Prognostic value of circadian blood pressure changes in relation to differing measures of day and night

Abstract Although the prognostic value of the day-night blood pressure (BP) changes is established, the most appropriate method for defining day and night is undefined. We assessed the prognostic value of the day-night BP changes by using three definitions of day and night in 2,934 initially untreated hypertensive subjects who underwent 24-hour ambulatory BP monitoring. Over a median follow-up period of 7 years, there were 356 cardiovascular events and 176 deaths. Total cardiovascular events and all-cause mortality were similarly higher in non-dippers (night/day ratio of systolic BP >10% or >0%) than in dippers regardless of the definition of day and night. In a receiver-operated characteristic (ROC) curve analysis of the night/day ratio of systolic BP on the occurrence of events, the area under the ROC curve did not differ among the different definitions of day and night (large fixed-clock intervals, narrow fixed-clock intervals, diary) for both total cardiovascular events (0.61 [95% confidence interval (CI): 0.58 to 0.64], 0.61 [95% CI: 0.57 to 0.63], 0.62 [95% CI: 0.58 to 0.65], respectively; P = 0.20) and all-cause mortality (0.65 [95% CI: 0.61 to 0.70], 0.64 [95% CI: 0.60 to 0.69], 0.65 [95% CI: 0.61 to 0.70], respectively; P = 0.78). The prognostic value of the diurnal BP changes is comparable when using different clock-dependent or independent definitions of day and night.