Experts proposed blood pressure (BP) load derived from 24-hour ambulatory BP recordings as a more accurate predictor of outcome than level, in particular in normotensive people. We analyzed 8711 subjects (mean age, 54.8 years; 47.0% women) randomly recruited from 10 populations. We expressed BP load as percentage (%) of systolic/diastolic readings ≥135/≥85 mm Hg and ≥120/≥70 mm Hg during day and night, respectively, or as the area under the BP curve (mm Hg×h) using the same ceiling values. During a period of 10.7 years (median), 1284 participants died and 1109 experienced a fatal or nonfatal cardiovascular end point. In multivariable-adjusted models, the risk of cardiovascular complications gradually increased across deciles of BP level and load (P<0.001), but BP load did not substantially refine risk prediction based on 24-hour systolic or diastolic BP level (generalized R(2) statistic ≤0.294%; net reclassification improvement ≤0.28%; integrated discrimination improvement ≤0.001%). Systolic/diastolic BP load of 40.0/42.3% or 91.8/73.6 mm Hg×h conferred a 10-year risk of a composite cardiovascular end point similar to a 24-hour systolic/diastolic BP of 130/80 mm Hg. In analyses dichotomized according to these thresholds, increased BP load did not refine risk prediction in the whole study population (R(2)≤0.051) or in untreated participants with 24-hour ambulatory normotension (R(2)≤0.034). In conclusion, BP load does not improve risk stratification based on 24-hour BP level. This also applies to subjects with normal 24-hour BP for whom BP load was proposed to be particularly useful in risk stratification.
Hypertension, 2014, Vol 63, Issue 5, p. 925-933
Journal Article; Research Support, Non-U.S. Gov't; Adult; Aged; Area Under Curve; Blood Pressure; Blood Pressure Monitoring, Ambulatory; Cardiovascular Diseases; Circadian Rhythm; Female; Humans; Hypertension; Incidence; Male; Middle Aged; Multivariate Analysis; Predictive Value of Tests; Prospective Studies; Retrospective Studies; Risk Factors