Relative Age Effect and Yo-Yo IR1 in Youth Soccer

Training & Testing Relative Age Effect and Yo-Yo IR1 in Youth Soccer Authors D. Deprez1*, R. Vaeyens1*, A. J. Coutts2, M. Lenoir1, R. Philippaerts1 Affiliations 1 Key words ▶ talent selection ● ▶ peak height velocity ● ▶ maturity ● ▶ endurance ● Movement and Sports Sciences, Ghent University, Ghent, Belgium School of Leisure, Sport & Tourism, University of Technology, Sydney, Lindfield, Australia Abstract ▼ The aims of the study were to investigate the presence of a relative age effect and the influence of birth quarter on anthropometric characteristics, an estimation of biological maturity and performance in the Yo-Yo Intermittent Recovery Test level 1 in 606 elite, Flemish youth soccer players. The sample was divided into 5 chronological age groups (U10– U19), each subdivided into 4 birth quarters. Players had their APHV estimated and height, weight and Yo-Yo IR1 performance were assessed. Differences between quarters were investigated using uniand multivariate analyses. Overall, significantly (P < 0.001) more players were born in the first Introduction ▼ accepted after revision March 28, 2012 Bibliography DOI http://dx.doi.org/ 10.1055/s-0032-1311654 Published online: 2012 Int J Sports Med © Georg Thieme Verlag KG Stuttgart · New York ISSN 0172-4622 Correspondence Dieter Deprez Movement and Sports Sciences Ghent University Watersportlaan 2 9000 Ghent Belgium Tel.: +32/9/2649 441 Fax: +32/9/2646 484 [email protected] Competition categories in most youth sports are organized into annual age groups with discrete cut-off dates. Whilst the intent of this approach is to provide equal competition, fair play and ageappropriate training for young athletes, these age-derived categories are responsible for creating subtle chronological age advantages [11]. This difference in chronological age is referred to as relative age, and its consequences are known as the relative age effect (RAE) [3, 33]. Being chronologically older within a/an (annual) sporting cohort provides significant attainment advantages when compared with those who are chronologically younger [3, 4]. In support, several authors have revealed skewed birth date distributions with overrepresentations of youth and professional level athletes born in the first part of the selection year in various sports [4, 11, 33]. Specifically, in soccer, players born in the first part of the selection year are likely to be more present at elite level [40]. It is generally considered that differences in growth and maturation *Authors with equal contributions. quarter (37.6 %) compared to the last (13.2 %). Further, no significant differences in anthropometric variables and Yo-Yo IR1 performance were found between the 4 birth quarters. However, there was a trend for players born in the first quarter being taller and heavier than players born in the fourth quarter. Players born in the last quarter tended to experience their peak in growth earlier, this may have enabled them to compete physically with their relatively older peers. Our results indicated selection procedures which are focused on the formation of strong physical and physiological homogeneous groups. Relative age and individual biological maturation should be considered when selecting adolescent soccer players. and the advantages of a greater physique are the major contributing factors to explain the increased success for players born earlier in the selection year [28, 33]. Since youth athletes with advanced biological maturation tend to have increased physical capacities compared to age-matched but less mature counterparts, coaches and talent scouts tend to favour the physically advanced players [26]. Several studies have shown that soccer players with increased biological maturity perform better in strength, power, speed and endurance, especially during the pubertal years (11– 15 years) [6, 7, 14, 15, 25, 27, 41]. Moreover, it has been shown that athletes born earlier in the selection year are taller and heavier than athletes born later in the selection year [6, 21]. Indeed, Sherar et al. [37] concluded that team selectors appear to preferentially select taller, heavier and early maturing male ice hockey players (aged 14–15 years) who have birth dates early in the selection year. In contrast, Hirose [21] reported no differences in height, body mass and skeletal age between the 4 birth quarters in 9–15-yearold elite Japanese soccer players selected into representative teams. Notably, however, the Deprez D et al. Relative Age Effect and … Int J Sports Med Downloaded by: UNIVERSITEIT GENT. Copyrighted material. 2 Training & Testing Deprez D et al. Relative Age Effect and … Int J Sports Med Materials and Methods ▼ Subjects and design Elite youth male soccer players from 2 professional soccer clubs from the Belgian first division participated in this mixed-longitudinal study. The age range of the players was 9.1–18.8 years. All players and their parents or legal representatives were fully informed of experimental procedures before giving their written informed consent to participate. The study was approved by the Ethics Committee of the Ghent University Hospital and the study was performed in accordance with the ethical standards of the International Journal of Sports Medicine [16]. The original data set contained 2 901 observations; however, to account for effect of familiarization on physical performance, the first Yo-Yo IR1 of each player was not included in the final data set. Additionally, age