Estimation of pressure pulse amplification between aorta and brachial artery using stepwise multiple regression models

F. Camacho*, A. Avolio, N. H. Lovell

*Corresponding author for this work

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

The pressure pulse is amplified between the aorta and peripheral sites. This study compares two methods to estimate pressure pulse amplification (PPA) between the aorta and the brachial artery. Method 1: PPA was determined from a multi-parameter linear regression of subject parameters (gender, age, height, weight, heart rate (HR), brachial systolic pressure (BSP), diastolic pressure (BDP), mean pressure (MP)). Method 2: PPA was calculated from central aortic pressure waveforms (CW) estimated from the same subject parameters. The sample population (1421 male, 992 female) was selected from a database where aortic pressure was estimated by mathematical transformation of a peripheral (radial) pulse calibrated to sphygmomanometric BSP and BDP. The two methods were consistent in showing HR and MP as the most important parameters to estimate PPA. Correlation coefficients (R2) of 0.48 (method 1) and 0.44 (method 2) were obtained using height, weight, HR, BSP, BDP and age. Inclusion of MP increased R2 to 0.77 (method 1) and 0.71 (method 2). This study shows that databases containing peripheral and central aortic pressure waveforms can be used to construct multiple regression models for PPA estimation. These models could be applied to studies of similar subject groups where peripheral waveforms may not be available.

Original languageEnglish
Pages (from-to)879-889
Number of pages11
JournalPhysiological Measurement
Volume25
Issue number4
DOIs
Publication statusPublished - Aug 2004
Externally publishedYes

Keywords

  • Central aortic blood pressure
  • Linear regression model
  • Pulse pressure amplification

Fingerprint Dive into the research topics of 'Estimation of pressure pulse amplification between aorta and brachial artery using stepwise multiple regression models'. Together they form a unique fingerprint.

Cite this