Prediction of peak oxygen pulse (O2Ppeak) without exercise testing in older adults

Abstract

Peak oxygen pulse has been considered a surrogate of cardiovascular function and an independent predictor of all cause mortality. However, O2Ppeak depends on maximal volitional effort which may limit its utility in older subjects. The aim of this study was to develop a model to estimate O2Ppeak without exercise in an elderly sample. This cross-sectional study enrolled 67 community-dwelling older adults (69.4±7.1 years; 41 men) for the non-exercise model development and 30 community-dwelling older adults (67.7±6.4 years; n=30; 17 men) for cross-validation. The non-exercise model was derived through hierarchical regression model and cross-validated by means of PRESS statistics and comparison against an independent sample. Classification accuracy of the model for tertiles of estimated and actual O2Ppeak was tested by gamma (γ) nonparametric correlation. The following prediction equation was generated: −3.416+0.137×weight (kg)+1.226×Veterans Specific Activity Questionnaire (VSAQ) (metabolic equivalents, METs)+1.987×gender (0=women, 1=men)−2.045×β-Blockers use (0=no, 1=yes)−0.044×resting heart rate (HR) (R2=0.83; standard error of estimate (SEE)=1.68mLbeat−1). Correlation in cross-validation group was 0.80 (P<0.001). A high probability was observed for the model to rank the values in the same tertile in validation and cross-validation groups (γ=0.98; γ=0.92, respectively, P<0.05). In conclusion, O2Ppeak can be estimated with reasonable precision without exercise testing, providing an alternative for elder subjects not capable to perform maximal effort.