categories younger than 9 ( < U10) and older than 18 years ( > U19) were also excluded because of low frequencies to assure sufficient statistical power. The final data set consisted of 1 253 data points of the Yo-Yo IR1 from 606 players who were classified into 5 age categories (U10–U11: n = 241; U12–U13: n = 271; U14–U15: n = 272; U16–U17: n = 269; U18– U19: n = 200). All players were born between 1988 and 2001 (e. g., players born in 1996 who were assessed in 2009 belong to the U14 age category). The data included in the present analysis was collected from 12 test occasions, between August 2007 and August 2010. Within each test year, 2 (in 2007 and 2010) to 4 (in 2008 and 2009) test periods were scheduled. Accordingly, a small number of players had several measures taken within each age category. To ensure that only one measure was taken for each player within each age category, the best performance on the Yo-Yo IR1 was taken. This approach ensured that each player only had one data point included within each age category and a maximum of 4 measures across different age categories (n players at one test result = 221; n players at 2 test results = 209; n players at 3 test results = 90; n players at 4 test results = 86). Birth date distribution To examine birth date distribution, players were divided into 4 birth quarters (BQ) and 2 semesters (S) according to their birth month (BQ1: January–March; BQ2: April– June; BQ3: July–September; BQ4: October–December and S1: January–June; S2: July–December). With a cut-off date of January 1, the selection year for youth soccer in Belgium runs from January 1 to December 31. Anthropometric measures Anthropometric measures of height (0.1 cm, Harpenden Portable Stadiometer, Holtain, UK), sitting height (0.1 cm, Harpenden Sitting Height Table, Holtain, UK) and body mass (0.1 kg, total body composition analyzer, TANITA BC-420SMA, Japan) were assessed according to previously described procedures (Lohman, 1988) and to manufacturer’s guidelines. Leg length was calculated by subtracting sitting height from stature. All anthropometric measures were taken by the same investigator to ensure test accuracy and reliability. The intra-class correlation coefficient for test-retest reliability and technical error of measurement (test-retest period of 1 h) in 40 adolescents were 1.00 (p < 0.001) and 0.49 cm for height and 0.99 (p < 0.001) and 0.47 cm for sitting height, respectively. Downloaded by: UNIVERSITEIT GENT. Copyrighted material. small number of players born later in the selection year also possessed advanced biological and physical maturation, which likely explain why these players were successfully selected into the elite representative teams. A similar trend was reported by Carling et al. [6], who suggested that the relative older age of soccer players (aged 14 years) may not always be linked to a significant advantage in physical and physiological components. Research from a variety of team sports, such as soccer, basketball and handball, have shown that the ability to perform intermittent high intensity activity seems to be an important discriminating factor between elite and sub-elite players [2]. Indeed, it is widely reported that soccer players from higher levels of competition (i. e., higher level professional leagues) travel greater distances during games at higher speeds than lower level counterparts [31]. Moreover, it has been suggested that increased aerobic fitness is an important physiological quality that allows players to recover faster between high intensity efforts and exercise at higher intensities during prolonged high intensity intermittent exercise [2, 20]. The Yo-Yo Intermittent Recovery Test Level 1 (Yo-Yo IR1) is a soccer specific field test that maximizes the aerobic energy system through intermittent exertion [1, 8, 23]. Several previous studies have shown that the Yo-Yo IR1 performance has a high level of reproducibility [23, 39] and is a valid measure of prolonged, high intensity intermittent running capacity [38]. Moreover, strong correlations have been reported between the Yo-Yo IR1 performance and the amount of high intensity running during a soccer match [2, 8, 23, 24, 39]. Whilst there is relatively little information available on Yo-Yo IR1 performance in elite youth soccer players, Rampinini et al. [34] and Castagna et al. [9, 10] reported distances of 2 150 ± 327 m (n = 16), 842 ± 352 m (n = 21) and 760 ± 283 m (n = 18) for elite soccer players, aged 17.6 ± 0.5 years, 14.1 ± 0.2 years and 14.4 ± 0.1 years, respectively. An experimental study by Hill-Haas et al. [20] reported Yo-Yo IR1 distances between 1 488 ± 345 m and 2 115 ± 261 m before and after the implementation of a soccer-specific preseason training program, respectively. Recently, a study by Markovic et al. [29] reported Yo-Yo IR1 performances of 106 elite, Croatian youth soccer players in 7 age-groups during adolescence varying from U13 to U19. The Yo-Yo IR1 distances ranged from 933 ± 241 m within U13players (n = 17) to 2 128 ± 326 m within U19-players (n = 15). However, at present there is little information on the changes in Yo-Yo IR1 performance in youth soccer players during adolescence. Such information may be useful for the process of monitoring development of physical capacity in gifted players. To our knowledge, there is little information on age related variance in performance in Yo-Yo IR1 in youth soccer players. Additionally, there have only been a few studies that have investigated the association