High blood pressure variability predicts 30-day mortality but not 1-year mortality in hospitalized elderly patients

Avraham Weiss, Yaron Rudman, Yichayaou Beloosesky, Amit Akirov, Tzippy Shochat, Alon Grossman

Research output: Contribution to journalArticlepeer-review

Abstract

Background: The association of blood pressure (BP) variability (BPV) in hospitalized patients, which represents day-to-day variability, with mortality has been extensively reported in patients with stroke, but poorly defined for other medical conditions. Aim and method: To assess the association of day-to-day blood pressure variability in hospitalized patients, 10 BP measurements were obtained in individuals ≥75 years old hospitalized in a geriatric ward. Day-to-day BPV, measured 3 times a day, was calculated in each patient as the coefficient of variation of systolic BP. Patients were stratified by quartiles of coefficient of variation of systolic BP, and 30-day and 1-year mortality data were compared between those in the highest versus the lowest (reference) group. Results: Overall, 469 patients were included in the final analysis. Mean coefficient of variation of systolic BP was 12.1%. 30-day mortality and 1-year mortality occurred in 29/469 (6.2%) and 95/469 (20.2%) individuals respectively. Patients in the highest quartile of BPV were at a significantly higher risk for 30-day mortality (HR =4.12, CI 1.12–15.10) but not for 1-year mortality compared with the lowest BPV quartile (HR =1.61, CI 0.81–3.23). Conclusions: Day-to-day BPV is associated with 30-day, but not with 1-year mortality in hospitalized elderly patients.

Original languageEnglish
Pages (from-to)259-263
Number of pages5
JournalBlood Pressure
Volume26
Issue number5
DOIs
StatePublished - 3 Sep 2017

Keywords

  • Blood pressure
  • elderly
  • mortality
  • variability

Fingerprint

Dive into the research topics of 'High blood pressure variability predicts 30-day mortality but not 1-year mortality in hospitalized elderly patients'. Together they form a unique fingerprint.

Cite this