International Journal of Sports Physiology and Performance, 2015, 10, 1048 -1051 http://dx.doi.org/10.1123/ijspp.2014-0463 © 2015 Human Kinetics, Inc.
brief report
Analysis of the Relative Age Effect in Elite Youth Judo Athletes David H. Fukuda The purpose of this investigation was to evaluate relative age effects (RAEs) in elite youth judo athletes from different chronological age groups, between sexes, and across weight categories. Data from 1542 place winners of the cadet (under 17 y, 2009–2013) and junior judo world championships (under 20/21 y, 1990–2013) were separated by birth month into quarters (Q1, Q2, Q3, and Q4). The observed values were compared with expected annual age distributions using χ2 analyses, and odd ratios (OR) were used to evaluate effect sizes between quarters. The observed frequency of place winners was significantly different from the expected frequency for the age-group and sex comparisons and all body-mass groups (P < .05) with the exception of the extra-light categories (P = .572). When comparing Q1 with Q4 (OR, 95% confidence interval), small effect sizes were observed for cadets (1.72, 1.12–2.66), juniors (1.54, 1.23–1.94), males (1.75, 1.32–2.33), females (1.39, 1.03–1.87), and the light- (1.79, 1.21–2.64) and middle-weight (1.80, 1.20–2.70) categories. RAEs are apparent in cadet and junior judo athletes. Thus, coaches and administrators should consider the potential for physical and/or competitive advantages while adopting strategies that encourage long-term participation in youth judo athletes. Keywords: adolescent, coaching, sport, sport management, physical performance Annual age-grouping strategies such as the separation of cohorts by chronological age may lead to a biased distribution of birth dates and potential for physical and/or competitive advantages, which are referred to as relative age effects (RAEs).1 In sports with selection criteria based on January 1, RAEs typically present a skewed annual distribution with a larger representative frequency of athletes in the first quarter of the year than in the fourth quarter. A negative consequence of RAEs is the possible early termination of sport involvement due to the selection disadvantage experienced by athletes born later in the selection year.1 In combat sports such as judo, greater training age and/or maturation status may be required to attain a sport-specific technical-tactical skill set.2 In a meta-analysis by Cobley et al,1 age, skill level, and sport context were identified as distinguishing factors in the propagation of RAEs. Athletes from older-age categories and elite-level competitors were shown to have lower effect sizes for RAEs than their younger and less competitive counterparts. However, only 2% of these data were related to female samples, and others have reported sex as a potential moderator of RAEs.1,3 Early research examining RAEs in athletes focused on team-based sports,1 but technicaltactical individual sports have also displayed this phenomenon, although the results remain equivocal.3–6 Specifically, the use of weight-class categories has been suggested to nullify or diminish RAEs in combat-sport athletes,5,6 but conflicting data exist3 and this hypothesis has yet to be thoroughly investigated. When separated by weight categories, only heavier Olympic-level judo athletes displayed an annual distribution significantly different from expected values.4 The absence or presence of the RAE in youth combat sports must be confirmed to determine the relevance of potential procedures aimed at improving or maintaining participation throughout adolescence. Therefore, the purpose The author is with the Inst of Exercise Physiology and Wellness, University of Central Florida, Orlando, FL. Address author correspondence to david. [email protected].
1048
of this investigation was to evaluate the RAE in elite youth judo athletes from different chronological age groups, between sexes, and across weight categories.
