this heterogenous Australian cohort. Further studies are required to investigate the applicability of the formula in other populations and age groups as well as accuracy in terms of estimating true arterial path-length.