Is repeated-sprint ability of soccer players predictable from field-based or laboratory physiological tests?

Aim. The aim of this study was to investigate multiple regression models with prediction equations that would enable a valid estimate of running repeated-sprint ability (RSA) in soccer players from the variables measured in field and laboratory physiological tests. Methods. Adolescent soccer players (N=33) performed five field performance tests and two laboratory tests for assessment of muscle strength of legs, sprint ability, anaerobic power and capacity, aerobic power and capacity, and running economy. These tests served as potential predictors of RSA. RSA was assessed by a intermittent anaerobic running test (IAnRT) consisted of ten 20 m sprints. Results. Multiple regression analysis revealed that the mean speed in the 20 m sprint test and the 2 km endurance running test accounted for 89% of total variation in the mean running speed in the IAnRT (V-IAnRT) as the indicator of capacity for multiple sprint work (R-2=0.89, SEM=0.09 m.s(-1)). Using the variables from the laboratory tests, the best prediction of the V-IAnRT was obtained from the running speed at the ventilatory threshold level (V-VT) and anaerobic power (Pmax.kg(-1)) (R-2=0.49, SEM=0.21 m.s(-1)). Conclusion. Performance in the multiple-sprint exercise as an indicator of RSA can be estimated by the regression equation with two predictors - mean speed in the 20 m sprint and 2 km running test with an error of 4%.