Methods Subjects and Design/Methodology Birthdate information for place winners (top 7) of the International Judo Federation cadet (under 17 y; 2009–2013) and junior world championships (under 20/21 y; 1990–2013) were compiled from an online source.7 After accounting for multiple place winners, with the first instance being included within each comparison, and individuals with incomplete information by deletion, 1,542 values were compared between age groups (cadets and juniors), and 1,495 values by sex (males and females), and between weight-class categories (extra light, light, middle, and heavy). This represents analyses of birthdates of 93% of cadet (2009–2013) and 86% of junior (1990–2013) place winners. Athletes were separated by birth month into quarter 1 (Q1: January to March), quarter 2 (Q2: April to June), quarter 3 (Q3: July to September), and quarter 4 (Q4: October to December). To maintain adequate sample sizes, and due to regulation changes in 1998 and differences between cadet and junior divisions, the weightclass categories were separated as follows: • Extra light: cadet male ≤46 kg and ≤55 kg, cadet female ≤40 kg and ≤44 kg, junior male ≤55 kg and ≤60 kg, junior female ≤44 kg and ≤48 kg • Light: cadet male ≤60 kg and ≤66 kg, cadet female ≤48 kg and ≤52 kg, junior male ≤65/66 kg and ≤71/73 kg, junior female ≤52 kg and ≤56/57 kg • Middle: cadet male ≤73 kg and ≤81 kg, cadet female ≤57 kg and ≤63 kg, junior male ≤78/81 kg and ≤86/90 kg, junior female ≤61/63 kg and ≤66/70 kg
RAE and Elite Youth Judoka 1049
• Heavy: cadet male ≤90 kg and >90 kg, cadet female ≤70 kg and >70 kg, junior male ≤95/100 kg and >95/100 kg, junior female ≤72/78 kg and >72/78 kg
Discussion
Statistical Analysis Chi-square goodness-of-fit tests were conducted to compare the observed frequency of place winners within each quarter compared with expected values evenly distributed across each quarter (25% for Q1, Q2, Q3, and Q4). A type I error rate of less than or equal to 5% (P ≤ .05) was considered statistically significant. Annual agedistribution values by quarter were calculated as percentages and 95% confidence intervals for graphical comparison. Odds ratios and 95% confidence intervals were calculated for observed versus expected values for Q1, Q2, and Q3 compared with Q4 as described previously.1,6 The odds ratios were interpreted as effect sizes with values of 1.22, 1.86, and 3.00 interpreted as small, medium, and large effects, respectively, as outlined by Olivier and Bell.8 Downloaded by University of Exeter on 09/25/16, Volume 10, Article Number 8
middle-weight categories and in the Q3–Q4 comparison for the light-weight categories.
Results The observed frequency of place winners was significantly different from the expected frequency for the age-group and sex comparisons and all body-mass groups with the exception of the extra-light categories (Table 1; Figure 1). For the Q1–Q4 comparison, small effect sizes were observed for cadets, juniors, males, females, and the light- and middle-weight categories. Small effect sizes in the Q2–Q4 comparison were observed for cadets, males, and the
Contrary to previous results in Olympic-level and professional weight-class athletes,5,6 RAEs are apparent in cadet and junior judo athletes. Furthermore, only the lightest athletes were unaffected by this phenomenon. Similar to previous investigations in other noncombat sports without weight classes,1 larger effects were observed for the younger age group and the male athletes. The results of this investigation support the previously described effects of age and elite status on the magnitude of RAEs. With increased competitive level and experience, dropout rates may be higher in Q1 than in Q4, thus leading to reduced RAEs or potentially a reversal of this effect.9 With the exception of Delorme,6 who reported no RAEs in amateur male and female boxers, data from weight-class athletes showing limited RAEs have primarily come from Olympic competition, where the overrepresentation of Q1 athletes may be decreased or no longer be evident.3–5 Albuquerque et al4 reported that RAEs were only present in the heaviest of Olympic judo competitors and hypothesized that this may be due to an altered physical demand of techniques used when compared with lighter athletes. The current results show that RAEs are present in all but the lightest weight classes, in which smaller stature and body mass could be considered prerequisite. Aesthetic sports with similar anthropometric requirements, such as gymnastics, also tend to display reduced RAEs.1 Previous explanations for
