Can ambulatory blood pressure measurements substitute assessment of subclinical cardiovascular damage?

Abstract

We have previously demonstrated that markers of subclinical organ damage (SOD) improve cardiovascular risk prediction in healthy individuals. We wanted to investigate whether this additive effect of SOD was due to inaccurate blood pressure (BP) measurement or whether ambulatory BP (AMBP) added further to risk prediction. In a population cohort of 1385 Danish individuals free of cardiovascular disease and diabetes, we recorded traditional risk factors, AMBP and pulse wave velocity (PWV), urine albumin/creatinine ratio (UACR), left ventricular mass index (LVMI) and carotid atherosclerotic plaques at baseline. A composite cardiovascular endpoint (CEP) consisting of cardiovascular death and nonfatal myocardial infarction and stroke was recorded in national registries. During a median follow-up of 12.8 years, a total of 119 CEPs occurred. In categorical analysis, presence of SOD as well as masked hypertension increased sensitivity of Systemic Coronary Risk Estimation from 73.9 to 89.1% (P < 0.001) and reduced specificity from 60.1 to 41.8% (P < 0.001). In continuous analysis, logUACR [hazard ratio = 1.20 (95% confidence interval [CI] 1.05-1.38), P = 0.009], atherosclerotic plaques [hazard ratio = 1.82 (95% CI 1.21-2.74), P = 0.004] and 24-h SBP [hazard ratio = 1.34 (95% CI 1.12-1.60), P = 0.002] but not logPWV or LVMI predicted CEP in a model with adjustments for age, sex, conventional BP, total cholesterol and smoking. Compared with a risk model using only traditional risk factors, adding PWV, UACR, plaques, LVMI and 24-h SBP increased C-index significantly from 0.76 to 0.79% and produced a net reclassification improvement of 23.3% (P = 0.001). UACR and plaques predicted cardiovascular events independently of AMBP and improved risk prediction.