between performance characteristics, biological maturity and the relative age effect in youth soccer players [6, 21, 37]. Therefore, the aims of this study were: (1) to describe the distribution of birth dates in elite Flemish youth soccer players (U10–U19) and (2) to examine the influence of relative age and an estimation of biological maturity on anthropometric characteristics and performance in Yo-Yo IR1 across the 4 birth quarters of the selection year in these elite youth soccer players. Training & Testing Table 1 Birth date distribution per quarter (BQ) and semester (S) by age group (n ( %)). BQ and S BQ 1 BQ 2 BQ 3 冧 S1 U10–U19 Flanders U10–U11 Flanders U12–U13 Flanders U14–U15 Flanders U16–U17 Flanders U18–U19 920 172 200 194 200 154 Flanders 346 (37.6 %) 272 (29.6 %) 618 (67.2 %) 81 921 (25.0 %) 83 539 (25.4 %) 60 (34.9 %) 50 (29.1 %) 110 (64.0 %) 15 582 (24.9 %) 15 926 (25.4 %) 82 (41.0 %) 59 (29.5 %) 141 (70.5 %) 15 827 (24.9 %) 16 135 (25.3 %) 84 (43.3 %) 48 (24.7 %) 132 (68.0 %) 16 292 (24.9 %) 16 687 (25.5 %) 66 (33.0 %) 64 (32.0 %) 130 (65.0 %) 16 999 (25.1 %) 17 214 (25.4 %) 54 (35.1 %) 51 (33.1 %) 105 (68.2 %) 17 221 (25.0 %) 17 577 (25.5 %) BQ 4 冧 n χ23 (BQ) χ21 (S) S2 181 (19.7 %) 121 (13.2 %) 302 (32.8 %) 84 741 (25.8 %) 78 124 (23.8 %) 41 (23.8 %) 21 (12.2 %) 62 (36.0 %) 16 162 (25.8 %) 14 937 (23.9 %) 33 (16.5 %) 26 (13.0 %) 59 (29.5 %) 16 525 (26.0 %) 15 178 (23.8 %) 35 (18.0 %) 27 (13.9 %) 62 (32.0 %) 16 816 (25.7 %) 15 610 (23.9 %) 43 (21.5 %) 27 (13.5 %) 70 (35.0 %) 17 502 (25.8 %) 15 997 (23.6 %) 29 (18.8 %) 20 (13.0 %) 49 (31.8 %) 17 736 (25.7 %) 16 402 (23.8 %) 122.1*** 103.3*** 17.8*** 12.7*** 38.9*** 32.9*** 38.7*** 24.0*** 18.5*** 16.7*** 20.1*** 19.2*** ***P < 0.001 Yo-Yo IR1 The Yo-Yo IR1 was conducted according to the methods of Krustrup et al. [23]. Participants were instructed to refrain from strenuous exercise for at least 48 h before the test sessions and to consume their normal pre-training diet before the test session. A standardized warming-up preceded each Yo-Yo IR1. All tests were completed on an indoor tartan running track at a temperature between 15–20 °C. The total duration of the test was 2–25 min and the individual scores were expressed as covered distance (m). All subjects ran the Yo-Yo IR1 test at least twice. In order to account for test familiarization, the first result was not taken into account. All players ran the test with running shoes. Maturity status An estimation of the biological maturity status from each player was calculated using equation 3 from Mirwald et al. [30]: Maturity offset = –9.236 + 0.0002708 (leg length × sitting height) – 0.001663 (decimal age × leg length) + 0.007216. (decimal age × sitting height) + 0.02292 (weight/height ratio) This non-invasive method, based on anthropometric variables, predicts years from peak height velocity as a measure of maturity offset. Consequently, age at peak height velocity (APHV) was calculated as the difference between chronological age (CA) and the predicted time (years) from peak height velocity (i. e., maturity offset). CA was calculated as the difference between the player’s birth date and the test date according to the table of Weiner and Lourie (1969). According to Mirwald et al. [30], equation 3 accurately estimates the maturity offset within an error of ± 1.14 years in 95 % of the cases in boys. This predictive equation was developed using data from 3 longitudinal studies (SGDS: Bailey, 1968; BMAS: Bailey, 1997; LLTS: Maes et al., 1996) on children who were 4 years from and 3 years after PHV (i. e., 13.8 years). Accordingly the age range from which the equation can be confidently applied is from 9.8 − 16.8 years. Therefore, in the present study the equation was only applied to players in the U10–U17 age categories. This equation was not applied to the U18 and U19 categories which included players aged 17.1–18.8 years. Statistical analyses All statistical analyses were completed using SPSS for windows (version 19.0). All results are presented as mean ± SD. First, differences between the observed and the expected birth date distributions were tested with chi-square statistics. Expected birth date distributions were calculated in accordance with the birth rate in Flanders between 1989 and 2001 (National Institute of Statistics) using weighted means. Second, within each age category, differences for chronological age (CA) and APHV were investigated between birth quarters (independent variable) using one-way analysis of variance (ANOVA). Multivariate analysis of covariance (MANCOVA) with CA and APHV as covariates and height, weight and Yo-Yo IR1 performance as dependent variables were used to examine differences between birth quarters (independent variable). Chronological age and APHV were controlled for as these are potential confounding factors in the analysis especially since significant differences in these variables were observed across birth quarters within each age category (U10–U11, Age: F = 14.393, P < 0.001, APHV: F = 3.781, P < 0.05; U12–U13, Age: F = 18.398, P < 0.001, APHV: F = 4.015, P < 0.01; U14–U15, Age: F = 10.195, P < 0.001; U16–U17, Age: F = 13.116, P < 0.001; U18–U19, Age: F = 14.778, P < 0.001). Within the U18– U19 age category, data were only adjusted for CA because the Mirwald equation had not previously been validated in these age groups. To interpret the results more distinct, partial eta squared (ŋ2) values were calculated. Threshold values for effect size statistics were 0.01, 0.06 and 0.14 for