Is "Usual" blood pressure a proxy for 24-h ambulatory blood pressure in predicting cardiovascular outcomes?

BACKGROUND The 24-h ambulatory blood pressure (ABP) is a stronger predictor of cardiovascular disease than conventional blood pressure (CBP), but it remains unclear how it compares with "usual" blood pressure (UBP), estimated after CBP has been corrected for regression dilution bias (RDB). METHODS We compared the associations of cardiovascular mortality (n = 50), cardiovascular events (n = 101), and cardiac events (n = 71) with systolic CBP, UBP, and ABP over 13 years of follow-up (median) in 1,167 randomly selected Belgians. We estimated the correction factor to compute UBP from CBP at the midpoint of follow-up (6.5 years) in 723 untreated individuals without cardiovascular disease. RESULTS Cardiovascular disease increased across quartiles of systolic CBP, UBP, and ABP (P for trend <= 0.02). For each 10 mm Hg increment in systolic ABP, the multivariate-adjusted hazard ratios for cardiovascular mortality and for cardiovascular and cardiac events were 1.38, 1.27, and 1.33, respectively (P < 0.001 for all). For CBP, the corresponding hazard ratios were 1.10 (P = 0.21),1.09 (P = 0.12), and 1.14 (P = 0.06); and for UBP, they were 1.18 (P = 0.21), 1.16 (P = 0.12), and 1.23 (P = 0.06), respectively. The risk function for cardiovascular disease in relation to ABP was significantly steeper than that for CBP, but not UBP. In Cox models, including CBP or UBP in the presence of ABP, only ABP predicted cardiovascular outcomes. CONCLUSIONS Correcting CBP for RDB resulted in a steeper slope of events on blood pressure than observed for CBP. The association with UBP was not statistically significant and did not enhance the prediction of outcome to the level of ABP.