Non-exercise VO2max estimation for physically active college students

This study sought to develop a maximal oxygen consumption (VO2max) regression model derived strictly from self-reported non-exercise (N-EX) predictor variables. The VO2max (mean +/- SD; 44.05 +/- 6.6 ml.kg-1.min-1) of 100 physically active college students (50 females, 50 males), aged 18 to 29 yr, was measured using a treadmill protocol and open circuit calorimetry. Questionnaire-based predictor variables used in the N-EX regression model included (a) the subject's perceived functional ability (PFA) to walk, jog, or run given distances, (b) habitual physical activity (PA-R) data, (c) body mass index (BMI), and (d) gender. BMI (kg.m-2) was computed from self-reported body weight in pounds and self-reported body height in feet and inches. The questionnaire-based N-EX regression model (R = 0.85, SEE = 3.44 ml.kg-1.min-1) developed in this study exceeded the accuracy of previously developed N-EX regression models and is comparable to many exercise-based regression models in the literature. Cross-validation using PRESS (predicted residual sum of squares) statistics demonstrated minimal shrinkage (R = 0.84, SEE = 3.60 ml.kg-1.min-1) of the present regression model. The PFA data were useful in explaining observed VO2max variance (squared partial r2 = 0.155, P < 0.0001) and enhanced the ability of the N-EX regression model to accurately predict criterion VO2max. These results suggest that a questionnaire-based N-EX regression model provides a valid and convenient method for predicting VO2max in physically active college students.