Accuracy of &OV0312;O2max Prediction Equations in Older Adults

Abstract
PETERSON, M. J., C. F. PIEPER, and M. C. MOREY. Accuracy of V̇O2max Prediction Equations in Older Adults. Med. Sci. Sports Exerc., Vol. 35, No. 1, pp. 145–149, 2003. Purpose We explored the accuracy and bias of prediction equations (ACSM and Foster) in older, deconditioned men and women. We also examined the predictors of V̇O2max to further understand which variables affect respiratory fitness in the elderly. Methods One hundred seventy-one community dwelling, men (72.6 ± 4.8 yr) and women (71.0 ± 5.1 yr) screened in a clinical trial were retrospectively examined. V̇O2max was measured using a standardized protocol with gas exchange measured. Measured V̇O2max values were compared with prediction equations via mean difference analyses, and bias was explored using Bland-Altman analyses. Regression analysis determined significant predictors of measured V̇O2max. Alpha was P ≤ 0.05. Results In men and women, Foster, 21.6 ± 5.7 and 17.9 ± 5.1 mL·kg−1·min−1, respectively, was not significantly different from measured V̇O2max, 21.7 ± 4.8 and 17.3 ± 4.0, respectively. ACSM overestimated V̇O2max in men and women, 26.3 ± 8.2 and 20.9 ± 7.3, respectively. By using Bland-Altman plots, ACSM showed significant overestimation bias in more fit women (r = 0.29), whereas Foster showed no estimation bias in either gender. Significant predictors of V̇O2max were gender, BMI, age, treadmill grade, and speed, with an equation R2 of 0.70. A measure of current activity levels did not make it into the final model (P = 0.0505) but is worthy of future consideration using more sensitive measures than ours. Conclusion ACSM is not appropriate for use when treadmill testing older adults. We believe the Foster equation’s V̇O2max prediction accuracy is acceptable, showing no bias along a continuum of aerobic capacity.