OR-11: Aortic pulse wave velocity as an integrated index of vascular function: relationship to mortality and the effect of ethnicity

Objective: To test the hypothesis that central aortic pulse wave velocity (PWV) as a measure of aortic compliance or distensibility, might be an independent risk factor for mortality, in African-Caribbean, Indian sub-continent (South Asian) and European origin people in Britain. Design and Methods: From 1987-1990 we used a Doppler ultrasound technique [Wright et al, Clin Sci 1990; 78:463-468] to measure aortic PWV, with probes at the root of the neck (aortic arch waveform) and just above the bifurcation, with the person lying flat for >5mins,averaged over 45-120 cardiac cycles. 3 BPs were measured before and 3 after the PWV. Random samples of type 2 diabetes patients attending out-patients and a community-based population from the 3 ethnic groups (77% response) who were all glucose tolerance (GT) tested were the controls, as previously described [Lancet 1991; 338: 842-847]. Participants’ death certificates were tagged and obtained via the NHS central register. Results: 232 Europeans (Eur.), 151 South Asians (SA) and 101 African-Caribbeans, were followed for 11 years, 242 women and 329 men; more men died (63.5% vs. 36.5%). At baseline, African- Caribbeans were slightly younger (58 yrs) than other groups (Eur. 62y, SA 60y) but despite higher BPs in both those still alive (144.6/ 83.2, vs. 138.5/ 77.1 (Eur.), 135.7/ 77.2 (SA) mmHg) and those who later died (156.3/ 87 vs. 150.3/ 79.8 and 153.3/ 80.4 mmHg), PWV was slower at 10.7 (95%CI 10-11.4) vs. 11.5 (11.1-12, Eur) and 11.3 (10.7-11.8, SA). At any given BP level, PWV was faster in those who died. In Cox regression models, adjusting for baseline differences in variables and follow-up, mortality was independently predicted by age (6, 4-9, % increase per year older, p