Usefulness of circadian amplitude of blood pressure in predicting hypertensive cardiac involvement.

Twenty-four-hour blood pressure (BP) profiles of 56 patients diagnosed as 'hypertensive' by WHO criteria were analyzed by the fit of a 24-hour cosine curve according to the single cosinor method. A left ventricular mass index (LVMI) was also assessed by two-dimensional echocardiography on each patient as a gauge of target organ involvement. LVMI and the BP MESOR correlates positively for systolic, S (r = 0.324), mean arterial, MA (r = 0.334) and diastolic, D (r = 0.267) BP (P less than 0.05), yet no statistically significant linear correlation between LVMI and the circadian BP amplitude (one-half of predictable change) was found. When a second-degree polynomial regression was fitted to the circadian BP amplitudes, an association was found (SBP: R2 = 0.138, P = 0.02; MAP: R2 = 0.167, P = 0.01; DBP: R2 = 0.128, P less than 0.01). The corresponding curves were characterized by peaks in the circadian amplitudes of SBP, MAP and DBP around a value of LVMI between 110 and 120 g/m2. For further scrutiny, three subgroups had been formed on the basis of literature, a priori with respect to the LVMI (group 1: LVMI less than 100); group 2: 100 less than LVMI less than 130; group 3: 130 less than LVMI). For MESORs, there was no difference between groups 1 and 2, whereas the MESOR of group 3 were larger than the other two groups. The circadian BP amplitudes of group 2 were larger than those of the other two groups for SBP, MAP and DBP. An increasing LVMI precedes a definitive increase of BP MESOR and coincides with an increase in the circadian BP amplitude; thus an increase in extent of circadian changes can alert the self-monitoring population of a target organ involvement.