Influence of Systolic Blood Pressure and Pulse Pressure on Heart Failure Risk in Elderly Treated Hypertensive Patients

It is unclear whether systolic blood pressure (BP) and pulse pressure (PP), either clinic or ambulatory, differ in their ability to predict heart failure (HF) in elderly hypertensive patients. To evaluate the influence of clinic and ambulatory systolic BP and PP on HF risk in elderly treated hypertensive subjects. In this retrospective study, we assessed 1314 individuals (aged ≥ 65 years) undergoing clinic and ambulatory BP measurement. Association of BP parameters with HF was investigated by univariable and multivariable Cox regression analyses. Differences between competing models were assessed by receiver operating characteristic curve (ROC) analysis, Akaike's Information Criterion (AIC) and Schwarz Bayesian Information Criterion (BIC). During the follow-up (7.0 ± 4.0 years), 126 HF events occurred. After adjustment for covariates, clinic systolic BP and PP were not significantly associated with HF risk, whereas daytime systolic BP, daytime PP, nighttime systolic BP and nighttime PP were, showing a hazard ratio (95% confidence interval), for each standard deviation increase, of 1.45 (1.21-1.73), 1.19 (1.01-1.43), 1.54 (1.31-1.80) and 1.29 (1.08-1.55), respectively. ROC analysis showed that models including daytime or nighttime systolic BP better predicted HF than those involving daytime or nighttime PP and that nighttime BP parameters better predicted HF than daytime ones (all P < 0.05). Comparison of competing models using AIC and BIC gave the same results. In elderly treated hypertensive patients, both ambulatory systolic BP and PP are predictors of HF, but systolic BP is superior to PP and nighttime parameters are superior to daytime ones.