Measuring Performance To Predict Performance - The Development of The Cycling Compass

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/359847001
Measuring performance to predict performance - the development of the

Cycling Compass -
Thesis · January 2022
CITATIONS READS
0 100
1 author:
Mireille Mostaert
Ghent University
16 PUBLICATIONS 233 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Talent identification within cycling View project
All content following this page was uploaded by Mireille Mostaert on 17 November 2022.
The user has requested enhancement of the downloaded file.

Measuring performance to predict
performance
– the development of the Cycling Compass –
Mireille Mostaert
Thesis submitted in fulfilment of the requirements for the degree of

“Doctor in Health Sciences”
Ghent, 2022
Supervisor
Prof. dr. Matthieu Lenoir Ghent University, Belgium
Co-supervisor
Prof. dr. Frederik Deconinck Ghent University, Belgium
Supervisory board
Prof. dr. Matthieu Lenoir Ghent University, Belgium
Prof. dr. Frederik Deconinck Ghent University, Belgium
Dr. Johan Pion HAN University of Applied Sciences, Netherlands
Ghent University, Belgium
Dr. Pieter Vansteenkiste Ghent University, Belgium
Examination board
Prof. dr. Jan Bourgois (Chairman) Ghent University, Belgium
Prof. dr. Roberto Codella University of Milan, Italy
Prof. dr. Marije Elferink-Gemser University of Groningen, Netherlands
Prof. dr. Reinout Van Schuylenbergh KU Leuven, Belgium
Prof. dr. Dirk De Clercq Ghent University, Belgium
Prof. dr. Jan Boone Ghent University, Belgium
Dr. Bram Constandt Ghent University, Belgium
Printed by University Press, Wachtebeke
@2022 Department of Movement and Sports Sciences, Watersportlaan 2, 9000 Gent, Belgium
ISBN: 9789078836100
All rights reserved. No part of this book may be reproduced, or published, in any form of in any way by print, photo
print, microfilm, or any other means without permission from the author.
Table of contents
List of abbreviations .....................................................................................................................................................................................................................................I

List of figures and tables ...................................................................................................................................................................................................................... III
Summary ............................................................................................................................................................................................................................................................ VII
Samenvatting .................................................................................................................................................................................................................................................. IX
PART 1 : GENERAL INTRODUCTION ................................................................................................................................................................................................ I
1. Project context............................................................................................................................................................................................................1
2. The challenging topic of talent identification...................................................................................................................................... 3
3. From Sports Compass to Cycling Compass..............................................................................................................................................5
3.1. The Sports Compass .......................................................................................................................................................................................... 5

3.2. Current approaches in cycling.................................................................................................................................................................. 6
3.3. The Cycling Compass ........................................................................................................................................................................................ 7
4. Potential components of the Cycling Compass................................................................................................................................... 9
4.1. Lessons from other sports .......................................................................................................................................................................... 9
4.1.1. Growth, maturation and age .................................................................................................................................................................... 9

4.1.2. Anthropometric tests ....................................................................................................................................................................................10
4.1.3. Physical tests .......................................................................................................................................................................................................10
4.1.4. Motor coordination tests ........................................................................................................................................................................... 12
4.1.5. Sport-specific tests ......................................................................................................................................................................................... 12
4.1.6. Additional constraints .................................................................................................................................................................................. 13
4.2. Focus on cycling ................................................................................................................................................................................................. 14
4.2.1. Features and athletic profile of road cycling ........................................................................................................................... 16

4.2.2. Features and athletic profile of track cycling .......................................................................................................................... 17
4.2.3. Features and athletic profile of cyclo-cross .............................................................................................................................18
4.2.4. Features and athletic profile of mountain bike ..................................................................................................................... 19
4.2.5. Features and athletic profile of BMX .............................................................................................................................................20
4.3. Talent identification based on competition performance ........................................................................................... 23
4.3.1. Competition results ....................................................................................................................................................................................... 23

4.3.2. Career pathway ................................................................................................................................................................................................. 24
4.3.3. Additional Constraints ................................................................................................................................................................................. 26
5. Hiatuses for the development of a Cycling Compass ................................................................................................................... 27
5.1. Talent orientation and transfer based upon discipline-specific profiles ........................................................ 27
5.2. Talent identification, selection and performance prediction .................................................................................... 27
5.2.1. Predictions based on a (non-)sport-specific test battery............................................................................................28

5.2.2. Predictions based on competition results .................................................................................................................................28
5.3. Talent development....................................................................................................................................................................................... 29
6. General overview and aims of this research ..................................................................................................................................... 30

7. References ................................................................................................................................................................................................................ 34
PART 2 : ORIGINAL RESEARCH .....................................................................................................................................................................................................45

Chapter 1 : Investigating the athletic profiles of cycling athletes.....................................................................................................................47
Study 1 - Discriminating performance profiles of cycling disciplines ...........................................................................................................49
Chapter 2 : Prediction of future performance in road cycling ............................................................................................................................. 75
Study 2 : Is motor coordination the key to success in youth cycling? .......................................................................................................... 77

Study 3 : The importance of performance in competitions as an indicator of future success in cycling ..................... 99
Study 4 : Do race results in youth competitions predict future success as a road cyclist? A retrospective study in
the Italian Cycling Federation ........................................................................................................................................................................................................ 119
PART 3 : GENERAL DISCUSSION ................................................................................................................................................................................................ 135
1. Summary of the main findings................................................................................................................................................................... 137
1.1. Investigating the athletic profiles of cycling athletes .................................................................................................. 137

1.2. Prediction of future performance in road cycling .............................................................................................................138
1.2.1. Predictions based on a (non-)sport-specific test battery..........................................................................................138

1.2.2. Predictions based on competition results...............................................................................................................................139
2. Overall discussion .............................................................................................................................................................................................. 140
2.1. Motor coordination ...................................................................................................................................................................................... 140
2.1.1. The role of motor coordination in the talent process .................................................................................................. 140
2.1.2. Additional analysis towards detecting motor coordination levels at a young age ............................. 141
2.2. Talent orientation and transfer ......................................................................................................................................................... 143

2.2.1. Road, track, cyclo-cross, and MTB .................................................................................................................................................... 143
2.2.2. Additional analysis towards the profile of BMX .................................................................................................................144
2.3. Talent identification, selection and performance prediction .................................................................................. 147
2.3.1. Longitudinal research in this dissertation............................................................................................................................... 147

2.3.2. Confounding factors....................................................................................................................................................................................149
2.4. Talent development...................................................................................................................................................................................... 151
3. Practical implications ....................................................................................................................................................................................... 153
3.1. Cycling compass .............................................................................................................................................................................................. 153
3.1.1. Talent detection.............................................................................................................................................................................................. 155

3.1.2. Talent orientation and transfer ........................................................................................................................................................ 155
3.1.3. Talent identification and performance prediction ...........................................................................................................156
3.1.4. Talent development .................................................................................................................................................................................... 157
4. Strengths, limitations, and suggestions for future research ................................................................................................ 158
4.1. Strengths ...............................................................................................................................................................................................................158
4.1.1. Adaptations of the test battery for future research .......................................................................................................158

4.1.2. Suggestions for measuring cycling performance for future research ............................................................165
4.2. Limitations ...........................................................................................................................................................................................................165
4.2.1. Main focus on male cycling ..................................................................................................................................................................165

4.2.2. Main focus on road cycling................................................................................................................................................................... 166
4.2.3. Limited multidimensional approach ............................................................................................................................................ 167
4.2.4. Limited information on training history .................................................................................................................................. 168
4.3. Suggestions for future research ...................................................................................................................................................... 169
4.3.1. Evolution in cycling..................................................................................................................................................................................... 169

4.3.2. Bio-banding ........................................................................................................................................................................................................170
4.3.3. Early specialisation or diversification/sampling ................................................................................................................170
4.3.4. Globalisation of the findings................................................................................................................................................................ 171
5. Conclusion................................................................................................................................................................................................................ 172
6. References ............................................................................................................................................................................................................... 173
ADDENDA ............................................................................................................................................................................................................................................... 181

Addendum A: pilot study 1 ................................................................................................................................................................................................................183
Addendum B: pilot study 2 .............................................................................................................................................................................................................. 189
Addendum C: Benchmarks ................................................................................................................................................................................................................ 192
CURRICULUM VITAE ........................................................................................................................................................................................................................ 199

ACKNOWLEDGEMENTS – DANKWOORD.............................................................................................................................................................................. 205
List of abbreviations
30” ABS = 30 seconds all-out cycling test absolute score

