Measurement agreement between estimates of aerobic fitness in youth: the impact of body mass index.

PURPOSE The purpose of this study was to examine the impact of body mass index (BMI) on the agreement between aerobic capacity estimates from different Progressive Aerobic Cardiorespiratory Endurance Run (PACER) equations and the Mile Run Test. METHOD The agreement between 2 different tests of aerobic capacity was examined on a large data set from 2 suburban school districts (n = 1,686 youth in Grades 3-10). Difference estimates between the Mile Run Test and several PACER equations were computed, and residuals were examined using cluster analysis. The implication of the discrepancy between these tests was also examined using FITNESSGRAM health-related standards for BMI. Comparisons were made against corresponding estimates of peak oxygen consumption from the Mile run because this equation is more established. RESULTS Results supported a 2-cluster solution. The discrepancy between tests was higher in participants with higher BMI scores (Z scores for residuals in this group ranged from -0.07 to 1.57). BMI was able to explain 30% to 34% of the disagreement between the Mile and different PACER equations of aerobic fitness. Classification analyses revealed that kappa scores were lower among PACER equations that do not include a BMI term (kappa = .12-.34 vs. .59-.81). Overall, the test-equating approach used in the Fitnessgram program to process PACER data had better agreement than alternative PACER equations that included BMI. CONCLUSION The results support the inclusion of BMI in prediction equations used to estimate aerobic capacity from the PACER.