Author(s): GUILLAUME, J.P., NOEL, C., MAURELLI, O., RENOUF, F., PRIOUX, J., Institution: CDFAS, Country: FRANCE, Abstract-ID: 807

INTRODUCTION:
Biological maturation and relative age influence selection process and performance in team sports (1,2). The asymmetry in the distribution of dates of birth favouring players born in a cohort refers to the notion of relative age effect (RAE) (3). The aims of this study were (i) to validate a new model (Multilevel regression analysis) for analysing the effects of relative age on the physical performance of elite youth handball players and (ii) to test the predictive capacity of this model
METHODS:
The data were collected over 17 years on 439 elite youth handball players. Players were classified by chronological age, maturity status (advanced, normal, delayed) and birth quartile (Q1:January-March;Q2:April-June;Q3:July-September;Q4:October-December). Each year, the 10 m speed test, the Counter Movement Jump (CMJ) test and agility test (T-test) were carried out. The relative age corresponded to the difference between the test date and the date of birth. The birth month of each participant was collected to categorize them into birth semesters: S1 (January to June) and S2 (July to December). Estimated maturity was expressed by a z-score relative to the age group and by the mean and standard deviation of the prediction of the groups adult height. Peak growth velocity was estimated using the Mirwald equation (4). In each annual-age category, a chi-square test was conducted to assess differences between the observed distribution of relative ages and the adjusted distribution based on a uniform distribution. To assess the models accuracy, weve selected a test group of 56 athletes who have taken the same tests under identical conditions but at different time periods, at least twice, the first measure is denoted the base data.
RESULTS:
Preliminary results for the 10 m speed test (p=0.003) and the T-test (p=0.014) show a significant influence of relative age on performance, while the maturity index for the 10 m speed (p=0.11) and the T-test (p=0.237) had no significant effect. For the CMJ test, none of the variables was significant (relative age: p=0.281; maturity index: p=0.205). RAE effects are observed across all annual-age categories, with more pronounced effects among the youngest participants, which diminish as age increases. Specifically, for the 10 m speed, the RAE effect ranges from large (e.g., 12 years, top 50%, χ^2=7.2, p=0.02, S1 vs. S2, odds ratio=4) to small, and for the t test, it ranges from medium to small.
Based on the test group the results for the 10-meter speed indicate a significant difference between the distribution of the base data and the observed (p=4.92e-07) and adjusted (p=6.054e-06) data, but not between the observed and adjusted data (p=0.4668). Concerning the T-test, the Wilcoxon test revealed significant differences between all the sample groups, particularly between the observed and adjusted data (p=5.555e-06).

CONCLUSION:
The model presented here makes it possible to reduce the effect of RAE on the results of physical tests and accurately predicts performance