Office, Home, and Ambulatory Blood Pressures as Predictors of Cardiovascular Risk

Ambulatory blood pressure (BP) is considered as the gold standard of BP measurement although it has not been shown to be more strongly associated with cardiovascular risk than is home BP. Our objective was to compare the prognostic value of office, home, and ambulatory BP for cardiovascular risk in 502 participants examined in 1992 to 1996. The end point was a composite of cardiovascular mortality, myocardial infarction, stroke, heart failure hospitalization, and coronary intervention. We assessed the prognostic value of each BP in multivariable-adjusted Cox models. The likelihood chi(2) ratio value was used to test whether the addition of a BP variable improved the model's goodness of fit. After a follow-up of 16.1 +/- 3.9 years, 70 participants (13.9%) had experienced >= 1 cardiovascular event. Office (systolic/diastolic hazard ratio per 1/1 mm Hg increase in BP, 1.024/1.018; systolic/diastolic 95% confidence interval, 1.009-1.040/0.994-1.043), home (hazard ratio, 1.029/1.028; 95% confidence interval, 1.013-1.045/1.005-1.052), and 24-hour ambulatory BP (hazard ratio, 1.033/1.049; 95% confidence interval, 1.019-1.047/1.023-1.077) were predictive of cardiovascular events. When all 3 BP variables were included in the model simultaneously, only systolic/diastolic ambulatory BP was a significant predictor of cardiovascular events (P=0.002/<0.001). Home systolic/diastolic BP improved the fit of the model only marginally when added to a model including office BP (chi(2)=3.0/4.0, P=0.09/0.047). Ambulatory BP, however, improved the fit of model more clearly when added to office and home BP (chi(2)=9.0/12.3, P=0.001/<0.001). Our findings suggest that ambulatory BP is prognostically superior to office and home BP.