Background Carotid ultrasound is widely used to measure haemodynamic parameters, such as intima-media thickness and blood flow velocities (i.e. peak-systolic velocity [PSV], end-diastolic velocity [EDV], and resistive index [RI]). However, the association between blood flow velocities and cardiovascular events remains unclear. Design and methods Baseline data, including quantitative ultrasonography, were obtained from 3146 adults as part of the Cardiovascular Diseases Risk Factor Two-Township Study. Occurrence of ischaemic heart disease (IHD) and stroke was determined from insurance claims and death certificates. The hazard ratio (HR) of CVD (IHD and stroke combined) was calculated for EDV and PSV of the common carotid artery using Cox models. Net reclassification index and integrated discrimination index were used to evaluate the capacity of EDV to predict IHD, stroke, and CVD. Results Median follow-up was 12.8 years. There were 220 cases of IHD and 247 cases of stroke. The HR (95% CI) for CVD from univariate analysis was 4.54 (3.51-5.85) for EDV <15 cm/s relative to EDV ≥ 20 cm/s (p < 0.0001), and 3.23 (2.51-4.15) for PSV < 65 cm/s relative to PSV ≥ 80 cm/s (p < 0.0001). The HR (95% CI) for CVD from multivariate analysis was 1.66 (1.22-2.26) for EDV < 15 cm/s relative to EDV ≥ 20 cm/s, and 1.39 (1.03-1.89) for PSV < 65 cm/s relative to PSV ≥ 80 cm/s. EDV slightly but significantly improved prediction of CVD (integrated discrimination index 0.56%, p = 0.016). Conclusions Low common carotid EDV and PSV were independently associated with future CVD, and EDV improved the prediction of future CVD. More prospective studies are required in different ethnic groups to understand the significance and implication of these findings.
ASJC Scopus subject areas