Table 1 Frequency for Each Comparison by Quarter (Q1, Q2, Q3, Q4), With Odds Ratios (OR) and 95% Confidence Intervals (CI)
Cadets
Age, y (mean ± SD)
Q1
Q2
16.1 ± 0.7
105
96
18
9
366
292
299
237
68
–7
1
–62
252
208
181
144
56
12
–15
–52
203
169
192
146
26
–9
15
–32
78
79
66
67
6
7
–7
–6
136
85
123
76
31
–20
18
–29
124
109
91
69
26
11
–7
–29
117
104
93
78
19
6
–5
–20
Δ Juniors
18.7 ± 1.1
Δ Males
18.5 ± 1.4
Δ Females
17.7 ± 1.4
Δ Extra light
17.8 ± 1.6
Δ Light
18.2 ± 1.4
Δ Middle
18.3 ± 1.4
Δ Heavy Δ
18.2 ± 1.4
Q4
Total
χ2
P
Q1 vs Q4, OR (CI)
Q2 vs Q4, OR (CI)
Q3 vs Q4, OR (CI)
86
61
348
12.437
.006
1.72 (1.12–2.66)
1.57 (1.02–2.44)
1.41 (0.91–2.19)
–1
–26 1194
28.077
brief report
Analysis of the Relative Age Effect in Elite Youth Judo Athletes David H. Fukuda The purpose of this investigation was to evaluate relative age effects (RAEs) in elite youth judo athletes from different chronological age groups, between sexes, and across weight categories. Data from 1542 place winners of the cadet (under 17 y, 2009–2013) and junior judo world championships (under 20/21 y, 1990–2013) were separated by birth month into quarters (Q1, Q2, Q3, and Q4). The observed values were compared with expected annual age distributions using χ2 analyses, and odd ratios (OR) were used to evaluate effect sizes between quarters. The observed frequency of place winners was significantly different from the expected frequency for the age-group and sex comparisons and all body-mass groups (P < .05) with the exception of the extra-light categories (P = .572). When comparing Q1 with Q4 (OR, 95% confidence interval), small effect sizes were observed for cadets (1.72, 1.12–2.66), juniors (1.54, 1.23–1.94), males (1.75, 1.32–2.33), females (1.39, 1.03–1.87), and the light- (1.79, 1.21–2.64) and middle-weight (1.80, 1.20–2.70) categories. RAEs are apparent in cadet and junior judo athletes. Thus, coaches and administrators should consider the potential for physical and/or competitive advantages while adopting strategies that encourage long-term participation in youth judo athletes. Keywords: adolescent, coaching, sport, sport management, physical performance Annual age-grouping strategies such as the separation of cohorts by chronological age may lead to a biased distribution of birth dates and potential for physical and/or competitive advantages, which are referred to as relative age effects (RAEs).1 In sports with selection criteria based on January 1, RAEs typically present a skewed annual distribution with a larger representative frequency of athletes in the first quarter of the year than in the fourth quarter. A negative consequence of RAEs is the possible early termination of sport involvement due to the selection disadvantage experienced by athletes born later in the selection year.1 In combat sports such as judo, greater training age and/or maturation status may be required to attain a sport-specific technical-tactical skill set.2 In a meta-analysis by Cobley et al,1 age, skill level, and sport context were identified as distinguishing factors in the propagation of RAEs. Athletes from older-age categories and elite-level competitors were shown to have lower effect sizes for RAEs than their younger and less competitive counterparts. However, only 2% of these data were related to female samples, and others have reported sex as a potential moderator of RAEs.1,3 Early research examining RAEs in athletes focused on team-based sports,1 but technicaltactical individual sports have also displayed this phenomenon, although the results remain equivocal.3–6 Specifically, the use of weight-class categories has been suggested to nullify or diminish RAEs in combat-sport athletes,5,6 but conflicting data exist3 and this hypothesis has yet to be thoroughly investigated. When separated by weight categories, only heavier Olympic-level judo athletes displayed an annual distribution significantly different from expected values.4 The absence or presence of the RAE in youth combat sports must be confirmed to determine the relevance of potential procedures aimed at improving or maintaining participation throughout adolescence. Therefore, the purpose The author is with the Inst of Exercise Physiology and Wellness, University of Central Florida, Orlando, FL. Address author correspondence to david. [email protected].
1048
of this investigation was to evaluate the RAE in elite youth judo athletes from different chronological age groups, between sexes, and across weight categories.
