Accuracy of Exercise-based Equations for Estimating Cardiorespiratory Fitness.

Area of Science:

• Exercise Physiology
• Clinical Assessment
• Biostatistics

Background:

• Cardiorespiratory fitness (CRF) is crucial for health and is often estimated using predictive equations based on exercise tests.
• Previous studies have not comprehensively compared various exercise-based prediction equations against directly measured CRF in a large cohort.
• Accurate CRF assessment is vital for clinical risk stratification and patient management.

Purpose of the Study:

• To compare the accuracy of different exercise-based prediction equations for estimating CRF.
• To evaluate the ability of these equations to correctly classify an individual's CRF status.
• To compare estimated CRF (eCRF) with directly measured CRF from cardiopulmonary exercise testing (CPET).

Main Methods:

• Analysis of 4871 tests from apparently healthy adults.
• Comparison of eCRF from 2 non-exercise, 3 submaximal, and 10 maximal exercise equations against directly measured CRF via CPET.
• Statistical analyses included correlations, standard error of estimate, intraclass correlation coefficients, and Cohen κ coefficients.

Main Results:

• All tested equations showed statistically significant associations with directly measured CRF (P < 0.01).
• Intraclass correlation coefficients ranged widely from 0.07 to 0.89, indicating variable agreement.
• Submaximal and maximal exercise equations achieved average correct classification rates of only 51% and 59%, respectively, for fitness tertiles.

Conclusions:

• Despite significant correlations, exercise-based equations demonstrate low accuracy in classifying individuals into fitness tertiles.
• This limited accuracy poses challenges for clinical risk stratification.
• Directly measuring CRF using cardiopulmonary exercise testing is recommended for accurate clinical assessment.