Application and interpretation of the yo-yo intermittent recovery test to the long-term physical development of girls association football players

We aimed to evaluate short- and long-term yo-yo intermittent recovery test, level 1 (YYIRL1) performance in girls football players. Mixed-linear modelling was used to determine within-season changes in YYIRL1 performance and between- and within-player variation over four years, at four time points (July, September, December and May) in 86 players (459 observations), mean age 13 ± standard deviation 1.8 years, 0.9 ± 1.5 years from peak height velocity. Twenty-three players, tested over three consecutive years were retained for further analysis. Magnitude-based inferences were used to quantify annual change in performance on a group and individual level. Within-player correlations were used to determine the association between YYIRL1 and maturation. ‘Very likely’ small (14, ±90% confidence interval 8.1–20%) improvements were observed between July and September and ‘possibly’ trivial (5.5, 0.41–11%) differences between September and May. Within-player variation throughout the in-season period, representing the typical error of the estimate was 23%, 22–25% and between-player, 38, 33–44%. We observed ‘most likely’ moderate improvements (32, 17–49%) over three years that were moderately associated with changes in maturation (r = 0.46, 0.13 to 0.70). A minimum change of ≥44% is required to detect ‘likely’ improvements in YYIRL1 performance on an individual basis.