Background

Carotid-femoral pulse wave velocity (PWV) is the gold-standard noninvasive measure of arterial stiffness. Data comparing tonometry-based devices such as the SphygmoCor XCEL to simpler brachial-cuff-based estimates of PWV, such as from the Mobil-O-Graph in African populations are sparse. We therefore aimed to compare PWV measured by the Mobil-O-Graph and the SphygmoCor XCEL device in a sample of South African women and children.

Methods

Women (n = 85) 29 years [interquartile range (IQR): 29-69] and their children/grandchildren (n = 27) 7 years (IQR: 4-11) were recruited for PWV measurement with Mobil-O-Graph and SphygmoCor XCEL on the same day. Wilcoxon signed-rank test, regression analysis, spearman correlation and Bland-Altman plots were used for PWV comparison between devices.

Results

For adults, the SphygmoCor XCEL device had a significantly higher PWV (7.3 m/s, IQR: 6.4-8.5) compared with the Mobil-O-Graph (5.9 m/s, IQR: 5.0-8.1, P = 0.001) with a correlation coefficient of 0.809 (P ≤ 0.001). Bland--Altman analysis indicated an acceptable level of agreement but significant bias (mean difference PWV: 0.90 ± 1.02 m/s; limits of agreement: -1.10 to 2.90). The odds of having a PWV difference more than 1 m/s decreased with a higher age [odds ratio (OR): 0.95, 95% confidence interval (95% CI) = 0.92-0.98] and increased with greater height (OR: 1.10, 95% CI = 1.01-1.21, P = 0.03) in multivariable analysis. In children, the Bland-Altman indicated an excellent level of agreement (-0.03 ± 0.63 m/s; limits of agreement: -1.26 to 1.21), but no correlation was found (rs = 0.08, P = 0.71).

Conclusion

Particularly in younger and taller women, the Mobil-O-Graph significantly underestimated PWV compared with the SphygmoCor. Although no correlation was found between the two devices for children, further research is required due to the small sample size. Furthermore, the clinical value of both methods in young African populations requires further investigation.