small, medium and large effect sizes, respectively [12]. Minimal statistical significance was set at P < 0.05. Follow-up univariate analyses using Bonferroni post hoc test were used where appropriate. Deprez D et al. Relative Age Effect and … Int J Sports Med Downloaded by: UNIVERSITEIT GENT. Copyrighted material. Age Category Training & Testing ▼ ▶ Table 1 shows the birth date distribution by quarter and ● semester for the total sample (U10–U19) and for the 5 age categories separately. Overall, 37.6 % of the players were born in the first quarter, while only 13.2 % of the players were born in the fourth (i. e., last) quarter. More detailed analysis within the age categories revealed that the percentage of players born in the first quarter of the selection year varied between 33.0 and 43.3 %, and 12.2–13.9 % for the last quarter. The birth date distribution of the soccer players differed significantly from the Flemish population (U10–U19, χ23 = 122.1, P < 0.001; U10–U11, χ23 = 17.8, P < 0.001; U12–U13, χ23 = 38.9, P < 0.001; U14–U15, χ23 = 38.7, P < 0.001; U16–U17, χ23 = 18.5, P < 0.001; U18–U19, χ23 = 20.1, P < 0.001). The distribution of players between semesters also demonstrated that a greater proportion of players were born in the first semester of the selection year (67.2 % for the total sample and 64.0– 70.5 % amongst the age categories). Similar to the quarterly distribution, there were significant differences from the Flemish population and the observed birth date distribution by semester (U10–U19, χ21 = 103.3, P < 0.001; U10–U11, χ21 = 12.7, P < 0.001; U12–U13, χ21 = 32.9, P < 0.001; U14–U15, χ21 = 24.0, P < 0.001; U16–U17, χ21 = 16.7, P < 0.001; U18–U19, χ21 = 19.2, P < 0.001). Anthropometric variables and Yo-Yo IR1 performance across the ▶ Table 2. 4 birth quarters for each age category are shown in ● The MANCOVA analysis demonstrated no significant main effect for birth quarter within all age categories: U10–U11 (F(9, 399) = 0.55, Wilks’ λ = 0.97), U12–U13 (F(9, 467) = 1.07, Wilks’ λ = 0.95), U14–U15 (F(9, 453) = 0.86, Wilks’ λ = 0.96), U16–U17 (F(9, 467) = 1.08, Wilks’ λ = 0.95) and U18–U19 (F(9, 355) = 1.13, Wilks’ λ = 0.93). Between-subjects effects for the covariates of age and APHV revealed a significant influence on height and weight in age categories U10–U17. Further, there was a significant effect of chronological age on the Yo-Yo IR1 performance in all age categories, except for age categories U10–U11 and U18– U19. Also, with the exception of the U10–U11 category, APHV did not influence the Yo-Yo IR1 performance in all age categories. In addition, the one way-ANOVA for APHV between the 4 birth quarters revealed significant differences within age categories U10–U11 (F = 3.781; P < 0.05) and U12–U13 (F = 4.015; P < 0.01). These results illustrate an earlier APHV for players born in the fourth birth quarter compared with players born in the first birth quarter. Discussion ▼ The aims of this study were to investigate the presence of a relative age effect and the influence of birth quarter on anthropometric variables, estimated biological maturation and Yo-Yo IR1 performance in 606 Flemish elite youth soccer players. The Table 2 Anthropometric variables, estimation of biological maturity and Yo-Yo IR1 performance of elite youth soccer players (U10–U19) across 4 birth quarters (BQ1-BQ4). Age Variable BQ1 BQ2 BQ3 BQ4 N = 60 9.7 ± 0.6a 13.0 ± 0.4 138.9 ± 5.2 32.2 ± 4.4 739 ± 270 N = 82 12.0 ± 0.6a 13.8 ± 0.4a 150.2 ± 6.5 38.4 ± 5.0 1186 ± 402 N = 84 13.8 ± 0.6a 14.0 ± 0.6 162.6 ± 9.2 49.0 ± 10.0 1565 ± 393 N = 66 15.8 ± 0.6a 14.1 ± 0.7 174.5 ± 6.5 61.9 ± 8.1 2012 ± 427 N = 54 17.7 ± 0.5a 177.6 ± 6.6 68.7 ± 6.7 2139 ± 462 N = 50 9.6 ± 0.6a 13.0 ± 0.4 138.0 ± 5.7 30.9 ± 4.2 797 ± 267 N = 59 11.5 ± 0.6b 13.6 ± 0.4b 148.8 ± 7.2 38.1 ± 5.9 1126 ± 351 N = 48 13.7 ± 0.5a, b 14.0 ± 0.5 161.5 ± 7.6 49.2 ± 8.4 1616 ± 422 N = 64 15.7 ± 0.6a, b 14.0 ± 0.6 174.0 ± 7.6 63.0 ± 8.8 1961 ± 416 N = 51 17.4 ± 0.5b 178.4 ± 6.9 70.0 ± 8.2 2187 ± 465 N = 41 9.1 ± 0.5b 12.8 ± 0.4 135.4 ± 4.9 29.6 ± 3.8 748 ± 275 N = 33 11.4 ± 0.5b 13.7 ± 0.3a, b 147.0 ± 5.5 36.5 ± 4.9 1008 ± 248 N = 35 13.4 ± 0.5b, c 14.0 ± 0.6 160.5 ± 8.3 47.5 ± 8.6 1410 ± 355 N = 43 15.5 ± 0.6b 14.0 ± 0.7 172.4 ± 7.6 60.7 ± 9.2 1900 ± 374 N = 29 17.3 ± 0.6b, c 175.6 ± 5.9 66.8 ± 7.8 2219 ± 402 N = 21 9.0 ± 0.6b 12.8 ± 0.3 134.3 ± 4.6 29.5 ± 3.3 705 ± 242 N = 26 11.3 ± 0.6b 13.6 ± 0.3a, b 145.7 ± 6.2 36.2 ± 4.8 1218 ± 363 N = 27 13.3 ± 0.6c 13.8 ± 0.5 160.8 ± 8.2 47.7 ± 8.2 1512 ± 184 N = 27 15.0 ± 0.6c 13.8 ± 0.6 173.6 ± 6.8 59.8 ± 6.1 1770 ± 416 N = 20 16.9 ± 0.6c 175.9 ± 7.0 68.4 ± 8.3 2210 ± 453 Covariates Category U10–U11 age (years) APHV (years) height (cm) weight (kg) Yo-Yo IR1 (m) U12–U13 age (years) APHV (years) height (cm) weight (kg) Yo-Yo IR1 (m) U14–U15 age (years) APHV (years) height (cm) weight (kg) Yo-Yo IR1 (m) U16–U17 age (years) APHV (years) height (cm) weight (kg) Yo-Yo IR1 (m) U18–U19 age (years) height (cm) weight (kg) Yo-Yo IR1 (m) F(Age) P F(APHV) P F(BQ) P – – 547.204 287.767 0.004 – – *** *** n. s. – – 498.247 345.655 5.255 – – *** *** * 14.393# 3.781# 0.954 0.296 0.492 *** * n. s. n. s. n. s. – – 448.446 241.065 9.347 – – *** *** ** – – 367.365 273.099 0.408 – – *** *** n. s. 18.398# 4.015# 0.483 0.627 1.940 *** ** n. s. n. s. n. s. – – 232.291 212.375 17.607 – – *** *** *** – – 833.955 697.117 0.647 – – *** *** n. s. 10.195# 0.674# 0.079 1.135 1.263 *** n. s. n. s. n. s. n. s. – – 113.074 89.093 5.398 – – *** *** * – – 432.137 347.692 0.012 – – *** *** n. s. 