|dc.description.abstract||This investigation aimed to: (i) assess the reliability of scores from an adapted 20 metre multistage fitness test (MSFT) performed on a dry turf rugby pitch, in which subjects wore studded boots, (ii) develop a calibration model using least squares linear regression which could predict VO2max scores (ml kg-1 min-1) using scores from the adapted MSFT (cumulative shuttles) and (iii) cross-validate the calibration model by quantifying the agreement between directly measured and predicted VO2max scores. A sample of 16 rugby players (12 males and 4 females) took part in the investigation (mean (SD), age 20.8 (±1.2) years, stature 1.81 (±0.08) m and body mass 76.6 (±13.4) kg). Each of the 16 subjects performed two MSFTs in order to establish the reliability of the adapted MSFT (initial test 67 (±21) shuttles, retest 68 (±19) shuttles). The agreement between test and retest scores was quantified using the 95% limits of agreement (LoA) method (bias = -3 (±2) shuttles, t = -0.66, P = 0.517; heteroscedasticity coefficient r = 0.012, P = 0.964; LoA = ±3 shuttles). Eleven subjects were then randomly selected from the group to form a calibration sample. An equation was then formed using average scores from the two MSFTs performed in the reliability stage (66 (±22) shuttles) and VO2max scores directly measured using cycle ergometry (44.2 (±8.5) ml kg-1 min-1). The equation formed was: VO2max (ml kg-1 min-1) = 22.3 + (0.327 x MSFT score (shuttles)). The standard error of estimate for the model was calculated as: sYX = ±4.9 ml kg-1 min-1.
The remaining five subjects not used in the calibration stage were used to cross-validate the calibration model. MSFT scores, obtained by averaging the two MSFT scores achieved by each subject in the reliability stage (70 (±18) shuttles), were entered into the calibration model and predicted VO2max scores were calculated (45.3 (±5.7) ml kg-1 min-1). These predicted scores were then compared to scores directly measured using cycle ergometry (44.3 (±7.1) ml kg-1 min-1). Their agreement quantified, again using the 95% LoA method (bias = -3.7 (±1.8) ml kg-1 min-1, t = -0.99, P = 0.380; heteroscedasticity coefficient r = 0.077, P = 0.902; LoA = ±3.5 ml kg-1 min-1). The findings of this study led to the conclusion that an MSFT performed on a dry turf rugby pitch whilst wearing studded boots provides repeatable scores within the sample group of rugby players used in the study. Scores from the adapted MSFT can predict VO2max scores using the devised calibration model and the standard error of estimate is lower when compared to the error in the original MSFT study by Léger and Lambert (1982). Despite this, the agreement between predicted and directly measured VO2max scores suggested that the calibration model predicted valid VO2max scores from adapted MSFT scores for the sample population. Because of the simplicity of the adapted MSFT and its ability to test multiple subjects simultaneously; it was decided that it is a useful test for assessing aerobic fitness of rugby players.||en