Estimating V˙O2peak from a Nonexercise Prediction Model

Abstract

Cardiorespiratory fitness is suggested to be an important marker of cardiovascular risk but is rarely evaluated in health care settings. In the present study, directly measured peak oxygen uptake (V·O 2peak) from a diverse population of 4637 healthy participants were used to develop and cross-validate a new nonexercise regression model of cardiorespiratory fitness for men and women. Multivariable regression analysis was used to develop a nonexercise model of cardiorespiratory fitness for men and women separately with V·O 2peak as the outcome. In the final models, 2067 men (mean age = 48.8 yr) and 2193 women (mean age = 47.9 yr) were included, respectively. Cross-validation of the models was done by standard data splitting procedures with evaluation of constant error and total error of a model developed on one sample and cross-validated on another sample. Age, waist circumference, leisure time physical activity, and resting HR, successively, were the most potent predictors of V·O 2peak for both men and women. Together, 61% and 56% of variance in V·O 2peak, for men and women, respectively, were explained by the full models. SEE was 5.70 and 5.14 for the models including men and women, respectively. The nonexercise regression model developed in the present study was fairly accurate in predicting V·O 2peak in this healthy population of men and women. The model might be generalized to other healthy populations and might be a valid tool for a rough assessment of cardiorespiratory fitness in an outpatient setting.