13.116# 1.246# 0.947 1.124 0.738 *** n. s. n. s. n. s. n. s. – 0.403 6.672 0.641 – n. s. * n. s. – – – – – – – – 14.778# 1.191 1.309 0.435 *** n. s. n. s. n. s. Means having a different subscript are significantly different at p < 0.05. Between-subjects effects for covariates and BQ are significant at:* p < 0.05; ** p < 0.01; *** p < 0.001; n. s. not significant. # F- and P-values for one way analysis of variance Deprez D et al. Relative Age Effect and … Int J Sports Med Downloaded by: UNIVERSITEIT GENT. Copyrighted material. Results results demonstrated an asymmetry in birth month distribution with ~40 % of players born in the first quarter of the selection year, which corresponds to ~1.5 times the expected frequency in the general Flemish population. Distribution of players in the first quarter within age categories U12–U13 and U14–U15 were more distinct (~42 %) than in age categories U10–U11, U16–U17 and U18–U19 (~34 %), while percentages of players born in the fourth quarter remained constant over the 5 age categories (~13 %). Further, there were no significant differences in anthropometric variables and Yo-Yo IR1 performance between the 4 birth quarters. However, there was a trend for players born in the first birth quarter being taller and heavier than players born in the fourth quarter. APHV did not influence the Yo-Yo IR1 performance. This finding supports the results of previous studies [6, 21, 28]. Notably, the values for APHV within the U10–U11 (9–10 years old) group in this study are lower than within the rest of the agegroups. This could be explained by the age of the verification samples (i. e., children between 11 and 16 years old) used for the development of Mirwald’s predictive equation [30]. Although Mirwald et al. [30] have reported that the formula is appropriate for athletes aged 10–16 years, it appears that the estimation is more accurate for athletes in the middle of this range. However, since the players in the present study were only compared within the same age-group these limitations of the predictive equation are not so important. The present Yo-Yo IR1 results are similar to Rampinini et al. [34] who reported a distance of 2150 ± 327 m in 17-year-old elite soccer players. Moreover, Hill-Haas et al. [20] also showed similar performance levels in talented 14-year-old Australian soccer players at the start of an experimental study (i. e., 1 488 ± 345 m for the experimental and 1764 ± 256 m for the control group). These comparisons show the high level of intermittent-endurance performance of the tested Belgian young elite players. Indeed, Bangsbo et al. [2] also reported lower Yo-Yo IR1 performance levels in an elite population of American and New Zealand youth soccer players aged 12–18 years (personal communication, unpublished observation). In addition, the present population had a considerably greater performance than that of 106 age-matched Croatian soccer players (e. g., Croatian U17 players: 1 581 ± 390 m vs. current U17 players: 1911 ± 408 m) [29]. The first aim of this study was to examine the presence of an RAE in elite Flemish youth soccer players. The findings revealed a skewed distribution of birth dates over the 5 age categories towards an earlier birth date which was in contrast to the evenly distributed general Flemish population. In agreement with many previous studies [4, 11], we observed that more youth soccer players were born in the first quarter of the selection year (from 33.0 to 43.3 %) compared with the fourth quarter (12.2 to 13.9 %). Indeed, several previous studies have shown that athletes who are relatively older within their age group are more likely to be selected to compete at the elite level in ice hockey, rugby, volleyball and basketball [4, 11]. Moreover, the relative proportion of players born in the first and last quarter of each selection year is similar to those previously reported in elite Spanish, Basque and Belgian youth soccer players (i. e., first quarter: 32.2 – 47.8 %, fourth quarter: 6.8–18.0 %) [13, 17, 19, 22, 32]. Similar to soccer, most sports that use annual age groupings to classify competition levels demonstrate subtle chronological age differences. Whilst the age-groups are intended to provide young athletes with better opportunities for instruction appropriate for their development, equal competition and fair play, it seems that these groupings create a positive selection bias for relatively older athletes. Indeed, in accordance with observations of others [18, 28, 40] the present results indicate that relatively older soccer players also receive early recognition from coaches and talent scouts. This has been suggested to be due to their larger anthropometric dimensions and increased physiological capacity, rather than advantages in technical or tactical skills, especially during puberty and adolescence [28]. Accordingly, it seems logical to assume that in sports such as soccer where an advanced physical development is advantageous, the relatively younger players are at a considerable disadvantage. However, in contrast, the present results showed no differences in anthropometric and physiological characteristics between players across all birth quarters in each category. Nonetheless, there was a trend with players born in the first quarter being taller and heavier than players born in the fourth quarter. This tendency was especially apparent in the younger age categories (further analysis revealed small to medium effect sizes for height (0.001–0.017) and weight (0.005–0.050) in all age categories). Whilst these tendencies in anthropometry are likely to be practically important (i. e., relatively older and thus taller players are likely to be more selected), they are most likely explained by increased chronological age. These observations agree with previous studies that also reported no differences across the 4 birth quarters in anthropometric and functional capacities in 160 French elite U14 soccer players [6] and 69 Portuguese 13–15 years old youth soccer players [28]. A possible explanation for the lack of differences between the birth quarters is that the talent identification and selection programs from which these players were selected, may have created homogenous groups of players possessing similar anthropometric characteristics and intermittent endurance capacity, whatever their birth month within an age group [6]. This may also explain the trends for differences in age at peak height velocity between the first and the last birth quarter. Indeed, whilst the players born in the fourth quarter are relatively younger, these players have compensated for this disadvantage through demonstrating an earlier age for onset of puberty (i. e., a younger age at peak height velocity). Hirose [21] reported similar findings in a study with 332 Japanese elite youth soccer players, aged 9–15 years, where the few players born late(r) in the selection year that were selected into the elite teams also showed advanced biological and physical characteristics. Collectively, these findings indicate an influence of a greater physique in the process of talent selection in soccer. In this study, it seems that players born later in the selection year have greater biological maturity or enter puberty earlier than players born earlier of the same age cohort to cope with the potential physical and physiological advantages of their relatively older peers. Therefore, coaches should be aware that physical and biological maturation are important components in the selection process. This could explain the homogeneity in anthropometric characteristics and intermittent endurance in the present sample of elite youth soccer players. Soccer players that are born later in the selection year and mature later are less present at elite youth level presumably due to physical disadvantages [33]. Nevertheless, several previous studies have shown that these players eventually achieve similar anthropometric dimensions, body mass, strength and power to those who mature earlier [5, 27, 35]. To compete with taller and stronger peers, these players may improve other qualities or strategies, such as technical and tactical skills and improve psyDeprez D et al. Relative Age Effect and … Int J Sports Med Downloaded by: UNIVERSITEIT GENT. Copyrighted material. Training & Testing chological characteristics such as mental toughness and resilience. If late born and late maturing players avoid early deselection and remain in their sport until late adolescence/ early adulthood (when the physical disadvantages disappear), they often outperform their early born or early mature counterparts. For instance, Carling et al. [6] reported that once players were selected into an elite youth academy (from the age of 13 years), their date of birth did not influence the opportunity to turn professional. Moreover, Vaeyens et al. [40] demonstrated no differences in the likelihood of being selected and playing minutes between early and late born adult Belgian semi-professional soccer players. Although whilst an RAE was observed in these Belgian semi-professional soccer players, it was suggested that early dropout of youth soccer players born later in the year accounted for the skewed birth date distribution. Indeed, there is evidence of a greater rate of dropout in youth soccer players [19] and ice hockey [4] from as early as 12 years. In accordance with these previous studies, the present results showed an RAE through all age categories (U10–U19), suggesting that many gifted, but relatively young players may be systematically overlooked simply because they are born late(r) in the selection year or are late maturing [28]. Additionally, within the last quarter late maturing boys seem no longer represented (dropout). In conclusion, it appears that the combination of being born later in a selection year and also having later maturation provide a significant disadvantage for being selected into elite youth soccer teams. Finally, the present study reported no differences in intermittent endurance performance between early and late born players. Several possible explanations may account for this observation. First, the amount of practice hours, irrespective of birth quarter, within the 2 professional soccer clubs examined in this study is similar. These similarities in physical training stimulus may have resulted in noticeable homogenous training outcome for all players participating in this study. It seems that the talent selection procedures focus on the formation of homogenous groups of players having similar intermittent endurance capacities. Further research is warranted on other physical and physiological parameters, such as speed and explosive strength. Additionally, even players who were not selected in the starting 11 for each match were prescribed additional physical conditioning to ensure that they received similar training stimuli as the starting players for each age group. Furthermore, it has previously been reported that early and late maturing soccer players do