Methods Subjects and Design/Methodology Birthdate information for place winners (top 7) of the International Judo Federation cadet (under 17 y; 2009–2013) and junior world championships (under 20/21 y; 1990–2013) were compiled from an online source.7 After accounting for multiple place winners, with the first instance being included within each comparison, and individuals with incomplete information by deletion, 1,542 values were compared between age groups (cadets and juniors), and 1,495 values by sex (males and females), and between weight-class categories (extra light, light, middle, and heavy). This represents analyses of birthdates of 93% of cadet (2009–2013) and 86% of junior (1990–2013) place winners. Athletes were separated by birth month into quarter 1 (Q1: January to March), quarter 2 (Q2: April to June), quarter 3 (Q3: July to September), and quarter 4 (Q4: October to December). To maintain adequate sample sizes, and due to regulation changes in 1998 and differences between cadet and junior divisions, the weightclass categories were separated as follows: • Extra light: cadet male ≤46 kg and ≤55 kg, cadet female ≤40 kg and ≤44 kg, junior male ≤55 kg and ≤60 kg, junior female ≤44 kg and ≤48 kg • Light: cadet male ≤60 kg and ≤66 kg, cadet female ≤48 kg and ≤52 kg, junior male ≤65/66 kg and ≤71/73 kg, junior female ≤52 kg and ≤56/57 kg • Middle: cadet male ≤73 kg and ≤81 kg, cadet female ≤57 kg and ≤63 kg, junior male ≤78/81 kg and ≤86/90 kg, junior female ≤61/63 kg and ≤66/70 kg
RAE and Elite Youth Judoka 1049
• Heavy: cadet male ≤90 kg and >90 kg, cadet female ≤70 kg and >70 kg, junior male ≤95/100 kg and >95/100 kg, junior female ≤72/78 kg and >72/78 kg
Discussion
Statistical Analysis Chi-square goodness-of-fit tests were conducted to compare the observed frequency of place winners within each quarter compared with expected values evenly distributed across each quarter (25% for Q1, Q2, Q3, and Q4). A type I error rate of less than or equal to 5% (P ≤ .05) was considered statistically significant. Annual agedistribution values by quarter were calculated as percentages and 95% confidence intervals for graphical comparison. Odds ratios and 95% confidence intervals were calculated for observed versus expected values for Q1, Q2, and Q3 compared with Q4 as described previously.1,6 The odds ratios were interpreted as effect sizes with values of 1.22, 1.86, and 3.00 interpreted as small, medium, and large effects, respectively, as outlined by Olivier and Bell.8 Downloaded by University of Exeter on 09/25/16, Volume 10, Article Number 8
middle-weight categories and in the Q3–Q4 comparison for the light-weight categories.
Results The observed frequency of place winners was significantly different from the expected frequency for the age-group and sex comparisons and all body-mass groups with the exception of the extra-light categories (Table 1; Figure 1). For the Q1–Q4 comparison, small effect sizes were observed for cadets, juniors, males, females, and the light- and middle-weight categories. Small effect sizes in the Q2–Q4 comparison were observed for cadets, males, and the
Contrary to previous results in Olympic-level and professional weight-class athletes,5,6 RAEs are apparent in cadet and junior judo athletes. Furthermore, only the lightest athletes were unaffected by this phenomenon. Similar to previous investigations in other noncombat sports without weight classes,1 larger effects were observed for the younger age group and the male athletes. The results of this investigation support the previously described effects of age and elite status on the magnitude of RAEs. With increased competitive level and experience, dropout rates may be higher in Q1 than in Q4, thus leading to reduced RAEs or potentially a reversal of this effect.9 With the exception of Delorme,6 who reported no RAEs in amateur male and female boxers, data from weight-class athletes showing limited RAEs have primarily come from Olympic competition, where the overrepresentation of Q1 athletes may be decreased or no longer be evident.3–5 Albuquerque et al4 reported that RAEs were only present in the heaviest of Olympic judo competitors and hypothesized that this may be due to an altered physical demand of techniques used when compared with lighter athletes. The current results show that RAEs are present in all but the lightest weight classes, in which smaller stature and body mass could be considered prerequisite. Aesthetic sports with similar anthropometric requirements, such as gymnastics, also tend to display reduced RAEs.1 Previous explanations for
Table 1 Frequency for Each Comparison by Quarter (Q1, Q2, Q3, Q4), With Odds Ratios (OR) and 95% Confidence Intervals (CI)
Cadets
Age, y (mean ± SD)
Q1
Q2
16.1 ± 0.7
105
96
18
9
366
292
299
237
68
–7
1
–62
252
208
181
144
56
12
–15
–52
203
169
192
146
26
–9
15
–32
78
79
66
67
6
7
–7
–6
136
85
123
76
31
–20
18
–29
124
109
91
69
26
11
–7
–29
117
104
93
78
19
6
–5
–20
Δ Juniors
18.7 ± 1.1
Δ Males
18.5 ± 1.4
Δ Females
17.7 ± 1.4
Δ Extra light
17.8 ± 1.6
Δ Light
18.2 ± 1.4
Δ Middle
18.3 ± 1.4
Δ Heavy Δ
18.2 ± 1.4
Q4
Total
χ2
P
Q1 vs Q4, OR (CI)
Q2 vs Q4, OR (CI)
Q3 vs Q4, OR (CI)
86
61
348
12.437
.006
1.72 (1.12–2.66)
1.57 (1.02–2.44)
1.41 (0.91–2.19)
–1
–26 1194
28.077
