Estimation of maximal oxygen uptake by bioelectrical impedance analysis

Previous non-exercise models for the prediction of maximal oxygen uptake (\( \dot V{\text{O}}_{{\text{2max}}} \)) have failed to accurately discriminate cardiorespiratory fitness within large cohorts. The aim of the present study was to evaluate the feasibility of a completely indirect method for predicting \( \dot V{\text{O}}_{{\text{2max}}} \) that was based on bioelectrical impedance analysis (BIA) in 66 young, healthy fit men and women. Multiple, stepwise regression analysis was used to determine the usefulness of BIA and additional covariates to estimate \( \dot V{\text{O}}_{{\text{2max}}} \) (ml min⁻¹). BIA was highly correlated to \( \dot V{\text{O}}_{{\text{2max}}} \) (r = 0.914; P < 0.001) and entered the regression equation first. The inclusion of gender and a physical activity rating further improved the model which accounted for 88% of the variance in \( \dot V{\text{O}}_{{\text{2max}}} \) and resulted in a relative standard error of the estimate (SEE) of 7.2%. Substantial agreement between the methods was confirmed by the fact that nearly all the differences were within ±2 SD. Furthermore, in contrast to previously published non-exercise models, no trend of a reduction in prediction accuracy with increasing \( \dot V{\text{O}}_{{\text{2max}}} \) values was apparent. It was concluded that a non-exercise model based on BIA might be a rapid and useful technique to estimate \( \dot V{\text{O}}_{{\text{2max}}} \), when a direct test does not seem feasible. However, though the present results are useful to determine the viability of the method, further refinement of the BIA approach and its validation in a large, diverse population is needed before it can be applied to the clinical and epidemiological settings.