APHV = Age at peak height velocity
AUC = Area under the curve
BB = Balance beam
BMI = Body mass index
BMX = Bicycle motocross
C = Continental
Cgs sports = Sports measured in centimetres, grams, and seconds
CI = Confidence interval
CMJ = Counter movement jump
ESHR = Endurance shuttle run
HR = Heart rate
JS = Jumping sideways
Lamax = Maximum blood lactate concentration
LTAD = Long-Term Athlete Development model
MANOVA = Multivariate analysis of variance
MANCOVA = Multivariate analysis of covariance
MS = Moving sideways
MTB = Mountain bike
npar= Nonparametric test
NON-PRO = Future non-professional cyclist
OR = Odds ratio
PC = Pro Continental
PRO = Future professional cyclist
Q1 = Quartile 1, January - March
Q2 = Quartile 2, April - June
Q3 = Quartile 3, July - September
Q4 = Quartile 4, October - December
RAE = Relative age effect
30” REL = 30 seconds all-out cycling test relative score
ROC = Receiver Operating Characteristic
I
SAR = Sit and reach
SBJ = Standing broad jump
SD = Standard deviation
SRM = Schoberer Rad Messtechnik
TD = Talent development
TID = Talent identification
TO = Talent orientation
U12 = Under 12, 8 – 11 years old
UCI = Union Cycliste Internationale
VO2max = Maximal oxygen consumption/uptake
Wmax = Maximal average power output
WT = World tour
X-C = Cross-country
II
List of figures and tables
FIGURES
PART 1
Figure 1.1. Schematic overview of the outline of the general introduction divided in five subchapters. 2
Figure 1.2. Gagné's Differentiated Model of Giftedness and Talent (DMGT 2.0; 2008 update) (Gagné, 2010). The model 3
illustrates that the development of athletes with gifts (top 10%, natural abilities) into talent (top 10%,
competencies) is influenced by different internal (intra-personal, chance) and external (environment, chance)
factors.
Figure 1.3. A visual representation of the Flemish Sports Compass (Victoris, 2020). 5
Figure 1.4. Execution of the standing broad jump measuring explosive leg power. 11
Figure 1.5. A visual representation of the interrelationship of the different cycling disciplines (LTAD-VOLUME, 2008). 14
Figure 1.6. The Athlete Development Triangle (ADT) developed by Gulbin et al. (2013a) represents the prevalence, 25
magnitude and direction of transitions between competition levels within sports.
Figure 1.7. Schematic overview of the original research. 33
PART 2
Figure 2.1. Shuttle bike circuit. 55
Figure 2.2. Discrimination of high performance athletes (≥16 years) from four different cycling disciplines based on 58
performance results on a non-sport-specific test battery (80.7% correctly classified; 42.1% cross-validated).
Figure 2.3. Graphical representation of the discriminant analysis of cyclists between 12.0 and 15.99 years of age based on 62
performance results on a non-sport-specific test battery.
Figure 2.4. Flowchart of the data collection. 81
Figure 2.5. Visual representation of the prediction equation in the U17 category (full model). 85
Figure 2.6. Visual representation of the prediction equation in the U19 category (full model). 88
Figure 2.7. Career trajectories of the achievers and non-achievers are displayed using success rate (mean and the positive 105
standard deviation) achieved in each year of the competition (U15= 3 competition years; U17= 2 competition
years; U19= 2 competition years). Significant differences in success rate between both groups are marked by
an arrow (p<0.05), while significant differences in success rate within one group between consecutive years
are indicated by an asterisk (p<0.05).
Figure 2.8. Overview of the birth date distribution (RAE) in the future achievers and future non-achievers group. The 106
success rate for athletes born in the different birth quarters are displayed for each competition year.
Significant differences in success rate between the different birth quarters are marked by an arrow (p<0.05).
Figure 2.9. Success score (mean + standard deviation) of future professional and non-professional in each competition 124
year. Significant main effects of time (***, p < 0.001) and group (###, p < 0.001) are reported.
PART 3
Figure 3.1. Learning how to bike (Ayelet-Keshet). 140
Figure 3.2. The spider plot displays the z-scores of the different (non-)sport-specific tests for the (non-)sporting reference 142
population, cyclists and BMX cyclists. A total sample of 348 children between 7.0 – 11.9 years of age (110
reference population no sport; 105 reference population sport; 86 cyclists, 47 BMX cyclists) were included in
spider plot of pilot
study 1.
Figure 3.3. The spider plot displays the z-scores of the different (non-)sport-specific tests for athletes between 145
12.0 – 15.9 years of age from five different cycling disciplines (BMX, road, cyclo-cross, track, and MTB cycling).
Figure 3.4. Graphical representation of the Cycling Compass. Four different levels could be distinguished: talent detection, 154
orientation, identification, and development.
Figure 3.5. Gagné's Differentiated Model of Giftedness and Talent (DMGT 2.0; 2008 update) (Gagné, 2010). 158
Figure 3.6. A visual representation of the interrelationship of the different cycling disciplines (LTAD-VOLUME, 2008). 169
Addenda
Figure A.1. The spider plot displays the z-scores of the different (non-)sport-specific tests for the (non-)sporting reference 184
population, cyclists and BMX. A total sample of 557 adolescents between 12.0 – 15.9 years of age (95 reference
population no sport; 101 reference population sport; 294 cyclists, 67 BMX cyclists) were included in pilot
study 1.
III
Figure A.2. Graphical representation of the discriminant analysis of cyclists, BMX and a (non-)sporting reference 185
population between 7.0 and 11.99 years of age based on performance results on a non-sport-specific test
battery.
Figure A.3. Discrimination of cyclists, BMX cyclists and the (non-)sporting reference population between 12.0 and 15.9 186
years of age based on performance results on a non-sport-specific test battery (71.6% correctly classified;
80.4% cross-validated).
Figure B.1. Graphical representation of the discriminant analysis of the different cycling disciplines (BMX, road, cyclo- 190
cross, track and MTB) between 12.0 and 15.99 years of age based on performance results on a non-sport-
specific test battery.
TABLES
PART 1
Table 1.1. The competition protocols of the various types of elite adult cyclists are displayed in the table. 15
Table 1.2. Overview of characteristics described in the literature for the different cycling disciplines per discipline. 22
PART 2
Table 2.1. Average performance on the anthropometric, physical, motor coordination, and cycling-specific tests (± SD) 57
of the cycling athletes ≥ 16 years of age from the four different disciplines and results of the MANOVA
(F-value, p-value, and effect size) and MANCOVA (F-value, p-value, and effect size).
Table 2.2. Results from the separate discriminant analyses are demonstrated, which discriminates characteristics 59
between one discipline compared to all the cycling disciplines together. A distinction is made between both
groups.
of the cycling athletes between 12.0 and 15.99 years of age from the four different disciplines and results of
the MANOVA (F-value, p-value, and effect size) and MANCOVA (F-value, p-value, and effect size).
Table 2.4. Detailed information about the bike formats used for the shuttle bike test and maximal cadence test. 70
Table 2.5. Detailed overview of the results from the multivariate analyses of covariance (MANCOVA) in the ≥ 16 year of 71
age group with maturity ratio, calendar age and training history as confounding variables.
Table 2.6. Detailed overview of the results from the multivariate analyses of covariance (MANCOVA) in the group 72
between 12.0 and 15.99 years of age with maturity ratio, calendar age and training history as confounding
variables.
Table 2.7. Scores awarded to performance results at provincial and national championships. 84
of the road cyclists.
Table 2.9. Results of the standardised hierarchical regression analyses for the U15 group. 87
Table 2.10. Results of the standardised hierarchical regression analyses for the U17 group. 89
Table 2.11. Average success rate (± SD) of the future achievers and non-achievers for each competition category and 104
results of the univariate analysis (Chi squared, degrees of freedom and p-values).
Table 2.12. Average success rate (± SD) of the future achievers and non-achievers for each competition category and birth 107
quarter (quarter1, quarter2, quarter3, quarter4). Additionally, the results of the univariate analysis (Chi
squared, degrees of freedom and p-values) are displayed.
Table 2.13. Average success rate (± SD) of the WT, PC, C, and future non-achievers for each competition category. The 109
results of the Kruskal-Wallis analyses (Chi squared, degrees of freedom, and p-values) are displayed. An
overview is provided of the birth date distribution (RAE) in the WT, PC, C, and future non-achievers group
together with the results of the crosstab analysis (Chi squared, degrees of freedom, and p-values). The bottom
part of the table represents the results from the logistic regression (Odds ratio, 95% confidence interval, ROC).
Table 2.14. Average success score (mean ± standard deviation) by future professional and non-professional cyclists for 125
each competition year.
Table 2.15. Average success score (mean ± standard deviation) of future professional and non-professional cyclists for 126
each competition year per birth quarter. Same superscript letter (a, b) indicates significant differences
between two groups (p < 0.05).
Abbreviations: PRO, future professional; NON-PRO, future non-professional; Q1, quartile 1; Q2, quartile 2; Q3,
quartile 3; Q4, quartile 4.
Table 2.16. Results of the logistic regression and probability of reaching the professional level. 127
Abbreviations: CI, confidence interval; AUC, area under the curve.
IV
PART 3
Table 3.1. An overview of the tests that could be included in the test batteries per age group (X = included; 159
(X) = optional).
Table 3.2. A critical review on the anthropometric measurements used in this dissertation. Suggestions for future 160
research.
Table 3.3. A critical review on the physical tests used in this dissertation. Suggestions for future research. 162
Table 3.4. A critical review on the motor coordination tests used in this dissertation. Suggestions for future research. 163
Table 3.5. A critical review on the cycling-specific measurements used in this dissertation. Suggestions for future 164
research.
Table 3.6. A critical review on cycling performance per age group. 165
Addenda
Table A.1. Average performance on the anthropometric, physical, and motor coordination tests (± SD) of the 187
(non-)sporting reference population, cyclists, and BMX cyclists between the age of 7.0 and 11.99 years and
results of the MANOVA (F-value, p-value, and effect size) and MANCOVA (F-value, p-value, and effect size).
Table A.2. Average performance on the anthropometric, physical, and motor coordination tests (± SD) of the 188
(non-)sporting reference population, cyclists, and BMX cyclists between the age of 12.0 and 15.99 years and
Table B.1. Average performance on the anthropometric, physical, motor coordination, and cycling-specific tests (± SD) 191
of the cycling athletes between 12.0 - 15.99 years of age from the four different disciplines and results of the
MANOVA (F-value, p-value, and effect size) and MANCOVA (F-value, p-value, and effect size).
Table C.1. Benchmarks (number of participants (N), mean, standard deviation (SD), percentile 25, percentile 50, and 192
percentile 75) of male track cyclists for the different age groups and tests of the (non-)sport-specific test
battery.
percentile 75) of male road cyclists for the different age groups and tests of the (non-)sport-specific test
battery.
percentile 75) of male cyclo-cross cyclists for the different age groups and tests of the (non-)sport-specific
test battery.
percentile 75) of male MTB cyclists for the different age groups and tests of the (non-)sport-specific test
battery.
percentile 75) of male BMX cyclists for the different age groups and tests of the (non-)sport-specific test
battery.
V
Summary
For endurance athletes, such as cyclists, having excellent physiological capabilities appears to be a prerequisite for
reaching the highest performance level. Nevertheless, even if youth athletes excel in competitions and/or
physiological tests, only a limited number of them will become professional cyclists at adult age. In other words,
the development process of athletes is complex and non-linear. Therefore, it is suggested that other performance
determining factors might contribute to future success in the athletic sports career. Currently, little is known about
the athletic profiles of youth cyclists, to what extent their future performance can be predicted, and how youth
cyclists develop during their athletic careers. Therefore, within this dissertation, it seemed challenging to
investigate the added value of measuring performance with a (non-)sport-specific test battery and cycling
performance based on competition results in youth cyclists, to eventually combine the information in one cycling-
specific practical tool, the Cycling Compass.
Study 1 examined the athletic profiles of the different cycling disciplines (road, track, cyclo-cross, MTB). The results
showed that the athletic profiles for the different cycling disciplines could only be discriminated from the young
adult level (≥ 16 years) on. The athletic profiles of young adult track, MTB and cyclo-cross athletes could be
distinguished from each other based on anthropometry, physical, motor coordination, and cycling-specific tests.
Only road cycling could not be clearly distinguished from the other disciplines. Despite the less pronounced
character of the different disciplines in the adolescent cycling population, the discriminatory characteristics were
in line with the results of the young adults.
The extent to which test performances measured with a (non-)sport-specific test battery can predict future cycling
performance was examined in study 2. This test battery consists of a combination of anthropometric, physical
performance, and coordination tests (the non-sport-specific part), supplemented with cycling specific tests (the
sport-specific part). The results showed that test performances in the U15 category could partially explain future
success two to three years later. Maturity and general motor coordination were the only variables that contributed
significantly to future success. From the U17 category onwards, neither maturity, relative age, motor coordination,
physical performance, nor cycling-specific performance could predict future cycling performance in the U19
category.
The aim of studies 3 and 4 was to gain insight into the extent to which current cycling performance and relative age
can predict future cycling performance. The results of study 3 showed that estimating potential based on current
cycling performance has no predictive power with respect to future cycling performance in the U15 category.
However, the importance of competition results increased with increasing age, while the influence of the relative
VII
age effect (RAE) decreased. The most accurate predictions for performance at the adult elite level were found in
the U19 category and for World Tour riders. Similar results were found in study 4 investigating an Italian population
of road cyclists. The findings showed that from the U17 category onwards, the competition results indicated the
chance of becoming a professional cyclist, with U23year1 giving the best predictive value. Both studies (studies 3 and
4) demonstrated decreased performance when transitioning to a new competition category and increased
performance after one year of experience within the same category.
Overall, our research showed that talent orientation and predicting future performance is challenging in youth
cycling. Nevertheless, general motor coordination seems to be an essential factor in the U15 category, while cycling
performance starts to be indicative for future success only from the U17 category onwards.
The findings were summarised into a practical tool for talent development in cycling, ‘the Cycling Compass’, which
can be considered a new applied and cycling-specific version of the previously developed Flemish Sports Compass
(Pion, 2015). The Cycling Compass provides recommendations for coaches, federations, athletes, and other
stakeholders for each component of the Cycling Compass and providing age-, and discipline-specific performance
benchmarks.
VIII
Samenvatting
Bij uithoudingsatleten zoals wielrenners, blijkt het hebben van excellente fysiologische parameters een
voorwaarde te zijn om het hoogste prestatieniveau te bereiken. Ook al blinken atleten uit tijdens wedstrijden en/of
op fysiologische proeven in de jeugdjaren, toch wordt slechts een klein aandeel van deze jongeren een
professionele atleet op volwassen leeftijd. Het ontwikkelingsproces is met andere woorden complex en niet-
rechtlijnig. Vermoedelijk spelen ook voor het vastleggen van toekomstig succes andere prestatiebepalende
factoren mee in de loopbaan van sporters. Er is echter nog maar weinig gekend over de atletische profielen van
jeugdrenners, in welke mate hun prestaties kunnen voorspeld worden en hoe jeugdrenners zich ontwikkelen tijdens
hun sportieve carrière. Daarom leek het binnen deze dissertatie een uitdaging om de meerwaarde te onderzoeken
van het meten van prestaties aan de hand van een (niet-)sport specifieke testbatterij en fietsprestaties op basis
van wedstrijdresultaten bij jeugdrenners, om uiteindelijk deze informatie te combineren in één fietsspecifieke
praktische tool, het ‘Cycling Compass’.
Studie 1 onderzocht de atletische profielen van de verschillende wielerdisciplines (weg, piste, cyclo-cross, MTB). De
resultaten toonden aan dat de atletische profielen voor de verschillende wielerdisciplines alleen konden worden
onderscheiden vanaf het jongvolwassen niveau (≥ 16 jaar). De atletische profielen van jongvolwassen
baanwielrenners, MTB en cyclo-cross atleten konden van elkaar onderscheiden worden op basis van
antropometrische, fysieke, motorische coördinatie en fiets-specifieke testen. Enkel wegwielrennen kon niet
duidelijk onderscheiden worden van de andere disciplines. Ondanks het minder uitgesproken karakter van de
verschillende disciplines in de adolescente wielerpopulatie, stemmen de discriminerende kenmerken overeen met
de resultaten van de jongvolwassenen.
De mate waarin prestaties op een (niet-)sport specifieke testbatterij toekomstige prestatie kan voorspellen, werd
bestudeerd in studie 2. Deze testbatterij bestaat uit een combinatie van antropometrische, fysieke en motorische
coördinatie testen (niet-sport specifiek onderdeel), aangevuld met wieler specifieke testen (sport specifiek
onderdeel). De resultaten toonden aan dat de prestaties op een (niet-)sport specifieke testbatterij in de U15
categorie gedeeltelijk toekomstig succes kunnen verklaren twee tot drie jaar later. Maturiteit en algemene
motorische coördinatie waren de enige variabelen die significant bijdroegen tot toekomstig succes. Vanaf de U17
categorie konden noch maturiteit, relatieve leeftijd, motorische coördinatie, fysieke prestatie, noch de wieler
specifieke prestaties het wedstrijdsucces in de U19jaar2 categorie voorspellen.
De derde en vierde studie hadden tot doel om inzicht te krijgen in de mate waarin prestatie tijdens competities en
de relatieve leeftijd toekomstig succes kunnen voorspellen. De resultaten van studie 3 toonden aan dat het
IX
inschatten van potentieel op basis van wedstrijdresultaten onvoorspelbaar is in de U15 categorie. Het belang van
wedstrijdresultaten nam echter toe met toenemende leeftijd, terwijl de invloed van het relatief leeftijdseffect
(RAE) daalde. De meest accurate voorspellingen voor prestaties op senioren elite niveau werden gevonden in de
U19 categorie en voor World Tour renners. Gelijkaardige bevindingen werden bekomen in studie 4 waarin een
populatie van Italiaanse wegwielrenners werd onderzocht. De resultaten toonden aan dat de wedstrijdprestaties
vanaf de U17 categorie een indicatie gaven over de kans om profwielrenner te worden, waarbij dit in de U23jaar1 de
beste voorspellende waarde gaf. In beide studies (studies 3 en 4) werd een daling in prestatie aangetoond bij de
overstap naar een nieuwe competitie categorie en een toename in prestatie na één jaar ervaring binnen dezelfde
categorie.
Algemeen gesteld toonde ons onderzoek aan dat talent oriëntatie en het voorspellen van toekomstige prestaties
binnen het jeugdwielrennen uitdagend is. Algemene motorische coördinatie blijkt een essentiële factor te zijn in
de U15 categorie, terwijl wedstrijdprestaties pas vanaf de U17 categorie een indicatie geven over toekomstig succes.
De bevindingen werden samengevat in een praktische tool voor talentontwikkeling in de wielersport, 'het Cycling
Compass', dat kan worden beschouwd als een nieuwe toegepaste en wielerspecifieke versie van het eerder
ontwikkelde Vlaamse Sportkompas (Pion, 2015). Het Cycling Compass voorziet in het aanleveren van aanbevelingen
voor coaches, federaties, atleten en andere belanghebbenden op elk niveau van het Cycling Compass en leeftijds-
en discipline-specifieke benchmarks.
X
PART 1 : GENERAL INTRODUCTION
1. Project context
My passion and motivation for sports have been present since childhood. My experiences as a competitive athlete
as well as a coach have led to the Master's degree of Movement and Sports Sciences - option Sports Training and
Coaching. The development and guidance of athletes and identifying athletes with potential are my key areas of
interest. During my Master thesis and especially during my PhD trajectory, I was able to gain numerous experiences
and insights into different sports and performance levels, such as recreation and competitive, artistic, ball, racket,
endurance, and combat sports. These understandings have helped me to maintain a broad view on talent research.
My interest in exact sciences and statistics were the foundation for approaching results precisely and challenging
myself with new statistical methods. Combining this PhD with a position as sport technical coordinator within the
Flemish Skating Union (federation) allowed me to better reflect on the practical relevance of the research findings.
The a priori choice to investigate cycling was at the request of Cycling Vlaanderen, who are eager to optimise their
federations' talent program. My motivation to fully dedicate my research to cycling was triggered in different ways.
Cycling and its different cycling disciplines play an essential role in the Belgian sports landscape. Belgium has
become one of the core countries within Europe for road cycling and cyclo-cross (Mignot, 2016). In particular,
Flanders has a cycling tradition and has delivered several top riders in the various cycling disciplines (except for
BMX and MTB) in the last decades. In contrast to other sports, an endurance sport such as cycling, in which the body
parts are physically constrained during exercise, has not been examined in talent research. Soon it became clear
that talent research in youth cycling was limited and that the talent identification program of Cycling Vlaanderen
was in its infancy. Consequently, I could contribute to a large-scale project with practical relevance by investigating
young cyclists who were already active members of a cycling club in the Flemish federation. My intention to develop
a relevant sport-specific application was two-fold: integrating the Flemish Sports Compass method in a population
of experienced young cyclists and evaluating the extent to which current cycling performance could predict future
cycling performance. Even though I have not been directly involved in cycling as an active participant, I saw this as
an advantage to interpret the research results more objectively. Close collaboration with the federation, coaches,
trainers, training/top sports/professional sports coordinators, and parents have contributed to bridge the gap
between science and practice.
The general introduction of the current dissertation is organised into five subchapters (Figure 1.1.). The first
subchapter emphasises the challenging topic of talent identification (subchapter 1: challenging topic of talent
identification). How one perceives talent, abilities, and the talent development pathways of the athletes remains a
key area of interest. In the literature, various methods are applied to investigate the athletic profiles. The use of a
1
(non-)sport-specific test battery appears to be suitable for this purpose; however, it has not previously been studied
and applied within the realm of endurance sports (subchapter 2: from Sports Compass to Cycling Compass). In the
third subchapter of the present dissertation, potential components for a Cycling Compass are listed and described
based on various tests that have already been used in talent development and identification research in various
sports (subchapter 3: potential components of the Cycling Compass). Focusing on cycling and its subdisciplines can
be considered as a case in the field of sports. To maintain the leading position in cycling in the long term and given
the increasing competitive pressure, it is also essential to further investigate talent identification based on cycling
performance in the youth categories. Based on previous findings in the literature, the hiatuses for developing a
Cycling Compass were discussed (subchapter 4: hiatuses for the development of a Cycling Compass). Finally, in the
fifth subchapter, the aims of this dissertation were explained in a general overview (subchapter 5: general overview
and aims of this research).
Potential
From Sports Hiatus for the General overview
Challenging topic of talent components of
Compass to development of a and aims of this
identification the Cycling
Cycling Compass Cycling Compass research
Compass
Figure 1.1.: Schematic overview of the outline of the general introduction divided in five subchapters.
2
2. The challenging topic of talent identification
The word ‘talent’ is well known within the sporting domain. Still, it is a controversial term in the realm of sports
sciences and some researchers prefer the wording ‘athletes with potential’ instead of talented athletes. Apart from
the terminology used, developing a talent identification program requires time and insights. Several researchers
have investigated the developmental pathway of athletes (Régnier et al., 1993; Balyi & Hamilton, 2004; Gagné,
2004; Côté et al., 2007). A few examples of well-known talent models are: (a) the Differentiated Model of Giftedness
and Talent (Gagné, 2004) (Figure 1.2), Development Model of Sports Participation (Côté et al., 2007), Long Term
Athlete Development model (Balyi & Hamilton, 2004), Youth Physical Development Model (Lloyd & Oliver, 2012),
and the FTEM framework (Foundations, Talent, Elite, Mastery) (Gulbin et al., 2013b).
Figure 1.2: Gagné's Differentiated Model of Giftedness and Talent (DMGT 2.0; 2008 update) (Gagné, 2010). The model
illustrates that the development of athletes with gifts (top 10%, natural abilities) into talent (top 10%, competencies) is
influenced by different internal (intra-personal, chance) and external (environment, chance) factors.
These different talent models have in common that they emphasise the complexity of the talent development
pathway of athletes. Most scientists agree that both biological or genetically constrained factors (i.e. nature) and
the end product of experience and learning (i.e. nurture) are important for talent; however, the nature versus
nurture debate continues to exist (Johnston et al., 2018). The development of athletes is considered a non-linear
3
process and is influenced by different internal (intra-personal, chance) and external (environment, chance) factors
(Gagné, 2004; Gulbin et al., 2013a) (Figure 1.2). This implies that individual differences between athletes need to be
considered in the development, and outperforming peers at a young age does not guarantee a professional sports
career as an adult.
From a scientific perspective, the talent developmental process can be divided into five different stages that
alternate or follow-up, such as (a) discovering potential performers who are currently not involved in the particular
sport (talent detection), (b) supporting athletes (with potential) in the search for the discipline(s) that best fits
their athletic profile (talent orientation and transfer), (c) identifying athletes with potential, who are already active
in the sport (talent identification), (d) providing athletes with potential extra support, training, and competition
possibilities allowing them to develop their full potential optimally throughout their athletic career (talent
development), and (e) selecting athletes with the best athletic profile for the essential competitions and team
composition (talent selection) (Vaeyens et al., 2008; Pion, 2015). Furthermore, there are considerable variations in
the interpretation of how one perceives talent or categorises an athlete as a talent or a high potential athlete.
Usually, the label of ‘talent’ will only be given after performance excellence has occurred (Nimmerichter, 2018).
4
3. From Sports Compass to Cycling Compass
3.1. The Sports Compass
A crucial step in developing a talent development program is the translation from talent to measurable variables.
This requires measuring the underlying multidimensional factors and characteristics that may contribute to
(excellent) performance (Régnier et al., 1993; Vandorpe, 2011; Vandorpe et al., 2012b). The ‘Flemish Sports Compass’
was developed between 2007 – 2010 using a valid and reliable (non-)sport-specific test battery with the aim to
(a) detect the ‘better movers’, (b) identify athletes with potential, and (c) orient and/or transfer athletes to the
sport that best fits with the athlete’s abilities (Lenoir & Philippaerts, 2011; Pion, 2015). It consists of anthropometric,
physical, motor coordination, and sport-specific tests. The Flemish Sports Compass has already been implemented
in several talent development programmes in different sports and countries, and it is considered a valuable tool
for sports federations, sport clubs, and other stakeholders (Pion, 2015). The Sports Compass is currently being used
in the Flemish primary schools (3rd and 4th grade). The Sports Compass uses the I Do (non-sport-specific test
battery), I Like (a questionnaire about preferences), and I Am tools (a questionnaire about motivation) to detect
and orient children towards the sports that best suit their motivation, preferences and physical/motor abilities
(Figure 1.3.).
Figure 1.3.: A visual representation of the Flemish Sports Compass (Victoris, 2020).
5
The advantages of using a (non-)sport-specific test battery for the generic screening of athletes have already been
demonstrated in sports such as artistic, ball, combat, and racquet sports (Deprez, 2015; Matthys et al., 2013; Mostaert
et al., 2020; Mostaert et al., 2016; Norjali Wazir et al., 2018; Robertson et al., 2021; Vandorpe et al., 2012b). The
(non-)sport-specific test battery is cheap and relatively easy to administer in small and large groups of youth
athletes. It is able to detect the athletes who have experienced a broad development (better movers), who might
also turn out to be effective champions. The test battery is used to distinguish and select a small group of youth
athletes with potential from a large pool of athletes. It can be applied to compare the athletic profiles and
characteristics of different sports, and disciplines. The performances can be monitored longitudinally and could be
related to future success in sports (Leyhr et al., 2018). Due to the non-sport-specific nature of the tests, the test
battery has the advantage of reducing the influence of bias resulting from differences in maturity, relative age,
training history, and/or sport-specific skills (Pion, 2015).
It should be noted that a (non-)sport-specific test battery only highlights one aspect of the athletic profile. Due to
the complexity of the athletes’ development, additional performance determinants should also be taken into
account to optimally evaluate the athletes’ potential. The test battery is mainly applicable in youth, but less
accurate to apply in the adult categories.
3.2. Current approaches in cycling
High potential athletes (talent) in cycling are often associated with athletes having excellent values on
physiological characteristics related to aerobic and anaerobic performance levels. It has been found that athletes
with a limited ability for endurance will never reach the top level (Bourgois, 2011). Studies on talent identification
in professional cycling have predominantly focused on adult athletes' morphological and physiological
performance characteristics (e.g., (an)aerobic performance variables) measured in field and/or laboratory settings
(McLean & Parker, 1989; Menaspà et al., 2012; Novak & Dascombe, 2014; Taeymans et al., 2016). For example, Sanders
& Van Erp (2020) indicated that power profiling, the assessment of field-derived power outputs, has been proven
useful to determine the adult athletes’ strengths and weaknesses, allow the comparison of adult cyclists, and
monitor their development. Also, non-invasive Muscle Talent Scans (proton magnetic resonance spectroscopy
method to quantify carnosine in the muscle) which measure the muscle typology of the athletes, have been used
for talent orientation and identification purposes in world-class adult cyclists (Lievens et al., 2021). Lievens and
colleagues (2021) suggest that the information on muscle typology can be a valuable addition to the
anthropometric, physical, and technical data.
6
Physiological tests have already proven their usefulness in adults; unfortunately, scientific evidence is still scarce
in youth cycling. In addition, these measures are expensive, require sophisticated equipment, are time-consuming,
and are not suitable for large groups of (youth) athletes (Pearson et al., 2006). It should be noted that the
physiological test results related to (an)aerobic performance levels might bias the potential of young athletes
during puberty (Moesch et al., 2011), since the development of youth into adult athletes is accompanied by many
changes in aerobic and anaerobic muscle metabolism markers partly due to the growth and maturation
experienced by the athlete. To date, it is not clear to what extent the higher physiological values measured in
professional cyclists can be related to the genetic predisposition or training history (Svendsen et al., 2018; Costa et
al., 2012). Research has indicated that having a high aerobic performance level (e.g. VO2 value) in the junior category
was practical to differentiate the competitive level within the junior category. However, these tests could not
predict whether a junior cyclist will progress to professional status in adulthood (Menaspà et al. 2010; Allen &
Hopkins, 2015). Therefore, it is assumed that other physiological, psychological, and environmental characteristics
may also contribute to future performance.
An alternative and common approach to talent identification within cycling is the use of competition results
(Svendsen et al., 2018). Coaches and scouts often apply competition results to evaluate cycling performance in the
short and/or long term (Van Bulck et al., 2021). In other sports, it has been suggested that results in junior
competitions are not always indicative for success in adult competition (Brouwers et al., 2012; Li et al., 2018a). It is
therefore questionable if this is the case for cycling.
3.3. The Cycling Compass
The rationale for developing a cycling-specific Sports Compass, the Cycling Compass, including test performances
measured with a (non-)sport-specific test battery and cycling performance based on competition results, is
multifactorial:
- The noticeable imbalance in research between adult and youth cycling must be tackled. Especially given the
increasing participation in youth sports and the pressure on coaches and their coaching teams
(Nimmerichter, 2018). They have a duty of care to ensure the health and well-being of their athletes, and
there is an increasing need for financial investment in long-term development programs.
- The potential benefit of applying a (non-)sport-specific test battery within an endurance sport such as
cycling has not been explored in the literature. Given the sport-specific characteristics of cycling, which
differ significantly from the sports already investigated in talent research (using the Sports Compass), it
7
is suggested that the previous findings could not be simply transferred to cycling. The added value of this
test battery is not yet investigated within cycling alongside the other performance determining factors.
- The added value of interpreting performance in youth competitions has not been sufficiently investigated
in cycling. Nevertheless, coaches and scouts are frequently applying competition results to evaluate youth
cycling performance in the short and/or long term (Van Bulck et al., 2021).
- Federations and other stakeholders are searching for more internal and external ‘scientific’ know-how to
optimise the talent programs. They aim to maximise sports participation at a recreational and competitive
level, which could result in larger talent pools. They also have the intention to reduce false positive and
false negative cases in the talent selection phase, and increase the probability of high performance.
Moreover, they are searching for solutions to reduce the total cost in order to invest extra financial
resources within the limited group of athletes with potential (Baker et al., 2018). Besides specialised
coaches who play a prominent role in identifying young athletes with potential, the use of a (non-)sport-
specific test battery and predictive models could be additional steps towards a cost and risk-reducing
talent identification process (Pion et al., 2015b).
The Cycling Compass, would be the first applied and cycling-specific version of the previously developed Flemish
Sports Compass (Pion, 2015). The main focus will be on cycling and its different disciplines, which can be considered
as a case in the field of sports. It will be developed for young cyclists, that are already active members of a cycling
club recognised by Cycling Vlaanderen.
8
4. Potential components of the Cycling Compass
4.1. Lessons from other sports
As previously mentioned, the Flemish Sports Compass uses a valid and reliable (non-)sport-specific test battery
consisting of anthropometric, physical, motor coordination tests, supplemented with sport-specific tests. A different
composition of tests can be chosen for each age group and sport to adapt to the diverse needs and performances
of the sport (Vandorpe, 2011). This corresponds to the sliding population approach of Régnier et al. (1993), which
showed that the anthropometric and physical characteristics become more important as the athletes grow older
and reach a higher performance level (Vaeyens et al., 2008; Pion, 2015). The test battery can be applied cross-
sectionally and/or longitudinally (Vaeyens et al., 2008). The latter involves several test moments across a more
extended period, while the measurements should always be carried out in the same way (Abbott et al., 2005;
Elferink-Gemser et al., 2011). Longitudinal follow-up provides better insights into the characteristics that contribute
to predicting future success (Abbott & Collins, 2002). It can diminish the risk of misinterpretation of the results and
may provide a more accurate interpretation of the potential and progress of the athletes (Johnston et al., 2018).
However, longitudinal studies are time- and energy-consuming. They also require a large sample size to
compensate for drop-out because young high potentials do not necessarily reach the highest performance level at
adult age (Johnson et al., 2008). As a result, only very few studies combine an extensive set of measures with a
longitudinal design. Below we give a brief overview of the various tests that have already been used to investigate
talent development and identification in various sports.
4.1.1. Growth, maturation and age
The timing, tempo, and status of growth, and therefore the degree of maturity, affects young athletes'
anthropometric, physical, and physiological characteristics during puberty (Malina et al., 2004; Pearson et al., 2006;
Teunissen et al., 2020). For the interpretation of the performances, it is not sufficient to consider calendar age since
this does not always correspond with the biological age of the athletes (Malina et al., 2004; Teunissen et al., 2020).
One athlete might outperform the others of the same age due to being earlier mature. Therefore, maturity is
frequently considered in talent research, using a non-invasive method estimating maturity based on
anthropometric parameters (Khamis & Roche, 1994; Mirwald et al., 2002; Moore et al., 2015; Fransen et al., 2018;
Teunissen et al., 2020). Besides, (subtle) differences in chronological age may occur within the same age group,
better known as the Relative Age Effect (RAE) (Vaeyens et al., 2005; Rees et al., 2016). Athletes born in the first
months of a selection year are on average more advanced in growth and development compared to their
9
counterparts born in the last months of that same selection year. As a result, these athletes are more likely to have
physical performance advantages (Vaeyens et al., 2005). This phenomenon occurs especially within sports where
speed and power are significant performance determinants for future success (Musch & Grondin, 2001; Cobley et
al., 2009). From that perspective, it is no surprise to observe that athletes from the first birth quartile of a selection
year are overrepresented in a given cohort of selected athletes, a phenomenon called the Relative Age Effect (RAE).
However, the advantages associated with being born in the first two quarters of the year or having early maturity
may disappear by the time the athletes reach adult elite level (Rees et al., 2016).
4.1.2. Anthropometric tests
In the literature, special attention is given to the anthropometric measurements, especially in sports such as
gymnastics and volleyball, where morphology is considered to be a determinant of performance (Di Cagno et al.,
2009; Pion et al., 2015a; Mostaert et al., 2020). The body morphology of athletes is investigated using
measurements of body height, sitting height, body weight, fat percentage, and BMI. These measurements are easy
to administer and are commonly used in talent research in various sports, measuring small and large populations
in a relatively short amount of time (Pion, 2015; Rommers et al., 2019). In addition, measurement of the skinfolds,
or a dual-X-ray absorptiometry scan to estimate fat percentage, or the measurement of the frontal area could also
provide useful information about the athletic profile of cyclists. Nevertheless, these approaches are more time
consuming, expensive, and are preferably applied in smaller groups (Wells & Fewtrell, 2006). Caution is warranted
when identifying and selecting athletes solely on anthropometric measurements since the timing and tempo of
growth and maturation differ inter-individually (Mostaert et al., 2016; Teunissen et al., 2020).
4.1.3. Physical tests
With the aim of searching high potential athletes, researchers and coaches became increasingly interested in the
athlete’s physical development, which refers to the development of strength, flexibility, endurance, speed and
motor skills. Physiologically, these characteristics are related and sometimes difficult to separate (Lloyd & Oliver,
2012). Speed, for example, requires strength. All these physical characteristics are partially innate and can be
partially trained at any time of the development, however, the response differs depending on the stage of growth
(Lloyd & Oliver, 2012; Tucker & Collins, 2012). The Flemish Sports Compass uses generic measurements such as the
endurance shuttle run (endurance), sit and reach (flexibility), 30m sprint (speed), standing broad jump (explosive
leg power) (Figure 1.4.), knee push ups (strength), curl ups (strength), and counter movement jump (explosive leg
10
power) to provide valuable information about the athletes' general physical development and physical
characteristics (Pion et al., 2015c).
Figure 1.4.: Execution of the standing broad jump measuring explosive leg power.
Previous findings showed that the generic test battery could be used to discriminate between athletes from
different sports and disciplines (Pion et al., 2015c). Triathletes, for example, could be discriminated from other
sports such as gymnastics, soccer, volleyball, etc., based on their performances on the endurance shuttle run test.
On the other hand, volleyball athletes outperformed the others on explosive jump power exercises (standing broad
jump, counter movement jump) and speed (30m sprint test) (Pion et al., 2015c). Consequently, these non-sport-
specific tests (endurance, speed, power flexibility, strength) are suggested valuable tests within talent research,
able to measure the characteristics of the different sports. Non-sport-specific tests have the advantage that they
are easy to apply and administer in small and large groups and the results can be generalised across different
sports. Moreover, the tests are non-sport-specific, which might minimise the influence of training history. Training
history may blur a clear interpretation of the tests: a moderately gifted athlete who is already involved in the same
sport for a long time might obtain the same score as a highly talented individual but only recently involved in the
sport. It is very plausible that the margin for improvement (“potential”) is more prominent in the latter athlete.
The parameters maturity and RAE should also be incorporated to accurately interpret the performances results.
11
4.1.4. Motor coordination tests
General motor coordination tests became a permanent component of the talent test batteries. These characteristics
seem to be an essential precursor for later success in sport (Williams & Ford, 2009; Vandorpe, 2011; Mostaert et al.,
2016; Toum et al., 2021). Although the phenomenon of ‘adolescent motor awkwardness’, especially in fastly growing
children, is regularly referred to by many researchers (Beunen & Malina, 1988; Davies & Rose, 2000). Recent studies
have shown that motor coordination tends to remain stable during puberty (Ahnert et al., 2010; Rommers et al.,
2019; Toum et al., 2021). Possessing good to excellent motor coordination skills is advantageous because sport-
specific skills can be learned faster (Vandorpe et al., 2012a; Rommers et al., 2019). Many standardised tools for the
evaluation of motor coordination have been developed and are frequently used in research, mostly in the area of
motor development, like the Bruininks-Oseretsky Test of Motor Proficiency – 2nd Edition (Bruininks & Bruininks,
2005; Fransen et al., 2014), the Motoriktest für Vier- bis Sechsjährige Kinder (Zimmer & Volkamer, 1987; Bardid et
al., 2016), the Movement Assessment Battery for Children (Fransen et al., 2014). Many of these tests have been
developed for the purpose of the identification of children with movement difficulties, and are limited in
discriminating between children and adolescents with an average or advanced level of development (O’Brien-Smith
et al., 2019). Most of these tests are also very time-consuming, which is problematic for the purpose of
multidimensional testing of relatively large samples. Since 2010, the (short version of the) KTK test battery (KTK3,
Körperkoordinationstest Für Kinder) consisting of three tests, (a) jumping sideways, (b) moving sideways, and (c)
balance beam (Kiphard & Schilling, 1974; Kiphard & Schilling, 2017; Novak et al., 2017; O’Brien-Smith et al., 2019;
Mostaert et al., 2021), is frequently used in talent research. This test battery measures a broad spectrum of motor
coordination skills, from locomotion to balance, and relies on components of physical fitness and motoric
coordination (Mostaert et al., 2021). Since these tests are not specifically trained skills during training sessions,
training history should have only a small effect on the scores (Rommers et al., 2019). Moreover, this valid tool is
applicable to measure the general motor coordination of youth athletes of different ages (Mostaert et al., 2016;
Mostaert et al., 2021).
4.1.5. Sport-specific tests
The generic test battery can be extended with sport-specific tests. For example, in gymnastics the test battery was
extended with a rope skipping and basic locomotion skills test, in volleyball with a spike jump and shuttle throw
task test, and in figure skating with a split and spin jump test (Vandorpe et al., 2011b; Mostaert et al., 2016; Mostaert
et al., 2020). Talent identification and development programs in soccer, use soccer-specific motor coordination (i.e.,
technical) skills (e.g., dribbling, T-tests, juggling, shooting, etc.) to distinguish between competition levels or players
12
positions (Höner et al., 2015). Vandendriessche et al. (2012) found that soccer-specific and non-specific motor
coordination tests are not related to biological maturity. Therefore, these tests seem to provide more insight into
the future potential of a young athlete when compared with fitness tests. However, the outcome measures depend
on the type of technical skill assessed (Deprez, 2015). These sport-specific tests should be easy to administer, time-
efficient, and have a strong link within the sports discipline (soccer: dribbling test with a ball; badminton: shuttle
throw with a shuttle; cycling: bicycle tests) (Den Hartigh et al., 2018). Future research is needed to develop and
evaluate cycling-specific tests which assess motor coordination skills within cycling.
4.1.6. Additional constraints
The holistic approach is increasingly being encouraged from a scientific point of view (Sieghartsleitner et al., 2019).
This approach considers an even broader spectrum of characteristics to identify talent, including physiological,
psychological, and cognitive characteristics, coach’s opinion, and familial support (emotional, financial, or
organisational factors) (Sieghartsleitner et al., 2019). Gaining an understanding into the psychological
characteristics of athletes seems essential for sports psychologists, coaches, and sports organisations to provide
optimal guidance and support for the athletes (Raglin, 2001; Collins et al., 2016; Nimmerichter, 2018). Recent
research indicates that validated measurements of personality and other psychological variables are not reliable
enough to use for talent selection (Allen & Hopkins, 2015; Nimmerichter, 2018; Hill et al., 2019; Laureys et al., 2021).
This might be since these variables continue to develop during the transition from childhood to early adulthood
(Roberts & DelVecchio, 2000; Nimmerichter, 2018). Moreover, the athletes may experience pressure while
completing the psychological questionnaire. As a result, answers will be formulated according to the expectations
of others.
In addition, cognitive characteristics, such as executive functions, can be related to youth sports participation.
Executive functions are cognitive processes required to control numerous daily-life tasks and are crucial for
cognitive, social, and psychological development (Diamond, 2013). De Waelle and colleagues (2021) indicated that
8-to-12 year old girls who participate in team sports outperform those who engage in self-paced sports (e.g., ‘sports
that allow the athlete time to prepare themselves for critical actions and perform at their own pace‘ (De Waelle et
al., 2021, p.2)), and non-elites on executive functions. More importantly, self-paced sports (e.g., cycling, swimming,
athletics) did not significantly outperform the non-athletes. The latter findings were inconsistent with research in
adults, where superior inhibition performance was found for self-paced sports (Jacobson & Matthaeus, 2014).
Inhibition refers to the deliberate, controlled suppression of pre-programmed responses (Miyake et al., 2000).
Future research is needed to determine whether cognitive characteristics linked to, for example, inhibitory control
at a later age, during adolescence, are able to distinguish athletes of different levels within cycling (Wong et al.,
13
2015; Martin et al., 2016; Nimmerichter, 2018; De Waelle et al., 2021). Although these characteristics might also be
used for talent identification in cycling, the above-mentioned characteristics were beyond the scope of this
dissertation.
4.2. Focus on cycling
Cycling is an overarching term for different cycling disciplines: road, track, cyclo-cross, mountain bike (MTB), bicycle
motocross (BMX), trial, artistic cycling, cyclo-ball, BMX freestyle, gravel racing, and cycling for athletes with physical
disabilities or intellectual impairments. In the context of talent orientation and transfer, this dissertation examines
the athletic profiles of the road cycling, track cycling, cyclo-cross, and MTB disciplines (in study 1). For the sake of
completeness, the athletic profile of BMX riders (Olympic discipline) is also investigated and will be further
discussed in the general discussion section. The road, track, cyclo-cross, MTB, and BMX disciplines can be
distinguished based on differences in competition duration, terrain characteristics, equipment, and number of
participants (Lucía et al., 2001; "UCI Regulations," 2018; Spindler et al., 2018) (Table 1.1.). As a result, the athletes
participating in these disciplines are expected to be characterised by their unique set of physical, motor
coordination, physiological, biomechanical, strategic, and technical performance determinants (Figure 1.5., Table
1.2.).
Figure 1.5.: A visual representation of the interrelationship of the different cycling disciplines (LTAD-VOLUME, 2008).
14
Furthermore, different rider profiles are assumed even within the same discipline, as each discipline is
characterised by different race types (Craig & Norton, 2001; Peinado et al., 2011). Whether cyclists should engage in
early specialisation in one specific discipline, or participate in different disciplines, given the high compatibility, has
not been formally investigated. The following section provides a brief overview of the athletic profiles of road, track
(endurance), cyclo-cross, MTB (cross-country), and BMX cyclists.
Table 1.1.: The competition protocols of the various types of elite adult cyclists are displayed in the table.
Track cycling MTB BMX
Road cycling Cyclo-cross
(endurance) (cross-country) (race)
Olympic/
non-Olympic Olympic Olympic Non-Olympic Olympic Olympic
discipline
Lap distance min 2.5 km Lap distance
Distance 80 – 280 km 4 – 50 km 300 m – 400 m
and max 3.5 km 4 km – 6 km
Duration 4–6h 4 min – 1 h 40 – 60 min 1h30 – 1h45 30 – 45 s
In or outdoor Outdoor Indoor Velodrome Outdoor Outdoor Outdoor
Mountain
Road, country, forest landscape on
Surface Road, cobbles, and gravel Wood, asphalt, or tarmac paths and meadowland forest roads, single Curvy and hilly trail
surfaces tracks, fields, earth
or gravel paths
Competition Spring and
Spring and summer Summer and winter Autumn and winter Spring and summer
season summer
Flat terrain
Specialist riders Time trialist Endurance
Cyclo-cross Cross-country Supercross
group All terrain Sprint
Uphill
Team and/or Team Team Team
Individual Individual
individual sport Individual Individual Individual
Individual pursuit*
Team pursuit
Points race*
Competition Stage races (4-21 days)*
Madison Cyclo-cross Cross-country BMX race
modalities One day races (1 day)
Scratch*
Time trial (1 day)
Omnium
Elimination*
U8: 15 min
U12: 4 - 5 km U12 (outdoor): 10 - 17 min U12: 10 - 17 min
Duration/ U12: 15 - 35 min
U15: 20 - 35 km U15: 20 - 22 min U15: 20 - 25 min No limitations on
distance youth U15: 35 - 45 min
U17: 60 - 80 km U17, U19: no limitations on U17: 30 min duration
races U17: 60 min
U19: 80 - 140 km duration U19: 40 min
U19: 75 min
Impellizzeri and Mateo et al. (2011);
Padilla et al. (2001); UCI Craig and Norton (2001); Carmichael et al. (2017);
References Marcora (2007); Cowell et al. (2012);
(2018) UCI (2018) UCI (2018)
UCI (2018) UCI (2018)
Note: * = non-Olympic competition formats
15
4.2.1. Features and athletic profile of road cycling
Road cycling is the best-known and most widespread competitive cycling speciality of all cycling disciplines. The
discipline became an Olympic discipline in 1896. Within Europe, Belgium is one of the leading countries in cycling,
together with countries such as France, Italy and Spain (Mignot, 2016; Van Der Hoeven et al., 2021). These countries
are very competitive in this discipline and organise international events. Cycling is becoming increasingly well
known worldwide, partly due to increasing media interest and financial rewards given to the winners of the
competitions. Road cycling is a physically demanding endurance sport and it is performed on an outdoor terrain,
mainly on paved roadways. Professional male road cyclists are often known to reach their average peak
performance at 29.5 +/- 4.3 years of age (Longo et al., 2016). Professional elite road cyclists train on average 900
– 1000 hours/year and participate at 90 – 100 competition days/year (Lucía et al., 2001; Svendsen et al., 2018). Road
cyclists can compete individually in individual time trial races or as part of a team in one-day races, stage races,
grand tour races, etc. (distance varying between 80 – 280 km; duration varying between 4 – 6 h with exception of
the time trial races) (Table 1.1.). There has been an increasing trend for professional riders to specialise in certain
types of races according to their own athletic abilities (flat terrain, uphill terrain, all terrain, sprinter) (Menaspà et
al., 2012). Youth road cycling consists of annual- and age-related competition categories, e.g., U12 (8-11 years), U15
(12-14 years), U17 (15-16 years), U19 (17-18 years), U23 (19-22 years) (according to the Belgian regulations) (Archive,
2019). Young cyclists can already compete at a relatively young age (> 7 years), and there are many competition
opportunities during the competitive season. Competition distance and –duration increase from the under 17 (U17),
under 19 (U19), under 23 (U23), and elite adult road (>U23) competition categories. Competition distances vary
between 60 – 80 km (U17), and 80 – 140 km (U19)(UCI, 2018).
Regarding anthropometric characteristics, elite cyclists generally have a body type that varies between a
mesomorph (larger skinfolds, larger girths, and a low frontal area per body mass) or short and tall meso-ectomorph
profile (lean, small girths and a small frontal area) (van der Zwaard et al. 2019; Foley et al., 1989; McLean & Parker,
1989). Anthropometric measures could be related to performance (endurance or sprint performance) (van der
Zwaard et al., 2019). Having a low-fat percentage can enhance cycling performance in climbing conditions (Lee et
al., 2002; Abbiss et al., 2018). Road cyclists have a predominance of slow-twitch muscle fibres and a broad typology
range was found, which could be explained by the specialisation of individual cyclists (Lievens et al., 2021).
With exception of the time trial races, overall, professional riders perform competitions at submaximal intensity
due to the longer duration of the races, and because they can decrease the energy cost by riding in a peloton (by
as much as ± 40%). Depending on the duration, race track, and distance of the race, the aerobic or anaerobic energy
16
system becomes increasingly important (Fernández-García et al., 2000; Faria et al., 2005). From a physiological
perspective, aerobic and anaerobic performance levels are important performance determinants (Lucia et al., 1998;
Lucia et al., 1999; Fernández-García et al., 2000; Padilla et al., 2001) (Table 1.2.). For example, riders who are able to
generate higher power and maximal cadence outputs at the end of a race, might have a chance of winning certain
races (Foley et al., 1989). Besides, from a biomechanical perspective, being light is especially important in time trial
and climbing races. Those characteristics are associated with reduced gravitational pull and aerodynamic drag of
the upper body (Foley et al., 1989; Mujika & Padilla, 2001; Menaspà et al., 2012; van der Zwaard et al., 2019). Moreover,
an increase in non-functional body weight lowers performance through a three-fold effect. It increases (a) the
energy cost of gearing, (b) rolling resistance, and (c) the cyclist's frontal surface area (Craig & Norton, 2001). The
intensity level during the races in the U17 and U19 categories is generally higher than the professional elite races
due to the shorter race duration and minor tactical nature of the race. Most of the competitions within these youth
categories are performed at a submaximal intensity and power output (Faria et al., 2005).
Drafting (reducing the energy cost by riding in a slipstream of others – peloton), positioning (positioning in the
beginning, middle, and/or end of the race), pacing (dividing the race according to the individual characteristics of
the athlete), breakaways (attacking cyclists forming a smaller group), and team dynamics are considered tactical
performance determinants in road cycling (Phillips & Hopkins, 2020).
4.2.2. Features and athletic profile of track cycling
Track cycling once dominated in cycling. However, road cycling quickly overruled track cycling due to the increasing
media and public attention in road cycling, a discipline tied to sports facilities (velodrome) (Mallon & Heijmans,
2011). The discipline became an Olympic discipline in 1896. Track cyclists often switch to road cycling during their
careers because they get more exposure, respect, and renumeration. However, Flanders has a track cycling tradition
(e.g. t'Kuipke) and often win international competitions. Track cycling distinguishes itself from the other cycling
disciplines because the events are carried out in velodromes (similar to other indoor sports) and these are
considered to be more ‘stable’ environments. As a result, factors such as surface and weather do not influence the
performance (Schumacher et al., 2001). Track cycling is performed on a track bicycle with fixed gear (only one drive
cog and one chainring). Elite male endurance track cyclists reach their average peak performance at 25.5 +/- 3.6
years of age (Longo et al., 2016). The discipline can be divided into two categories, namely sprint (< 1 km) and
endurance events (distance: > 1 km - < 50 km) (Craig & Norton, 2001). In this dissertation, we only focus on the
athletic profile of cyclists performing in endurance track events (Table 1.1.). The race duration of the youth categories
17
(U17, U19) are shorter compared to the elite adult categories. For example, riders from the U19 category perform
an individual pursuit of 3 km, compared to 4 km in the elite adult category.
In track cycling, anthropometric measures are determinants of performance. Due to the constant conditions, these
factors are even more crucial in track cycling than in other cycling disciplines (Schumacher et al., 2001). The mean
somatotype of the elite endurance track cyclists could be defined as short or tall meso-ectomorph (van der Zwaard
et al., 2019; Foley et al., 1989; McLean & Parker, 1989). Male endurance track cyclists show a low BMI, muscle mass,
and limb girth (Muros-Molina et al., 2022). Having a low-fat percentage can enhance cycling performance in indoor
conditions (Lee et al., 2002). Elite track endurance cyclists show an intermediate muscle typology (high proportion
of slow- and fast-twitch muscle fibres) (Lievens et al., 2021).
As the riders perform with a fixed gear the cycling cadence is one of the few variables a rider can adjust to control
performance and fatigue (Abbiss et al., 2009; Ansley & Cangley, 2009). The optimal cadence for riders is relatively
stable; however, it can change during the development of young athletes due to changes in muscle fibre
composition and muscle coordination (Hautier et al., 1996). The stable environment of the velodrome limits the
influence of environmental factors. Therefore, physiological and biomechanical aspects such as aerodynamic
characteristics (e.g., the position of the rider on the bicycle, being light) are the keys to success (Schumacher et al.,
2001). Endurance track cyclists require a high contribution of both the aerobic and anaerobic energy system to
excel in these shorter-duration races compared to road cyclists (Craig & Norton, 2001) (Table 1.2.). Given the shorter
events in the U17 and U19 categories, it is expected that the percentage of anaerobic energy delivered and the
average power outputs will be higher in youth track cyclists. Dependent on the race type, positioning (elimination,
mass-start) and/or pacing (sprint) becomes more performance determining in track cycling (Phillips & Hopkins,
2020).
4.2.3. Features and athletic profile of cyclo-cross
Despite being absent in the Olympic Games, cyclo-cross is Belgium's second most popular cycling discipline. It is
considered the winter sport of Belgium, and it is very popular with the Belgian public and media. (West) Europeans
dominated the competitions, but most top riders have come from Belgium in the past decades. Most international
races are organised in Europe (Mallon & Heijmans, 2011). Its World Championships have been contested since 1950
(Bossi et al., 2018). Since cyclo-cross is an outdoor winter sport, the performances can be strongly influenced by
the weather conditions. The course consists of road, country and forest paths, (man-made) obstacles, and
meadowland surfaces, requiring the rider to quickly dismount, carry the bike while navigating the obstruction and
18
remount. The skills required for the races are based on road cycling, MTB, and running (Carmichael et al., 2017). The
duration of events is approximately 60 min for elite adults consisting of several laps of roughly 3 km (UCI, 2018)
(Table 1.1.). In the youth categories, the duration of events are approximately 30 min for U17, and 40 min for U19
cyclo-cross cyclists, consisting of several laps of roughly 3 km (UCI, 2018).
A pilot study by Taeymans et al. (2016) indicated that professional cyclo-cross athletes have a mesomorph–
ectomorph somatotype profile and a low fat mass. However, further research is needed to get an understanding
into the morphological profile of cyclo-cross cyclists. Additionally, Lievens et al. (2021) found that elite cyclo-cross
cyclists have a slow muscle typology.
Although, competitions are performed at high intensity, cyclo-cross cyclists must have the ability to produce high
power outputs and have to reach a high anaerobic performance level (Carmichael et al., 2017) (Table 1.2.).
Furthermore, a well-developed aerobic energy system is required for top-level performance to recover more easily
between high-intensity periods (Carmichael et al., 2017; Bossi et al., 2018). In the adult elite, the U17 and U19
categories, competitions are performed at high intensity. Besides, an explosive start, good positioning skills, and
excellent bike skills to control and stabilise the bicycle are needed to succeed in the adult elite cyclo-cross category
(Carmichael et al., 2017; Phillips & Hopkins, 2020).
4.2.4. Features and athletic profile of mountain bike
Mountain bike (MTB) started in the US in the 1970s and is now one of the world's most popular outdoor recreational
activities. MTB was included for the first time in the Olympic summer games in 1996 (Mallon & Heijmans, 2011).
During the last decade, off-road cycling (cross-country) has enjoyed exponential growth in popularity. Mountain
bike (cross-country) is an outdoor discipline, in which competitions are performed with a bicycle designed for off-
road cycling, in a mountain landscape on forest roads, tracks, fields, and earth or gravel paths. It includes significant
climbing and descending. In contrast with cyclo-cross the entire track can be completed on the bike, without the
need to dismount (Novak et al., 2018). Male MTB cyclists reach their average peak performance at 28.2 +/- 4.7 years
of age (Longo et al., 2016). A race lasts between 90-105 minutes, and typically consists of a sequence of 4 – 6 km
laps (Impellizzeri & Marcora, 2007; UCI, 2018; Mirizio et al., 2021) (Table 1.1.). Youth competitions differ in duration
of the race and length of the track but are generally similar to international races (Fornasiero et al., 2018). Junior
mountain bike competitions have a lap length of 4 – 6 km and an average climbing altitude of about 500 – 1000
m (Mirizio et al., 2021).
19
The anthropometric profile of a cross-country rider corresponds to that of a road cyclist (climber). As climbing is
part of an MTB competition, having low body mass and body fat percentage values can be beneficial (Lee et al.,
2002; Impellizzeri et al., 2005; Impellizzeri & Marcora, 2007; Impellizzeri et al., 2008). The mean somatotype of the
elite MTB riders could be defined as ecto-mesomorph (Sánchez-Muñoz et al., 2017). Mirizio and colleagues' (2021)
research showed that anthropometric variables do not correlate with race time. Thus, as long as athletes have an
appropriate body composition, minor inter-individual variations do not appear to determine performance (Mirizio
et al., 2021). Elite cross-country athletes have a slow muscle typology (Lievens et al., 2021).
Elite MTB cycling demands a well-developed aerobic performance levels to compete at at a high level (Impellizzeri
& Marcora, 2007; Bejder et al., 2019; Lievens et al., 2021). Nevertheless, the contribution of the anaerobic system is
essential to generate high power during specific phases in the trail (start, climbing) (Fornasiero et al., 2018; Bejder
et al., 2019). The physiological profile of road cyclists (climbers) and MTB cyclists are comparable (Novak &
Dascombe, 2014). For MTB, the combination of aerobic and anaerobic performance level variables, and skill-based
characteristics are strong predictors of the performance, in both young and adult cyclists (Impellizzeri et al., 2005;
Novak et al., 2018; Mirizio et al., 2021) (Table 1.2.). Furthermore, cross-country MTB is characterised by a fast start
followed by an even pacing strategy and its need for technical skills to control and stabilise the bicycle, especially
during descents, where obstacles need to be avoided at very high speed (Impellizzeri & Marcora, 2007; Granier et
al., 2018).
4.2.5. Features and athletic profile of BMX
Bicycle motocross (BMX) is a relatively young discipline that originated in the USA in the late 1960s (Cowell et al.,
2012). A few years later, the sport became known in Belgium through the Netherlands. Since 2008, it has been an
Olympic discipline, and with each passing year, BMX is becoming more popular. Step by step, the number of BMX
facilities is increasing in Flanders. BMX is characterised by being a short duration and explosive sport. Elite athletes
reach their peak performances at an average age of 23.2 +/- 2.5 years (Longo et al., 2016). The races have a duration
of 30 and 45 seconds, with a 30-minute recovery period in between the races (Mateo-March et al., 2012) (Table 1.1.).
The riders who reach the final (max. 8) have completed at least six races (Mateo-March et al., 2012). Each race is
performed on a 300m – 400m course that includes turnouts and a series of consecutive and irregular hills (rhythm
sections). The start takes place at the top of a steep ramp of 2.5m to 8m high depending on the competition
category (Rylands & Roberts, 2019). Given the technical character of the discipline, athletes start practicing BMX
already at a young age (around the age of five years old).
20
With respect to anthropometry, BMX athletes have mesomorph or mesomorph-endomorph profile (de Azambuja
Pussieldi et al., 2010; Daneshfar et al., 2020). They possess a higher body fat percentage compared to MTB and road
cyclists (de Azambuja Pussieldi et al., 2010), and are characterised by a faster muscle typology (high proportion of
fast-twitch muscle fibres) compared to the other cycling disciplines (Lievens et al., 2021).
According to Mateo et al. (2012) and Zabala et al. (2008) most of the energy is provided from the anaerobic
metabolism (Table 1.2.). Even though a BMX race consists of repeated short efforts at high intensity with high power
outputs, a well-developed aerobic performance level seems important (Louis et al., 2013). A better aerobic condition
ensures better recovery between the heats and will therefore have a favourable influence on the final performance
(Tomlin & Wenger, 2001). Furthermore, BMX races consist of 30 to 45-second efforts in which the oxidative
pathways play an essential role (Ferguson et al., 2021). Coaches and riders consider the start and positioning
performance determining factors for the final finishing position (Mateo et al., 2011). More specifically, the first 8-10
seconds of a race are statistically correlated with the finish position of the riders (Rylands et al., 2014). Therefore,
BMX riders require well-developed muscular power of the lower limbs, and since they perform with fixed gear the
pedalling frequency is a performance determinant (Rylands et al., 2017). Subsequently, the races consist of a mix
of acyclic and cyclic phases, and the acyclic work is also known to consist of overcoming obstacles, where balance,
coordination and isometric strength are crucial.
21
Table 1.2.: Overview of characteristics described in the literature for the different cycling disciplines per discipline.
Track cycling MTB BMX
Road cycling Cyclo-cross#
(endurance) (cross-country) (race)
Age at peak performance 29.5 ± 4.3 years 25.5 ± 3.6 years / 28.2 ± 4.7 years 23.2 ± 2.5 years
Body composition elites
Body length 183 ± 4 173.0 ± 184.6 171.5 ± 9.08 184 ± 6 180 ± 4
Body weight 72.3 ± 3.5 65.25 ± 85.35 68.33 ± 9.20 79.1 ± 13.4 81.3 ± 4.9
Body morphology mesomorph - ectomorph meso - ectomorph meso–ectomorph meso - ectomorph meso - endomorph
Physiological characteristics elites
Wmax (W) 452 ± 35* 373 ± 31** 455 ± 35* 353 ± 48*
Wmax (W/kg) 6.3 ± 0.6* 5.8 ± 0.6 4.4 ± 0.7*
VO2max (l/min) 5.13 ± 0.36** 4.15 ± 0.80***
VO2max (ml/min/kg) 69.6 ± 11.5* 68.5 ± 6.4** 60.16 ± 9.98*** 65.3 ± 7* 52.4 ± 5.9*
Lamax (mmol/l) 10.04 ± 1.81** 12.0 ± 1.8***
% Time in LOW (HR) 38 0.4 ± 0.4 18 ± 10
% Time in MODERATE (HR) 38 6.1 ± 6.7 51 ± 9
% Time in HIGH (HR) 24 93.6 ± 6.7 31 ± 16
CMJ (cm) 49 ± 7 58.6 ± 7.7
Muscle fibre type Slow muscle typology Intermediate muscle typology Slow muscle typology Slow muscle typology Fast muscle typology
Technical characteristics elites
Drafting
Positioning
Positioning Positioning Positioning
Pacing Positioning
Technical characteristics Pacing Pacing Pacing
Breakaway Bike skills
Aerodynamics Bike skills Bike skills
Team dynamics
Aerodynamics
Mujika and Padilla (2001); Novak and Craig et al. (1993); Craig and Impellizzeri and Marcora (2007); Novak Novak and Dascombe (2014);
Carmichael et al. (2017);
References Dascombe (2014); Longo et al. (2016); Norton (2001); Longo et al. and Dascombe (2014); Longo et al. (2016); Longo et al. (2016); Petruolo et
Lievens et al. (2021)
Lievens et al. (2021) (2016); Lievens et al. (2021) Bejder et al. (2019); Lievens et al. (2021) al. (2020); Lievens et al. (2021)
Note: * = incremental test 100W+ 30W/min; ** = incremental cycling test 200W +25W/min; *** = 60W + 35W/3min; # = the research of Carmichael et al. (2017) experienced (non ± elite) cyclo-cross cyclists, while the other studies
examined professional adult elite cyclists.
22
4.3. Talent identification based on competition performance
In a more traditional talent identification process, athletes, who are already active in the sport, are selected based
on achievement measures (race results, rankings), expertise of coaches, and talent scouts (Staff et al., 2021). This
methodology is the predominant pathway for identifying potential. However, additional insights should be gained
about the age at which cycling performance in youth road cycling become useful for talent identification.
Professional cycling teams, cycling federations, and nations also show an increasing interest in the performance
characteristics of young cyclists with potential, since they experience a power struggle to win medals in the major
international competitions mainly due to the increasing media awareness worldwide (Vaeyens et al., 2009; Van
Reeth & Larson, 2016). Understanding the indicators that predict success at the elite level remains a critical concern
for many sport stakeholders (Brouwers et al., 2012; Gulbin et al., 2013b). Cycling performance can be analysed in
two ways: based on single performances (competition results), or based on career pathway (development in
competition results compared to peers).
4.3.1. Competition results
Within time trial sports such as swimming and speed skating relative season best scores are used to compare race
performances (Stoter et al., 2019; Post et al., 2020). All the season-best times are related to the overall world record
of the corresponding category and sex. This method enables to compare results across multi-generations and
corrects for the evolutionary changes in the sport (de Koning, 2010). In tennis, researchers and coaches use official
rankings of the international federation from the junior category onwards (Brouwers et al., 2012; Li et al., 2018b).
However, these rankings are not available in the younger youth categories. Therefore, the competitive
performances of the players are calculated using the sum of the index scores achieved during a limited number of
competitions and divided by the number of tournaments (Brouwers et al., 2012). Nevertheless, quantifying
performance data is even more challenging in road cycling compared to swimming, speed skating, and tennis. Race
outcomes in road cycling are additionally dependent on contextual factors such as the race track (e.g., varying
distance, duration, altimeter, surface) and weather conditions (Svendsen et al., 2018). Further research is required
to create more uniformity into the methodologies to quantify performance data to get a better understanding into
the career pathways of the athletes.
Competitive performances are an essential source of information within cycling and help determine cyclists'
development in two different ways. First, the development of athletes is influenced by the competitive system
itself. For example, the competitive system for road cyclists in Belgium is constructed in a way that only the better
23
performers in the youth competitions, such as the provincial championships, may participate in the national
competitions such as the Belgian championships. The performances during races are subject to contextual, e.g.,
weather conditions, opponents, crashes, puncture, health problems and personal factors, e.g., maturity, motivation,
personality (Phillips & Hopkins, 2020). Factors such as these can often impede an athlete’s performance, and thus
lead to missed opportunities of participation in prestigious competitions. Furthermore, this exclusion can cause an
athlete to gain less experience and have fewer chances to become talent-spotted (Svendsen et al., 2018).
Additionally, repeatedly obtaining bad results and/or encountering bad luck, injuries, and crashes may discourage
athletes from continuing sports participation, which may cause drop-out (Svendsen et al., 2018).
Second, coaches and professional cycling teams use competition results to evaluate cycling performance to identify
and select athletes. Whether or not an athlete is selected, and therefore gets the opportunity to compete, influences
his/her development. Indeed, selection may accelerate and optimise development, whereas not being selected may
slow down the development or discourage athletes from continuing in the sport (Gagné, 2010; Johnston & Baker,
2020). Therefore, it is important to avoid selection errors, especially concerning athletes with future potential
(Brouwers et al., 2012).
4.3.2. Career pathway
Gulbin and colleagues (2013a) investigated the specific career pathways and transitions across competition
categories of athletes from different sports, taking individual variability into account (Figure 1.6.). The main finding
was that athletes experience a non-linear career pathway (Gulbin et al., 2013a). The early high achievers do not
necessarily reach the highest performance level at a later age. In contrast, others who seem less successful at a
young age may even reach the highest performance level at a later age, as well as athletes who transfer later in
their careers to a new complementary sport (talent transfer) (Pion et al., 2021). As mentioned before, different
internal and external factors such as environment, motivation, personality of the athlete, chances and/or obstacles
(e.g., injuries, infrastructure, education, sponsorship, coaches) can positively or negatively affect the development
of the athlete (Gagné, 2004). In addition, the sport continues to change (Lath et al., 2020). Someone who wins the
Tour de France today with a particular athletic profile may not be able to do so ten years later (Baker et al., 2018;
Stoter et al., 2019; Lath et al., 2020). Nevertheless, coaches are expected to make short- and long-term predictions.
24
Figure 1.6.: The Athlete Development Triangle (ADT) developed by Gulbin et al. (2013a) represents the prevalence, magnitude
and direction of transitions between competition levels within sports.
Note: Percentages reflect relative transition from all those at that respective competition level; Jnr = junior; Snr = senior;
LOC = Local club and/or school; REG = Regional; AUS = Australia
As mentioned before, road cyclists can start competing from the age of 8 years. The transition between categories
occurs automatically based on the age of the cyclists and within road cycling there are multi-year age categories.
As a result, cyclists transitioning to the next category will have to compete against cyclists of the same and older
ages (1 to 3 years age difference), better known as a ‘between year effect’ (Faber et al., 2020). The transition from
the junior to the senior category is considered the most challenging transition in the career pathway of athletes
(Stambulova et al., 2012; Drew et al., 2019). Between 18 and 24 years of age, these athletes may experience
difficulties in athletic and non-athletic domains (Morris, 2013; Drew et al., 2019). At this stage of life, athletes are
confronted with many changes, for example, the academic transition from secondary school to high school or work,
the transition from adolescence to young adulthood, and the increasing volume, intensity, and difficulty of training
sessions and competitions. The extent to which athletes experience this transition as challenging is individually
different and depends on various external and contextual factors (Drew et al., 2019). If athletes possess an accurate
perception of the transition process, they are more likely to experience greater success (Morris et al., 2016; Drew
et al., 2019). However, there is a need to expand the knowledge about road cyclists' career pathways to guarantee
sufficient support for athletes when transitioning to a new competition category.
25
The ultimate goal of a road cyclist in their sports career is to be recruited for a World tour (WT, first division), Pro
Continental (PC, second division), or Continental (C, third division) team, respectively in descending order of
importance (Van Bulck et al., 2021). Riders hoping for early selection should perform well at competitions from 19
years onwards, since they are mainly selected based on the cycling performance (Svendsen et al., 2018). Riders who
are not selected for one of these teams perform at an amateur level or drop-out of the sport. The UCI regulations
(Union Cycliste Internationale) determine which teams may participate in which races (UCI, 2018). World tour (WT)
teams perform at the highest elite level in road cycling and are invited to the prestigious races. There is a limited
number of WT teams (19 in 2021) and the teams consist of a small number of very talented riders. In contrast, there
is no limit to the total number of Pro Continental (PC) teams and Continental (C) teams. PC teams can earn an
invitation to the prestigious races, but only a few riders/teams may participate. In the lower division competitions,
PC teams and C teams are allowed to participate (Miller & Susa, 2018). Teams are composed of riders with different
athletic profiles, and for each competition, the team selects a limited number of riders with the appropriate athletic
profiles with the aim of winning the races (Miller & Susa, 2018). However, the career pathway to amateur, C, PC, and
WT is long and complex.
4.3.3. Additional Constraints
Athletes would not be able to reach their top level without the support of the coaches, family, sports federation,
and/or other stakeholders. Coaches play a significant role in the development of athletes, as they monitor them
closely and are responsible for important decisions during their careers. Experienced coaches use an intuitive
approach, and evaluations are based on the athlete as a whole (holistic character), taking into account information
from different dimensions (Buekers et al., 2015; Sieghartsleitner et al., 2019). Therefore, coaches' ratings appear to
be relatively reliable and valid. Nevertheless, their decisions are frequently biased by subjective emotions, and
some coaches do not work within the generally accepted talent models. Moreover, the athletes' maturity status
may bias the coaches' decisions. Nevertheless, experienced coaches would assess maturity better and take it into
account (Romann & Fuchslocher, 2013; Sieghartsleitner et al., 2019).
26
5. Hiatuses for the development of a Cycling Compass
5.1. Talent orientation and transfer based upon discipline-specific profiles
As previously mentioned, male road, track, and MTB cyclists (no information available on cyclo-cross) reach their
average peak performance at an age later than 25 years (Longo et al., 2016) and are considered ‘relatively’ late
specialization sports compared to, e.g., BMX, where peak performance is achieved at an earlier age (23.2 +/- 2.5
years) (Longo et al., 2016). To date, there is no consensus that early diversification would be beneficial in both early
and late specialization sports (Moesch et al., 2011). Diversification, a process that implies low value of domain
specificity (Sieghartsleitner et al., 2018) corresponds to participating in different sports and activities during
athletic development, which may have a positive influence on the development of more diverse physical, social and
psychological skills (Moesch et al., 2011; Brouwers et al., 2012). The study of Güllich and colleagues (2014) showed
that world-class athletes trained and competed in various sports during childhood, adolescence, and adulthood
and that they specialised in their domain sport later in their careers. The volume of domain-specific training did
not differ between world-class and national athletes. By specializing too early in one sports discipline, athletes
may miss opportunities to develop their potential and motor skills optimally (Jayanthi et al., 2013; Mostafavifar et
al., 2013). Moreover, these athletes are more likely to experience physical and psychological burn-out and drop-out
(Strachan et al., 2009; Sieghartsleitner et al., 2018). Such an approach appears to be very beneficial for many sports,
except for the disciplines that require early specialisation due to the specific demands of the sport or age-related
regulations (e.g., artistic gymnastics). For coaches and federations, it is essential to know when talent orientation
and/or transfer can occur in the development of athletes with the aim of lifelong sports participation and improved
performance at the adult professional level.
Knowledge about the different cycling profiles and their similarities and differences is needed to support athletes
in their choice and search for the cycling discipline that best suits their abilities. Comparative studies are scarce in
the literature and are currently limited to examining the physiological characteristics of professional adult cyclists
(e.g. (an)aerobic performance variables, power-profiling, muscle typology, etc.).
5.2. Talent identification, selection and performance prediction
The following section provides an overview of the current insights and hiatuses of the predictive value of test and
cycling performances.
27
5.2.1. Predictions based on a (non-)sport-specific test battery
Several studies in volleyball (Pion et al., 2015a; Mostaert et al., 2020), gymnastics (Vandorpe et al., 2012b), and
soccer (Deprez, 2015; Johnston et al., 2018) indicated that general motor coordination is a significant predictor of
future success. Moreover, these studies suggest that the physical test results are less reliable for predicting
performance later on. The latter findings may be due to research that is often conducted with homogeneous groups
selected on these physical characteristics for development programs in the Top Sport Schools (Pion et al., 2015a).
However, one might question the use of such a non-sport-specific test battery and especially motor coordination
tests in the context of cycling. Indeed, during cycling, the hands, feet, and hips of the athlete are fixated most of
the time. From that perspective cycling is completely different from many other sports where athletes ‘move freely’
in space. The motor coordination tests are included in this work for the following reasons:
a) A generic coordination test has been proven to be predictive for performance in sports where technical
skills are very different from the content of the test (e.g., Pion et al. (2015a) in volleyball). As such, the KTK
test can be considered as a tool for the very generic concept of ‘learning potential’ in sports.
b) Although the body parts are physically constrained during cycling, motor coordination is still crucial. The
athlete must develop coordination of the athlete-bicycle complex instead of focusing on body
coordination. Especially during curves, abrupt changes of position in the peloton, keeping balance,
avoiding falls, negotiating obstacles, or dealing with different surfaces, coordination is important for the
athlete-bicycle complex (Zeuwts, 2016).
Future research should examine whether test performances measured with a non-sport-specific test battery,
supplemented with sport-specific tests can be helpful for talent identification and prediction of future performance
within cycling.
5.2.2. Predictions based on competition results
Predicting future success based on competition results is challenging (Elferink-Gemser et al., 2011; Brouwers et al.,
2012; Güllich, 2014). The accuracy of prediction is, amongst other factors, dependent on the type of sport (Brouwers
et al., 2012). For so-called cgs sports, where performance is measured in centimetres, grams, and seconds (e.g., long
jump, weight lifting, and triathlon), performance prediction seems more straightforward than non-cgs sports, such
as artistic-, team-, and combat sports (Güllich & Emrich, 2014). Cgs sports are mainly individual or closed sports.
They are characterised by a high focus on physical abilities and a limited focus on technical and tactical aspects.
Additionally, the accuracy of the prediction depends on the timing. The younger the athletes and the further away
28
from their peak performance, the more challenging the prediction becomes. Therefore, it is suggested that cycling
may have lower prediction values in the younger competition categories since peak performance is reached at a
later age (>25 years) (Brouwers et al., 2012). Svendsen et al. (2018) and Schumacher et al. (2006) were one of the
first to relate competitive performance at the junior level (U19) with performances at the senior elite level in road
cycling. Svendsen and colleagues (2018) found that better performances at the junior level provide more
opportunities for riders to reach the world-class level in adulthood. In contrast, Schumacher and colleagues (2006)
reported that only 1/3rd of all riders who participated in the world elite championships were successful in the junior
category. Similar results were found within sports such as soccer, swimming, volleyball, and judo (Barreiros &
Fonseca, 2012). Successful athletes in the junior category are more likely to obtain better financial support, team
contracts, and race opportunities (Svendsen et al., 2018). As a result, they might be more easily spotted for their
talent. Further research in cycling may provide new knowledge into cyclists' developmental pathways and the
predictability of cycling performance across several cycling categories.
5.3. Talent development
Age- and sex-specific benchmarks of test performances are frequently reported in the literature (Norjali Wazir et
al., 2018; Mostaert et al., 2021; Robertson et al., 2021). These benchmarks may inform and help coaches and sports
federations involved in talent identification, development, orientation, and selection (Pion, 2015). In addition,
benchmarks allow comparing athlete performances (Carey & Delaney, 2010; Colan, 2013; Pion, 2015). An extensive
database consisting of cross-sectional and/or longitudinal data is needed to develop reliable benchmarks (Lenoir
& Philippaerts, 2011; Pion, 2015). Benchmarks can be reported as raw scores or normalised scores. Raw scores are
straightforward to interpret and are most commonly used by coaches, athletes, and sports organisations.
Normalised scores describe a value's relationship to the mean of a group of values. Z-scores, for example, express
the individual score relative to the mean and standard deviation of a reference group. Within cycling, benchmarks
of a (non-)sport-specific test battery are lacking. However, these benchmarks could shed light on how individual
riders, disciplines or age group perform and develop relative to each other. Moreover, benchmarks offer several
advantages: (a) benchmarking can offer coaches and athletes new understandings into the strengths and
weaknesses of the athlete, allowing them to set realistic goals (Pion, 2015), (b) it can motivate and improve
performances of athletes at short- and long-term, (c) it can be used as a tool to guide athletes to the discipline(s)
that best fit their qualities (talent orientation) (Pion et al., 2021), (d) benchmarks can support (national) coaches,
sports organisations (focused on elite sports), professional cycling teams to identify, select, and develop athletes
with potential (Stoter et al., 2019).
29
6. General overview and aims of this research
The overall dissertation title contains four underlying constructs: (1) youth cycling, (2) performance, (3) talent
prediction, and (4) the Cycling Compass. First, the main focus of this dissertation is on youth cycling. Currently, there
is a lack of information on youth sports participation in cycling compared to data obtained from professional adult
cycling studies. Given the complexity of the development of young athletes that have already been highlighted in
previous research, this dissertation seeks to explore the athletic characteristics of young cycling performers of
different cycling disciplines. Second, performance within this research refers to test and cycling performances.
Talent test batteries are already applied and considered valuable in talent research in various sports, and
competition results are regularly used by coaches to evaluate cycling performance. Nevertheless, within cycling,
the practical relevance and value of both components have not yet been examined. Third, this dissertation focuses
on talent prediction, which is based on the longitudinal follow-up of athletes. This method is preferred as it provides
a better understanding into the characteristics that contribute to predicting future success rather than youth
athletes’ current performance. Finally, the Cycling Compass refers to a new applied and cycling-specific version of
the previously developed Flemish Sports Compass (Pion, 2015). The Cycling Compass would be the first construct in
youth cycling that combines all these different components and features.
The first objective (chapter 1) of this dissertation was to investigate the athletic profiles of cyclists from four
different cycling disciplines. The cycling disciplines (road, cyclo-cross, MTB, and track cycling) show similarities and
differences based on, e.g., the competition format, physiological, and morphological characteristics of the cyclists.
However, research is limited to the comparison of the athletic profiles of professional adult cyclists (Lucía et al.,
2001; Peinado et al., 2011; Novak & Dascombe, 2014; Spindler et al., 2018; Lievens et al., 2021; Sanders & Van Erp,
2020). Consequently, the first study of this dissertation focuses on the athletic profiles of adolescent and young
adult cyclists in four different disciplines. It was investigated to what extent test performance measured with a
generic test battery, consisting of a combination of anthropometric, physical performance, and coordination tests
(the non-sport-specific part), supplemented with cycling specific tests (the sport-specific part), can distinguish and
discriminate athletic profiles of these disciplines (study 1) (Figure 1.7.).
The second objective of this dissertation, presented in chapter two, is to gain insight into the extent to which test
performances measured with a (non-)sport-specific test battery can predict future cycling performance. In the
literature, physiological parameters such as aerobic and anaerobic performance levels, power profiles, cycling
economy have already been extensively studied (Menaspà et al., 2010; Nimmerichter, 2018; Van Erp & Sanders, 2020;
Gallo et al., 2021). These parameters are helpful to identify the better performers within the junior and senior
30
categories. However, research has shown that these characteristics alone are not sensitive enough to predict future
competitive success at adult age (Menaspà et al., 2010). Thus, other performance-determining features appear to
contribute to future cycling peformance (Phillips & Hopkins, 2020). The use of a (non-)sport-specific test battery,
measuring characteristics such as strength, agility, explosivity, motor coordination, has not been introduced before
within cycling. Nevertheless, previous research has already shown that this test battery can be a valuable tool in
the search for talent and for the prediction of performance (Johnson et al., 2008; Vandorpe et al., 2012b; Deprez,
2015; Mostaert et al., 2020). Subsequently, the second study aimed to assess the extent to which test performances
on the (non-)sport-specific test battery can predict future cycling performance two to three years later within road
cycling (study 2) (Figure 1.7.).
The third aim of this dissertation, presented in chapter two, is to gain an understanding into the extent to which
current cycling performance and relative age can predict future cycling performance. Retrospective studies of
competition results can be a powerful tool to objectify the career pathway of riders who may or may not reach the
elite level (Schumacher et al., 2006). Schumacher et al. (2006) and Svendsen et al. (2018) were among the first to
relate juniors' competitive performance to seniors' competitive performance in road cycling. While Svendsen et al.
(2018) found that good race performance at junior age was a strong predictor of later success in senior elite cycling,
Schumacher et al. (2006) found only limited predictive value. Further research on this subject appears to be
necessary. Therefore, the third study aimed to investigate to what extent current cycling performance during
national and provincial Belgian competitions in the youth categories (U15, U17, U19) can predict future cycling
performance, with particular attention to the relative age and career development of the riders (study 3) (Figure
1.7.). Additionally, the third study's findings were verified with research results from abroad, e.g., Italy. The career
trajectories and the predictability of cycling performance were examined, with the difference that all races in the
youth years (U17, U19) and U23 were included (study 4) (Figure 1.7.).
Finally, the general discussion summarises the findings into a practical tool for talent development in cycling, ‘the
Cycling Compass’, which can be considered a new applied and cycling-specific version of the previously developed
Flemish Sports Compass (Pion, 2015). The Cycling Compass provides recommendations for coaches, federations,
athletes, and other stakeholders for each component in the Cycling Compass and provides performance
benchmarks. Age- and sex-specific performance benchmarks on all kinds of test batteries are frequently reported
in the literature to inform coaches and sports organisations involved in talent identification, development, and
selection (Vandorpe et al., 2011a; Pion, 2015; Mostaert et al., 2021). Benchmarks, consisting of raw scores or
normalised scores, allow comparing athletes' performances (Carey & Delaney, 2010; Colan, 2013; Pion, 2015).
31
In other words, benchmarks could be a valuable and supportive tool when optimizing the developmental pathway
for cycling.
32
Research question Sample size Research group Methods Analyses
To what extent can test performances measured with a 243 male cyclists
(non-)sport-specific test battery distinguish and between 12.0 – 15.9 years ; Road, Track, cyclo- (Non-)sport-specific MAN(C)OVA
discriminate athletic profiles of four different cycling 63 male cyclists cross, MTB cyclists test battery Discriminat analyses
Chapter 1 disciplines? between 16.0 – 27.5 years
Study 1
111 male cyclists