not differ in running economy [36]. Indeed, in the 2 teams investigated in the present study, specific coordination programs were implemented and there was specific focus to ensure that each player was trained to move efficiently in soccer specific movements (i. e., change of direction and regular acceleration/decelerations). It was therefore likely that most players had similar movement proficiency which also may explain the lack of differences in the Yo-Yo IR1 performance. Finally, since APHV was not a confounding factor for the performance in the Yo-Yo IR1, the relative advantages of maturation were likely to have a relatively small influence on the Yo-Yo IR1 results. In conclusion, the present findings provide no rationale for identifying and selecting primarily players born in the first quarter of the selection year. Our data revealed no differences in the Yo-Yo IR1 which assesses the soccer-specific aerobic capacity, one of the most important performance determinants. Searching for soccer players who display greater physical dominance (i. e., taller and heavier) over their peers during the selection process Deprez D et al. Relative Age Effect and … Int J Sports Med is likely to delimit selected players to early maturers or those who are relatively older than their peers. Since selection into elite development pathways for youth players often provides increased development and coaching opportunities, these older and more physically mature players are often inappropriately identified as being ‘gifted’. Indeed, there is the risk that players who are equally gifted but physically less mature at younger ages may be deselected on the basis of their poorer physical characteristics and not on their adult potential. At present, few programs that identify and develop young soccer players have the ability to account for these advantages in age and maturational status. Therefore, to overcome these limitations we suggest that greater consideration should be given to assessing individual biological maturation in the selection of adolescent players. The present study indicated identification and development procedures that are focused on the formation of strong physical and physiological homogeneous groups. In elite youth soccer, within a specific age-group, a higher chronological age is not associated with a better Yo-Yo IR1 performance which suggests that the relative age of the players does not provide a significant advantage in terms of soccer-specific endurance. Therefore, coaches and talent scouts should understand that a player who is born late(r) in the selection year is not always a late maturing boy (conversely, a player who is born early in the selection year is not per definition early maturing). Therefore, coaches and talent scouts should aim to identify players with the potential for success in the long term, and focus on the holistic potential of players, including technical, tactical and psychological skills whilst also accounting for relative age and maturational status. The present observations may change the current selection policies in elite soccer and facilitate the selection of greater number of players born in the late part of the selection year. Acknowledgements ▼ Sincere thanks to the parents and children who consented to participate in this study and to the directors and coaches of the participating Flemish soccer clubs, SV Zulte Waregem and KAA Gent. The authors would like to thank the participating colleagues, Job Fransen, Stijn Matthys, Johan Pion, Barbara Vandorpe and Joric Vandendriessche, for their help in collecting data. References 1 Bangsbo J. Fitness Training in Football: A Scientific Approach. August Krogh Institute, Copenhagen University, 1994; 1–336 2 Bangsbo J, Iaia MF, Krustrup P. The Yo-Yo Intermittent Recovery Test: A useful tool for evaluation of physical performance in intermittent sports. Sports Med 2008; 38: 37–51 3 Barnsley HR, Thompson HR, Barnsley PE. Hockey Success and Birthdate: The Relative Age Effect. Canadian Association of Health. Physical Education and Recreation 1985; 51: 23–28 4 Barnsley HR, Thompson AH. Birthdate and success in minor hockey: The key to the NHL. Can J Beh Sci 1988; 20: 167–176 5 Beunen G, Ostyn M, Simons J, Renson R, Claessens AL, Vanden Eynde B, Lefevre J, Vanreusel B, Malina RM, Van t Hof MA. Development and tracking in fitness components: Leuven longitudinal study on lifestyle, fitness and health. Int J Sports Med 1997; 18: 171–178 6 Carling C, Le Gall F, Reilly T, Williams AM. Do anthropometric and fitness characteristics vary according to birth date distribution in elite youth academy soccer players? Scand J Med Sci Sports 2009; 19: 3–9 7 Carvalho HM, Coelho e Silva MJ, Figueiredo AJ, Conçalves CE, Castagna C, Philippaerts RM, Malina RM. Cross-validation and reliability of the Downloaded by: UNIVERSITEIT GENT. Copyrighted material. Training & Testing 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 line-drill test of anaerobic performance in basketball players 14–16 years. J Strength Cond Res 2011; 25: 1113–1119 Castagna C, Impellizzeri FM, Chamari K, Carlomagno D, Rampinini E. Aerobic fitness and yo-yo continuous and intermittent tests performances in soccer players: A correlation study. J Strength Cond Res 2006; 20: 320–325 Castagna C, Impellizzeri FM, Cecchini E, Rampinini, Barbero Alvarez JC. Effects of intermittent-endurance fitness on match performance in young male soccer players. J Strength Cond Res 2009; 23: 1954–1959 Castagna C, Manzi V, Impellizzeri