A re-appraisal of the reliability of the 20 m multi-stage shuttle run test

Most studies that have set out to quantify the test-retest reliability of the multi-stage shuttle run test (MSSRT) have typically used inappropriate statistics (correlation coefficients) and/or one of two possible calculation methods for deriving predictions of maximal oxygen uptake (VO2max), and have not considered the impact of habituation on performance in the MSSRT and whether its reliability improves beyond a single repeat trial. In this context, the aim of this study was to assess the reliability of the MSSRT over three trials. Thirty-five healthy and active university games players (22 males and 13 females) performed the MSSRT on three occasions, 1 week apart. Individual total numbers of completed shuttles were recorded and estimated VO2max values were derived via established equation and table methods. Analysis revealed that the overall mean VO2max score from the equation method (52.5 +/- 7.8 ml kg(-1) min(-1)) was significantly higher than that for the table (46.9 +/- 8.9 ml kg(-1) min(-1)), whilst the mean trial 2 and trial 3 scores were significantly higher than that for trial 1, but not different to each other. The Limits of Agreement for the table method were -1.4 +/- 5.0 (trial 1-trial 2) and 0.0 +/- 5.5 ml kg(-1) min(-1) (trial 2-trial 3), and for the equation method -1.1 +/- 4.7 (trial 1-trial 2) and 0.0 +/- 5.0 ml kg(-1) min(-1) (trial 2-trial 3). These results suggest that systematic bias is eliminated after the first trial (due to habituation), but a considerable amount of random error remains, regardless of the type of score calculated. Hence, among our sample, the MSSRT does not appear to be reliable enough for the purpose of monitoring changes in VO2max due to non-random reasons.