To what extent can test performances measured with a (Non-)sport-specific
between 13.0 – 14.9 years;
(non-)sport-specific test battery predict future cycling Road cyclists test battery and Hierarchical multiple regression
67 male cyclists
Chapter 2 performance two to three years later within road cycling?
between 15.0 – 16.9 years
competition results
Study 2
To what extent can current cycling performance during 307 Belgian male cyclists Non-parametric analysis of
Belgian youth competitions predict future cycling born between 1990 – 1993: Competition results longitudinal data, crosstabs and
Road cyclists
performance, with particular attention to the relative age 32 future achievers, 275 future (U15, U17, U19) Kruskal-Wallis
Chapter 2
and career development of the riders? non-achievers Logistic regression
Study 3
To what extent can current cycling performance of Italian 1345 Italian male cyclists Non-parametric analysis of
riders predict future cycling performance, with particular born between 1992 – 1995: Competition results longitudinal data, crosstabs and
Road cyclists
attention to the relative and career development of the 43 future professionals, 1302 (U17, U19, U23) Kruskal-Wallis
Chapter 2
riders? future non-professionals Logistic regression
Study 4
Cycling
Compass
Figure 1.7.: Schematic overview of the original research.
33
7. References
Abbiss, C., Peiffer, J., & Menaspà, P. (2018). Performance assessment in laboratory and field conditions. In Elite Youth
Cycling (pp. 73-87): Routledge.
Abbiss, C., Quod, M., Levin, G., Martin, D., & Laursen, P. (2009). Accuracy of the Velotron ergometer and SRM power
meter. International journal of sports medicine, 30(02), 107-112.
Abbott, A., Button, C., Pepping, G.-J., & Collins, D. (2005). Unnatural selection: Talent identification and development
in sport. Nonlinear dynamics, psychology, and life sciences, 9(1), 61-88.
Abbott, A., & Collins, D. (2002). A theoretical and empirical analysis of a'state of the art'talent identification model.
High ability studies, 13(2), 157-178.
Ahnert, J., Schneider, W., & Bös, K. (2010). Developmental changes and individual stability of motor abilities from the
preschool period to young adulthood. In Human development from early childhood to early adulthood
(pp. 45-72): Psychology Press.
Allen, S. V., & Hopkins, W. G. (2015). Age of peak competitive performance of elite athletes: a systematic review.
Sports Medicine, 45(10), 1431-1441.
Ansley, L., & Cangley, P. (2009). Determinants of “optimal” cadence during cycling. European Journal of Sport Science,
9(2), 61-85.
Archive. (2019). Retrieved December 2019, from Belgian Cycling https://www.uitslagen.kbwb-rlvb.com/
Bailey, R., Cope, E. J., & Pearce, G. (2013). Why do children take part in, and remain involved in sport? A literature
review and discussion of implications for sports coaches. International Journal of Coaching Science, 7(1),
56-75.
Baker, J., Schorer, J., & Wattie, N. (2018). Compromising talent: Issues in identifying and selecting talent in sport.
Quest, 70(1), 48-63.
Balyi, I., & Hamilton, A. (2004). Long-term athlete development: Trainability in childhood and adolescence. Olympic
coach, 16(1), 4-9.
Bardid, F., Huyben, F., Deconinck, F. J., De Martelaer, K., Seghers, J., & Lenoir, M. (2016). Convergent and divergent
validity between the KTK and MOT 4-6 motor tests in early childhood. Adapted Physical Activity Quarterly,
33(1), 33-47. doi:10.1123/APAQ.2014-0228
Barreiros, A. N., & Fonseca, A. M. (2012). A retrospective analysis of Portuguese elite athletes' involvement in
international competitions. International Journal of Sports Science & Coaching, 7(3), 593-600.
Bejder, J., Bonne, T. C., Nyberg, M., Sjøberg, K. A., & Nordsborg, N. B. (2019). Physiological determinants of elite
mountain bike cross-country Olympic performance. Journal of sports sciences, 37(10), 1154-1161.
Beunen, G., & Malina, R. M. (1988). Growth and physical performance relative to the timing of the adolescent spurt.
Exercise and sport sciences reviews, 16(1), 503-540.
Bourgois, J. (2011). Uithoudingsvermogen. In Basis voor verantwoord trainen (Vol. 47, pp. 90-151): Publicatiefonds
voor Lichamelijke Opvoeding (PVLO).
Bossi, A. H., O’Grady, C., Ebreo, R., Passfield, L., & Hopker, J. G. (2018). Pacing strategy and tactical positioning during
cyclo-cross races. International journal of sports physiology and performance, 13(4), 452-458.
Brouwers, J., De Bosscher, V., & Sotiriadou, P. (2012). An examination of the importance of performances in youth
and junior competition as an indicator of later success in tennis. Sport Management Review, 15(4), 461-
475.
Bruininks, R. H., & Bruininks, B. D. (2005). BOT2: Bruininks-oseretsky test of motor proficiency: manual: Pearson
Assessments.
Buekers, M., Borry, P., & Rowe, P. (2015). Talent in sports. Some reflections about the search for future champions.
Movement & Sport Sciences-Science & Motricité(88), 3-12.
Carey, J. J., & Delaney, M. F. (2010). T-scores and Z-scores. Clinical reviews in bone and mineral metabolism, 8(3),
113-121.
34
Carmichael, R., Heikkinen, D., Mul-lin, E., & McCall, N. (2017). Physiological response to cyclocross racing. Sports
Exercise Medicine Open Journal, 3(2), 74-80.
Cobley, S., Baker, J., Wattie, N., & McKenna, J. (2009). Annual age-grouping and athlete development. Sports Medicine,
39(3), 235-256.
Colan, S. D. (2013). The why and how of Z scores. Journal of the American Society of Echocardiography, 26(1), 38-40.
Collins, D., MacNamara, Á., & McCarthy, N. (2016). Super champions, champions, and almosts: important differences
and commonalities on the rocky road. Frontiers in Psychology, 6, 2009. doi:10.3389/fpsyg.2015.02009
Costa, A. M., Breitenfeld, L., Silva, A. J., Pereira, A., Izquierdo, M., & Marques, M. C. (2012). Genetic inheritance effects on
endurance and muscle strength. Sports Medicine, 42(6), 449-458.
Côté, J., Baker, J., & Abernethy, B. (2007). Practice and play in the development of sport expertise. Handbook of sport
psychology, 3, 184-202.
Cowell, J. F., McGuigan, M., & Cronin, J. (2012). Strength training considerations for the bicycle Motocross athlete.
Strength & Conditioning Journal, 34(1), 1-7.
Craig, N. P., Norton, K., Bourdon, P., Woolford, S., Stanef, T., Squires, B., Olds, T., Conyers, R., & Walsh, C. (1993). Aerobic
and anaerobic indices contributing to track endurance cycling performance. European journal of applied
physiology and occupational physiology, 67(2), 150-158.
Craig, N. P., & Norton, K. I. (2001). Characteristics of track cycling. Sports Medicine, 31(7), 457-468.
Crane, J., & Temple, V. (2015). A systematic review of dropout from organized sport among children and youth.
European physical education review, 21(1), 114-131.
Cycling Vlaanderen, v. (2017-2020). Strategic plan of Cycling Vlaanderen.
Cycling Vlaanderen, v. (2020-2021). Jeugdgids 2020-2021.
Cycling Vlaanderen, v. (2021-2024). Strategic plan of Cycling Vlaanderen 2021-2024.
Daneshfar, A., Petersen, C., Miles, B., & Gahreman, D. (2020). Prediction of track performance in competitive BMX
riders using laboratory measures. Journal of Science and Cycling, 9(1), 44-56.
Davies, P. L., & Rose, J. D. (2000). Motor skills of typically developing adolescents: awkwardness or improvement?
Physical & occupational therapy in pediatrics, 20(1), 19-42.
de Azambuja Pussieldi, G., Santos, B. L., Pereira, L. A., & Marins, J. C. B. (2010). Comparação do perfil antoprométrico
e somatotípico de Ciclistas de elite em diferentes modalidades. Fitness & performance journal(3), 9-14.
de Koning, J. J. (2010). World records: how much athlete? How much technology? International journal of sports
physiology and performance, 5(2), 262-267.
De Waelle, S., Laureys, F., Lenoir, M., Bennett, S. J., & Deconinck, F. J. (2021). Children involved in team sports show
superior executive function compared to their peers involved in self-paced sports. Children, 8(4), 264.
Den Hartigh, R. J., Niessen, A. S. M., Frencken, W. G., & Meijer, R. R. (2018). Selection procedures in sports: Improving
predictions of athletes’ future performance. European Journal of Sport Science, 18(9), 1191-1198.
Deprez, D. (2015). Anthropometrical, physical fitness and maturational characteristics in youth soccer:
methodological issues and a longitudinal approach to talent identification and development. Ghent
University, Ghent, Belgium.
Di Cagno, A., Baldari, C., Battaglia, C., Monteiro, M. D., Pappalardo, A., Piazza, M., & Guidetti, L. (2009). Factors
influencing performance of competitive and amateur rhythmic gymnastics—Gender differences. Journal
of Science and Medicine in Sport, 12(3), 411-416.
Diamond, A. (2013). Executive functions. Annual review of psychology, 64, 135-168.
Drew, K., Morris, R., Tod, D., & Eubank, M. (2019). A meta-study of qualitative research on the junior-to-senior
transition in sport. Psychology of Sport and Exercise, 45, 101556. doi:10.1016/j.psychsport.2019.101556
Elferink-Gemser, M. T., Jordet, G., Coelho-E-Silva, M. J., & Visscher, C. (2011). The marvels of elite sports: how to get
there? British Journal of Sports Medicine, 45(9), 683-684.
Faber, I. R., Liu, M., Cece, V., Jie, R., Martinent, G., Schorer, J., & Elferink-Gemser, M. T. (2020). The interaction between
within-year and between-year effects across ages in elite table tennis in international and national
contexts–A further exploration of relative age effects in sports. High ability studies, 31(1), 115-128.
35
Faria, E. W., Parker, D. L., & Faria, I. E. (2005). The science of cycling. Sports Medicine, 35(4), 285-312.
Ferguson, H. A., Zhou, T., Harnish, C., & Chase, J. G. (2021). Model of 30-s sprint cycling performance: Don’t forget the
aerobic contribution! IFAC-PapersOnLine, 54(15), 316-321.
Fernández-García, B., Pérez-Landaluce, J., Rodríguez-Alonso, M., & Terrados, N. (2000). Intensity of exercise during
road race pro-cycling competition. Medicine and Science in Sports and Exercise, 32(5), 1002-1006.
Foley, J., Bird, S., & White, J. (1989). Anthropometric comparison of cyclists from different events. British Journal of
Fornasiero, A., Savoldelli, A., Modena, R., Boccia, G., Pellegrini, B., & Schena, F. (2018). Physiological and anthropometric
characteristics of top-level youth cross-country cyclists. Journal of sports sciences, 36(8), 901-906.
Fransen, J., Bush, S., Woodcock, S., Novak, A., Deprez, D., Baxter-Jones, A. D., Vaeyens, R., & Lenoir, M. (2018). Improving
the prediction of maturity from anthropometric variables using a maturity ratio. Pediatric exercise science,
30(2), 296-307. doi:10.1123/pes.2017-0009
Fransen, J., D’Hondt, E., Bourgois, J., Vaeyens, R., Philippaerts, R. M., & Lenoir, M. (2014). Motor competence assessment
in children: Convergent and discriminant validity between the BOT-2 Short Form and KTK testing batteries.
Research in developmental disabilities, 35(6), 1375-1383. doi:10.1016/j.ridd.2014.03.011
Gagné, F. (2004). Transforming gifts into talents: The DMGT as a developmental theory. High ability studies, 15(2),
119-147. doi:10.1080/1359813042000314682
Gagné, F. (2010). Motivation within the DMGT 2.0 framework. High ability studies, 21(2), 81-99.
Gagné, F. (2013). The DMGT: Changes within, beneath, and beyond. Talent Development & Excellence, 5(1), 5-19.
Gallo, G., Filipas, L., Tornaghi, M., Garbin, M., Codella, R., Lovecchio, N., & Zaccaria, D. (2021). Thresholds power profiles
and performance in youth road cycling. International journal of sports physiology and performance, 16(7),
1049-1051.
Granier, C., Abbiss, C. R., Aubry, A., Vauchez, Y., Dorel, S., Hausswirth, C., & Le Meur, Y. (2018). Power output and pacing
during international cross-country mountain bike cycling. International journal of sports physiology and
performance, 13(9), 1243-1249.
Gulbin, J., Weissensteiner, J., Oldenziel, K., & Gagné, F. (2013a). Patterns of performance development in elite
athletes. European Journal of Sport Science, 13(6), 605-614.
Gulbin, J. P., Croser, M. J., Morley, E. J., & Weissensteiner, J. R. (2013b). An integrated framework for the optimisation
of sport and athlete development: A practitioner approach. Journal of sports sciences, 31(12), 1319-1331.
Güllich, A. (2014). Many roads lead to Rome–Developmental paths to Olympic gold in men's field hockey. European
Journal of Sport Science, 14(8), 763-771.
Güllich, A., & Emrich, E. (2014). Considering long-term sustainability in the development of world class success.
European Journal of Sport Science, 14(sup1), S383-S397. doi:10.1080/17461391.2012.706320
Hautier, C., Linossier, M., Belli, A., Lacour, J., & Arsac, L. (1996). Optimal velocity for maximal power production in non-
isokinetic cycling is related to muscle fibre type composition. European journal of applied physiology and
occupational physiology, 74(1), 114-118.
Hill, A., MacNamara, Á., & Collins, D. (2019). Development and initial validation of the Psychological Characteristics
of Developing Excellence Questionnaire version 2 (PCDEQ2). European Journal of Sport Science, 19(4), 517-
528.
Höner, O., Votteler, A., Schmid, M., Schultz, F., & Roth, K. (2015). Psychometric properties of the motor diagnostics in
the German football talent identification and development programme. Journal of sports sciences, 33(2),
145-159.
Impellizzeri, F., Ebert, T., Sassi, A., Menaspà, P., Rampinini, E., & Martin, D. (2008). Level ground and uphill cycling
ability in elite female mountain bikers and road cyclists. European journal of applied physiology, 102(3),
335-341.
Impellizzeri, F. M., & Marcora, S. M. (2007). The physiology of mountain biking. Sports Medicine, 37(1), 59-71.
Impellizzeri, F. M., Rampinini, E., Sassi, A., Mognoni, P., & Marcora, S. (2005). Physiological correlates to off-road
cycling performance. Journal of sports sciences, 23(1), 41-47.
36
Jacobson, J., & Matthaeus, L. (2014). Athletics and executive functioning: How athletic participation and sport type
correlate with cognitive performance. Psychology of Sport and Exercise, 15(5), 521-527.
Jayanthi, N., Pinkham, C., Dugas, L., Patrick, B., & LaBella, C. (2013). Sports specialization in young athletes: evidence-
based recommendations. Sports health, 5(3), 251-257.
Johnson, M. B., Castillo, Y., Sacks, D. N., Cavazos Jr, J., Edmonds, W. A., & Tenenbaum, G. (2008). “Hard Work Beats
Talent until Talent Decides to Work Hard”: Coaches' Perspectives regarding Differentiating Elite and Non-
Elite Swimmers. International Journal of Sports Science & Coaching, 3(3), 417-430.
Johnston, K., & Baker, J. (2020). Waste reduction strategies: factors affecting talent wastage and the efficacy of
talent selection in sport. Frontiers in Psychology, 10, 2925.
Johnston, K., Wattie, N., Schorer, J., & Baker, J. (2018). Talent identification in sport: a systematic review. Sports
Medicine, 48(1), 97-109.
Khamis, H. J., & Roche, A. F. (1994). Predicting adult stature without using skeletal age: the Khamis-Roche method.
Pediatrics, 94(4), 504-507.
Kiphard, E. J., & Schilling, F. (1974). Körperkoordinations Test für Kinder. Beltz Test GmbH. In: Weinheim.
Kiphard, E. J., & Schilling, F. (2017). Körperkoordinationstest für Kinder : KTK.
Lath, F., den Hartigh, R., Wattie, N., & Schorer, J. (2020). Talent Selection: Making Decisions and Prognoses about
Athletes. In Talent Identification and Development in Sport (pp. 50-65): Routledge.
Laureys, F., Collins, D., Deconinck, F. J., & Lenoir, M. (2021). Exploring the use of the Psychological Characteristics of
Developing Excellence (PCDEs) in younger age groups: First steps in the validation process of the PCDE
Questionnaire for Children (PCDEQ-C). PloS one, 16(11), e0259396.
Lee, H., Martin, D. T., Anson, J. M., Grundy, D., & Hahn, A. G. (2002). Physiological characteristics of successful mountain
bikers and professional road cyclists. Journal of sports sciences, 20(12), 1001-1008.
Lenoir, M., & Philippaerts, R. (2011). Van detectie tot medaille. Eindrapport van het project Vlaams Sport Kompas in
opdracht van de Vlaamse Overheid (2007-2011).
Leyhr, D., Kelava, A., Raabe, J., & Höner, O. (2018). Longitudinal motor performance development in early adolescence
and its relationship to adult success: An 8-year prospective study of highly talented soccer players. PloS
one, 13(5), e0196324.
Li, P., De Bosscher, V., Pion, J., Weissensteiner, J. R., & Vertonghen, J. (2018a). Is international junior success a reliable
predictor for international senior success in elite combat sports? European Journal of Sport Science, 18(4),
550-559.
Li, P., De Bosscher, V., & Weissensteiner, J. R. (2018b). The journey to elite success: a thirty-year longitudinal study of
the career trajectories of top professional tennis players. International Journal of Performance Analysis
in Sport, 18(6), 961-972.
Lievens, E., Bellinger, P., Van Vossel, K., Vancompernolle, J., Bex, T., Minahan, C., & Derave, W. (2021). Muscle typology
of world-class cyclists across various disciplines and events. Medicine and Science in Sports and Exercise,
53(4), 816-824.
Lloyd, R. S., & Oliver, J. L. (2012). The youth physical development model: A new approach to long-term athletic
development. Strength & Conditioning Journal, 34(3), 61-72.
Longo, A. F., Siffredi, C. R., Cardey, M. L., Aquilino, G. D., & Lentini, N. A. (2016). Age of peak performance in Olympic
sports: A comparative research among disciplines. Journal of Human Sport and Exercise, 11(1), 31-41.
doi:10.14198/jhse.2016.111.03
LTAD-VOLUME, C. (2008). Long-Term Athlete Development - Cycling. 1.
Lucia, A., Hoyos, J., Carvajal, A., & Chicharro, J. (1999). Heart rate response to professional road cycling: the Tour de
France. International journal of sports medicine, 20(03), 167-172.
Lucía, A., Hoyos, J., & Chicharro, J. L. (2001). Physiology of professional road cycling. Sports Medicine, 31(5), 325-337.
Lucia, A., Pardo, J., Durantez, A., Hoyos, J., & Chicharro, J. (1998). Physiological differences between professional and
elite road cyclists. International journal of sports medicine, 19(05), 342-348.
Malina, R. M., Bouchard, C., & Bar-Or, O. (2004). Growth, maturation, and physical activity (2nd ed.): Human kinetics.
37
Mallon, B., & Heijmans, J. (2011). Historical dictionary of cycling: Scarecrow Press.
Martin, K., Staiano, W., Menaspà, P., Hennessey, T., Marcora, S., Keegan, R., Thompson, K. G., Martin, D., Halson, S., &
Rattray, B. (2016). Superior inhibitory control and resistance to mental fatigue in professional road cyclists.
PloS one, 11(7), e0159907. doi:10.1371/journal.pone.0159907
Mateo, M., Blasco-Lafarga, C., & Zabala, M. (2011). Pedaling power and speed production vs. technical factors and
track difficulty in bicycle motocross cycling. The Journal of Strength & Conditioning Research, 25(12), 3248-
3256.
Matthys, S. P., Vaeyens, R., Fransen, J., Deprez, D., Pion, J., Vandendriessche, J., Vandorpe, B., Lenoir, M., & Philippaerts,
R. (2013). A longitudinal study of multidimensional performance characteristics related to physical
capacities in youth handball. Journal of sports sciences, 31(3), 325-334.
McLean, B. D., & Parker, A. W. (1989). An anthropometric analysis of elite Australian track cyclists. Journal of sports
sciences, 7(3), 247-255.
Menaspà, P., Rampinini, E., Bosio, A., Carlomagno, D., Riggio, M., & Sassi, A. (2012). Physiological and anthropometric
characteristics of junior cyclists of different specialties and performance levels. Scandinavian Journal of
Medicine & Science in Sports, 22(3), 392-398. doi:10.1111/j.1600-0838.2010.01168.x
Menaspà, P., Sassi, A., & Impellizzeri, F. M. (2010). Aerobic fitness variables do not predict the professional career of
young cyclists. Medicine and Science in Sports and Exercise, 42(4), 805-812.
Mignot, J.-F. (2016). The history of professional road cycling. In The economics of professional road cycling (pp. 7-
31): Springer.
Miller, J., & Susa, K. (2018). Comparison of anthropometric characteristics between world tour and professional
continental cyclists. Journal of Science and Cycling, 7(3), 3-6.
Mirizio, G. G., Muñoz, R., Muñoz, L., Ahumada, F., & Del Coso, J. (2021). Race Performance Prediction from the
Physiological Profile in National Level Youth Cross-Country Cyclists. International Journal of
Environmental Research and Public Health, 18(11), 5535.
Mirwald, R. L., Baxter-Jones, A. D., Bailey, D. A., & Beunen, G. P. (2002). An assessment of maturity from
anthropometric measurements. Medicine & Science in Sports & Exercise, 34(4), 689-694.
Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D. (2000). The unity and diversity of
executive functions and their contributions to complex "frontal lobe" tasks: A latent variable analysis.
Cognitive Psychology, 41(1), 49-100. doi:10.1006/cogp.1999.0734
Moesch, K., Elbe, A. M., Hauge, M. L., & Wikman, J. M. (2011). Late specialization: the key to success in centimeters,
grams, or seconds (cgs) sports. Scandinavian Journal of Medicine & Science in Sports, 21(6), e282-e290.
doi:10.1111/j.1600-0838.2010.01280.x
Moore, S. A., McKay, H. A., Macdonald, H., Nettlefold, L., Baxter-Jones, A. D., Cameron, N., & Brasher, P. M. (2015).
Enhancing a somatic maturity prediction model. Med Sci Sports Exerc, 47(8), 1755-1764.
Morris, R. (2013). Investigating the youth to senior transition in sport: From theory to practice. Aberystwyth
University,
Morris, R., Tod, D., & Oliver, E. (2016). An investigation into stakeholders’ perceptions of the youth-to-senior transition
in professional soccer in the United Kingdom. Journal of Applied Sport Psychology, 28(4), 375-391.
Mostaert, M., Coppens, E., Laureys, F., D'Hondt, E., Deconinck, F. J., & Lenoir, M. (2021). Validation of a motor competence
assessment tool for children and adolescents (KTK3+) with normative values for 6-to 19-year-olds.
Frontiers in physiology, 12, 652952. doi:10.3389/fphys.2021.652952
Mostaert, M., Deconinck, F., Pion, J., & Lenoir, M. (2016). Anthropometry, physical fitness and coordination of young
figure skaters of different levels. International journal of sports medicine, 37(07), 531-538. doi:10.1055/s-
0042-100280
Mostaert, M., Pion, J., Lenoir, M., & Vansteenkiste, P. (2020). A Retrospective Analysis of the National Youth Teams in
Volleyball: Were They Always Faster, Taller, and Stronger? Journal of Strength and Conditioning Research,
1-7. doi:10.1519/jsc.0000000000003847
38
Mostafavifar, A. M., Best, T. M., & Myer, G. D. (2013). Early sport specialisation, does it lead to long-term problems?
British Journal of Sports Medicine, 47, 1060-1061.
Mujika, I., & Padilla, S. (2001). Physiological and performance characteristics of male professional road cyclists.
Muros-Molina, J. J., Mateo-March, M., Zabala, M., & Sánchez-Muñoz, C. (2022). Anthropometric differences between
world-class professional track cyclists according to speciality (endurance vs. sprint). The Journal of sports
medicine and physical fitness.
Musch, J., & Grondin, S. (2001). Unequal competition as an impediment to personal development: A review of the
relative age effect in sport. Developmental review, 21(2), 147-167.
Nimmerichter, A. (2018). Elite Youth Cycling (1st ed.): Routledge.
Norjali Wazir, M. R. W., Mostaert, M., Pion, J., & Lenoir, M. (2018). Anthropometry, physical performance, and motor
coordination of medallist and non-medallist young fencers. Archives of Budo, 14, 33-40. doi:1854/LU-
8599260
Novak, A. R., Bennett, K. J., Beavan, A., Pion, J., Spiteri, T., Fransen, J., & Lenoir, M. (2017). The applicability of a short
form of the Körperkoordinationstest für Kinder for measuring motor competence in children aged 6 to 11
years. Journal of Motor Learning and Development, 5(2), 227-239. doi:10.1123/jmld.2016-0028
Novak, A. R., Bennett, K. J., Fransen, J., & Dascombe, B. J. (2018). Predictors of performance in a 4-h mountain-bike
race. Journal of sports sciences, 36(4), 462-468.
Novak, A. R., & Dascombe, B. J. (2014). Physiological and performance characteristics of road, mountain bike and BMX
cyclists. Journal of Science and Cycling, 3(3), 9.
O’Brien-Smith, J., Tribolet, R., Smith, M., Bennett, K., Fransen, J., Pion, J., & Lenior, M. (2019). The use of the
Körperkoordinationstest für Kinder in the talent pathway in youth athletes: A systematic review. Journal
of Science and Medicine in Sport, 22(9), 1021-1029. doi:10.1016/j.jsams.2019.05.014
Padilla, S., Mujika, I., Orbananos, J., Santisteban, J., Angulo, F., & Goiriena, J. J. (2001). Exercise intensity and load during
mass-start stage races in professional road cycling. Medicine & Science in Sports & Exercise, 33(5), 796-
802.
Pearson, D., Naughton, G. A., & Torode, M. (2006). Predictability of physiological testing and the role of maturation
in talent identification for adolescent team sports. Journal of Science and Medicine in Sport, 9(4), 277-287.
Peinado, A. B., Benito, P. J., Díaz, V., González, C., Zapico, A. G., Álvarez, M., Maffulli, N., & Calderón, F. J. (2011).
Discriminant analysis of the speciality of elite cyclists. Journal of Human Sport and Exercise, 6(3), 480-
489.
Petruolo, A., Connolly, D. R., Bosio, A., Induni, M., & Rampinini, E. (2020). Physiological profile of elite Bicycle Motocross
cyclists and physiological-perceptual demands of a Bicycle Motocross race simulation. The Journal of
sports medicine and physical fitness, 60(9), 1173-1184. doi:10.23736/s0022-4707.20.10855-7
Phillips, K. E., & Hopkins, W. G. (2020). Determinants of Cycling Performance: a Review of the Dimensions and
Features Regulating Performance in Elite Cycling Competitions. Sports Medicine-Open, 6, 1-18.
Pion, J. (2015). The Flemish sports compass: From sports orientation to elite performance prediction. Ghent
Pion, J., Fransen, J., Deprez, D., Segers, V., Vaeyens, R., Philippaerts, R. M., & Lenoir, M. (2015a). Stature and jumping
height are required in female volleyball, but motor coordination is a key factor for future elite success.
The Journal of Strength & Conditioning Research, 29(6), 1480-1485. doi:10.1519/JSC.0000000000000778
Pion, J., Lenoir, M., Vandorpe, B., & Segers, V. (2015b). Talent in female gymnastics: a survival analysis based upon
performance characteristics. International journal of sports medicine, 94(11), 935-940.
Pion, J., Segers, V., Fransen, J., Debuyck, G., Deprez, D., Haerens, L., Vaeyens, R., Philippaerts, R. M., & Lenoir, M. (2015c).
Generic anthropometric and performance characteristics among elite adolescent boys in nine different
sports. European Journal of Sport Science, 15(5), 357-366. doi:10.1080/17461391.2014.944875
39
Pion, J., Teunissen, J. W., Ter Welle, S., Spruijtenburg, G., Faber, I., & Lenoir, M. (2021). How similarities and differences
between sports lead to talent transfer: A process approach In J. Baker, S. Cobley, & J. Schorer (Eds.), Talent
Identification and Development in Sport: International Perspectives (2nd ed.). London: Routledge.
Post, A. K., Koning, R. H., Visscher, C., & Elferink‐Gemser, M. T. (2020). Multigenerational performance development
of male and female top‐elite swimmers–A global study of the 100 m freestyle event. Scandinavian Journal
of Medicine & Science in Sports, 30(3), 564-571.
Raglin, J. S. (2001). Psychological factors in sport performance. Sports Medicine, 31(12), 875-890.
Rees, T., Hardy, L., Güllich, A., Abernethy, B., Côté, J., Woodman, T., Montgomery, H., Laing, S., & Warr, C. (2016). The great
British medalists project: a review of current knowledge on the development of the world’s best sporting
talent. Sports Medicine, 46(8), 1041-1058.
Régnier, G., Salmela, J., & Russell, S. (1993). Talent detection and development in sport. A Handbook of Research on
Sports Psychology (edited by R. Singer, M. Murphey, and LK Tennant); 290-313. In: New York: Macmillan.
Roberts, B. W., & DelVecchio, W. F. (2000). The rank-order consistency of personality traits from childhood to old age:
a quantitative review of longitudinal studies. Psychological bulletin, 126(1), 3.
Robertson, K., Laureys, F., Mostaert, M., Pion, J., Deconinck, F. J., & Lenoir, M. (2021). Mind, body, and shuttle:
multidimensional benchmarks for talent identification in male youth badminton. Biology of sport, 38(4),
79-94.
Romann, M., & Fuchslocher, J. (2013). Relative age effects in Swiss junior soccer and their relationship with playing
position. European Journal of Sport Science, 13(4), 356-363.
Rommers, N., Mostaert, M., Goossens, L., Vaeyens, R., Witvrouw, E., Lenoir, M., & D’Hondt, E. (2019). Age and maturity
related differences in motor coordination among male elite youth soccer players. Journal of sports
sciences, 37(2), 196-203.
Sánchez-Muñoz, C., Muros, J. J., & Zabala, M. (2017). World and Olympic mountain bike champions' anthropometry,
body composition and somatotype. The Journal of sports medicine and physical fitness, 58(6), 843-851.
Sanders, D., & van Erp, T. (2020). The Physical Demands and Power Profile of Professional Men’s Cycling Races: An
Updated Review. International journal of sports physiology and performance, 1(aop), 1-10.
Schumacher, Y., Mueller, P., & Keul, J. (2001). Development of peak performance in track cycling. Journal of sports
medicine and physical fitness, 41(2), 139.
Schumacher, Y. O., Mroz, R., Mueller, P., Schmid, A., & Ruecker, G. (2006). Success in elite cycling: A prospective and
retrospective analysis of race results. Journal of sports sciences, 24(11), 1149-1156.
Sieghartsleitner, R., Zuber, C., Zibung, M., & Conzelmann, A. (2018). “The Early Specialised Bird Catches the Worm!”–A
Specialised Sampling Model in the Development of Football Talents. Frontiers in Psychology, 9, 188.
Sieghartsleitner, R., Zuber, C., Zibung, M., & Conzelmann, A. (2019). Science or coaches’ eye?–both! Beneficial
collaboration of multidimensional measurements and coach assessments for efficient talent selection in
Elite Youth Football. Journal of sports science & medicine, 18(1), 32.
Spindler, D. J., Allen, M. S., Vella, S. A., & Swann, C. (2018). The psychology of elite cycling: a systematic review. Journal
of sports sciences, 36(17), 1943-1954.
Staff, T., Gobet, F., & Parton, A. (2021). Early Specialization and Critical Periods in Acquiring Expertise: A Comparison
of Traditional Versus Detection Talent Identification in Team GB Cycling at London 2012. Journal of Motor
Learning and Development, 9(2), 296-312.
Stambulova, N., Franck, A., & Weibull, F. (2012). Assessment of the transition from junior-to-senior sports in Swedish
athletes. International Journal of Sport and Exercise Psychology, 10(2), 79-95.
Stoter, I. K., Koning, R. H., Visscher, C., & Elferink-Gemser, M. T. (2019). Creating performance benchmarks for the
future elites in speed skating. Journal of sports sciences, 37(15), 1770-1777.
Strachan, L., Côté, J., & Deakin, J. (2009). “Specializers” versus “samplers” in youth sport: comparing experiences and
outcomes. The sport psychologist, 23(1), 77-92.
40
Svendsen, I. S., Tønnesen, E., Tjelta, L. I., & Ørn, S. (2018). Training, Performance and Physiological Predictors of a
Successful Elite Senior Career in Junior Competitive Road Cyclists. International journal of sports
physiology and performance, 13(10), 1287-1292. doi:10.1123/ijspp.2017-0824
Taeymans, J., Roeykens, J., Neels, H., Gielen, J., & Vissers, D. (2016). Anthropometry and exercise testing in
professional cyclo-cross athletes. Paper presented at the Congress of the European College of Sports
Science, Vienna, Austria.
Teunissen, J. W., Rommers, N., Pion, J., Cumming, S. P., Rössler, R., D’Hondt, E., Lenoir, M., Savelsbergh, G. J., & Malina,
R. M. (2020). Accuracy of maturity prediction equations in individual elite male football players. Annals of
Human Biology, 47(4), 409-416.
Toum, M., Tribolet, R., Watsford, M. L., & Fransen, J. (2021). The confounding effect of biological maturity on talent
identification and selection within youth Australian football. Science and Medicine in Football, 5(4), 263-
271.
Tucker, R., & Collins, M. (2012). What makes champions? A review of the relative contribution of genes and training
to sporting success. British Journal of Sports Medicine, 46(8), 555-561.
UCI. (2018). UCI Regulations. Retrieved from http://www.uci.ch/inside-uci/rules-and-regulations/regulations/
Vaeyens, R., Güllich, A., Warr, C. R., & Philippaerts, R. (2009). Talent identification and promotion programmes of
Olympic athletes. Journal of sports sciences, 27(13), 1367-1380. doi:10.1080/02640410903110974
Vaeyens, R., Lenoir, M., Williams, A. M., & Philippaerts, R. M. (2008). Talent identification and development
programmes in sport. Sports Medicine, 38(9), 703-714.
Vaeyens, R., Philippaerts, R. M., & Malina, R. M. (2005). The relative age effect in soccer: A match-related perspective.
Journal of sports sciences, 23(7), 747-756.
Van Bulck, D., Vande Weghe, A., & Goossens, D. (2021). Result-based talent identification in road cycling: discovering
the next Eddy Merckx. Annals of Operations Research, 1-18. doi:10.1007/s10479-021-04280-0
Van Der Hoeven, S., Constandt, B., Schyvinck, C., Lagae, W., & Willem, A. (2021). The grey zone between tactics and
manipulation: The normalization of match-fixing in road cycling. International Review for the Sociology of
Sport, 10126902211038414.
van der Zwaard, S., De Ruiter, C. J., Jaspers, R. T., & De Koning, J. J. (2019). Anthropometric clusters of competitive
cyclists and their sprint and endurance performance. Frontiers in physiology, 10, 1276.
Van Erp, T., & Sanders, D. (2020). Demands of professional cycling races: Influence of race category and result.
European Journal of Sport Science, 21(5), 666-677. doi:10.1080/17461391.2020.1788651
Van Reeth, D., & Larson, D. J. (2016). The economics of professional road cycling: Springer.
Vandendriessche, J. B., Vaeyens, R., Vandorpe, B., Lenoir, M., Lefevre, J., & Philippaerts, R. M. (2012). Biological
maturation, morphology, fitness, and motor coordination as part of a selection strategy in the search for
international youth soccer players (age 15–16 years). Journal of sports sciences, 30(15), 1695-1703.
Vandorpe, B. (2011). The contribution of motor coordination to sports participation in school-aged children and to
talent identification in young female gymnasts. Ghent University, Ghent, Belgium.
Vandorpe, B., Vandendriessche, J., Lefèvre, J., Pion, J., Vaeyens, R., Matthys, S., Philippaerts, R., & Lenoir, M. (2011a). The
Körperkoordinationstest für kinder: Reference values and suitability for 6–12‐year‐old children in
Flanders. Scandinavian Journal of Medicine & Science in Sports, 21(3), 378-388. doi:10.1111/j.1600-
0838.2009.01067.x
Vandorpe, B., Vandendriessche, J., Vaeyens, R., Pion, J., Lefevre, J., Philippaerts, R., & Lenoir, M. (2011b). Factors
discriminating gymnasts by competitive level. International journal of sports medicine, 32(08), 591-597.
doi:1854/LU-1949420
Vandorpe, B., Vandendriessche, J., Vaeyens, R., Pion, J., Matthys, S., Lefevre, J., Philippaerts, R., & Lenoir, M. (2012a).
Relationship between sports participation and the level of motor coordination in childhood: a longitudinal
approach. Journal of Science and Medicine in Sport, 15(3), 220-225.
41
Vandorpe, B., Vandendriessche, J. B., Vaeyens, R., Pion, J., Lefevre, J., Philippaerts, R. M., & Lenoir, M. (2012b). The value
of a non-sport-specific motor test battery in predicting performance in young female gymnasts. Journal
of sports sciences, 30(5), 497-505. doi:10.1080/02640414.2012.654399
Victoris, G. U. (2020). Sportkompas: a sport orientation tool for children - infographic. Retrieved from
https://www.victoris.be/wp-content/uploads/2021/06/SportKompas_Infographic_EN.pdf
Wells, J., & Fewtrell, M. (2006). Measuring body composition. Archives of disease in childhood, 91(7), 612-617.
Williams, A. M., & Ford, P. R. (2009). Promoting a skills-based agenda in Olympic sports: The role of skill-acquisition
specialists. Journal of sports sciences, 27(13), 1381-1392.
Wong, C. N., Chaddock-Heyman, L., Voss, M. W., Burzynska, A. Z., Basak, C., Erickson, K. I., Prakash, R. S., Szabo-Reed, A.
N., Phillips, S. M., & Wojcicki, T. (2015). Brain activation during dual-task processing is associated with
cardiorespiratory fitness and performance in older adults. Frontiers in aging neuroscience, 7, 154.
Zeuwts, L. (2016). Understanding and stimulating the development of perceptual-motor skills in child bicyclists.
Ghent University, Ghent, Belgium.
Zimmer, R., & Volkamer, M. (1987). Motoriktest für vier-bis sechsjährige Kinder: Beltz Test.
42
PART 2 : ORIGINAL RESEARCH
Chapter 1 : Investigating the athletic profiles of
cycling athletes
Study 1 - Discriminating performance profiles of cycling disciplines
Mireille Mostaert¹*, Felien Laureys¹, Pieter Vansteenkiste¹, Johan Pion¹,², Frederik J.A. Deconinck¹,
Matthieu Lenoir¹
¹ Department of Movement and Sport Sciences, Ghent University, Ghent, Belgium
² Sport and Exercise Studies, HAN University of Applied Sciences, Nijmegen, The Netherlands
-
International Journal of Sport Sciences and Coaching
Abstract
The aim of this study was to document the athletic profiles of adolescent and young adult high performance
cycling athletes and to investigate to what extent different cycling disciplines can be discriminated from each other
based on a generic test battery. A total of 243 adolescent (12.0-15.99 y) and 63 young adult (≥16 y) male cyclists
from road cycling, track cycling, cyclo-cross, and mountain bike (MTB) participated in the study. All participants
performed four anthropometric, five physical, three motor coordination, and two cycling-specific tests. Using
discriminant analyses, the young adult athletes could be correctly classified to their discipline for 80.7%. Track
cyclists outperformed the other disciplines in explosive and coordinative skills while for cyclo-cross, and
particularly MTB, the performance on the shuttle bike test seemed to be a distinctive variable. Road cyclists
however, showed a significant overlap in performance characteristics with the other disciplines. In spite of the less
pronounced discriminative character in the adolescent cycling population (51.0%), the discriminative characteristics
are in line with the results of the young adults. This study allows to orient cyclists towards their best-fitted
discipline in young adulthood. The relevance of these findings for coaches, experts and federations with respect to
early/late specialization, and talent orientation are discussed.
Keywords: Anthropometry, cyclo-cross, mountain biking, motor coordination, road cycling, talent development,
youth sport
50
Introduction
Engaging in a variety of different sports, related sub-disciplines, and/or play-like activities allows young
athletes to experience different physical, cognitive, affective, and psychosocial challenges (Côté et al., 2009). This
phenomenon, called early diversification, has positive influences on the development of intrinsic motivation, motor
coordination, and sport-specific skills (Balyi & Hamilton, 2004; Côté et al., 2007; Côté et al., 2009). Côté et al. (2007)
showed that, in general, 12-year-old athletes gradually reduce involvement in other sports and/or related
disciplines and begin to focus on their main sport, a process that evolves into highly deliberate practice around the
age of 16. Providing talent orientation advice has been suggested to facilitate the search for the main discipline, in
which the athlete has maximal chances to excel. This process could subsequently lead to better performances,
satisfaction, higher motivation, and increased enjoyment (Pion, 2015; Hohmann et al., 2018), all of which contribute
to persistence in sports activity (Crane & Temple, 2015). Although early diversification is suggested, especially in
late specialization sports such as cycling (Moesch et al., 2011; Longo et al., 2016), this does not alter the fact that
young athletes may already prepare themselves in the discipline that suits him/her the most.
Appropriate talent programs can help federations and coaches to gain insight in the specific athletic profiles
of different disciplines and sports. It is expected that cyclists in talent programs should develop to an athletic
profile that fits the requirements needed to succeed at adult competitive level (Vandorpe et al., 2012b). Like in
many other sports, the coach plays a crucial role in the important decisions of the young athlete’s career. He/she
will often decide whether a young cyclist has enough potential to be provided with better training facilities, should
be selected for a training camp or an important competition. However, there is a lack of reference values of young
cyclists. A coach therefore needs to make that decision on a limited amount of information, for example competitive
performance or subjective opinions. Currently, it is not clear to what extent the high performance adult cycling
profiles are already present and reliably measurable during adolescence and young adulthood.
Current knowledge about the different competition characteristics and athletic profiles of cyclists is mainly
based upon information from adult and junior professional athletes (Nimmerichter, 2018). In literature, the
morphological characteristics of professional male cyclists are regularly investigated in relation to their
physiological profile and role in competition (Impellizzeri et al., 2008). The body types of the cycling disciplines
vary between body length and body composition (muscularity and fat percentage) (Craig & Norton, 2001; Lee et al.,
2002; Menaspa et al., 2010; Peinado et al., 2011; Menaspa et al., 2012). Physiological performance parameters such
as peak power output and ventilatory/lactate thresholds are considered to be important characteristics to
determine the level of performance of cyclists (Padilla et al., 2001; Impellizzeri & Marcora, 2007; Menaspa et al.,
2010; Louis et al., 2012; Menaspa et al., 2012; Novak & Dascombe, 2014). Although, the use of physiological tests is
inviting, within talent research, in which large young populations are tested, it is of great importance to strive for
51
alternative strategies (Pion, 2015). Especially as these physiological tests come with considerable financial and
practical constraints (Pearson et al., 2006).
At first glance, coordination might seem to be of lesser significance in cycling, especially when compared to
sports in which technical skill is a main performance determinant such as in gymnastics or pole vaulting.
Nevertheless, keeping balance during cornering, climbing, and braking in a peloton or on slippery roads without
losing time is technically challenging, particularly within disciplines such as MTB, BMX and cyclo-cross. In addition,
it has been demonstrated that general motor coordination (Vandorpe et al., 2011b) is related to future progression
of an athlete (Vandorpe et al., 2012a; Pion, 2015). Although several studies have underlined the value of this
characteristic in talent orientation research (Vandorpe et al., 2012b; Pion et al., 2015c), there is a dearth of
knowledge on young cycling athletes.
In summary, insights on the general and sport-specific characteristics of young cycling athletes are currently
lacking, in spite of the potential value for supporting coaches’ decisions with respect to orientation, selection, and
training. The current study therefore focuses on the profiles of adolescent and young adult cyclists, in four different
cycling disciplines: road cycling, track cycling, mountain bike (MTB), and cyclo-cross. The main purpose of this study
is to investigate to what extent a generic test battery for cyclists can discriminate performance characteristics
between different cycling disciplines. In doing so, this study will explore similarities and differences between the
cycling disciplines to provide a scientific basis for early diversification and talent orientation. In line with Pion et
al. (2014) and Moesch et al. (2011) it is hypothesised that the differences between the cycling disciplines will be
more pronounced in the older age groups due to ‘natural’ (de-)selection, development, and individual response to
specific training.
Materials and methods
Participants
Data was collected in collaboration with the Flemish Cycling Federation (Cycling Vlaanderen). Participants
were recruited in two ways. First, to identify and monitor young competitive cyclists, Cycling Vlaanderen organises
five test days per year. The participants in these events are adolescent cyclists, who are delegated by the qualified
club trainers based on their potential to excel compared to an average cycling athlete of the same age. For the
current study, the data collected on these test days in the period 2016-2019 were used. A total of 1267 cyclists were
tested between 2016 and 2019. Only the first complete test moment for each participant between 12.0 and
15.99 years of age, with at least one full year of training experience in cycling, and not active in BMX as their main
and/or secondary discipline were included in the study. Consequently, 243 athletes were retained in this research.
52
Online informed consent was obtained from all participants and none of them refused participation. Second, to
identify the performance characteristics of high performance athletes, Cycling Vlaanderen organises separate test
days. 63 male young adult high performance athletes between 16.0 and 27.5 years of age (average age: 18.9 +/- 3.5
years) volunteered to participate in the current research. This group consisted of athletes active in the national
cycling academy (training volume: 10-16h/week), national selection, and professional riders.
Prior to the study, the cyclists were surveyed about their competitive disciplines. The majority of the
participants was active in more than one cycling discipline. Therefore, the 306 cyclists were classified according to
their self-reported main discipline: 98 road cyclists, 115 track cyclists (endurance races), 58 cyclo-cross, and 35
mountain bike (cross-country). Training history, expressed as the number of years the athlete competes in cycling
competitions, was registered using a demographic questionnaire. The study was conducted in accordance with the
Helsinki Declaration and was approved by the Ethics Committee of the Ghent University Hospital under registration
no. PA 2016/021.
Procedure and measurements
The generic test battery consisted of four anthropometric, five physical, and three motor coordination tests.
The generic tests were administered in accordance with the standardised protocols (Kiphard & Schilling, 1974;
Council of Europe, 1988; Matthys et al., 2013; Novak et al., 2017). Based on the advice of an experienced trainer, two
cycling-specific tests which assess crucial aspects of cycling were included. A brief description of all tests is
provided below.
Anthropometry
Body height and sitting height were measured to the nearest 0.1 cm using a calibrated stadiometer (Seca
and Harpenden, Holtain Ltd., UK). In addition, body weight to the nearest 0.1 kg was obtained using a digital balance
scale with a foot-to-foot bioelectrical impedance system (Tanita, BC420SMA, Weda B.V., Holland). Weight and height
values were used to calculate body mass index (BMI; kg/m²). Age of peak height velocity (APHV) and maturity ratio
were predicted by a gender-specific regression equation based upon a non-invasive technique using chronological
age and height, weight and leg length (subtract sitting height from standing height) measurements (Fransen et al.,
2018).
53
Physical performance tests
Five physical performance tests were executed. The athletes performed the tests barefooted with the
exception of the sprint and endurance shuttle run tests. These latter two tests were preceded by a short warm-up
(running, acceleration, and active stretching exercises). Unless stated otherwise the procedures described in the
manual of the EUROFIT test battery were followed (Council of Europe, 1988; Matthys et al., 2013). The explosive leg
power of the participants was determined by performing a standing broad jump (SBJ; to the nearest 1 cm) (Council
of Europe, 1988). The best performance of two trials was registered for the SBJ. The trunk and hamstring flexibility
were assessed through the sit and reach (SAR) to the nearest 0.5 cm (the best score of two trials) (Council of
Europe, 1988). An endurance plank test was measured by executing a plank position for as long as possible (one
trial, to the nearest 1 s) (Asplund & Ross, 2010; Tong et al., 2014). A Seca stadiometer was used to control the position
of the trunk (the athlete has to touch the stadiometer with his back, while holding the position). A 30-meter sprint
test was performed twice with split times at 5 m, 10 m, 20 m and 30 m. The best time of the two trials was assessed
to measure speed (accuracy of 0.001 s, MicroGate Racetime2 chronometry and Polifemo Light Photocells;
MicroGate, Italy) (Matthys et al., 2013). Endurance was measured by a 20-meter endurance shuttle run test (ESHR)
(Council of Europe, 1988), to the nearest of 0.5 min.
Motor coordination tests
The motor coordination tests were performed barefooted. Three validated and reliable subtests of the
Körperkoordinationstest für Kinder (KTK) (Kiphard & Schilling, 1974; Vandorpe et al., 2011b; Kiphard & Schilling, 2017)
were used in the study: 1) moving sideways on wooden boards for 20 seconds (MS; sum of two attempts ), 2) two-
legged jumping from side to side, 15 seconds long (JS; sum of two attempts), and 3) Walking backward along a
balance beam of 6 cm, 4.5 cm, and 3 cm wide (BB; three trials on each balance beam, total number of steps with a
maximum 72). The fourth test of hopping for height was not performed because of the risk of ankle distortions
(Prätorius & Milani, 2004). According to the study of Novak et al. (2017) KTK with only three items remains valid
and shows a relatively high accuracy.
Cycling-specific tests
In addition to the generic test battery, two specific cycling tests were added: a shuttle bike and a cycling-
specific maximal cadence test. The shuttle bike test was performed on a BMX bicycle and aims to evaluate agility
and cycling skill (2 different formats). Detailed information about the bike formats can be found in Addendum 2.4..
The riders needed to complete an eight-shaped circuit (Figure 2.1.) as quickly as possible without hitting the markers
and without losing control over the bicycle. Prior to the test, the cyclists got the opportunity to familiarise with the
54
BMX bicycle and the trail (two attempts). The riders started from a standing position with the help of the test leader
by holding the rider at the back seat (both feet on the pedals) and departed on a signal of the test leader. The
shuttle bike test was performed twice. The best time of both attempts, timed through time gates (accuracy of 0.001
s, MicroGate Racetime2 chronometry and Polifemo Light Photocells; MicroGate, Italy), was used for the analysis.
Test-retest reliability was investigated for the shuttle bike test, using a sub-sample of 23 cyclists who performed
the tests twice within 6 months. The shuttle bike test is a reliable test with an intra class correlation of 0.763
(95% CI 0.440 – 0.899).
Figure 2.1.: Shuttle bike circuit
The maximal cadence test was performed on a calibrated Flanders SR1 Ultimate racing device or
standardised SRM cycle ergometer (SRM - Schoberer Rad Messtechnik; Germany). Athletes with a body height of
>1,55m performed the test on the SRM cycle ergometer. Detailed information about the bike formats can be found
in Addendum 2.4.. The seat and handlebar height were adjusted for each participant to guarantee a bike position
the athletes felt comfortable in. Experienced trainers administered the test. The cyclists got the opportunity to
familiarise with the bike position and resistance while pedalling, without reaching maximal cadence in advance.
For ten seconds the athlete had to perform this maximal cadence (amount) after the following instruction was
given: “Pedal as fast as possible, until the test leader gives a signal to stop”. The cyclists began in a flying start
(resistance 1) while seated and with their hands placed on the drop handlebars. The highest score (peak frequency)
was recorded for one attempt. However, when the trainers observed that the cyclist did not perform at his best,
had not understood the instructions, or the test was interrupted, the test was repeated after a short instruction.
Test-retest reliability was investigated for the maximal cadence test, using the same sub-sample of 23 cyclists who
55
performed the tests twice within 6 months. The maximal cadence test is a reliable test with an intra class
correlation of 0.936 (95% CI 0.863 – 0.970).
Data analysis
The cross-sectional data were analysed with SPSS Statistics version 25. A top-down approach was used,
validating the differences in a young adult group followed by adolescent performers. Separate multivariate
analyses of variance (MANOVA) were conducted, for anthropometric, physical, motor coordination and cycling
specific results. This was followed by separate multivariate analyses of covariance (MANCOVA) with maturity ratio,
calendar age and training history as confounding variables. Post hoc analyses (Scheffe) were applied. The results
are presented in Table 2.1. and 2.3. and in Addendum 2.5. and 2.6., however, the results of the MANCOVA analysis are
discussed below. Discriminant analyses were used to discriminate the athletic profiles of the four different cycling
disciplines (for both young adults and adolescents). The analysis was executed in two ways: discrimination between
the four disciplines, and discrimination of each discipline separately from the other disciplines together. The
classification coefficients are adjusted for group size. For all analyses the significance level was set at a p-value of
< 0.05. Effects with p<0.1 were considered marginally non-significant. Partial eta squared was computed to obtain
effect size. Cohen (1988) cut-offs were used, meaning that effect sizes for partial eta squared smaller than 0.01 are
reported and interpreted as trivial, those between 0.01 and 0.059 are small, between 0.06 and 0.139 are moderate,
and values larger than 0.14 are large.
Results
Young adults
The covariates of maturity ratio, calendar age, and training history marginally affected the results. The
detailed results are presented in Table 2.1. and Addendum 2.5. In the young adult group (≥ 16 years), the MANCOVA
analyses revealed a multivariate significant effect for the cycling disciplines based on the anthropometric
measurements (F=1.879, p=0.042, eta-squared=0.123), physical (F=1.697, p=0.037, eta-squared=0.191), motor
coordination (F=2.182, p=0.027, eta-squared=0.107), and cycling specific tests (F=2.440, p=0.031,
eta-squared= 0.130). Apart from (marginally non) significant effects on body weight and balance beam test, the
BMI, sprint test and shuttle bike test resulted in significant differences between the disciplines.
56
Table 2.1.: Average performance on the anthropometric, physical, motor coordination, and cycling-specific tests (± SD) of the cycling athletes ≥ 16 years of age from the four different disciplines
and results of the MANOVA (F-value, p-value, and effect size) and MANCOVA (F-value, p-value, and effect size).
Road Track Cyclo-cross MTB Effect size MANCOVA Effect size

