Abstract

Objective: Vascular disease is highly prevalent in the elderly. This study aimed to evaluate arterial phenotype in elderly men and compare carotid-femoral pulse wave velocity (cfPWV) assessed by two techniques (Sphygmocor (S)and Vicorder (V)). Methods: 1722 men (72-92 years), participants in the British Regional Heart Study, underwent ultrasound assessment of carotid intima-media thickness (cIMT), carotid distensibility coefficient and presence of carotid plaque. cfPWV and ankle brachial pressure index (ABPI) were also assessed. 123 men returned for between visit reproducibility assessments. Results: Good reproducibility was demonstrated in all measures (Gwet's agreement=0.8 for plaque, intraclass correlation=0.65 for ABPI and coefficient of variation <13% for all other measures). Measurements were obtained in >90% of men for all measures except cfPWV(S) and ABPI. In 1122 men with both cfPWV(V) and cfPWV(S) data, cfPWV(S) was greater than cfPWV(V) (mean difference=0.23,95%CI 0.10 to 0.37 m/s). cfPWV(V) was higher at low cfPWV values and cfPWV(S) was higher at high cfPWV values. Correlation of V transit time (TT) against S carotid and femoral TT demonstrated that the slope of the regression line for femoral TT was steeper than for carotid TT, resulting in a proportionally greater subtraction of carotid TT from femoral TT at higher PWVs. Conclusions: Reproducible, satisfactory quality non-invasive measurements of vascular phenotype were obtainable in a large proportion of elderly men. The discrepancy in results between the two PWV measures may partly be due to the differential impact of subtracting carotid TT when deriving cfPWV(S) across the clinical PWV range.