An Evaluation of the Predictive Validity and Reliability of Ventilatory Threshold

Abstract

To identify a valid and reliable method to determine 40-km time trial (40K) performance in a laboratory setting. Part 1: Ventilatory threshold (VT) and 40K performance were determined on two occasions (February/September) using two subsets of cyclists (N = 15 each; VO(2max) 67.6 +/- 4.2/71.5 +/- 3.0 mL x kg(-1) x min(-1)) to determine the predictive validity of VT assessments. Variables of interest were power output at VT, peak power output (MaxVT(w)), and average power output during 40K (40K(avgwatts)). For VT determination we used: breakpoint of VE/VO2; breakpoint of VE/VCO2; V-slope; RER = 1; and RER = 0.95. In part 2, test-retest reliability of VT and MaxVT(w) were examined in 20 subjects (VO(2max) 64.8 +/- 8.0 mL x kg(-1) x min(-1)) on two occasions, separated by 48 h. Regression analyses between power outputs at VTs and 40K(avgwatts) showed significant predictive validity for (February/September): V-slope (r = 0.79/0.84; SEE 155/13.3W), VE/VO2 (r = 0.80/0.81; SEE 15.2/14.2W), RER0.95 (r = 0.73/0.58; SEE 17.4/21.2W), RER1 (r = 0.75/0.74; SEE 16.8/16.7W), and MaxVT(w) (r = 0.81/0.73; SEE 15.0/17.1W). Paired t-tests between power outputs at VTs and the 40K(avgwatts) indicated that mean power outputs at VE/O2 (261 +/- 29W; P = 0.33) and RER0.95 (274 +/- 55W; P = 0.93) in February and VE/VO2 (274 +/- 37W; P = 0.79) in September were not significantly different from the respective 40K(avgwatts) (277 +/- 30W/281 +/- 30W). Test-retest reliability analysis yielded the following intraclass correlation and relative test-retest errors: V-slope: 0.98, 2.6%; VE/VO2: 0.95, 5.3%; RER0.95: 0.87, 9.8%; RER1: 0.94, 5.7%; VE/VCO2: 0.87, 12.1%; MaxVT(w): 0.98, 2.6%. The high test-retest reliability and consistent ability to accurately predict athletes' 40K(avgwatts) across a competitive season indicated that VE/VO2 was superior to the other evaluated methods.