MANOVA
≥16 years (Mean ± SD) (Mean ± SD) (Mean ± SD) (Mean ± SD) p-value Partial Eta overall p-value Partial Eta
F-value
N=22 N=26 N=8 N=7 Squared F-value Squared
Age (years) 19.8 ± 4.2 17.7 ± 2.5 16.5 ± 0.5 23.5 ± 1.5
APHV (years) 15.2 ± 1.2 14.5 ± 0.8 14.2 ± 0.5 16.1 ± 0.4
Training history (years) 7.6 ± 3.3 7.2 ± 3.4 7 ± 2.1 9.9 ± 2.9
ANTROPOMETRICS 1.629 0.089 0.103 1.879 0.042 0.123
Height (cm) 181.3 ± 6 179.6 ± 5.5 175.3 ± 8 180.5 ± 5 2.016 0.121 0.093 1.297 0.284 0.065
Sitting Height (cm) 94.8 ± 3.1 94.4 ± 2.7 92.5 ± 4.8 95.9 ± 2.6 1.537 0.214 0.072 0.733 0.537 0.038
Weight (kg) 69.1 ± 7.8 69.5 ± 7.3 61.1 ± 8.6 70.2 ± 8.7 2.718 0.053 0.121 2.587 0.062 0.122
BMI (kg/m²) 21 ± 1.8 21.5 ± 1.6 19.8 ± 1.3 21.5 ± 2.2 2.295 0.087 0.104 3.600 0.019 0.162
PHYSICAL CHARACTERISTICS 1.750 0.029 0.187 1.697 0.037 0.191
Planking (sec) 215 ± 89 169 ± 70 174 ± 60 190 ± 74 1.562 0.208 0.074 1.131 0.345 0.057
SBJ (cm) 213 ± 17 214 ± 27 219 ± 22 203 ± 17 0.651 0.586 0.032 0.176 0.912 0.009
SAR (cm) 27.3 ± 10 27 ± 7.8 27.7 ± 7.9 26.3 ± 4.3 0.041 0.989 0.002 0.091 0.965 0.005
Sprint5m (sec) 1.216 ± 0.085b 1.097 ± 0.101a.c.d 1.242 ± 0.118b 1.199 ± 0.078b 8.420 <0.001 0.300 9.274 <0.001 0.332
Sprint10m (sec) 2.022 ± 0.101b 1.86 ± 0.113a.c 2.037 ± 0.132b 1.984 ± 0.107 10.702 <0.001 0.352 10.154 <0.001 0.352
Sprint20m (sec) 3.435 ± 0.143b 3.226 ± 0.141a.c 3.432 ± 0.204b 3.402 ± 0.136 9.294 <0.001 0.321 7.870 <0.001 0.297
Sprint30m (sec) 4.789 ± 0.205b 4.533 ± 0.209a 4.763 ± 0.291 4.752 ± 0.172 6.551 0.001 0.250 5.038 0.004 0.213
MOTOR COORDINATION 2.599 0.008 0.119 2.182 0.027 0.107
Sum Moving Sideways (N) 72 ± 12 75 ± 14 74 ± 9 65 ± 6 1.293 0.285 0.062 0.268 0.848 0.014
Sum Jumping Sideways (N) 93 ± 12 101 ± 12 99 ± 9 90 ± 7 3.071 0.035 0.135 1.431 0.243 0.071
Sum Balance Beam (N) 58 ± 9 58 ± 12 64 ± 8 64 ± 8 1.299 0.283 0.062 2.469 0.071 0.117
CYCLING SPECIFIC TESTS N=21 N=21 N=8 N=7 3.548 0.003 0.170 2.440 0.031 0.130
Shuttle bike (sec) 11.53 ± 0.935d 11.526 ± 1.293 10.584 ± 0.481 9.928 ± 0.503a 6.168 0.001 0.259 4.725 0.006 0.221
Maximal cadence (N) 224 ± 16 228 ± 10 227 ± 10 218 ± 10 1.122 0.349 0.060 0.221 0.881 0.013
Note: Road = a, Track = b, Cyclo-cross = c, MTB= d, which indicates significant differences (p<0.05) between the different groups based on the MANCOVA (post hoc analyses).
57
The discriminant analyses revealed that 80.7% of the cyclists could be correctly assigned to their current cycling
discipline (42.1% cross-validated). Function 1 was found to be significant (rcan = 0.797, Wilks’ Lambda = 0.161 and
p < 0.001) and determined 62.7% of the variance between the disciplines. Functions 2 and 3 were not significant
(pfunction2= 0.155 ; pfunction3= 0.444). In this model of original grouped cases 76.2% of the road cyclists, 85.7% of the
track cyclists, 62.5% of the cyclo-cross cyclists, and 100% of the MTB cyclists were correctly assigned to their
discipline. Figure 2.2. visualises the discrimination and the overlap between the different cycling disciplines. Visual
inspection of the figure demonstrates that MTB, cyclo-cross, and track cycling can be distinguished relatively well
from each other, while road cycling considerably overlaps with the other disciplines.
Figure 2.2.: Discrimination of high performance athletes (≥16 years) from four different cycling disciplines based on
performance results on a non-sport-specific test battery (80.7% correctly classified; 42.1% cross-validated). Note: the scatter
plot has the canonical discriminant function coefficients as its axes, with Function 1 on x-axis and Function 2 on the y-axis.
58
Table 2.2.: Results from the separate discriminant analyses are demonstrated, which discriminates characteristics between one discipline compared to all the cycling disciplines together. A
distinction is made between both groups.
Road Track Cyclo-cross MTB Road Track Cyclo-cross MTB

