AMBULATORY BLOOD PRESSURE VARIABILITY INDICES AS INDEPENDENT PREDICTORS OF ALL CAUSE MORTALITY: 8C.01

Abstract

Listen

Translate article

Objective: To evaluate new indices, derived from ambulatory BP variability, for associations with clinical characteristics and survival. Design and Methods: Correlation and outcome analyses of an ABPM database. BP variability (BPV), was defined as the weighted average of awake and sleep SD of BP. The following were defined using systolic (SDs) and diastolic (SDd) BPV: BP Variability Ratio (BPVR), defined as SDs/SDd, estimates the slope of the systolic-vs-diastolic pressure relationship using symmetric regression (the corresponding symmetric Ambulatory Arterial Stiffness Index is 1–1/BPVR); dV = SDs2-SDd2; U = (SDs-SDd)SDs; and dS/PP = (SDs-SDd)/PP, where PP is the pulse pressure. Results: Between 1991 and 2009, 4690 patients (age 55 ± 16 years; 52% women) underwent ABPM. 58% were treated for hypertension and 9% for diabetes. Ambulatory BP was 137 ± 16/79 ± 10 mmHg, and BPV was 13 ± 4/9 ± 2 mmHg. The median values (and IQR) of the variability-derived indices were: BPVR 1.36 (1.15–1.61); dS/PP 0.057 (0.027–0.090); U 37 mmHg2 (14–79); and dV 64 mmHg2 (26–128). BPVR and dS/PP were more dependent on age and hypertension treatment status compared to U, dV and BPV, but less dependent on gender and systolic BP. BPVR and dS/PP were entirely independent of mean arterial pressure, while U and dV had mild positive correlations with it. The figure (top) shows standardized values of BPV and its derivatives according to PP tertiles, adjusted for covariates. During median follow-up of 9.0 years (IQR 5.1–13.4 years) 524 patients died. BPV and its derivatives predicted all-cause mortality. The figure (bottom) shows decreasing multivariable-adjusted HR associated with BPV indices as PP increases. Conclusions: BPV and derived indices have meaningful associations with clinical characteristics, but are nonetheless independently linked with outcome. Prognosis might be directly affected by BPV, or indirectly through associated deranged arterial properties. Indeed, arterial rigidity as assessed by PP appears to markedly modulate the prognostic significance of ambulatory BPV. This suggests a specific role for prediction of outcome in subjects with low PP hypertension (isolated diastolic hypertension) or prehypertension.