FM, Weston M, Barbero Alvarez JC. Relationship between endurance field tests and match performance in young soccer players. J Strength Cond Res 2010; 24: 3227–3233 Cobley S, Baker J, Wattie N, McKenna J. Annual age-grouping and athlete development: a meta-analytical review of relative age effects in sport. Sports Med 2009; 39: 235–256 Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale (NJ): Lawrence Erlbaum, 1988: 567 Del Campo DG, Vicedo JC, Villora SG, Jordan OR. The relative age effect in youth soccer players from Spain. J Sports Sci Med 2010; 9: 190–198 Figueiredo AJ, Gonçalves CE, Coelho e Silva MJ, Malina RM. Youth soccer players, 11–14 years: Maturity, size, function, skill and goal orientation. Ann Hum Biol 2009; 36: 60–73 Figueiredo AJ, Gonçalves CE, Coelho e Silva MJ, Malina RM. Characteristics of youth soccer players who drop out, persist or move up. J Sports Sci 2009; 27 (2009): 883–891 Harriss DJ, Atkinson G. Update – ethical standards in sport and exercise science research. Int J Sports Med 2011; 32: 819–821 Helsen WF, Starkes JL, Van Winckel J. The influence of relative age on success and dropout in male soccer players. Am J Hum Biol 1998; 10: 791–798 Helsen WF, Hodges NJ, Van Winckel J, Starkes JL. The roles of talent, physical precocity and practice in the development of soccer expertise. J Sports Sci 2000; 18: 727–736 Helsen WF, Van Winckel J, Williams AM. The relative age effect in youth soccer across Europe. J Sports Sci 2005; 23: 629–636 Hill-Haas SV, Coutts AJ, Rowsell GJ, Dawson BT. Generic versus smallsided game training in soccer. Int J Sports Med 2009; 30: 636–642 Hirose N. Relationships among birth-date distribution, skeletal age and anthropometric characteristics in adolescent elite soccer players. J Sports Sci 2009; 27: 1159–1166 Jimenez IP, Pain MTG. Relative age effect in Spanish association football: Its extent and implications for wasted potential. J Sports Sci 2008; 26: 995–1003 Krustrup P, Mohr M, Amstrup R, Rysgaard T, Johansen J, Steensberg A, Pedersen KP, Bangsbo J. The Yo-Yo Intermittent Recovery Test: physiological response, reliability and validity. Med Sci Sports Exer 2003; 35: 697–705 24 Krustrup P, Mohr M, Nybo L, Jensen JM, Nielsen JJ, Bangsbo J. The yo-yo IR2 test: Physiological response, reliability and application to elite soccer. Med Sci Sports Exerc 2006; 38: 1666–1673 25 Le Gall F, Carling C, Williams AM, Reilly T. Anthropometric and fitness characteristics of international, professional and amateur male graduate soccer players from an elite youth academy. J Sci Med Sport 2010; 13: 90–95 26 Malina RM, Peňa Reyes ME, Eisenmann JC, Horta L, Rodrigues J, Miller R. Height, mass and skeletal maturity of elite Portuguese soccer players aged 11–16 years. J Sport Sci 2000; 18: 685–693 27 Malina RM, Eisenmann JC, Cumming SP, Ribeiro B, Aroso J. Maturityassociated variation in the growth and functional capacities of youth football (soccer) players 13–15 years. Eur J Appl Physiol 2004; 91: 555–562 28 Malina RM, Ribeiro B, Aroso J, Cumming SP. Characteristics of youth soccer players aged 13–15 years classified by skill level. Br J Sports Med 2007; 41: 290–295 29 Markovic G, Mikulic P. Discriminate ability of the Yo-Yo Intermittent Recovery Test (level 1) in prospective young soccer players. J Strength Cond Res 2011; 25: 2931–2934 30 Mirwald RL, Baxter-Jones ADG, Bailey DA, Beunen GP. An assessment of maturity from anthropometric measurements. Med Sci Sport Exerc 2002; 34: 689–694 31 Mohr M, Krustrup P, Bangsbo J. Match performance of high-standard soccer players with special reference to develompent of fatigue. J Sport Sci 2003; 21: 519–528 32 Mujika I, Vaeyens R, Matthys SPJ, Santisteban J, Goiriena J, Philippaerts RM. The relative age effect in a professional football club setting. J Sport Sci 2009; 27: 1153–1158 33 Musch J, Grondin S. Unequal competition as an impediment to personal development: A review of the relative age effect in sport. Dev Review 2001; 21: 147–167 34 Rampinini E, Impellizzeri FM, Castagna C, Azzalin A, Brabo DF, Wisløff U. Effect of match-related fatigue on short-passing ability in young soccer players. Med Sci Sport Exerc 2007; 40: 934–942 35 Rösch D, Hodgson R, Peterson L, Graf-Baumann T, Junge A, Chomiak J, Dvorak J. Assessment and evaluation of football performance. Am J Sports Med Suppl 2000; 28: S29–S39 36 Segers V, De Clercq D, Janssens M. Running economy in early and late maturing youth soccer players does not differ. Br J Sports Med 2008; 42: 289–294 37 Sherar LB, Baxter-Jones ADG, Faulkner RA, Russell KW. Do physical maturity and birth date predict talent in male youth ice hockey players? J Sports Sci 2007; 25: 879–886 38 Sirotic AC, Coutts AJ. Physiological and performance test correlates of prolonged, high–intensity, intermittent running performance in moderately trained women team sport athletes. J Strength Cond Res 2007; 21: 138–144 39 Thomas A, Dawson B, Goodman C. The Yo-Yo test: reliability and association with a 20-m shuttle run and VO2max. Int J Sports Physiol Perf 2006; 1: 137–149 40 Vaeyens R, Philippaerts RM, Malina RM. The relative age effect in soccer: A match-related perspective. J Sports Sci 2005; 23: 747–756 41 Vaeyens R, Malina RM, Janssens M, Van Renterghem B, Bourgois J, Vrijens, Philippaerts RM. A multidisciplinary selection model for youth soccer: the Ghent Youth Soccer Project. Br J Sports Med 2006; 40: 928–934 Deprez D et al. Relative Age Effect and … Int J Sports Med Downloaded by: UNIVERSITEIT GENT. Copyrighted material. Training & Testing