≥ 16 years ≥ 16 years ≥ 16 years ≥ 16 years 12-15 years 12-15 years 12-15 years 12-15 years
ANTROPOMETRICS
Height (cm) - + + - +
Sitting Height (cm) + + - -
Weight (kg) - + - -
BMI (kg/m²) - + -
PHYSICAL CHARACTERISTICS
Planking (sec) + -
SBJ (cm)
SAR (cm)
Endurance (beep test/ mins)
Sprint5m (sec) - + - + -
Sprint10m (sec) - + - + -
Sprint20m (sec) - + + -
Sprint30m (sec) - + + -
MOTOR COORDINATION
Sum Moving Sideways (N) + -
Sum Jumping Sideways (N) - + -
Sum Balance Beam (N) - +
CYCLING SPECIFIC TESTS
Shuttle bike (sec) - - + - - + +
Maximal cadence (N)
Note 1: + represents better performances/higher scores observed for the particular discipline compared to the three other disciplines together, while - indicated poorer performances/lower scores. Note 2: a colour
code is added to show the level of significance. Dark grey = Wilks Lambda significant (p<0.05); Light grey = Wilks Lambda marginally non-significant (p<0.10).
59
Subsequently, separate discriminant analyses were used to contrast each discipline with the three other
cycling disciplines together (see Table 2.2.). This way, 78.9% of the road cyclists, 87.7% of the track cyclists, 91.2% of
the cyclo-cross cyclists, and 94.7% of the MTB could be correctly classified to their discipline. Compared to the others
disciplines, road cyclists had higher scores on the planking test (p<0.1), but a lower performance on the sprint test,
jumping sideways (p<0.1), and shuttle bike test (p<0.1). Track cyclists performed better on the 30m sprint test and
on two motor coordination tests (moving sideways and jumping sideways) compared to the others, but slower
times were observed on the shuttle bike test (p<0.1). Cyclo-cross cyclists seem to differentiate themselves from
the others based on anthropometric measurements (being smaller and lighter). Significant higher scores were
absent, and a lower performance was found on the 30m sprint test (p<0.1). Finally, the MTB cyclists performed
significantly better on the shuttle bike test but had lower scores on the moving sideways and jumping sideways
test.
Adolescents
Similar analyses were carried out for the adolescent cyclists (See Table 2.3. and Addendum 2.6.). The
covariates maturity ratio and age had a significant effect on the results of the anthropometric measurements.
Significant differences between the disciplines were absent for anthropometry in the MANCOVA analyses (F=1.442,
p=0.142, eta-squared=0.024). No multivariate significant effect was found for the physical measurements (F=1.110,
p=0.326, eta-squared=0.037). For motor coordination (F=1.943, p=0.044, eta-squared=0.024) and the cycling
specific tests (F=3.756, p=0.001, eta-squared=0.046), multivariate significant differences in performance were
found on the balance beam test and shuttle bike test only.
A significant overlap in performance characteristics of the cycling disciplines was observed in the discriminant
analysis. Only an average of 51.0% of the cyclists could be correctly classified (39.9% cross-validated) using the
generic test battery (Figure 2.3.). The four-group cluster within the three-dimensional space (rcan = 0.527,
Wilks’ Lambda = 0.620 and pfunction1 <0.001; pfunction2= 0.306; pfunction3= 0.676) indicates that the functions discriminate
between the four cycling disciplines. 43.4% of the road cyclists, 64.0% of the track cyclists, 62.0% of the cyclo-cross
cyclists, and 10.7% of the MTB cyclists were correctly assigned to their discipline.
60
Table 2.3.: Average performance on the anthropometric, physical, motor coordination, and cycling-specific tests (± SD) of the cycling athletes between 12.0 and 15.99 years of age from the four
different disciplines and results of the MANOVA (F-value, p-value, and effect size) and MANCOVA (F-value, p-value, and effect size).
Road Track Cyclo-cross MTB Effect size MANCOVA Effect size

MANOVA F-
12.0 – 15.99 years (Mean ± SD) (Mean ± SD) (Mean ± SD) (Mean ± SD) p-value Partial Eta overall p-value Partial Eta
value
N=76 N=89 N=50 N=28 Squared F-value Squared
Age (years) 14 ± 1.2 14 ± 1.2 13.4 ± 0.9 13.8 ± 1.3
APHV (years) 13.8 ± 0.6 13.6 ± 0.6 14.1 ± 0.5 14 ± 0.6
Training history (years) 3.7 ± 2.1 3.9 ± 1.8 4.1 ± 1.8 4.2 ± 1.9
ANTROPOMETRICS 4.268 <0.001 0.067 1.442 0.142 0.024
Height (cm) 167.3 ± 10.3 167.9 ± 11.5 159 ± 9.8 161 ± 10 9.942 <0.001 0.111 1.637 0.182 0.020
Sitting Height (cm) 86 ± 5.9 86.7 ± 6.1 81.1 ± 4.9 83 ± 5.7 11.988 <0.001 0.131 2.183 0.091 0.027
Weight (kg) 51.8 ± 10.8 55 ± 11.1 44.7 ± 8.2 47.8 ± 8.9 11.961 <0.001 0.131 2.434 0.066 0.030
BMI (kg/m²) 18.3 ± 2.2 19.3 ± 2.2 17.5 ± 1.6 18.3 ± 1.7 9.310 <0.001 0.105 3.733 0.012 0.045
PHYSICAL CHARACTERISTICS 1.099 0.339 0.036 1.110 0.326 0.037
Planking (sec) 148 ± 74 131 ± 62 152 ± 63 143 ± 68 1.423 0.237 0.018 1.254 0.291 0.016
SBJ (cm) 188 ± 23 185 ± 25 184 ± 18 187 ± 19 0.369 0.775 0.005 2.160 0.093 0.027
SAR (cm) 20.7 ± 7.1 20.5 ± 7.2 21 ± 6.1 19.1 ± 6.8 0.495 0.686 0.006 1.857 0.138 0.023
Endurance (beep test/ mins) 9.6 ± 1.9 9.4 ± 2 9.5 ± 1.6 9.6 ± 1.9 0.181 0.909 0.002 0.883 0.451 0.011
Sprint5m (sec) 1.217 ± 0.108 1.186 ± 0.108 1.234 ± 0.099 1.229 ± 0.109 2.734 0.044 0.033 1.965 0.120 0.024
Sprint10m (sec) 2.05 ± 0.171 2.026 ± 0.156 2.089 ± 0.139 2.085 ± 0.134 2.221 0.086 0.027 0.895 0.445 0.011
Sprint20m (sec) 3.558 ± 0.238 3.523 ± 0.262 3.619 ± 0.211 3.601 ± 0.217 1.993 0.116 0.024 0.734 0.533 0.009
Sprint30m (sec) 5.002 ± 0.352 4.964 ± 0.386 5.106 ± 0.293 5.086 ± 0.315 2.151 0.094 0.026 0.576 0.632 0.007
MOTOR COORDINATION 2.839 0.003 0.035 1.943 0.044 0.024
Sum Moving Sideways (N) 64 ± 8 65 ± 9 64 ± 9 62 ± 9 0.888 0.448 0.011 0.699 0.554 0.009
Sum Jumping Sideways (N) 86 ± 11 87 ± 11 87 ± 11 88 ± 11 0.550 0.648 0.007 1.242 0.295 0.016
Sum Balance Beam (N) 53 ± 12c 54 ± 13 60 ± 9a 57 ± 11 4.874 0.003 0.058 4.088 0.007 0.049
CYCLING SPECIFIC TESTS 5.564 <0.001 0.066 3.756 0.001 0.046
Shuttle bike (sec) 12.334 ± 1.156c 12.412 ± 1.417c 11.433 ± 1.075a.b 11.664 ± 1.475 8.210 <0.001 0.093 7.298 <0.001 0.085
Maximal cadence (N) 205 ± 14 206 ± 15 202 ± 16 202 ± 15 1.088 0.355 0.013 0.693 0.557 0.009
Note: Road = a, Track = b, Cyclo-cross = c, MTB= d, which indicates significant differences (p<0.05) between the different groups based on the MANCOVA (post hoc analyses).
61
Figure 2.3.: Graphical representation of the discriminant analysis of cyclists between 12.0 and 15.99 years of age based on
performance results on a non-sport-specific test battery. Note: the scatter plot has the canonical discriminant function
coefficients as its axes, with Function 1 on x-axis and Function 2 on the y-axis.
Separate discriminant analysis revealed that the road cyclists were significantly taller, had lower scores on
the balance beam test, and a marginally non-significant slower time on the shuttle bike (p<0.1) compared to the
others (69.5% correctly classified). The track cyclists could be correctly classified for 70.4%. They were significantly
taller and heavier and outperformed to the others on the 30m sprint test. Significant lower performances were
observed for the shuttle bike test and planking test. The cyclo-cross sub-sample (83.1% correctly classified) had an
anthropometric profile of being smaller and lighter compared to the others. Better performances were registered
on the balance beam and shuttle bike test, while they seem to perform slower on the sprint test (p<0.1). 89.3% of
the MTB cyclists could be correctly classified to their actual discipline. The MTB cyclists seem to perform better on
the shuttle bike test (p<0.1). An overview of these findings is displayed in Table 2.2..
Discussion
The current study investigated the athletic profiles of high performing adolescent and young adult cyclists
in four cycling disciplines. Based on their performance on the test battery, a discriminant analysis could correctly
62
classify 80.7% of the young adult cyclists, but only 50.1% of the adolescents. Nevertheless, for both adolescents and
young adults, track cyclists seemed to be characterised by higher sprint speed and better motor coordination, while
mountain bikers seemed to outperform the other disciplines in a shuttle bike test.
Discriminating the cycling disciplines
Due to the distinct differences in competition format at professional level, we expected to find differences
in athletic profiles between the different cycling disciplines (Foley et al., 1989; Padilla et al., 2001; Novak &
Dascombe, 2014). The present discriminant models indeed suggest a high degree of discriminative power of the
test battery in the young adult group, correctly classifying 80.7% of high performing athletes from the various
cycling disciplines. These results therefore could provide reference values for the athletic profiles of young adult
high performance cyclists in these four disciplines. In contrast to the young adult group however, the adolescent
age group was featured by an overlap of roughly 50% between the four disciplines. The performance on the current
test battery could not adequately discriminate between the various disciplines. This is in line with studies in other
sports such as combat sports, which showed that profiles of athletes become more specific as they grow older (Li
et al., 2018a).
The result that the four disciplines could not be discriminated yet, based on the athletic profiles of high
performing adolescents, underlines the dynamic developmental pathway, which suggests that there is no need for
early specialization (Baker, 2003; Gulbin et al., 2013a). The findings of the current study seem to be in line with the
strategy of “specialised sampling” (Sieghartsleitner et al., 2018). “Specialised sampling ” addresses a focus on the
primary sport domain and the positive effect of participating within diversified forms of domain-specific free play.
Accordingly, we should expect that encouraging free play and sports participation in the different cycling disciplines
is promising for later success within cycling. Additionally, participation in other sports with a similar profile may
also be beneficial to help maintain sporting involvement, motivation and reduce injuries and drop-out (Pion et al.,
2021). Anecdotally, several cyclists have matured into champions after practicing multiple other sports or
disciplines in their youth career (i.e. Peter Sagan, Cadel Evans, Mark Cavendish, Greg van Avermaet) (Nimmerichter,
2018). However, further research is needed to investigate the benefits of applying the “specialised sampling”
strategy within cycling. Especially the need for high focus on the primary sport domain at a young age is
questionable.
63
Profiles of the four cycling disciplines
Although the discriminating power of the test battery was much lower for the adolescents than for the
young adults, in both groups results indicate similar differences between the disciplines. Based on these
differences, excelling in some characteristics could be linked to talent.
In both the adolescent and the young adult group, track cyclists were found to outperform the other disciplines in
explosiveness (sprint). This finding is in line with the nature of track cycling, characterised by shorter events at high
intensity, therefore requiring more explosive power (Jeukendrup et al., 2000; Craig & Norton, 2001). In addition, for
the young adults, motor coordination also seems to be an important and differentiating performance characteristic
for track cyclists. Possibly, good motor coordination is beneficial for the dynamic nature of track cycling, with
regular changes between high and low speed (while using a single gear), and the use of the tilted track. These
circumstances provide a challenge with respect to motor learning. As such this result is in line with the idea that
generic motor coordination is a cornerstone for learning new motor skills(Clark & Metcalfe, 2002).
In the young adult group, high performance cyclo-cross athletes were characterised mostly by a different
body composition. On average, cyclo-cross athletes were smaller and lighter than the other disciplines. However,
this was not evident yet in the younger group. Cyclo-cross cyclists might benefit from these characteristics as they
are required to dismount and remount their bikes throughout a race, perform rapid changes in intensity and
technical elements on the course (Carmichael et al., 2017). Whilst previous research indicates that anthropometric
differences should be evident for each discipline (Foley et al., 1989; McLean & Parker, 1989; Lee et al., 2002;
Impellizzeri et al., 2008; Peinado et al., 2011), the body composition of athletes in the three other cycling disciplines
were similar in the current study. It should be noted, however, that the current test battery only includes the basic
anthropometric measurements. Using additional measures (e.g., skinfolds) can provide a more detailed indication
of an athletic body composition, and reveal other differences between disciplines.
Already from adolescence, performance on the shuttle bike test seems to be a distinctive variable especially
for MTB and for cyclo-cross. These two disciplines outperform road and track cyclists on this test, with MTB also
outperforming cyclo-cross. As the shuttle bike test requires agility and bike handling skills it reflects the
characteristics of MTB and cyclo-cross (Impellizzeri & Marcora, 2007; Novak & Dascombe, 2014; Carmichael et al.,
2017).
In spite of the evident differences between the other disciplines, the profile of road cyclists overlapped
significantly with the other three disciplines, both for adolescents as for young adults. This might be due to the fact
that road cycling is regularly combined with other disciplines. The seasonal schedule of road cycling is compatible
with the competition season for the three other disciplines, resulting in a lot of cyclo-cross, track and MTB cyclists
combining two disciplines. Furthermore, apart from combining disciplines, no specific information on the emerging
64
specialisation in this group (sprinting, climbing, etc.) was included. In other words, the current analysis did not take
into account the multiple profiles, whereas there are specific profiles for cyclo-cross, track cycling, and for MTB,
there might be multiple profiles within road cycling(Padilla et al., 2001; Impellizzeri et al., 2008). This may explain
the high inter-variability within road cycling and the relative similarity with the other disciplines (Mujika & Padilla,
2001; Peinado et al., 2011).
Practical implications
The present cross-sectional study provides reference values for road, track, cyclo-cross and MTB high
performance cyclists. These objective, scientifically substantiated data, may benefit coaches and federations when
evaluating and comparing the current performances of their young adult athletes. This adds opportunities for
monitoring and guiding athletes through their talent development. Furthermore, talent orientation advice can be
given to the adolescent cyclists (12.0 – 15.9 years) who gradually gain more experience on the bike. This is supported
by the finding that, although there is no discrimination possible between the disciplines, the separate athletic
profiles of the adolescents revealed performance characteristics that are similar to those of the young adult group
(≥ 16.0 years) for each discipline. Exploring different disciplines that best reflect the individual’s talent make-up
(see Table 2.2.), may prepare the adolescent cyclists in the discipline(s) that is (are) the most suitable on the long
term (early diversification). Since the disciplines can be more accurately differentiated from each other from the
age of 16 onwards, in combination with the fact that the physiological and performance characteristics are fully
expressed at the end of growth, better advice can be provided for specialization in their main discipline from that
age on.
Although the measurement of peak power output and ventilatory/lactate thresholds could potentially
improve the discriminative power of the test battery, these tests are more expensive and cumbersome to carry out.
Applying these tests for large groups of young potential cyclists is therefore problematic. The current generic test
battery however is cheap and relatively easy to administer in large groups and could be used as a talent orientation
tool for adolescent and young adult cyclists. In the future it might also be valuable to include test data from other
domains relevant for talent identification such as executive functions or psychological characteristics and to test
whether this test battery can discriminate between future achievers and future non-achievers. However, to test the
predictive power of this test battery, a longitudinal/retrospective study is needed based on a large number of
participants.
65
Limitations
Although this study was able to reveal novel and useful information about the athletic profiles of adolescent
and young adult high performance cyclists, it is important to consider some limitations. First, the current study
divided the athletes into four groups based on the disciplines in which the athletes were already active. Although
the choice of the discipline is usually supported by the coaches’ expertise, this classification is based on the
assumption that everyone practices the discipline that fits his athletic profile. Due to natural (de-)selection, the
young adult group consisted most likely of more talented cyclists than the adolescent group. Longitudinal research
is recommended to determine to what extent the superior results on several of the non-sport-specific and cycling
specific tests are due to natural (de-)selection, development and/or individual response to specific training
(Vaeyens et al., 2008). A second limitation of the current study is that it did not control for earlier sport
diversification of the participants. Using logs of deliberate practice would have given a more detailed picture of
training history and volume. Further research is needed to investigate the influence of earlier sport diversification
on talent orientation at a later stage.
Conclusion
The present study is the first to document the anthropometric, physical, motor coordination, and cycling
specific profile of young athletes from four different cycling disciplines. The results indicate that athletic profiles
for different cycling disciplines only clearly emerge when athletes reach young adult level, with differences evident
for anthropometry, physical, motor coordination, and cycling specific test(s). These findings indicate that such a
test battery may provide valuable orientation and development information to coaches and federations in late
specialisation sports such as cycling. While these initial results support the theory of early diversification for
athletes amongst cycling disciplines, it warrants further research to explore the advantages and disadvantages of
such practices.
Acknowledgement
The authors wish to thank Cycling Vlaanderen for their collaboration. Special thanks to Koen Beeckman
(technical director of topsport) and Jan Vancompernolle (coach of the Flemish topsportschool of cycling) for the
realization of this project.
66
References
Asplund, C., & Ross, M. (2010). Core stability and bicycling. Current sports medicine reports, 9(3), 155-160.
doi:10.1249/JSR.0b013e3181de0f91
Baker, J. (2003). Early specialization in youth sport: A requirement for adult expertise? High ability studies, 14(1),
85-94.
Balyi, I., & Hamilton, A. (2004). Long-Term Athlete Development: Trainability in Childhood and Adolescence. Olympic
coach, 16(1), 4-9.
Carmichael, R., Heikkinen, D., Mul-lin, E., & McCall, N. (2017). Physiological response to cyclocross racing. Sports
Exercise Medicine Open Journal, 3(2), 74-80.
Clark, J. E., & Metcalfe, J. S. (2002). The mountain of motor development: A metaphor. Motor development: Research
and reviews, 2(163-190), 183-202.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, N.J.: L. Erlbaum Associates.
Côté, J., Lidor, R., & Hackfort, D. (2009). ISSP position stand: To sample or to specialize? Seven postulates about youth
sport activities that lead to continued participation and elite performance. International Journal of Sport
and Exercise Psychology, 7(1), 7-17. doi:10.1080/1612197x.2009.9671889
Council of Europe, E. (1988). EUROFIT : handbook for the EUROFIT tests of physical fitness: Rome : Council of Europe.
Committee for the Development of Sport.
Craig, N. P., & Norton, K. I. (2001). Characteristics of track cycling. Sports Medicine, 31(7), 457-468.
Crane, J., & Temple, V. (2015). A systematic review of dropout from organized sport among children and youth.
European physical education review, 21(1), 114-131.
Foley, J., Bird, S., & White, J. (1989). Anthropometric comparison of cyclists from different events. British Journal of
30(2), 296-307. doi:10.1123/pes.2017-0009
Gulbin, J., Weissensteiner, J., Oldenziel, K., & Gagné, F. (2013). Patterns of performance development in elite athletes.
European Journal of Sport Science, 13(6), 605-614.
Hohmann, A., Siener, M., & He, R. (2018). Prognostic validity of talent orientation in soccer. German Journal of Exercise
and Sport Research, 48(4), 478-488.
Impellizzeri, F., Ebert, T., Sassi, A., Menaspa, P., Rampinini, E., & Martin, D. (2008). Level ground and uphill cycling
335-341.
Impellizzeri, F. M., & Marcora, S. M. (2007). The physiology of mountain biking. Sports Medicine, 37(1), 59-71.
Jeukendrup, A. E., Craig, N. P., & Hawley, J. A. (2000). The bioenergetics of world class cycling. Journal of Science and
Medicine in Sport, 3(4), 414-433.
Lee, H., Martin, D. T., Anson, J. M., Grundy, D., & Hahn, A. G. (2002). Physiological characteristics of successful mountain
bikers and professional road cyclists. Journal of sports sciences, 20(12), 1001-1008.
Li, P., De Bosscher, V., Pion, J., Weissensteiner, J. R., & Vertonghen, J. (2018). Is international junior success a reliable
550-559.
doi:10.14198/jhse.2016.111.03
67
Louis, J., Billaut, F., Bernad, T., Vettoretti, F., Hausswirth, C., & Brisswalter, J. (2012). Physiological demands of a
simulated BMX competition. International journal of sports medicine, 10, 0032-1327657.
McLean, B. D., & Parker, A. W. (1989). An anthropometric analysis of elite Australian track cyclists. Journal of sports
sciences, 7(3), 247-255.
Menaspa, P., Rampinini, E., Bosio, A., Carlomagno, D., Riggio, M., & Sassi, A. (2012). Physiological and anthropometric
characteristics of junior cyclists of different specialties and performance levels. Scandinavian Journnal of
Medicine and Science in Sports, 22(3), 392-398. doi:10.1111/j.1600-0838.2010.01168.x
Menaspa, P., Sassi, A., & Impellizzeri, F. M. (2010). Aerobic fitness variables do not predict the professional career of
doi:10.1111/j.1600-0838.2010.01280.x
Padilla, S., Mujika, I., Orbananos, J., Santisteban, J., Angulo, F., & Goiriena, J. J. (2001). Exercise intensity and load during
mass-start stage races in professional road cycling. Medicine & Science in Sports & Exercise, 33(5), 796-
802.
489.
Pion, J., Fransen, J., Lenoir, M., & Segers, V. (2014). The value of non-sport-specific characteristics for talent
orientation in young male judo, karate and taekwondo athletes. Archives of Budo, 10(1), 147-154.
doi:10453/94313
Pion, J., Segers, V., Fransen, J., Debuyck, G., Deprez, D., Haerens, L., Vaeyens, R., Philippaerts, R. M., & Lenoir, M. (2015).
Prätorius, B., & Milani, T. (2004). Motorische Leistungsfähigkeit bei Kindern: Koordinations-und
Gleichgewichtsfähigkeit: Untersuchung des Leistungsgefälles zwischen Kindern mit verschiedenen
Sozialisationsbedingungen. Deutsche Zeitschrift für Sportmedizin, 55(7/8), 172-176.
68
Tong, T. K., Wu, S., & Nie, J. (2014). Sport-specific endurance plank test for evaluation of global core muscle function.
Physical Therapy in Sport, 15(1), 58-63. doi:10.1016/j.ptsp.2013.03.003
Vandorpe, B., Vandendriessche, J., Lefèvre, J., Pion, J., Vaeyens, R., Matthys, S. a., Philippaerts, R., & Lenoir, M. (2011).
The KörperkoordinationsTest für Kinder: reference values and suitability for 6–12‐ year‐ old children in
0838.2009.01067.x
Vandorpe, B., Vandendriessche, J., Vaeyens, R., Pion, J., Matthys, S., Lefevre, J., Philippaerts, R., & Lenoir, M. (2012a).
Relationship between sports participation and the level of motor coordination in childhood: a longitudinal
approach. Journal of Science and Medicine in Sport, 15(3), 220-225.
Vandorpe, B., Vandendriessche, J. B., Vaeyens, R., Pion, J., Lefevre, J., Philippaerts, R. M., & Lenoir, M. (2012b). The value
69
Supporting information
Addendum a
Table 2.4.: Detailed information about the bike formats used for the shuttle bike test and maximal cadence test.
Format 1 Format 2
Frame Type Haro Annex Mini BMX UMF part of Merida Pro XL
Toptube length (cm) 52 45
Handlebar Handlebar width (cm) 50 70
Tyre Type Kenda K1054 Kenda Krackpot (50 406)
Shuttle Bike
Tyre size (in inches ") 20x1-3/2 20x1,95

