DEVELOPMENT AND VAILIDATION OF THE “GORILLA RUN TEST”

Lane C. MadisonƗ1, Kylie J. BrownƗ1, Nicholas B. DrakeƗ1, Derek A. Crawfordǂ1, and Michael. J. Carperǂ1. 1Applied Physiology Laboratory. Department of Health, Human Performance and Recreation, Pittsburg State University, Pittsburg, KS, USA Maximal oxygen consumption (VO2max) testing is widely used in laboratories and requires expensive pieces of equipment. There are numerous prediction equations used to determine VO2max, but none are based on the anaerobic treadmill test (AnTT). PURPOSE: The purpose of this investigation was to develop and validate a regression models to predict maximal oxygen consumption (VO2max) from AnTT time. METHODS: A total of 100 college-aged males and females participated in this investigation. Subjects performed a VO2max treadmill (TM) test utilizing the Bruce protocol. A metabolic cart was used for collection of expired gases and determination of VO2max. 48 hours later subjects performed an AnTT where the treadmill was set at a 20% grade and speed was determined based off their VO2max classification (males: 8mph trained, 7mph untrained; females: 7mph trained, 6mph untrained). Time to exhaustion and VO2max were collected during this test. Reliability of VO2max estimates were compared to Bruce VO2max values using interclass correlation coefficients (ICC) while Bland-Altman (BA) plots were used to determine the agreement (presence of bias and number of values falling within ±2SD limits of agreement, LOA) between the two methods of VO2max determination. RESULTS: Separating trained (N=43) and untrained (N=57) subjects, significant correlations are present between VO2max and AnTT time (trained: r=.622, p=.000; untrained: r=.541, p=.000). Regression equations for both groups (trained: y=38.82+0.23*x; untrained: y=28.64+0.2*x) demonstrate moderate reliability (trained: ICC=.718, p=.000; untrained: ICC=.629, p=.000) for estimating individuals’ VO2max based on AnTT time. BA plot analysis reveal no significant bias from the line of equality for either equation (trained: mean=.18, p=.793; untrained: mean=-.05, p=.929) and acceptable agreement between the two methods (95% of values within LOA). CONCLUSION: Based on these results, we have created reliable regression models to predict VO2max from AnTT times that are in agreement with direct assessment during the Bruce protocol. These models need to be tested using a unique sample to confirm this agreement and to identify standard error of the estimates in order to determine their validity for use in practice