Tyre pressure (bar) 4 4
Tyre threaded or Slick Threaded Threaded
Bike Total bicycle weight (kg) 7,5 10,5
Front area cog size 36 44
Back area cog size 14 16
Crank length (mm) 155 172,5
Surface Type Tempomark II Tempomark II
Format 1 for athletes <1,55m body height Format 2 for athletes >1,55m body height
Frame Type Flanders SR1 Ultimate Racing SRM ergometer
Frame length (cm) 50 -
Toptube length (cm) 40 -
Handlebar Handle bar width (cm) 38 41
Tyre Type Bontrager R2 plus 700x23cc -
Cadence test
Tyre diameter (mm) 700 -

Tire thickness (mm) 23 -
Pressure (bar) 8 -
Bike Front area cog size 34 53
Back area cog size 25 15
Crank length (mm) 155 172,5
Ergometer Type Tackx Flow SRM
Resistance gradient -4 1
70
Addendum b
Table 2.5.: Detailed overview of the results from the multivariate analyses of covariance (MANCOVA) in the ≥ 16 year of age group with maturity ratio, calendar age and training history as
confounding variables.
Covariate MANCOVA Covariate MANCOVA Covariate MANCOVA
≥ 16 years Maturity ratio p-value Maturity ratio p-value Age p-value Age p-value Training history p-value Training history p-value
F-value F-value F-value F-value F-value F-value
ANTROPOMETRICS 7.307 <0.001 1.846 0.046 4.737 0.002 1.845 0.046 0.570 0.685 1.595 0.099
Height (cm) 1.333 0.253 1.401 0.252 1.190 0.280 1.377 0.259 0.292 0.591 1.905 0.139
Sitting Height (cm) 10.100 0.002 0.493 0.689 3.969 0.051 0.524 0.667 0.177 0.676 1.357 0.265
Weight (kg) 15.923 <0.001 2.701 0.054 12.850 0.001 2.574 0.063 1.185 0.281 2.547 0.065
BMI (kg/m²) 23.655 <0.001 3.728 0.016 17.403 <0.001 3.337 0.025 1.048 0.310 2.172 0.101
PHYSICAL CHARACTERISTICS 1.189 0.325 1.771 0.026 1.696 0.130 1.836 0.020 2.087 0.061 1.810 0.022
Planking (sec) 0.006 0.938 1.387 0.256 <0.001 0.986 1.378 0.258 0.228 0.635 1.585 0.203
SBJ (cm) 1.358 0.249 0.141 0.935 1.680 0.200 0.129 0.942 3.219 0.078 0.309 0.819
SAR (cm) 0.252 0.617 0.096 0.962 0.005 0.945 0.041 0.989 0.337 0.564 0.022 0.996
Sprint5m (sec) 2.533 0.117 9.419 <0.001 3.374 0.071 9.800 <0.001 2.957 0.091 9.115 <0.001
Sprint10m (sec) 1.202 0.277 11.137 <0.001 1.201 0.278 11.119 <0.001 1.359 0.248 11.026 <0.001
Sprint20m (sec) 0.051 0.822 8.833 <0.001 0.012 0.914 8.641 <0.001 0.048 0.827 9.143 <0.001
Sprint30m (sec) 0.266 0.608 5.783 0.002 0.666 0.418 5.545 0.002 0.569 0.454 6.266 0.001
MOTOR COORDINATION 1.667 0.184 2.236 0.023 1.842 0.150 2.123 0.032 5.879 0.001 2.874 0.004
Sum Moving Sideways (N) 3.150 0.081 0.238 0.870 3.671 0.060 0.204 0.893 5.335 0.024 0.648 0.587
Sum Jumping Sideways (N) 1.098 0.299 1.702 0.177 1.855 0.178 1.449 0.238 2.480 0.121 2.510 0.068
Sum Balance Beam (N) 3.707 0.059 1.929 0.135 4.100 0.047 1.993 0.125 15.623 <0.001 2.586 0.062
CYCLING SPECIFIC TESTS 2.217 0.119 2.457 0.029 2.089 0.134 2.449 0.030 0.673 0.515 3.310 0.005
Shuttle bike (sec) 0.599 0.443 5.036 0.004 0.784 0.380 5.006 0.004 0.258 0.614 5.225 0.003
Maximal cadence (N) 3.437 0.069 0.071 0.975 2.966 0.091 0.075 0.973 1.211 0.276 1.495 0.227
71
Addendum c
Table 2.6.: Detailed overview of the results from the multivariate analyses of covariance (MANCOVA) in the group between 12.0 and 15.99 years of age with maturity ratio, calendar age and
training history as confounding variables.
Covariate MANCOVA Covariate MANCOVA Covariate MANCOVA
12.0 – 15.99 years Maturity ratio p-value Maturity ratio p-value Age p-value Age p-value Training history p-value Training history p-value
F-value F-value F-value F-value F-value F-value
ANTROPOMETRICS 590.395 <0.001 2.315 0.007 60.242 <0.001 3.763 <0.001 1.575 0.182 4.198 <0.001
Height (cm) 779.827 <0.001 2.135 0.096 204.425 <0.001 7.076 <0.001 1.087 0.298 9.622 <0.001
Sitting Height (cm) 2189.486 <0.001 3.328 0.020 237.178 <0.001 9.573 <0.001 0.029 0.865 11.843 <0.001
Weight (kg) 605.480 <0.001 4.222 0.006 174.205 <0.001 10.081 <0.001 0.246 0.621 11.789 <0.001
BMI (kg/m²) 94.796 <0.001 4.809 0.003 43.593 <0.001 7.758 <0.001 0.052 0.821 9.285 <0.001
PHYSICAL CHARACTERISTICS 23.781 <0.001 1.235 0.202 25.643 <0.001 1.044 0.405 0.984 0.449 1.104 0.332
Planking (sec) 0.944 0.332 1.666 0.175 2.158 0.143 1.649 0.179 <0.001 0.987 1.417 0.238
SBJ (cm) 103.193 <0.001 3.201 0.024 103.936 <0.001 0.955 0.415 0.503 0.479 0.342 0.795
SAR (cm) 33.342 <0.001 1.957 0.121 25.205 <0.001 1.046 0.373 1.326 0.251 0.537 0.658
Endurance (beep test/ mins) 51.857 <0.001 2.274 0.081 82.185 <0.001 1.316 0.270 0.485 0.487 0.187 0.905
Sprint5m (sec) 24.694 <0.001 1.161 0.325 35.476 <0.001 1.956 0.121 0.080 0.777 2.706 0.046
Sprint10m (sec) 51.517 <0.001 0.432 0.730 64.408 <0.001 1.121 0.341 0.234 0.629 2.166 0.093
Sprint20m (sec) 103.893 <0.001 0.517 0.671 122.251 <0.001 0.896 0.444 0.581 0.447 1.923 0.126
Sprint30m (sec) 127.016 <0.001 0.515 0.673 142.847 <0.001 0.810 0.489 0.247 0.619 2.087 0.103
MOTOR COORDINATION 10.420 <0.001 2.233 0.019 11.500 <0.001 2.652 0.005 0.791 0.500 2.787 0.003
Sum Moving Sideways (N) 23.060 <0.001 0.697 0.554 23.609 <0.001 0.800 0.495 2.100 0.149 0.945 0.419
Sum Jumping Sideways (N) 19.030 <0.001 1.755 0.157 27.068 <0.001 1.468 0.224 1.256 0.264 0.464 0.707
Sum Balance Beam (N) 0.402 0.526 4.886 0.003 1.778 0.184 5.389 0.001 1.151 0.284 4.631 0.004
CYCLING SPECIFIC TESTS 43.498 <0.001 4.640 <0.001 31.695 <0.001 5.281 <0.001 3.668 0.027 5.366 <0.001
Shuttle bike (sec) 2.680 0.103 9.159 <0.001 6.160 0.014 9.738 <0.001 6.548 0.011 7.771 <0.001
Maximal cadence (N) 86.988 <0.001 0.767 0.513 62.979 <0.001 0.420 0.739 2.277 0.133 1.218 0.304
72
Chapter 2 : Prediction of future performance in
road cycling
Study 2 : Is motor coordination the key to success in youth cycling?
Mireille Mostaert¹*, Pieter Vansteenkiste¹, Felien Laureys¹, Nikki Rommers2, Johan Pion¹,3,
Frederik J.A. Deconinck¹, Matthieu Lenoir¹
¹ Department of Movement and Sport Sciences, Ghent University, Ghent, Belgium

2
Department of Movement and Sports Sciences, Vrije Universiteit Brussel, Brussels, Belgium
3
Sport and Exercise Studies, HAN University of Applied Sciences, Nijmegen, The Netherlands
Submitted to International Journal of Sports Physiology and Performance

Abstract
To evaluate the predictive value of a (non-)sport-specific test battery on the future success of young cyclists, test
scores were compared to the competition performances two to three years later. Three motor coordination, five
physical performance, and two cycling-specific measurements were collected in 111 U15 (13.0 – 14.9 years) and 67
U17 (15.0 – 16.9 years) male road cyclists. In addition, maturity status and relative age were assessed. National and
provincial competition results two to three years later, in the U17year2 and U19year2 categories, were submitted to two
separate four-stage hierarchical regressions. The results of the model of the U15 group revealed that maturity,
relative age, motor coordination, physical performance, and cycling-specific performance accounted for 22.6% of
the variance in competitive success. For the U15 category, only maturity and motor coordination were significant
predictors of competitive success in the U17year2 category. Maturity and motor coordination each uniquely explained
±5% of the variance. However, for the U17 group, neither maturity, relative age, motor coordination, physical
performance, nor cycling-specific performance could predict competitive success in the U19year2 category. The
current study underlines the importance of general motor coordination as a building block necessary for optimal
development in youth cycling. However, considering the lack of predictive value from the U17 category onwards,
other features may determine further development of youth athletes. Nevertheless, it argued why athletes need
to possess a minimum level of all physical, motor coordination, and cycling-specific characteristics to experience
success and enjoyment in their sport.
Keywords: generic test battery, youth, road cycling, motor coordination, talent prediction
78
Introduction
As is the case for all sports, in cycling, sports institutions and (inter)national federations are responsible
for providing athletes the possibility to enjoy their sport in a recreational and/or competitive way. They are facing
the challenge of 'developing healthy, capable and resilient young athletes, while attaining widespread, inclusive,
sustainable and enjoyable participation and success for all levels of individual athletic achievement' (Bergeron et
al., 2015). An effective integrated approach to athlete development in competitive athletes begins with an
assessment of potential talent, followed by a structured training/development programme (Vaeyens et al., 2009;
Gulbin et al., 2013a; Bergeron et al., 2015). However, identifying potential in youth athletes is difficult and requires
further insights into the performance characteristics and athletic profile of cyclists.
Within road cycling, competition results are frequently used as the main criterion to identify and select
youth cyclists as potential cyclists (Schumacher et al., 2006; Svendsen et al., 2018; Mostaert et al., 2021b). For
example, the predictive value of youth competition results from the U17, U19, and U23 category onwards was
recently studied and confirmed, with the highest success rates reported in the second year of the categories
(Svendsen et al., 2018; Mostaert et al., 2021b; Mostaert et al., 2022). However, competitive success in the U15
category could not differentiate future achievers from future non-achievers, nor predict future performance at the
senior elite level (Mostaert et al., 2021b). Consequently, since competition results alone do not seem to be a
sufficient indicator of future success in the U15 category, other factors may contribute to competition outcomes in
the younger categories. These findings are in line with the results of Staff et al. (2021), who indicated that the
traditional talent identification method, using achievement measures, expert assessment of performance by
coaches, and talent scouts at an early age, is inferior to a detection talent identification methodology based on a
range of generic physical and skill based tests between the ages of 11 to 16 years. This underlines the need for a
multidimensional approach to talent identification in cycling.
Morphological and physiological features have been commonly tested in function of talent identification
in road cycling (Phillips & Hopkins, 2020). In both youth (U17-U19) and professional (U23 – elite level) cyclists,
success is related to high levels of physiological characteristics (Lucía et al., 2001; Perez-Landaluce et al., 2002;
Atkinson et al., 2003; Van Schuylenbergh et al., 2004; Aagaard et al., 2011; Rodríguez-Marroyo et al., 2011; Menaspa
et al., 2012; Sanders et al., 2017; Svendsen et al., 2018; Van Erp & Sanders, 2020; Gallo et al., 2021). However, the
traditional physiological measures of aerobic fitness that differentiate high- from low-level youth (U19) cyclists do
not predict their future competitive level in adulthood (Menaspa et al., 2010). The lack of predictive validity of these
measures of aerobic fitness is probably mainly due to the multifactorial nature of cycling performance (Menaspa
et al., 2010). Besides VO2max, which is considered a prerequisite in road cycling, other sport-specific tests measuring,
e.g., technical skills should be further investigated in youth cycling (Menaspa et al., 2010; Phillips & Hopkins, 2020).
79
Skill-based tests, such as general motor coordination tests seem able to discover the athlete's real
potential to pick up new technical skills (Vandorpe et al., 2012b; Pion et al., 2015a; Rommers et al., 2019). In other
sports such as gymnastics, soccer, and volleyball, this low-resource assessment approach has been related to future
success in sports (Leyhr et al., 2018). Recently, Mostaert et al. (2020a) showed that general motor coordination
(measured using the KTK3 test battery), in combination with measures of anthropometry, physical performance,
and cycling-specific tests, could accurately (80,7%) discriminate between young adult athletes of different cycling
disciplines. However, it is not clear if motor coordination could also contribute to talent identification in youth
cycling. Although motor coordination has been shown to benefit coaches and federations when evaluating
young/adult athletes (Pion, 2015), future research is needed within youth cycling athletes.
It is generally known that the selection process of potential youth athletes is biased by the young athlete's
maturational timing and relative age (Cumming et al., 2018), especially in sports that strongly rely on
characteristics such as speed, strength, and power (Malina et al., 2015). Children who mature earlier and/or are
relatively older than their peers, benefit from a margin in body size and physiological development, and could
therefore falsely be identified as more potential than their peers (Baker et al., 2010b; Malina et al., 2015). However,
after puberty (i.e. around the age of 17 years for boys), maturity and age-related differences are reduced and largely
eliminated in non-athletes and athletes (Malina et al., 2004; Hollings et al., 2014; Malina et al., 2015). Similar results
were found for road cycling as the relative age effect (RAE) was present in the U15 category, but seems to fade
away as athletes reach the U19 category (Mostaert et al., 2021b). In the U19 category, the athletes reached
adulthood which may explain that the RAE on inter-individual differences may decrease (Musch & Grondin, 2001;
Vaeyens et al., 2009). Becoming physically and mentally stronger at a later age might be of a greater importance
in endurance sports to develop into successful adult elite athletes (Schumacher et al., 2006; Mostaert et al., 2021b).
Therefore, it is suggested that future research should consider both the RAE and maturation in youth cycling
research, especially when they affect parameters that are crucial in identifying and selecting young athletes.
The purpose of this study was to examine if a (non-)sport-specific test battery has a predictive value for
the competition results of youth cyclists two to three years later. It is hypothesised that the motor coordination,
physical, and cycling-specific tests are predictors for future performance.
80
Methods
Participants
In collaboration with the Flemish Cycling Federation (Cycling Vlaanderen), five test days per year were
organised to identify and monitor young competitive cyclists. A sample of 620 male road cyclists participated in
the test days between 2015 and 2017. Of those, 372 cyclists were between 13.0 and 16.9 years of age (in the U15 or
U17 category). Next, excluding partial test results and duplicates resulted in a total sample of 306 cyclists (Figure
2.4.). Only the 111 cyclists who performed the test battery between 13.0 - 14.9 years of age (U15), and competed in
at least one of the selected provincial and/or national competitions in the second year of the U17 category (two to
three years later) were included in the first cohort. The second cohort consisted of 67 male road cyclists who
performed the test battery between the age of 15.0 - 16.9 years (U17), and registered at least one competition result
in the second year of the U19 category (two to three years later). The study was conducted in accordance with the
Helsinki Declaration and was approved by the Ethics Committee of the Ghent University Hospital under registration
no. PA 2016/021.
2015-2017
Road cyclist
n = 620 Excluded due to
not between 13 and 17 years
old at time of testing
Tested in U15 Tested in U17 n = 248
n = 209 n = 163
Excluded due to
more than one participation
n = 66
Tested in U15 Tested in U17
n = 169 n = 137
Excluded due to
no competition results in U17 or U19
2017-2019
n = 128
Competing in Competing in
U17year2 U19year2
n = 111 n = 67
Figure 2.4.: Flowchart of the data collection.
81
Procedure and measurements
Prior to testing, anthropometric measurements body height, weight, and sitting height measurements were
assessed to estimate the biological maturity using a sex-specific regression equation (Fransen et al., 2018). Age of
peak height velocity (APHV; years) was applied in the analyses. Additionally, the relative age effect (RAE) was also
included in the statistics. Within road cycling, an annual age-grouping competition strategy from January until
December is used. For the purpose of this study, the athletes were divided into four birth quarters according to
their month of birth in relation to the yearly cut-off date to examine the birth date distribution and relative age
effect (RAE): Q1: January – March, Q2: April – June, Q3: July – September, Q4: October – December. Furthermore, the
cyclists completed 10 (non-)sport-specific tests, consisting of five physical, three motor coordination, and two
cycling-specific measurements. The data were collected at the same indoor venue and assessed by a team of 10 to
12 experienced examiners affiliated with Ghent University. The athletes performed the tests barefooted with the
exception of the sprint, endurance shuttle run, and cycling-specific tests. The barefooted tests were performed in
a randomised circuit manner, followed by a short warm-up (running, acceleration, and active stretching exercises)
in order to perform the sprint, shuttle bike, endurance shuttle run, and maximal cadence test, respectively. To assess
if the (non-)sport-specific test battery could indicate competition performance, race results from national and
provincial competitions were retrospectively collected (Callens, 2002-2011; Archive, 2019).
Motor coordination tests
To evaluate general gross motor coordination, the Körperkoordinationstest für Kinder ( KTK3) test battery
was used (Kiphard & Schilling, 1974; Kiphard & Schilling, 2007; Vandorpe et al., 2011a; Novak et al., 2017). The first
subtest was moving sideways (MS) on a straight-line handling two wooden platforms for 20 seconds (sum of two
attempts). The second subtest was jumping sideway (JS) performing two-legged jumping over a wooden slat for
15 seconds (sum of two attempts). The third subtest was walking backwards (BB) along a balance beam of 6 cm,
4.5 cm, and 3 cm wide (sum of three trials on each balance beam, with a maximum of 8 steps per trail). The KTK3
test battery has a good reliability and validity, shown by a test-retest reliability ranging from 0.80 to 0.95 for the
specific subtests (Vandorpe et al., 2011b; Novak et al., 2017; Mostaert et al., 2021a).
Physical performance tests
The best performance of two trials on the standing broad jump (SBJ), which measures the explosive leg
power of the participants, was registered (to the nearest 1 cm) (Council of Europe, 1988). To assess the hamstring
and lower back flexibility, a sit and reach (SAR) was performed (to the nearest 0.5 cm; the best score of two trials)
(Council of Europe, 1988). Core muscle endurance was determined using an endurance plank test by executing a
82
plank position for as long as possible (one trial, to the nearest 1 s) (Asplund & Ross, 2010; Tong et al., 2014) . The
best time of two trials on the 30-meter sprint test, which measures speed, was registered (accuracy of 0.001 s,
MicroGate Racetime2 chronometry and Polifemo Light Photocells; MicroGate, Italy) (Matthys et al., 2013).
Cardiovascular endurance was obtained by a 20-meter endurance shuttle run test (ESHR), to the nearest 0.5 min
(Council of Europe, 1988). The test-retest reliability and validity has been described elsewhere (Council of Europe,
1988; Matthys et al., 2013)
Cycling-specific tests
The procedures of the cycling-specific shuttle bike and maximal cadence test are in accordance with the test
procedures described in the study of Mostaert et al. (2020a). To assess agility and cycling skills, a shuttle bike test
was performed on a BMX bicycle (2 different sizes depending on body height). The riders needed to complete an
eight-shaped circuit as quickly as possible without hitting the markers and without losing control over the bicycle.
The shuttle bike test was performed twice, with the best time used for analyses. The time was recorded using time
gates (accuracy of 0.001 s, MicroGate Racetime2 chronometry and Polifemo Light Photocells; MicroGate, Italy). The
shuttle bike test has an intra-class correlation of 0.763 (95% CI 0.440 – 0.899) (Mostaert et al., 2020a).
Agility was assessed using a maximal cadence test, performed on a calibrated Flanders SR1 Ultimate racing
device or standardised SRM cycle ergometer (SRM - Schoberer Rad Messtechnik; Germany). Athletes with a body
height of > 1,55 m performed the test on the SRM cycle ergometer. The cyclists began in a flying start (resistance 1)
while seated and with their hands placed on the drop handlebars. For ten seconds the athlete had to perform their
maximal cadence. The highest score (peak frequency) was recorded for one attempt. The maximal cadence test has
an intra-class correlation of 0.936 (95% CI 0.863 – 0.970) (Mostaert et al., 2020a).
Competition results
As an outcome variable, the competitive performance level was determined. Results from four national
(Belgian championship road and time trial, and two national climb races) and two provincial competitions
(provincial championship road and time trial) were retrospectively gathered (Callens, 2002-2011; Archive, 2019).
The competitive performances of the first cohort were determined at the time the cyclists were active in the U17year2.
For the second cohort, the competitive performances in the U19year2 were established. To set the actual overall
performance level of the cyclists, scores were given according to the place achieved for each competition (in line
with Mostaert et al. (2022)). As participation in national competitions is only authorised for the best competitors
from each province, the national championships results weighed double compared to the provincial championships
(Table 2.7.).
83
Table 2.7.: Scores awarded to performance results at provincial and national championships.
Placement Scores Provincial Championships Scores National Championships

1 20 40
2 18 36
3 16 32
4 14 28
5 12 24
>5 10 20
>10 8 16
>15 6 12
>20 4 8
>30 2 4
DNF/DNS* 1 2
*
DNF = did not finish; DNS = did not start
Data analysis
The data were analysed with SPSS Statistics version 25. Descriptive statistics were assessed and there were
no outliers in the data. Standardised z-scores were calculated with the following formula, using means and
standard deviations for each age group (13, 14, 15, and 16 years) and test item (Pion, 2015):
(𝑟𝑎𝑤𝑠𝑐𝑜𝑟𝑒𝑡𝑒𝑠𝑡 − 𝑚𝑒𝑎𝑛𝑡𝑒𝑠𝑡 )
𝑧 − 𝑠𝑐𝑜𝑟𝑒𝑡𝑒𝑠𝑡 =
𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛𝑡𝑒𝑠𝑡
Overall mean z-scores were calculated for all tests and averaged for the three performance categories; i.e.,
physical performance (i.e. z-scores from SBJ, SAR, plank test, 30m sprint, and ESHR), motor coordination
(i.e. z-scores from jumping sideways, moving sideways, and balance beam) and cycling-specific performance (i.e.
z-scores from shuttle bike and maximal cadence). Prior to conducting a hierarchical multiple regression, the
relevant assumptions of this statistical analysis, such as multicollinearity, normality of residuals, homoscedasticity,
linearity, and independence of errors, were tested (Jeong & Jung, 2016). A four-stage hierarchical regression
analysis was performed separately for the U15 and U17 groups to explain the variance of the outcome variable,
competitive performance level. APHV and RAE were entered at stage one of the regression. Motor coordination was
included at stage two, physical performance at stage three, and cycling-specific performance at stage four.
Statistical significance, size, direction of the standardised regression coefficients, and change in R-square were
reported. For all analyses, the significance level was set at a p-value of < 0.05.
84
Results
Performance prediction in the U17 category
Descriptive statistics are presented as mean ± standard deviation (SD) in Table 2.8.. The four-stage
hierarchical regression statistics are displayed in Table 2.9. and Figure 2.5.. The results revealed that at stage one
the model accounted for 7.8% of the variance in competition results for APHV and RAE. APHV was the only significant
predictor uniquely explaining 7.5% of the variance. Adding motor coordination revealed a significant predictive
relationship between motor coordination and future competitive success (accounted for 20.5% of the variance).
Motor coordination uniquely explained an additional 12.8% of the variance in competitive success. Introducing
physical performance and the cycling-specific tests in the third and final stage, respectively, did not increase the
variance in competitive success. The full model, including APHV and RAE, and all three performance variables, was
statistically significant (F(5,105) = 6.131; p < 0.001) and accounted for 22.6% of the variance in competitive success.
Figure 2.5.: Visual representation of the prediction equation in the U17 category (full model).
85
Table 2.8.: Average performance on the anthropometric, physical, motor coordination, and cycling-specific tests (± SD) of the
road cyclists.
U15 U17
(N=111) (N=67)
Mean SD Mean SD
Age (years) 13.7 0.6 15.5 0.6
Competitive performance (score) 27.4 26.6 20.2 25.1
APHV (years) 13.7 0.7 13.9 0.5
Body height (cm) 164.5 9.8 175.13 7.2
Body weight (kg) 50.5 9.6 60.54 8.1
Motor Coordination (z-score) -0.02 0.79 0.00 0.76

Moving sideways (N) 62 8 65 9
Jumping sideways (N) 87 11 92 10
Balance beam (N) 54 11 54 11
Physical Performance (z-score) 0.00 0.66 -0.01 0.64

Planking (sec) 123 50 140 45
Standing broad jump (cm) 183 21 194 20
Sit and reach (cm) 24.4 7.6 25.9 7.2
Endurance Shuttle Run (min) 9.6 1.7 11.0 1.5
30m Sprint (sec) 4.983 0.299 4.784 0.266
Cycling Specific (z-score) -0.01 0.73 0.01 0.79

Shuttle bike (sec) 12.564 1.476 12.290 1.444
Maximal cadence (N) 200 13 210 12
Birth quarters (N) Q1 Q2 Q3 Q4

U15 36 (32%) 39 (35%) 22 (20%) 14 (13%)
U17 14 (21%) 19 (28%) 17 (25%) 17 (25%)
Maturity status (N) Early Average Late

U15 13 (12%) 90 (81%) 8 (7%)
U17 0 (0%) 65 (97%) 2 (3%)
86
Table 2.9.: Results of the standardised hierarchical regression analyses for the U15 group.
Variable B SE β R² F ∆R² ∆F²

Stage 1 0.078 4.540 * 0.078 4.540 *
Constant 165.907 *** 46.066
APHV -9.836 ** 3.309 -0.275
Relative age effect -5.787 9.483 -0.056
Stage 2 0.205 9.204 *** 0.128 17.172 ***

Constant 172.118 *** 42.988
APHV -10.111 ** 3.087 -0.283
Motor coordination 12.146 *** 2.931 0.359
Stage 3 0.215 7.250 *** 0.010 1.310

Constant 158.405 ** 44.566
APHV -9.071 ** 3.214 -0.254
Motor coordination 10.006 ** 3.473 0.296
Physical performance 4.906 4.286 0.121
Stage 4 0.226 6.131 *** 0.011 1.512

Constant 152.255 ** 44.739
APHV -8.587 ** 3.230 -0.240
Motor coordination 9.006 * 3.559 0.266
Cycling-specific performance 4.463 3.629 0.122
Note: *p<0.05; **p<0.01; ***p<0.001
87
Performance predictions in the U19 category
The four-stage hierarchical regression revealed that neither APHV, RAE, motor coordination, physical
performance, nor cycling-specific performance in the U17 category could predict competitive success in the U19 year2
category (F(5,61)=0.521; p=0.759; R²=0.041). The descriptive statistics and results from the hierarchical regression are
displayed in Table 2.8., Table 2.10., and Figure 2.6., respectively.
Figure 2.6.: Visual representation of the prediction equation in the U19 category (full model).
88
Table 2.10.: Results of the standardised hierarchical regression analyses for the U17 group.
Variable B SE β R² F ∆R² ∆F²

Stage 1 0.013 0.421 0.013 0.421
Constant -52.460 83.750
Maturity ratio 5.154 6.086 0.108
Relative age effect 1.817 11.162 0.021
Stage 2 0.019 0.416 0.006 0.415

Constant -58.071 84.585
Motor coordination 2.664 4.133 0.081
Stage 3 0.021 0.336 0.002 0.113

Constant -67.421 89.597
Stage 4 0.041 0.521 0.020 1.257

Constant 126.310 77.078
Maturity ratio -97.634 69.468 -0.182
Cycling-specific performance -5.682 4.878 -0.178
Note: *p<0.05; **p<0.01; ***p<0.001
89
Discussion
The present study offers a better understanding of the usefulness of a (non-)sport-specific test battery
for young cyclists, to predict competition performance two to three years later. Results showed that a (non-)sport-
specific test battery performed in the U15 accounted for 22.6% of the variance in the competition performance as
a U17 athlete. The findings underline the importance of maturity and general motor coordination in youth cycling,
each of which uniquely explained ±5% of the variance. However, in the U17 category, neither the APHV, RAE, motor
coordination, physical performance, nor cycling-specific performance could predict competitive success in the
U19year2 category.
In the U15 category, the best predictions on future success could be made based on the full model,
consisting of maturity and RAE, motor coordination, physical performance, and cycling-specific performance. These
findings confirm research by Staff et al. (2021), indicating that identifying athletic potential based on a range of
generic, physical, and skill-based tests is beneficial within talent research. The results showed that within youth
road cycling it is important to consider the possible impact of maturity (APHV) on future race results. Earlier mature
athletes are more likely to achieve better competitive results two to three years later, which is in line with the
findings in previous research (Malina et al., 2013). However, caution is warranted for coaches and federations when
selecting and/or excluding athletes at this age, since the advantages that the earlier mature athletes experience
may disappear after puberty (Malina et al., 2015). A possible solution to better predict the impact of the physical
ability on future success is to use bio-banding in youth athletes (Malina et al., 2019). Bio-banding allows a better
comparison between athletes of the same biological age, instead of only comparing athletes based on
chronological age. Based on the findings of Mostaert et al. (2021b), a RAE was expected within the U15 category.
However, within the current study RAE did not contribute to competition performance two to three years later.
Motor coordination was the only performance variable that significantly contributed to future success.
Excelling in motor coordination in the U15 category resulted in a 4.7% higher chance to perform at a high level two
to three years later. Similar results were found in other sports for motor coordination applying the KTK3 test battery
(Vandorpe et al., 2012b; Pion et al., 2015b; Mostaert et al., 2020b). This motor coordination test battery seems to
measure a general trait underlying a wide variety of skills needed in various sports and is sensitive enough to
predict future success (Vandorpe et al., 2012b; Pion, 2015; Höner & Votteler, 2016). To obtain higher motor
coordination levels in athletes, previous studies suggested that athletes may benefit from early diversification
including broad exposure to formal and informal play, practice and games, in both sport-specific and non-sport-
specific ways (Gulbin et al., 2013a; Staff et al., 2021). Staff et al. (2021) indicated that a later engagement in the
development of expertise resulted in faster motor learning and development. This is in line with the previous
findings of Mostaert et al. (2020a), who suggested that there is no need for early specialisation within cycling as
90
the different cycling profiles of high performing adolescents between 12.0 and 15.99 years of age (road, track, cyclo-
cross, and mountain bike) could only be correctly discriminated for 50.1%. Accordingly, we should expect that
encouraging free play, diversified training, and sports participation in the different cycling disciplines is promising
for later success within cycling.
Physical performance and cycling-specific performance did not significantly contribute to the prediction
of competitive success. This finding is in agreement with those of Philippaerts et al. (2008) and Vandorpe et al.
(2012b) who reported that general physical performance characteristics could determine current ability, but are
less suitable as an indication for the potential to excel in the future. Moreover, training, growth, and maturation
may have a possible impact on the physical variables (Malina et al., 2004; Pearson et al., 2006; Vandorpe et al.,
2012b), which might complicate the prediction of future performances. The cycling-specific tests consisted mainly
of skill tests performed on a bicycle. Including cycling-specific tests is similar to other sports federations that focus
on talent identification, who also usually include sport-specific tests (soccer: dribbling tests; gymnastics: rope
skipping and climbing; volleyball: spike jump) in their test battery. The extent to which these tests can distinguish
athletes of different levels and predict future performance is dependent on the sport and the type of test. However,
sport-specific tests have in common that they are more subject to the training history of the athletes (Sattler et al.,
2012). One might argue that, in contrast to the physical performance tests, the cycling-specific tests, similar to the
soccer-specific tests (i.e. dribbling, shooting…), are less affected by maturity (Rommers et al., 2019), because they
mainly consist of tests with a motor component. Further research is necessary to determine if these cycling-specific
tests are influenced by maturity and can measure the current ability of the athletes.
The present study was able to predict short-term cycling success over two to three years in young male
road cyclists (U15). However, when applying this approach to older age groups (i.e. U17 category), it seems that a
modified test battery should be developed to predict future performance. This is in line with the 'sliding
populations' approach of Régnier et al. (1993), suggesting that specific test batteries are needed for different age
groups in order to predict whether athletes can or cannot reach the top level in the following age group. The current
results showed that from the U17 category onwards, neither motor coordination, physical performance, nor cycling-
specific performance could predict competitive success in the U19year2 category. One reason might be the
homogeneity of the current sample since the cyclists were past puberty. In parallel with the 'proficiency barrier'
hypothesis in the research field of motor development (Seefeldt, 1980; Stodden et al., 2013), it could be assumed
that at this stage in the development, the athletes need to possess a minimum level of general motor coordination
to be able to experience success and enjoyment in their sport (De Meester et al., 2018). It may be possible that
when athletes gain this level of development, generic testing no longer holds much or any discriminative value for
athlete selection. This now means, that there is room, however, for individual profiles to vary with the perspective
91
that they have achieved the necessary proficiency levels in these general characteristics. Another reason might be
that the demands of road cycling change. Around the age of 16, cyclists begin to focus on their main discipline, a
process that evolves into highly deliberate practice (Côté et al., 2007). Furthermore, cyclists begin to compete in
national and international competitions, and the distance of the races increase from the U19 category onwards (20
– 35 km in the U15 category, 60 – 80 km in the U17 category, 80 – 140 km in the U19 category) (Svendsen et al.,
2018). Therefore, it is assumed that other features may become performance determining. The systematic review
of Phillips and Hopkins (2020) emphasises the importance of not only focusing on the individual dimensions such
as physiological, morphological, emotional, and cognitive features. Tactical features emerging from the inter-
personal dynamics between cyclists, strategic features related to competition format and the race environment,
and global features related to societal and organisational constraints should also be taken into account (Phillips &
Hopkins, 2020). Future research is necessary to examine a more holistic approach within youth cycling, as the
current test battery seems not sensitive enough to predict the cycling-specific performance at these older ages.
Despite the usefulness of these findings, some limitations should be addressed in future research. First,
the current study consisted of two different cohorts, making predictions of performance two to three years later.
Future research should span longer periods of longitudinal follow-up with one group of cycling athletes. This could
provide additional insights into the development and the predictive value of the test battery on performance.
Second, it would be valuable to include detailed information on training history and volume, using logs of deliberate
practice. Third, considering the findings of Menaspa et al. (2012), who found that juniors (U19) are specialised in the
same way as professional cyclists, it seems valuable to differentiate for the different road cycling specialities
(uphill, flat terrain, all-terrain, and sprint riders) in the future. Finally, there is a significant potential for selection
bias when selecting cyclists for the talent test days. Additionally, only cyclists who registered at least one
competition result in the second year of the U17 or U19 category were included in the study. Consequently, it is
suggested that the riders who experienced less success during competitions may have dropped out from cycling,
and the analyses may have consisted of only the better riders from the U15 and U17 categories.
Practical implications
The current study provides reference values and gives a valid overview of the predictive value and
usefulness of a (non-)sport-specific test battery within road cycling. These findings can aid coaches and federations
when they consider (de)selection in function of talent identification. Selected youth cyclists (U15) should
outperform their peers in one or more of the motor coordination, physical and sport-specific features, taking into
account maturity and RAE. Special attention should be paid to the performances on the motor coordination tests
and the possible impact of maturity. However, the identification of potential talent is complex (Baker & Horton,
92
2004; Gagné, 2013). Therefore, once the cyclists participate in the U17 category, it is suggested to use this test
battery to monitor development rather than to identify potential talent.
Conclusion
The present study showed that a (non-)sport-specific test battery could be used within a U15 cycling
population to predict future success two to three years later. Maturity and general motor coordination were the
only variables that significantly contributed to future success. The findings emphasise that high levels of motor
coordination in cycling can be promising for later success. Considering the reduced predictive value from the U17
category onwards, these tests alone do not seem to be sufficient to reliably predict future success.
Acknowledgements
The authors wish to thank Cycling Vlaanderen for their collaboration. Special thanks to Koen Beeckman (technical
director of Top Sport) and Jan Vancompernolle (training coordinator of track cycling) for the realisation of this
project.
93
References
Aagaard, P., Andersen, J., Bennekou, M., Larsson, B., Olesen, J., Crameri, R., Magnusson, S., & Kjaer, M. (2011). Effects of
resistance training on endurance capacity and muscle fiber composition in young top‐level cyclists.
Scandinavian Journal of Medicine & Science in Sports, 21(6), e298-e307.
Asplund, C., & Ross, M. (2010). Core stability and bicycling. Current sports medicine reports, 9(3), 155-160.
doi:10.1249/JSR.0b013e3181de0f91
Atkinson, G., Davison, R., Jeukendrup, A., & Passfield, L. (2003). Science and cycling: current knowledge and future
directions for research. Journal of sports sciences, 21(9), 767-787.
Baker, J., & Horton, S. (2004). A review of primary and secondary influences on sport expertise. High ability studies,
15(2), 211-228.
Baker, J., Schorer, J., & Cobley, S. (2010). Relative age effects. Sportwissenschaft, 40(1), 26-30.
Bergeron, M. F., Mountjoy, M., Armstrong, N., Chia, M., Côté, J., Emery, C. A., Faigenbaum, A., Hall, G., Kriemler, S., &
Léglise, M. (2015). International Olympic Committee consensus statement on youth athletic development.
British Journal of Sports Medicine, 49(13), 843-851.
Callens, B. (2002-2011). Wielerjaarboek.
Cumming, S. P., Searle, C., Hemsley, J. K., Haswell, F., Edwards, H., Scott, S., Gross, A., Ryan, D., Lewis, J., & White, P.
(2018). Biological maturation, relative age and self-regulation in male professional academy soccer
players: A test of the underdog hypothesis. Psychology of Sport and Exercise, 39, 147-153.
De Meester, A., Stodden, D., Goodway, J., True, L., Brian, A., Ferkel, R., & Haerens, L. (2018). Identifying a motor
proficiency barrier for meeting physical activity guidelines in children. Journal of Science and Medicine in
Sport, 21(1), 58-62.
30(2), 296-307. doi:10.1123/pes.2017-0009
Gagné, F. (2013). The DMGT: Changes within, beneath, and beyond. Talent Development & Excellence, 5(1), 5-19.
1049-1051.
Hollings, S. C., Hume, P. A., & Hopkins, W. G. (2014). Relative-age effect on competition outcomes at the World Youth
and World Junior Athletics Championships. European Journal of Sport Science, 14(sup1), S456-S461.
Höner, O., & Votteler, A. (2016). Prognostic relevance of motor talent predictors in early adolescence: A group-and
individual-based evaluation considering different levels of achievement in youth football. Journal of
sports sciences, 34(24), 2269-2278.
Jeong, Y., & Jung, M. J. (2016). Application and interpretation of hierarchical multiple regression. Orthopaedic
Nursing, 35(5), 338-341.
Kiphard, E. J., & Schilling, F. (2007). Körperkoordinationstest für kinder: KTK: Beltz-Test.
Leyhr, D., Kelava, A., Raabe, J., & Höner, O. (2018). Longitudinal motor performance development in early adolescence
and its relationship to adult success: An 8-year prospective study of highly talented soccer players. PloS
one, 13(5), e0196324.
94
Malina, R., e Silva, M. C., & Figueiredo, A. (2013). Growth and maturity status of youth players. In Science and soccer
(pp. 319-344): Routledge.
Malina, R. M., Cumming, S. P., Rogol, A. D., Coelho-e-Silva, M. J., Figueiredo, A. J., Konarski, J. M., & Kozieł, S. M. (2019).
Bio-banding in youth sports: background, concept, and application. Sports Medicine, 49(11), 1671-1685.
Malina, R. M., Rogol, A. D., Cumming, S. P., e Silva, M. J. C., & Figueiredo, A. J. (2015). Biological maturation of youth
athletes: assessment and implications. British Journal of Sports Medicine, 49(13), 852-859.
Menaspa, P., Rampinini, E., Bosio, A., Carlomagno, D., Riggio, M., & Sassi, A. (2012). Physiological and anthropometric
Mostaert, M., Coppens, E., Laureys, F., D'Hondt, E., Deconinck, F. J., & Lenoir, M. (2021a). Validation of a motor
competence assessment tool for children and adolescents (KTK3+) with normative values for 6-to 19-
year-olds. Frontiers in physiology, 12, 652952. doi:10.3389/fphys.2021.652952
Mostaert, M., Gallo, G., Faelli, E., Ruggeri, P., Delbarba, S., Codella, R., Vansteenkiste, P., & Filipas, L. (2022). Do race
results in youth competitions predict future success as a road cyclist? A retrospective study in the Italian
Cycling Federation. International journal of sports physiology and performance, 1(aop), 1-6.
Mostaert, M., Laureys, F., Vansteenkiste, P., Pion, J., Deconinck, F. J., & Lenoir, M. (2020a). Discriminating performance
profiles of cycling disciplines. International Journal of Sports Science & Coaching, 16(1), 110-122.
doi:10.1177/1747954120948146
Mostaert, M., Pion, J., Lenoir, M., & Vansteenkiste, P. (2020b). A Retrospective Analysis of the National Youth Teams
in Volleyball: Were They Always Faster, Taller, and Stronger? Journal of Strength and Conditioning
Research, 1-7. doi:10.1519/jsc.0000000000003847
Mostaert, M., Vansteenkiste, P., Pion, J., Deconinck, F. J., & Lenoir, M. (2021b). The importance of performance in
competitions as an indicator of future success in cycling. European Journal of Sport Science, 1-10.
doi:10.1080/17461391.2021.1877359
Perez-Landaluce, J., Fernandez-Garcia, B., Rodriguez-Alonso, M., García-Herrero, F., García-Zapico, P., Patterson, A., &
Terrados, N. (2002). Physiological differences and rating of perceived exertion (RPE) in professional,
amateur and young cyclists. Journal of sports medicine and physical fitness, 42(4), 389-395.
Philippaerts, R., Coutts, A., & Vaeyens, R. (2008). Physiological perspectives of the identification and development
of talented performers in sport. In Talent identification and development-the search for sporting
excellence (pp. 49-68): H&P Druck.
95
Pion, J., Fransen, J., Deprez, D., Segers, V., Vaeyens, R., Philippaerts, R. M., & Lenoir, M. (2015a). Stature and jumping
height are required in female volleyball, but motor coordination is a key factor for future elite success.
The Journal of Strength & Conditioning Research, 29(6), 1480-1485. doi:10.1519/JSC.0000000000000778
Pion, J., Lenoir, M., Vandorpe, B., & Segers, V. (2015b). Talent in female gymnastics: a survival analysis based upon
performance characteristics. International journal of sports medicine, 94(11), 935-940.
Rodríguez-Marroyo, J. A., Pernía, R., Cejuela, R., García-López, J., Llopis, J., & Villa, J. G. (2011). Exercise intensity and
load during different races in youth and junior cyclists. The Journal of Strength & Conditioning Research,
25(2), 511-519.
Rommers, N., Mostaert, M., Goossens, L., Vaeyens, R., Witvrouw, E., Lenoir, M., & D’Hondt, E. (2019). Age and maturity
related differences in motor coordination among male elite youth soccer players. Journal of sports
sciences, 37(2), 196-203.
Sanders, D., Abt, G., Hesselink, M. K., Myers, T., & Akubat, I. (2017). Methods of monitoring training load and their
relationships to changes in fitness and performance in competitive road cyclists. International journal of
sports physiology and performance, 12(5), 668-675.
Sattler, T., Sekulic, D., Hadzic, V., Uljevic, O., & Dervisevic, E. (2012). Vertical jumping tests in volleyball: reliability,
validity, and playing-position specifics. The Journal of Strength & Conditioning Research, 26(6), 1532-1538.
Seefeldt, V. (1980). Developmental motor patterns: Implications for elementary school physical education.
Psychology of motor behavior and sport, 36(6), 314-323.
Stodden, D. F., True, L. K., Langendorfer, S. J., & Gao, Z. (2013). Associations among selected motor skills and health-
related fitness: indirect evidence for Seefeldt's proficiency barrier in young adults? Research quarterly for
exercise and sport, 84(3), 397-403.
Tong, T. K., Wu, S., & Nie, J. (2014). Sport-specific endurance plank test for evaluation of global core muscle function.
Physical Therapy in Sport, 15(1), 58-63. doi:10.1016/j.ptsp.2013.03.003
Van Schuylenbergh, R., Eynde, B. V., & Hespel, P. (2004). Correlations between lactate and ventilatory thresholds and
the maximal lactate steady state in elite cyclists. International journal of sports medicine, 25(06), 403-
408.
Vandorpe, B., Vandendriessche, J., Lefèvre, J., Pion, J., Vaeyens, R., Matthys, S., Philippaerts, R., & Lenoir, M. (2011a). The
0838.2009.01067.x
Vandorpe, B., Vandendriessche, J., Lefèvre, J., Pion, J., Vaeyens, R., Matthys, S. a., Philippaerts, R., & Lenoir, M. (2011b).
The KörperkoordinationsTest für Kinder: reference values and suitability for 6–12‐year‐old children in
0838.2009.01067.x
96
Vandorpe, B., Vandendriessche, J. B., Vaeyens, R., Pion, J., Lefevre, J., Philippaerts, R. M., & Lenoir, M. (2012). The value
97
Study 3 : The importance of performance in competitions as an indicator of
future success in cycling
Mireille Mostaert¹*, Pieter Vansteenkiste¹, Johan Pion¹,², Frederik J.A. Deconinck¹, Matthieu
Lenoir¹
¹ Department of Movement and Sports Sciences, Ghent University, Ghent, Belgium
² Sport and Exercise Studies, HAN University of Applied Sciences, Nijmegen, The Netherlands
-
European Journal of Sport Science
Abstract
In sports, outperforming your peers at young age does not guarantee a professional sports career as an adult.
To gain more insight in how performance in youth cycling competitions can be an indicator for future success, the
current study (a) examined differences in success rate between future achievers and future non-achievers,
(b) investigated whether the relative age effect affects these career pathways, and (c) explored to what extent
youth competitive performance can predict success at adult age. The sample consisted of 307 male road cyclists
who achieved at least one top ten result during national and provincial youth competitions (U15, U17, U19). Thirty-
two were classified as future achievers because they reached senior elite level; the remaining 275 were classed as
future non-achievers. Non-parametric analyses of longitudinal data, Kruskal-Wallis, and logistic regressions were
applied. These analyses revealed that the future achievers started to outperform the future non-achievers from
U17 onwards. While the relatively older cyclists have an advantage over other cyclists in U15, this effect was smaller
in U17, and was absent in U19. Finally, the competitive success rate of U15 cyclists could not predict success at adult
age. However, for U17 and U19 cyclists each additional top 10 result was associated with 3-5%, and 6% higher chance
to reach elite level at adult age, respectively. Overall, these results demonstrate that estimating potential based
on competition results is unpredictable in the U15 category, however may provide a better indication of future
success from U17 onwards.
Keywords: competition, talent, youth, prediction, performance
100
Introduction
As sports expertise develops in a rather non-linear and dynamic way (Abbott et al., 2005), searching for
young athletes with the potential to excel in the future is difficult. Moreover, a complex interplay of factors related
to both the athlete and the environment, together with the component of chance, makes long-term prediction of
success problematic (Gagné, 2004; Elferink-Gemser et al., 2011; Li et al., 2018a). Nevertheless, sports organisations
start at an increasingly younger age with talent identification and the implementation of elite athlete development
programmes to increase the chances of success for their athletes on the international stage (Barreiros et al., 2014;
Li et al., 2018b). Currently, talent development programmes are often reserved for youth athletes who perform well
in youth competitions (e.g., through selection for regional or national teams). Therefore, it is important to consider
the age at which performance in competitions becomes important as a predictor of future success.
The relationship between youth success and success at senior level has been studied in several sports
(Barreiros & Fonseca, 2012; Li et al., 2018a). It has been demonstrated that early involvement in international events,
and selection of talent at young age, is not a prerequisite of future success in tennis (Brouwers et al., 2012; Li et al.,
2020), volleyball (Barreiros & Fonseca, 2012; Barreiros et al., 2014), swimming (Barreiros & Fonseca, 2012; Barreiros
et al., 2014), or soccer (Barreiros & Fonseca, 2012; Barreiros et al., 2014). Only a third of international pre-junior
athletes reappeared as senior athletes, confirming the difficulties of predicting late success based on early
identification and selection (Barreiros et al., 2014). In contrast, research in track and field (Hollings, 2006; Scholz,
2006) and combat sports (Li et al., 2018a) argues that early success is an important contributor to predicting future
performance level. In general, the extent to which success in youth competitions is a predictor for future success
seems to vary by gender and type of sport (Li et al., 2018a).
How well an athlete performs in a specific sport during development depends on the interplay between the
specific demands of a sport and the individual capacities of the athlete (Gagné, 2004; Vaeyens et al., 2009; Elferink-
Gemser et al., 2011). Some of these characteristics are modifiable through training (e.g., endurance), but others, such
as date of birth, are out of control of the athlete (Vaeyens et al., 2005). As youth competitions are usually divided
in age categories, athletes who are born earlier in the year experience significant advantages when competing
against those who are chronologically younger (Hollings et al., 2014). As a result, youth and professional level
athletes who are chronologically older are overrepresented within an annual age cohort. This phenomenon is called
the Relative Age Effect (RAE), and is frequently found among participants in youth sports such as soccer, swimming
(Musch & Grondin, 2001; Delorme et al., 2010; Deprez et al., 2012; Hollings et al., 2014; Rees et al., 2016). In extreme
cases, the advantage of the RAE can be close to 12 months of developmental differences in the areas of
anthropometry, physical performance, experience, motor, social and/or cognitive development (Cumming et al.,
101
2018). Due to this phenomenon, the age at which performance in competitions becomes a reliable predictor for
future success in a specific sport is dependent on the importance of the RAE on youth performance.
For physically demanding sports such as cycling, it is common that athletes are likely to experience an
ascending trajectory of development (Gulbin et al., 2013a), characterised by maintaining the same level of
competition or advancing to a higher level of competition. This phenomenon might be attributed to the well-
developed age-specific competition opportunities and to the nature of the sport that focuses on physical
capabilities (peak performance at 29.5 +/- 4.3 years)(Longo et al., 2016). Nevertheless, only around 30% of cycling
athletes who had participated at the Junior World Championships (U19 category) also participated in the same kind
of events as adult athletes (Schumacher et al., 2006). Due to the limited selection at major international events,
many high-quality athletes may miss qualification for several different reasons (Schumacher et al., 2006). To our
knowledge, there is only limited empirical research that investigates the athletic careers of young cycling athletes
from earlier to later stages and considers performances at national events.
In the context of the preceding, the current study retrospectively analysed competition results of cyclists
who did or did not reach the elite level as an adult athlete. First, differences in success rate between future
achievers and future non-achievers across three consecutive cycling competition categories (spanning the
categories U15, U17, and U19) were analysed. Second, we investigated whether RAE affects the career pathways of
future achievers and future non-achievers. Finally, the extent to which youth competitive performance can predict
success at adult age was investigated. As such, the current study might provide opportunities for coaches to
optimise the identification and selection process of athletes.
Materials and methods
Participant details
A total of 307 Belgian male road cyclists, born between 1990 and 1993, were included in the study. The
cyclists were categorised into future achievers (N=32) and future non-achievers (N=275). The future achievers
needed to be part of a professional World Tour (WT: N=13), Pro Continental (PC: N=8) or Continental (C: N=11) team
for at least four years during their sports career. On at least three moments during their professional career, the
achievers attained a year-end UCI World ranking, whereas this was not the case for the non-achievers. The UCI-
ranking is calculated based on all races from World Tour level down to 1.2 and 2.2 races. The UCI rankings ensure
that the athletes participated regularly and successfully in professional races (Mujika & Padilla, 2001). Information
was gathered via ProCyclingStats (Lip, 2020) and the UCI ranking(Archive, 2020). The non-achievers did not reach
102
the highest competition level at adult age, however, achieved at least one top 10 result at a competition in the
youth categories.
Data acquisition
Top 10 results from national and provincial competitions were gathered retrospectively(Callens, 2002-2011;
Archive, 2019) for three different consecutive youth categories the U15 (12 – 14 years; 3 competition years), U17
(15 – 16 years; 2 competition years), U19 (17 – 18 years; 2 competition years). These categories are organised using
the annual age-grouping strategy from January until December. The competitions included for each year in the U15
category were the Belgian, Flemish, and provincial road championship. For the U17 and U19 category results from
six championships per year were taken into account: the Belgian and provincial championship road and time trial,
and two national climb races. Only athletes who started in the U15 category or were already active before the U15
category were included in the study. Subsequently, the success rate was computed for each competition category
and/or competition year, taking into account the number of top 10 results achieved and the number of competitions
listed (U15total: N=9; U17 total: N=12; U19 total: N=12).
n of top 10 results
Success rate ∶ ( ) ∗ 100
n of competitions
To examine birth date distribution and RAE, the athletes were divided into four birth quarters according to their
month of birth: Q1: January – March, Q2: April – June, Q3: July – September, Q4: October – December.
Statistical analysis
Firstly, the difference in success rate between the future achievers and future non-achievers for the three
consecutive competition categories, were investigated through a non-parametric analysis of longitudinal data
(nparLD – F1 LD F1 design). The non-parametric analyses were preferred as they do not require the normality
assumption (Noguchi et al., 2012; Nahm, 2016), the assumption of variance, and equal sample size. A second non-
parametric analysis of longitudinal data was used to examine the developmental trajectories and competitive
transitions between and within both groups. The Wald-Type statistic was interpreted, and interaction effects
between group and category were reported. Post hoc analyses were performed with Kruskal-Wallis (Chi-squared)
and Dunn-test analyses, and p-values were adjusted with the Benjamini-Hochberg method.
Second, the RAE on the performance results of future achievers and future non-achievers was examined
using crosstabs. Kruskal-Wallis analyses were established for each cycling category separately (U15, U17, U19).
Finally, to explore the predictive value of youth competitive performance for each separate cycling category as an
indicator of future success, logistic regressions were established. The data were controlled for multi-collinearity.
Odds ratio (OR) and the 95% confidence interval (CI) were reported. Furthermore, Receiver Operating Characteristic
103
(ROC) analyses were performed to evaluate the predictive power. Values of Area Under the Curve (AUC) were
considered to reflect poor (0.5–0.7), acceptable (0.7–0.8), excellent (0.8–0.9) or outstanding (>0.9) model accuracy
(Hosmer Jr et al., 2013). The analyses were accomplished using R version 3.6.1 with a p-value of <0.05.
Additionally, as there might be considerable differences between the three categories of future achievers,
subsequent analyses were carried out to investigate the developmental trajectories and the predictive value of
youth competitive performance separately for WT, PC, and C, versus future non-achievers. Although the sample sizes
of these three groups are small, these exploratory analyses might shed some extra insights in the developmental
trajectories of successful youth cyclists. Post hoc analyses with Kruskal-Wallis (Chi-squared) and Dunn-test
analyses were utilised, and p-values were adjusted with the Benjamini-Hochberg method. The RAE on the
performance results of these four groups were examined using crosstabs. Logistic regressions were performed to
investigate the predictive value separately for WT, PC, C versus future non-achievers.
Results
Success rate per category and transition between categories
The career trajectories of future achievers and non-achievers across the 7 years of competition are displayed
in Figure 2.7. and Table 2.11.. A significant main effect for the groups (Wald χ2=10.412, df=1, p=0.001) suggests that
across the 7 competition years, the future achievers had an overall higher success rate compared to the non-
achievers. A significant main effect for the competition years (Wald χ2=90.054, df=6, p<0.001) indicated that the
success rate changed across the 7 years. A significant interaction effect (Wald χ2=45.300, df=6, p<0.001) suggests
that the success rate of both groups did not change similarly over time. With exception of the U15 category, future
achievers outperformed the future non-achievers in every competition year of the U17 and U19 category.
Table 2.11.: Average success rate (± SD) of the future achievers and non-achievers for each competition category and results of the
univariate analysis (Chi squared, degrees of freedom and p-values).
Achievers Non-Achievers Univariate
Mean ± SD Mean ± SD Chi squared df p
U15year1 14.29 ± 30.67 16.06 ± 28.3 0.253 1 0.615
U15year2 21.11 ± 34.45 18.72 ± 28.21 0.005 1 0.945
U15year3 22.92 ± 34.33 17.33 ± 27.51 0.206 1 0.650
U17year1 18.23 ± 19.1 7.39 ± 11.49 12.136 1 <0.001
U17year2 26.56 ± 24.27 15.33 ± 18.35 6.685 1 0.010
U19year1 20.83 ± 19.4 6.69 ± 11.67 22.775 1 <0.001
U19year2 27.06 ± 18.75 8.94 ± 14.2 39.151 1 <0.001
104
The success rate of the non-achievers significantly decreased when making the transition to a higher category
(from U15year3 to U17year1 p=0.015, and from U17year2 to U19year1 p<0.001). Within the U17 category, an increase in success
rate was observed after one year of experience for the future non-achievers group (from U17year1 to U17year2 p<0.001).
No significant change in success rate was noticed for the achievers between and within the different competition
categories (from U15year3 to U17year1 p=0.657, from U17year1 to U17year2 p=0.394, and from U17year2 to U19year1 p=0.590).
However, the baseline and end level differed significantly within the future achievers’ group (from U15year1 to U19year2
p<0.001), however not in the non-achievers group (from U15year1 to U19year2 p=0.372).
Figure 2.7.: Career trajectories of the achievers and non-achievers are displayed using success rate (mean and the positive standard
deviation) achieved in each year of the competition (U15= 3 competition years; U17= 2 competition years; U19= 2 competition years).
Significant differences in success rate between both groups are marked by an arrow (p<0.05), while significant differences in success
rate within one group between consecutive years are indicated by an asterisk (p<0.05).
Relative age effect
The birth date distribution (RAE) of the future achievers and non-achievers in the four birth quarters across
the different competition years are provided in Figure 2.8. and Table 2.12.. The non-parametric analysis of
longitudinal data revealed a significant main effect for group (Wald χ2=11.507, df=1, p<0.001) and competition years
(Wald χ2=108.047, df=6, p<0.001), which are explained in the previous paragraph. No significant main effect was
found for RAE (Wald χ2=2.795, df=3, p=0.424), demonstrating that the overall success rate was not affected by birth
quarter. A three-way interaction effect was found between group, birth quarter, and the competition years
(Wald χ2= 75.495, df=18, p<0.001). Competition years and group (Wald χ2= 69.964, df=6, p<0.001), and competition
years and birth quarter (Wald χ2= 92.111, df=18, p<0.001) showed significant two-way interaction effects, which
indicated that success rate changed differently over time. A significant two-way interaction effect between the
105
competition years and birth quarter was found within the achievers group (Wald χ2=154.070, df=18, p<0.001) as
well as in the non-achievers group (Wald χ2=29.897, df=18, p<0.038). Success rate of achievers in the first year of
the U15 category was significantly better for cyclists born in the second quarter compared to the riders from the
first (p=0.014), third (p=0.014) and fourth (p=0.017) birth quarter. There was no RAE on the competition results in
the U17 and U19 category. In the non-achievers group, athletes born in the first birth quarter demonstrated better
success rates compared to athletes born in the third (py1=0.006; py3=0.006) and fourth (py1=0.015; py3=0.015) birth
quarter in the first and third year of the U15 category. In the second year of the U17 category the influence was still
significant but less pronounced in the non-achievers group as in the U15 category. A significantly higher success
rate was found for the cyclists born in the first quarter compared to the cyclists born in the second quarter
(p=0.017). In the U19 category, the influence of RAE on the performance results was absent in the non-achievers
group (p>0.005).
Figure 2.8.: Overview of the birth date distribution (RAE) in the future achievers and future non-achievers group. The success rate for
athletes born in the different birth quarters are displayed for each competition year. Significant differences in success rate between the
different birth quarters are marked by an arrow (p<0.05).
106
Table 2.12.: Average success rate (± SD) of the future achievers and non-achievers for each competition category and birth quarter (quarter1,
quarter2, quarter3, quarter4). Additionally, the results of the univariate analysis (Chi squared, degrees of freedom and p-values) are
displayed.
Q1 Q2 Q3 Q4 Univariate
Mean ± SD Mean ± SD Mean ± SD Mean ± SD Chi squared df p
Non-Achievers (N) 84 82 62 47
a a
U15year1 20.63 ± 32.27 12.60 ± 24.92 6.45 ± 18.93 7.09 ± 19.58a 13.616 3 0.003
U15year2 20.24 ± 30.18 19.11 ± 28.70 12.37 ± 20.23 11.35 ± 26.26 6.105 3 0.107
a a a
U15year3 21.83 ± 29.50 19.11 ± 29.64 13.98 ± 22.22 10.64 ± 25.16 7.981 3 0.046
U17year1 10.12 ± 13.45 5.89 ± 10.27 6.18 ± 10.98 6.38 ± 9.55 6.402 3 0.094
U17year2 18.85 ± 20.34a 11.38 ± 18.12a 15.32 ± 16.34 15.96 ± 16.65 9.786 3 0.020
U19year1 8.93 ± 14.96 5.69 ± 9.52 5.38 ± 9.91 6.03 ± 10.09 2.336 3 0.506
U19year2 10.32 ± 15.29 7.52 ± 13.13 8.06 ± 13.41 9.93 ± 15.01 2.059 3 0.560
Achievers (N) 8 8 6 10
a a a
U15year1 0.00 ± 0.00 37.50 ± 37.53 0.00 ± 0.00 10.00 ± 31.62a 12.383 3 0.006
U15year2 20.83 ± 39.59 41.67 ± 42.72 5.56 ± 13.61 10.00 ± 22.50 5.199 3 0.158
U15year3 20.83 ± 30.54 29.17 ± 41.55 22.22 ± 34.43 20.00 ± 35.83 0.227 3 0.973
U17year1 25.00 ± 19.92 10.42 ± 19.80 16.67 ± 10.54 20.00 ± 21.94 3.182 3 0.364
U17year2 20.83 ± 23.15 31.25 ± 22.60 25.00 ± 29.34 28.33 ± 26.12 0.984 3 0.805
U19year1 31.25 ± 13.91 16.67 ± 23.57 19.44 ± 19.48 16.67 ± 19.25 4.159 3 0.245
U19year2 27.08 ± 19.80 27.08 ± 21.71 38.89 ± 20.18 21.67 ± 13.72 2.911 3 0.406
a
Note: indicates significant differences (p<0.05) between the different groups based on the post hoc analyses.
Predictive value of youth competitive performance
The logistic regression analysis indicated that the success rate in the three competition years of the U15
category could not predict future success at adult age (p>0.005). The model was statistically significant in both
competition years of the U17 (p<0.001) and U19 category (p<0.001). With every additional top 10 result, cyclists in
the first and second year of the U17 category have 5% (ORU17year1=1.052, 95% CI=1.027-1.077, AUC=0.661) and 3%
(ORU17year2=1.025, 95% CI=1.009-1.042, AUC = 0.631) more chance to reach elite level at adult age, respectively. Cyclists
in both years of the U19 category have 6% more chance to excel in the sport at adult age with every additional top
10 result (ORU19year1=1.059, 95% CI=1.034-1.084, AUC = 0.716 ; ORU19year2=1.058, 95% CI=1.037-1.080, AUC = 0.799).
Subdivision of the future achievers
Additional analyses revealed that especially the World Tour (WT) riders performed significantly better in the
U17 and U19 category compared to the non-achievers. The results are displayed in Table 2.13.. The difference
between Pro Continental (PC) or Continental (C) riders and the non-achievers was less pronounced, but still
significant from U19 category onwards. Riders from the WT, PC, C and non-achievers group were equally distributed
across the different birth quarters (Chi-squared=7.364 ; df=9; p=0.599). The logistic regression analysis indicated
107
that with every additional top ten result, cyclists of the U17 and U19 category had 6.4% (U17 year1=8.07%;
U17year2=4.76%) and 8.8% (U19year1=9.02%; U19year2=8.60%) more chance to reach WT level at adult age, respectively.
The success rate of future PC and C riders in the three competition categories could not predict future success at
adult age (p>0.005).
108
Table 2.13.: Average success rate (± SD) of the WT, PC, C, and future non-achievers for each competition category. The results of the Kruskal-Wallis analyses (Chi squared, degrees of freedom, and p-values) are
displayed. An overview is provided of the birth date distribution (RAE) in the WT, PC, C, and future non-achievers group together with the results of the crosstab analysis (Chi squared, degrees of freedom, and p-
values). The bottom part of the table represents the results from the logistic regression (Odds ratio, 95% confidence interval, ROC).
WT Pro Continental Continental Non-achievers

(N=13) (N=8) (N=11) (N=275) Univariate
Competition years Mean ± SD Mean ± SD Mean ± SD Mean ± SD Chi squared df p
U15year1 7.69 ± 19.97 8.33 ± 23.57 21.21 ± 40.2 12.73 ± 26.02 1.1628 3 0.762
U15year2 17.95 ± 32.25 20.83 ± 39.59 21.21 ± 34.23 16.61 ± 27.22 0.14269 3 0.986
U15year3 30.77 ± 31.8 20.83 ± 39.59 15.15 ± 34.52 17.33 ± 27.51 3.7142 3 0.294
U17year1 25.64 ± 19.97a 14.58 ± 13.91 12.12 ± 19.85 7.33 ± 11.47a 17.501 3 <0.001
U17year2 41.03 ± 24.17a 16.67 ± 17.82 16.67 ± 21.08a 15.33 ± 18.35a 15.401 3 0.002
U19year1 33.33 ± 20.41a 14.58 ± 10.68b 10.61 ± 15.41a 6.67 ± 11.66a,b 32.039 3 <0.001
U19year2 41.03 ± 18.78a 16.67 ± 15.43 19.7 ± 10.05b 8.91 ± 14.18a,b 44.199 3 <0.001
RAE N N N N Chi squared df p
RAE 7.364 9 0.599

Q1 (N) 2 3 3 84
Q2 (N) 3 1 4 82
Q3 (N) 3 1 2 62
Q4 (N) 5 3 2 47
Predictive value OR (95% CI) ROC OR (95% CI) ROC OR (95% CI) ROC
U17year1 1.081* (1.043 - 1.119) 0.770 1.042 (0.993 - 1.093) 0.656 1.028 (0.980 - 1.078) 0.534
U17year2 1.048* (1.024 - 1.072) 0.803 1.004 (0.968 - 1.041) 0.534 1.004 (0.968 - 1.041) 0.501
U19year1 1.090* (1.053 - 1.129) 0.854 1.037 (0.997 - 1.079) 0.712 1.022 (0.982 - 1.063) 0.555
U19year2 1.086* (1.052 - 1.120) 0.907 1.029 (0.997 - 1.061) 0.675 1.038* (1.006 - 1.071) 0.761
a,b
Note: indicate significant differences (p<0.05) between the different groups based on the post hoc analyses.
* indicates significant (p<0.05) predictions based on the logistic regression analyses.
109
Discussion
The present study aimed to investigate the career trajectories of future achievers and future non-achievers
using retrospective analyses of competition results. Results showed that with increasing age, the importance of
competition results increased, while the influence of birthdate distribution decreased. The most accurate
predictions for performance at the senior elite level were found in the U19 category.
Success rate of achievers and non-achievers
While the success rate of future achievers and non-achievers does not differ in the U15 category, future
achievers outperform the non-achievers in the U17, and even more so in the U19 category. These findings are in line
with the results of Güllich and Emrich (2014) showing a low contribution of juvenile success (U15 category) and
individual differences in success achieved at senior age. As was the case for tennis (Brouwers et al., 2012; Li et al.,
2020), and volleyball (Barreiros & Fonseca, 2012; Barreiros et al., 2014), swimming (Barreiros & Fonseca, 2012;
Barreiros et al., 2014), and soccer (Barreiros & Fonseca, 2012; Barreiros et al., 2014), early involvement in
competitions and selection of talent during early ages do not seem to be a prerequisite of success within cycling.
From U17 onwards, future achievers do seem to outperform their peers more and more often. The additional
analyses of the achievers group suggest that the widening gap between achievers and non-achievers is mostly
caused by the future World Tour riders who outperform their peers. Although this analysis was carried out on a
small-sized subsample, the results are in line with those of Svendsen et al. (2018), who reported that in
competitions at the age of 18, the future WT riders already outperform riders who will not reach the WT level.
The widening gap in success rate can be attributed to different mechanisms. First, the impact of biological
development on the competitive performance of the younger athletes (Li et al., 2020) can play a part. Athletes may
become physically and mentally stronger at a later age, which might be essential in endurance sports to develop
into successful elite athletes later (Schumacher et al., 2006). A second possible cause of the increasing superiority
of the achievers is training load and experience (Brouwers et al., 2012). However, the amount of practice at a young
age does not seem to be as important as the accumulated training volume from the age of 18 onwards (Moesch et
al., 2011). Furthermore, effect of training volume might play a different role in female cycling than in male cycling
as female cyclists were found to compensate their lower training volume by training at a higher intensity (Van Erp
et al., 2019). Next to physical improvements, this might also lead to better improved race preparation, which also
positively affects race outcomes. Third, the changes in race format, more particularly the increase in race distance,
from the U15 to the U19 category may also explain the widening gap in success rate. Cyclists with more ‘talent’ for
endurance may only excel in the later categories with longer races (Lucía et al., 2001; Van Erp & Sanders, 2020).
110
It is notable that future non-achievers showed a significant decrease in success rate when transitioning to
a new competition category, which was not the case for the future achievers. The increase in success rate after one
year experience in the non-achievers group suggests that there is a need to provide active assistance for athletes
to make a smoother transition from one level to the next to mitigate potential dropout (Green, 2005). However, a
proper understanding of the transition process is required as it involves a number of transition variables (i.e.
individual characteristics, external factors, environment) that influence the experience prior to, during, and after
transition (Drew et al., 2019). When comparing competition results of athletes, it is therefore suggested that
coaches should consider the performance in all years of a competition category. However, this is not the case the
future achievers’ group, as their performance level seems to be maintained when transitioning to a higher
competition category. Further research is needed to investigate the individual career pathways to support these
findings.
Influence of Relative Age Effect on success rate
Continued participation and/or success at youth level in a particular sport can be associated with the
physical and biological development of the athletes (Vaeyens et al., 2009). Our results demonstrate that relatively
older cyclists performed significantly better compared to their younger peers, but only in the U15 category of both
groups. This is consistent with previous findings (Cobley et al., 2009; Baker et al., 2010a; Baker et al., 2010b; Deprez
et al., 2012; Gibbs et al., 2012; van den Honert, 2012; Cumming et al., 2018; Doyle & Bottomley, 2018) that observed
RAEs exist before puberty. To our knowledge, there is currently no consensus in the literature regarding the
persistence of RAE into adulthood (Vaeyens et al., 2005; Cumming et al., 2018; Jones et al., 2018). Some studies
found that the RAE on inter-individual differences may decrease when the athletes reach adulthood (Musch &
Grondin, 2001; Vaeyens et al., 2009), which may explain why the RAE faded as athletes reached the U19 category.
Becoming physically and mentally stronger at a later age might be of a greater importance in endurance sports to
develop into successful adult elite athletes (Schumacher et al., 2006). RAEs are frequently attributed to the direct
and indirect advantages afforded by advanced maturity (Musch & Grondin, 2001; Cumming et al., 2018), potentially
explaining the increased success of athletes born in the first birth quarters. However, RAE and maturation should
be considered individually as the RAE is a phenomenon independent of being an early or late mature (Cumming et
al., 2018).
Closer examination revealed that a biased distribution of birthdates was absent in this particular sample.
The WT, PC, C riders as well as the non-achievers were equally distributed over the different birth quarters. This is
in contrast with several studies (Musch & Grondin, 2001; Delorme et al., 2010; Deprez et al., 2012; Hollings et al.,
2014; Rees et al., 2016) that found an overrepresentation of athletes born at the beginning of the competitive year.
111
The present study indicated that within the current sample, athletes were not eliminated based on their relative
age during adolescence, reducing the risk of systematic exclusion of athletes with potential (Tribolet et al., 2019).
Predictive value of youth competition results for later success
The current study illustrated the complexity of predicting future success based on competition results
(Güllich & Emrich, 2014). The further from professional-level at adult age, the more difficult it is to make
assumptions about career progression, what can be expected since road cyclists reach peak performance at a
relatively late age (29.5 +/- 4,3 years) (Brouwers et al., 2012). From the U17 and U19 competition category onwards,
the success rate becomes a better indicator of later success. The predictive value of success rate was most
pronounced based on the career pathway of the future WT riders. Similar findings were reported by Svendsen et al.
(2018) indicating the importance of race performance at the age of 18. However, to what extent better
performances at young age set the stage for better subsequent opportunities such as team contracts, financial
support, and racing opportunities needs to be further investigated (Svendsen et al., 2018). Consequently, we can
assume that high training workloads and competition-related psychological pressure at an early age is not required
to increase the chances of reaching the professional level (Schumacher et al., 2006; Güllich & Emrich, 2014;
Sieghartsleitner et al., 2018). Further research is needed to link these findings to the results of Côté et al. (2009)
and Baker (2003), who state that childhood sport specialization is not an essential ingredient for exceptional sport
performance as an adult.
Limitations
There are a number of limitations that should be considered. First, the current study only used competition
results from national and provincial races as indicator of performance level. Given the many national and
international races in which riders can participate, the athletes did not all compete in the same championships. As
a result, the level of a competition may vary from year to year. Additionally, external factors such as the type of the
race, training history, and weather conditions may also affect the performances and determine whether or not an
athlete achieves a top 10 placement. Therefore, the results should be interpreted with caution.
Second, while differences in average success rate clearly shows the future achievers perform better, large
standard deviations suggest that results should be interpreted with caution. When analysed on an individual level,
the career pathways show to be highly variable, with highly successful young cyclists ending up without a
professional contract, and future achievers being unsuccessful throughout their youth. This supports the idea that
the athletic developmental pathway (Gulbin et al., 2013a) is dynamic and a myriad of factors may influence success,
in particular, the socio-cultural environment and changes in physical and physiological characteristics (Gagné,
112
2004). This may contribute to the high inter-individual variability in performances (Gulbin et al., 2013a). Therefore,
a more holistic approach is suggested for efficient talent selection, using a combined strategy of multidimensional
measurements, such as competitive performance, physical, mental, and tactical skills at sport-specific level (Abbott
et al., 2005; Schumacher et al., 2006), as well as coach assessments (Sieghartsleitner et al., 2019). Further research
should focus on the contribution of combined factors and indicators of later success, together with a long-term
follow up in the development of peak performance in professional cycling (Schumacher et al., 2006).
Conclusion
The present study demonstrated that the career trajectories differ between achievers and non-achievers.
Future achievers had higher success rates from the U17 category onwards and benefited more whilst transitioning
into a new competitive category. The competitive performance is influenced by relative age effect in the U15
category, however, this effect disappeared as the athletes progressed into older age categories. The chance of
reaching elite level in adulthood increased as the success rate of the cyclists was higher in the U19 category.
However, the large standard deviations suggest that the career pathways are variable and the results should be
interpreted with caution. In general, this study provides insights into the career trajectories of young cyclists until
(elite) adult level and serves as a reference frame for coaches and federations when interpreting competition
results.
Acknowledgement
The authors wish to thank Belgian Cycling and Cycling Vlaanderen for their collaboration. Special thanks to
Frederik Broché (technical director Belgian Cycling), Koen Beeckman (technical director of topsport Cycling
Vlaanderen), and Jan Vancompernolle (performance analyst Cycling Vlaanderen) for the realization of this project.
113
References
Archive. (2020). Retrieved September 2020, from UCI ranking https://www.uci.org/road/rankings/
85-94.
Baker, J., Cobley, S., Montelpare, W. J., Wattie, N., & Faught, B. E. (2010a). Exploring proposed mechanisms of the
relative age effect in Canadian minor hockey. International Journal of Sport Psychology, 41(2), 148-159.
Baker, J., Schorer, J., & Cobley, S. (2010b). Relative age effects. Sportwissenschaft, 40(1), 26-30.
Barreiros, A., Côté, J., & Fonseca, A. M. (2014). From early to adult sport success: Analysing athletes' progression in
national squads. European Journal of Sport Science, 14(sup1), S178-S182. doi:10.1080/17461391.2012.671368
Barreiros, A. N., & Fonseca, A. M. (2012). A retrospective analysis of Portuguese elite athletes' involvement in
international competitions. International Journal of Sports Science & Coaching, 7(3), 593-600.
475.
Callens, B. (2002-2011). Wielerjaarboek.
39(3), 235-256.
Côté, J., Lidor, R., & Hackfort, D. (2009). ISSP position stand: To sample or to specialize? Seven postulates about youth
sport activities that lead to continued participation and elite performance. International Journal of Sport
and Exercise Psychology, 7(1), 7-17. doi:10.1080/1612197x.2009.9671889
Cumming, S. P., Searle, C., Hemsley, J. K., Haswell, F., Edwards, H., Scott, S., Gross, A., Ryan, D., Lewis, J., & White, P.
(2018). Biological maturation, relative age and self-regulation in male professional academy soccer
players: A test of the underdog hypothesis. Psychology of Sport and Exercise, 39, 147-153.
Delorme, N., Boiché, J., & Raspaud, M. (2010). Relative age effect in elite sports: Methodological bias or real
discrimination? European Journal of Sport Science, 10(2), 91-96.
Deprez, D., Vaeyens, R., Coutts, A., Lenoir, M., & Philippaerts, R. (2012). Relative age effect and Yo-Yo IR1 in youth
soccer. International journal of sports medicine, 33(12), 987-993.
Doyle, J. R., & Bottomley, P. A. (2018). Relative age effect in elite soccer: More early-born players, but no better
valued, and no paragon clubs or countries. PloS one, 13(2), e0192209.
Drew, K., Morris, R., Tod, D., & Eubank, M. (2019). A meta-study of qualitative research on the junior-to-senior
transition in sport. Psychology of Sport and Exercise, 45, 101556. doi:10.1016/j.psychsport.2019.101556
119-147. doi:10.1080/1359813042000314682
Gibbs, B. G., Jarvis, J. A., & Dufur, M. J. (2012). The rise of the underdog? The relative age effect reversal among
Canadian-born NHL hockey players: A reply to Nolan and Howell. International Review for the Sociology
of Sport, 47(5), 644-649.
Green, B. C. (2005). Building sport programs to optimize athlete recruitment, retention, and transition: Toward a
normative theory of sport development. Journal of Sport Management, 19(3), 233-253.
114
Hollings, S. (2006). World junior success is a prerequisite for world senior success. Modern Athlete and Coach, 44(1),
14-17.
Hollings, S. C., Hume, P. A., & Hopkins, W. G. (2014). Relative-age effect on competition outcomes at the World Youth
and World Junior Athletics Championships. European Journal of Sport Science, 14(sup1), S456-S461.
Hosmer Jr, D. W., Lemeshow, S., & Sturdivant, R. X. (2013). Applied Logistic Regression (Vol. 398): John Wiley & Sons.
Jones, B. D., Lawrence, G. P., & Hardy, L. (2018). New evidence of relative age effects in “super-elite” sportsmen: a
case for the survival and evolution of the fittest. Journal of sports sciences, 36(6), 697-703.
Li, P., De Bosscher, V., Pion, J., Weissensteiner, J. R., & Vertonghen, J. (2018a). Is international junior success a reliable
550-559.
Li, P., De Bosscher, V., & Weissensteiner, J. R. (2018b). The journey to elite success: a thirty-year longitudinal study of
the career trajectories of top professional tennis players. International Journal of Performance Analysis
in Sport, 18(6), 961-972.
Li, P., Weissensteiner, J. R., Pion, J., & Bosscher, V. D. (2020). Predicting elite success: Evidence comparing the career
pathways of top 10 to 300 professional tennis players. International Journal of Sports Science & Coaching,
15(5-6), 793-802. doi:10.1177/1747954120935828
Lip, B. v. d. Z. S. (2020). Retrieved September 2020, from ProCyclingStats https://www.procyclingstats.com/
doi:10.14198/jhse.2016.111.03
doi:10.1111/j.1600-0838.2010.01280.x
Nahm, F. S. (2016). Nonparametric statistical tests for the continuous data: the basic concept and the practical use.
Korean Journal of Anesthesiology, 69(1), 8.
Noguchi, K., Gel, Y. R., Brunner, E., & Konietschke, F. (2012). nparLD: an R software package for the nonparametric
analysis of longitudinal data in factorial experiments. Journal of Statistical Software, 50(12).
Scholz, W. (2006). The throwing events at the IAAF World Junior Championships: A whistle stop on the journey to
elite athletics. New Studies in Athletics, 21(2), 7.
115
Tribolet, R., Watsford, M. L., Coutts, A. J., Smith, C., & Fransen, J. (2019). From entry to elite: The relative age effect in
the Australian football talent pathway. Journal of Science and Medicine in Sport, 22(6), 741-745.
Vaeyens, R., Philippaerts, R. M., & Malina, R. M. (2005). The relative age effect in soccer: A match-related perspective.
Journal of sports sciences, 23(7), 747-756.
van den Honert, R. (2012). Evidence of the relative age effect in football in Australia. Journal of sports sciences,
30(13), 1365-1374.
Van Erp, T., Sanders, D., & De Koning, J. J. (2019). Training characteristics of male and female professional road
cyclists: a 4-year retrospective analysis. International journal of sports physiology and performance, 15(4),
534-540. doi:10.1123/ijspp.2019-0320
116
Study 4 : Do race results in youth competitions predict future success as a road
cyclist? A retrospective study in the Italian Cycling Federation
Gabriele Gallo1,2*, Mireille Mostaert3*, Emanuela Faelli1,2, Piero Ruggeri1,2, Sundeep Delbarba4,
Roberto Codella4,5, Pieter Vansteenkiste3**, Luca Filipas4,5**.
1
DINOGMI, Università degli Studi di Genova, Genova, Italy
2
Centro Polifunzionale di Scienze Motorie, Università degli Studi di Genova, Genova, Italy
3
Department of Movement and Sport Sciences, Ghent University, Ghent, Belgium
4
Department of Biomedical Sciences for Health, Università degli Studi di Milano, Milan, Italy
5
Department of Endocrinology, Nutrition and Metabolic Diseases, IRCCS MultiMedica, Milan, Italy
-
International Journal of Sports Physiology and Performance
Abstract
The aim of this study was to investigate the relationship between youth road cycling success and becoming
a professional cyclist. Specifically, we sought to analyze: (i) the differences success score in youth categories
between future professional and future non-professional cyclists; (ii) whether relative age effect influences youth
road cycling career pathways; (iii) whether youth competition success could predict the career as professional
cyclist. The number of points gathered in the annual national ranking of 1345 Italian cyclists in the U17, U19 and
U23 category were retrospectively analyzed. Participants were divided into two groups: future professional (PRO;
n = 43) and future non-professional (NON-PRO; n = 1302), depending on whether they reached the professional
level or not. PRO outperformed NON-PRO in all the youth categories considered (i.e. U17, U19 and U23). Relatively
older cyclists, within the same annual age group, were not overrepresented in PRO and do not have an advantage
over younger cyclists within all the competition years. The number of points gathered in youth competitions
provides an indication of probability of becoming professional cyclists from the U17 onwards with predictive value
increasing with age-category. Handling the transition to a new age group well (especially the U19-U23 transition),
and therefore having success competing against older and more experienced cyclists, is an important factor for
talent identification in youth cycling.
Key words: youth cycling; talent identification; endurance; career pathway
120
Introduction
Talent identification (i.e. predicting the possible future performance level of an athlete)(Abbott & Collins,
2002) has become a topic of great importance for different sports stakeholders such as national sports federations,
professional teams, and sponsors. Currently, youth athletes are often regarded as being “talented” when they
perform well in youth competitions. However, as youth performance is sensitive to initial conditions, transitions,
and exponential behavioral distributions, talent development (i.e. the progression of performance of an
athlete)(Abbott et al., 2005) seems to be a nonlinear dynamic process (Abbott et al., 2005). Therefore, it is unclear
whether the best youth athletes are effectively the best potential elite athletes. Previous studies have already
shown that youth success does not necessarily predict success in adulthood (Vaeyens et al., 2009; Pizzuto et al.,
2017). Yet the recruitment for national federations’ development programs is usually on the basis of youth athletes’
current performance rather than their possible future level of performance (Elferink-Gemser et al., 2011). This
means that future performers might miss out on extra training facilities, highly certified trainers, and medical
supervision, because they underperformed in youth competitions. One of the possible factors contributing to the
discrepancy between youth and elite success is the relative age effect (RAE), which is the performance’s
consequence of differences in day of birth between individuals in the same age group (Musch & Grondin, 2001). A
meta-analytical review reported an overall consistent small RAE across sports, which is moderated by several
factors such as chronological age differences, age category, skill level and sport context (Cobley et al., 2009).
In road cycling there have been few studies investigating the association between youth- and professional
cycling success. Schumacher and colleagues (2006) reported that only about one-third of all participants in
professional world championships were successful at juniors world championships, which suggests an only limited
importance of performance at young age for future success. Nevertheless, their data also suggested that those
who were successful as a junior athlete (17-18 years old), were more successful as an adult. Similarly, Menaspà and
colleagues (2010) found that being selected in the juniors Italian national team increases the odds for future
success, but it does not guarantee becoming a professional cyclist. Svendsen and colleagues (2018) on the other
hand showed that performance on the Norwegian national road championships at 18-years old has a good
predictive value for whether athletes would reach World Tour level as an adult (i.e. the highest cycling professional
competition level). Finally, a recent retrospective study of Mostaert and colleagues (Mostaert et al., 2021b) extended
the investigation across three different youth categories: Under 15 (U15; 13-14 years old), Under 17 (U17; 15-16 years
old) and Under 19 (U19; 17-18 years old). They reported that future professional cyclists obtained better results in
Belgian provincial and national championships compared to future non-professional cyclists in U17 and U19, but
not in U15. Success in the U17 and in the U19 category increased the odds of reaching the World Tour level as an
121
adult cyclist. Furthermore, they showed that race results were affected by the RAE in the U15 category, but that this
effect disappeared by the time cyclists reached U19.
The current study takes this topic a step further by retrospectively analyzing cyclists’ performance on a
wider range of races (limiting the influence of contextual factors such as crashes, punctures, health problems, and
tactical factors, on the success in youth competitions), and by including the ‘oldest’ youth category, U23 (which,
following the results of Mostaert and colleagues should provide best predictive value for future success). Firstly,
we investigated differences in youth race results between future professional and future non-professional cyclist
across the U17, U19 and U23 cycling categories. Secondly, we investigated the RAE on race results in the different
categories in both future professional and future non-professional cyclists. Thirdly, we analyzed whether youth
race results can predict obtaining a professional cycling contract as an adult.
Methods
Participants
A total of 1345 Italian road cyclists born between 1992 and 1995 were included in this study. Inclusion criteria
were being affiliated with a team of the Italian national federation and having scored at least one point (see section
below) in U17, U19 or U23 categories. Participants were divided in two groups: future professional (PRO; n = 43) and
future non-professional (NON-PRO; n = 1302). PRO were riders who performed at least three years as professional
cyclists. This selection criterion was already applied by a previous study (Menaspa et al., 2010) and takes into
consideration that the first professional cycling contract has to last at least for two years. Hence, a third year as a
professional cyclist could be considered the confirmation of the qualities for being professional. Using this criterion,
we have excluded riders that could have become professionals only for a short period. Conversely, NON-PRO were
riders who scored at least one point in U17, U19 or U23 category, but failed to reach the professional level, or had
a professional contract for less than three years. An online database (www.procyclingstats.com) was used to
classify participants as PRO or NON-PRO.
Data acquisition
Success in youth categories was determined as the number of points gathered in the annual Italian national
ranking of the Italian Cycling Federation (success score). Three consecutive youth categories were considered: U17
(15-16 years old – two years of competitions), U19 (17-18 years old – two years of competitions), U23 (19-22 years
old – four years of competitions). These categories are organised using the traditional age-grouping from January
until December. Success score in each category was retrospectively collected via an online database
122
(www.ciclismo.info). The scoring system awards points for each race in the national calendar in the following way:
1st place = 5 points; 2nd place = 4 points; 3rd place = 3 points; 4th place = 2 points; 5th place = 1 point. In the U19 and
U23 categories, higher points were awarded for the most importance races: national races, 10 - 8 - 6 - 4 - 2 points;
international races and Italian Championship, 15 - 12 - 9 - 6 - 3 points; European Championship, 20 - 16 - 12 - 8 - 4
points; World Championship, 30 - 24 - 18 - 12 - 6 points. To examine birth date distribution and RAE, the athletes
were divided into four birth quarters according to their birth month: Q1, January – March; Q2, April – June; Q3, July
– September; Q4, October – December.
Statistical analysis
The current dataset was analyzed using similar statistical procedures as were used by Mostaert and
colleagues (2021b). Differences in points scored between PRO and NON-PRO were investigated using a non-
parametric analysis of longitudinal data (nparLD – F1 LD F1 design). The Wald-Type statistic was interpreted, and
interaction effects between group and category (U17year1, U17year2, U19year1, U19year2, U23year1, U23year2, U23year3, U23year4)
were reported. Post-hoc analyses were performed with Kruskal-Wallis and Dunn-test analyses, and p-values were
adjusted with the Benjamini-Hochberg method. Distribution of participants across the birth quarters was checked
using crosstabs. The RAE on success score was examined using another non-parametric analysis of longitudinal
data but with “birth quarter” as additional between-factor. Post-hoc analyses were carried out using Kruskal-Wallis
tests. Finally, to explore the predictive value of competitive performance in each cycling category as an indicator
for future success, logistic regressions were carried out. The data were controlled for multi-collinearity, and Odds
Ratio (OR), and the 95% confidence interval (CI) were reported. Furthermore, Receiver Operating Characteristic
(ROC) analyses were performed to evaluate the predictive power. Values of Area Under the Curve (AUC) were
considered to reflect poor (0.5 – 0.7), acceptable (0.7 – 0.8), excellent (0.8 – 0.9) or outstanding (> 0.9) model
accuracy (Hosmer Jr et al., 2013). The analyses were accomplished using R version 3.6.1 with a significant set as
p < 0.05. Probability to obtain a professional contract within the current population was calculated in Excel based
on the results of the logistic regression using the following formula:
𝑂𝑑𝑑𝑠
P = 1+𝑂𝑑𝑑𝑠 ; with Odds = EXP(Log Odds), and Log Odds = intercept + (Coefficient x Success score)
123
Results
Points scored by future professional and non-professional
The career trajectories of PRO and NON-PRO across the eight years of competition are presented in
Figure 2.9.. A significant main effect between the two groups (Wald χ2 = 401.329, df = 1, p < 0.001) shows that the
PRO overall had a higher success score than the NON-PRO. Post-hoc tests show that this was the case in all eight
of the competition years (Table 2.14.).
Figure 2.9.: Success score (mean + standard deviation) of future professional and non-professional in each competition year. Significant
main effects of time (***, p < 0.001) and group (###, p < 0.001) are reported.
The significant within-subjects effect (Wald χ2 = 660.900, df = 7, p < 0.001) indicates that success score
changed across the eight years, however, a significant interaction effect (Wald χ2 = 221.068, df = 7, p < 0.001) shows
that changes over time were not the same for both groups (Figure 2.9.). In both PRO and NON-PRO, success score
decreased significantly (apart for one case) when making the transition to a higher category (PRO: from U17 year2 to
U19year1, p = 0.068, from U19year2 to U23year1, p < 0.001; NON-PRO: U17year2 to U19year1, p < 0.001, U19year2 to U23year1,
p < 0.001), however, it seems to decrease more for PRO when transitioning from U19 to U23. For both PRO and
NON-PRO, success score was higher in the second than in the first year in U17 and U19 (PRO: U17 year1 to U17year2,
p = 0.001; U19year1 to U19year2, p = 0.003; NON-PRO: U17year1 to U17year2, p < 0.001; U19year1 to U19year2, p = 0.007). However,
in both cases, the improvements of PRO seem larger than those of NON-PRO. In the U23 category, NON-PRO did not
increase their success score over the different years. PRO recorded a significant higher success score in year 2 than
in year 1 (p = 0.010), while not from year 2 to year 3 and from year 3 to year 4.
124
Table 2.14.: Average success score (mean ± standard deviation) by future professional and non-professional cyclists for each
competition year.
Competition year PRO NON-PRO χ2 df p

U17year1 18.72 ± 23.25 6.55 ± 10.16 22.584 1 < 0.001
U17year2 43.35 ± 31.66 15.47 ± 17.54 44.068 1 < 0.001
U19year1 29.37 ± 27.13 4.76 ± 11.31 72.346 1 < 0.001
U19year2 57.44 ± 41.58 8.10 ± 18.68 109.460 1 < 0.001
U23year1 7.33 ± 12.66 0.46 ± 3.50 258.950 1 < 0.001
U23year2 22.81 ± 29.51 1.08 ± 6.74 307.450 1 < 0.001
U23year3 31.53 ± 39.41 1.56 ± 9.42 230.710 1 < 0.001
U23year4 49.19 ± 45.35 1.16 ± 6.87 245.720 1 < 0.001
Abbreviations: PRO, future professional; NON-PRO, future non-professional.
Relative age effect
Riders of the PRO and NON-PRO were equally distributed across the different birth quarters (χ2 = 3.241;
df = 3; p = 0.356). As was the case in the previous analysis, the non-parametric analysis of longitudinal data revealed
a significant main effect for group (Wald χ2 = 347.397, df = 1, p < 0.001) and for competition years (Wald χ2 = 672.449,
df = 7, p < 0.001), and a significant interaction effect between competition years and group (Wald χ2 = 291.231,
df = 7, p < 0.001). However, no significant main effect was found for RAE (Wald χ2 = 2.592, df = 3, p = 0.459),
demonstrating that the overall success score was not affected by birth quarter.
Nevertheless, a three-way interaction effect between group, birth quarter, and the competition years
(Wald χ2 = 51.727, df = 21, p < 0.001), and a two-way interaction effect between competition years and birth quarter
(Wald χ2 = 54.258, df = 21, p < 0.001) were found. The interaction effect between competition years and birth quarter
was significant for both the PRO (Wald χ2 = 48.961, df = 21, p = 0.001) and NON-PRO (Wald χ2 = 43.686, df = 28,
p < 0.003). However, for PRO, no RAE on the success score was found in any of the competition years. In the NON-
PRO, a significant effect of birth quarter was found for U17year1, but not for any other age category. For an overview
of success score per group, category, and birth quarter, see Table 2.15..
125
Table 2.15.: Average success score (mean ± standard deviation) of future professional and non-professional cyclists for each
competition year per birth quarter. Same superscript letter (a, b) indicates significant differences between two groups
(p < 0.05).
Abbreviations: PRO, future professional; NON-PRO, future non-professional; Q1, quartile 1; Q2, quartile 2; Q3, quartile 3;
Q4, quartile 4.
Q1 Q2 Q3 Q4 χ2 df p
PRO n=8 n = 12 n = 15 n=8
U17year1 20.75 ± 29.41 18.00 ± 20.28 22.53 ± 28.02 10.63 ± 6.99 0.600 3 0.896
U17year2 43.25 ± 34.40 44.67 ± 30.89 48.07 ± 34.30 32.63 ± 28.19 1.435 3 0.697
U19year1 23.88 ± 30.72 32.75 ± 27.17 28.60 ± 30.93 31.25 ± 18.44 2.102 3 0.552
U19year2 56.13 ± 56.95 73.08 ± 48.71 44.93 ± 30.35 58.75 ± 28.23 2.716 3 0.438
U23year1 10.13 ± 20.57 10.92 ± 15.77 5.00 ± 6.22 3.50 ± 3.30 0.602 3 0.896
U23year2 7.50 ± 10.20 37.08 ± 42.64 21.93 ± 25.74 18.38 ± 17.26 5.450 3 0.142
U23year3 13.75 ± 15.07 38.75 ± 47.72 37.53 ± 44.08 27.25 ± 32.86 1.070 3 0.784
U23year4 43.00 ± 40.09 61.67 ± 59.68 44.20 ± 37.59 46.00 ± 43.98 0.454 3 0.929
NON-PRO n = 381 n = 340 n = 328 n = 254
U17year1 7.09 ± 10.11a 7.14 ± 10.45b 6.55 ± 11.17 4.96 ± 8.19a,b 15.478 3 0.001
U17year2 15.54 ± 17.06 15.61 ± 17.13 16.93 ± 19.45 13.28 ± 16.04 5.553 3 0.136
U19year1 4.90 ± 12.60 4.34 ± 8.31 4.89 ± 13.49 4.93 ± 9.60 1.698 3 0.637
U19year2 7.43 ± 18.11 7.27 ± 14.99 8.86 ± 21.36 9.22 ± 20.23 4.344 3 0.227
U23year1 0.34 ± 2.87 0.25 ± 1.65 0.54 ± 3.41 0.81 ± 5.64 4.281 3 0.233
U23year2 0.64 ± 4.17 1.14 ± 6.12 1.28 ± 8.45 1.42 ± 8.04 6.885 3 0.076
U23year3 0.83 ± 4.89 1.29 ± 5.49 1.88 ± 12.01 2.61 ± 13.84 6.421 3 0.093
U23year4 0.86 ± 6.89 1.13 ± 5.55 1.02 ± 5.83 1.83 ± 9.25 5.474 3 0.140
Predictive value of youth success
The logistic regression analysis indicated that the success score in the competition years of the U17, U19,
and U23 categories may predict becoming professional (p < 0.005). Table 2.16. shows results of logistic regression
analysis per category. With every additional point in success score, cyclists in the first and second year of the U17
category have 4.8 % and 4.2 % higher odds to reach professional level, respectively. Cyclists in the first year of the
U19 category have 5.1 %, and in the second year 3.8 % higher odds to reach professional level with every additional
point in success score. With every additional point in success score, cyclists in the first year of the U23 category
have 10.0 %, in the second year 7%, in the third year 5 %, and in the fourth year 8 % higher odds to reach professional
level, respectively. See table 2.16. for how the success score relates to the probability to reach the professional level
in current population.
126
Table 2.16.: Results of the logistic regression and probability of reaching the professional level.
Abbreviations: CI, confidence interval; AUC, area under the curve.
U17year1 U17year2 U19year1 U19year2 U23year1 U23year2 U23year3 U23year4

Intercept -3.920 -4.500 -4.008 -4.245 -3.616 -3.807 -3.815 -4.355
Coefficient 0.047 0.041 0.050 0.037 0.095 0.066 0.046 0.079
Odds Ratio 1.048 1.042 1.051 1.038 1.100 1.068 1.047 1.083
CI (95 %) 1.031 - 1.065 1.032 - 1.052 1.038 - 1.065 1.029 - 1.047 1.064 - 1.137 1.049 - 1.088 1.035 - 1.059 1.065 - 1.100
AUC 0.709 0.797 0.844 0.930 0.815 0.904 0.856 0.846
Success
Probability of becoming professional (%)
Score
0 1.95 1.10 1.78 1.41 2.62 2.17 2.16 1.27
1 2.04 1.15 1.87 1.47 2.87 2.32 2.25 1.37
5 2.45 1.35 2.28 1.70 4.15 3.00 2.70 1.87
10 3.07 1.66 2.91 2.04 6.53 4.12 3.37 2.76
15 3.85 2.03 3.71 2.45 10.12 5.65 4.20 4.05
20 4.82 2.49 4.71 2.94 15.35 7.69 5.22 5.91
30 7.48 3.72 7.54 4.22 32.02 13.89 8.02 12.21
40 11.43 5.52 11.86 6.02 55.02 23.80 12.12 23.52
60 24.77 11.82 26.81 11.91 89.19 53.93 25.65 60.08
80 45.64 23.51 49.93 22.22 98.23 81.44 46.32 88.04
100 68.17 41.35 73.08 37.63 99.73 94.27 68.35 97.30
120 84.52 61.78 88.08 56.04 99.96 98.40 84.38 99.44
140 93.30 78.76 95.26 72.92 99.99 99.57 93.11 99.88
160 97.26 89.48 98.21 85.05 100.00 99.88 97.13 99.98
180 98.91 95.12 99.33 92.32 100.00 99.97 98.83 100.00
200 99.57 97.81 99.75 96.21 100.00 99.99 99.53 100.00
127
Discussion
The main finding of this study is that PRO cyclists outperformed the NON-PRO in all the youth categories
that were considered (U17, U19 and U23). From U17 category onwards, results of road cycling races may therefore
provide an indication of the probability of becoming a professional cyclist: with every additional point in success
score, cyclists have 4 to 10 % higher odds to turn professional, with U23 year1 presenting the best prediction value. In
addition, the relatively older cyclists do not have any advantage over other cyclists within the same age group in
all competition years (except U17year1 for the NON-PRO group): both PRO and NON-PRO were equally distributed
across birth quarters, indicating no RAE in Italian cycling youth categories from the U17 category onwards.
Differences success between future professional and future non-professional
PRO cyclists gathered more points in the annual national ranking than NON-PRO in all the competition years
that have been analyzed. This finding is in line with Mostaert and colleagues (2021b), which found that future
professional Belgian cyclists in U17year1, U17year2, U19year1 and U19year2 achieved better race results in provincial and
national championships compared to cyclists who did not reach professional level. Accordingly, also Svendsen and
colleagues (2018) found that World Tour (i.e. the highest road cycling competition level) Norwegian cyclists placed
significantly better in U19 National Road Championships compared to Continental (i.e. an lower competition level
than World Tour), club, and retired athletes. Therefore, evidence suggests that superior race results from U17
onwards are associated with a future professional road cycling career. In a recent study, Gallo and colleagues (2021)
reported that U17 riders who scored at least one point in the annual Italian national ranking possess significantly
higher absolute and relative power output at 2 mmol·L-1 and 4 mmol·L-1 of blood lactate concentration compared
to riders who did not score points. Menaspà and colleagues (2010) found that higher classified cyclists in U19 annual
Italian national ranking have higher physiological aerobic characteristics compared to cyclist who scored less
points. On the same line, Svendsen and colleagues (2018) showed that, in the U19 category, cyclists who reach the
professional World Tour level possess a higher maximal aerobic power compared to cyclist who not compete at
World Tour level. Linking our finding with those of these studies, it seems that possessing superior aerobic
physiological parameters in youth categories is associated with a future professional road cycling career. Whether
these physiological characteristics are genetically inherited or gained throughout superior training strategies
and/or others environmental interactions during childhood and adolescence, remains unknown.
Transitions across youth categories
When athletes transitioned to a new competition category (i.e. U17year2 to U19year1 and U19year2 to U23year1), a
significant decrease in success in both PRO and NON-PRO was observed. Conversely, we found an increase in success
128
score after one-year experience within the same category (i.e. U17 year1 to U17year2, U19year1 to U19year2 and, only in PRO,
U23year1 to U23year2). These results are in line with the findings of Mostaert and colleagues (2021b), who also reported
a significant decrease in success score when athletes move to a new age-category. However, in their study, this
effect reached significance only in the non-achiever group. The current study shows that this decrease in success
score is also present in the future-achiever group. The difference between these two studies could be the result of
the current study adopting a more extensive scoring system and analyzing a wider range of races. This limited the
influence of contextual factors (e.g., crashes, punctures, health problems, etc.) on success score and potentially
provided a more correct representation of the athletes’ performance.
Giving the decrease in success score when moving to the next category, our study highlights the importance
of an appropriate psychological support when changing the category. Specifically, in order to prevent negative
psychological and behavioral consequences (e.g., decline of self-confidence and/or overtraining), which could lead
to a premature dropout from competitive cycling (Cervelló et al., 2007), athletes must be made aware and
conscious that the lower competition success is a natural consequence of racing with competitors with superior
training and racing experience. Interestingly, we found that the higher decrease in success occurs when
transitioning from U19year2 to U23year1 in PRO cyclists. This is not surprising, as the U23 category comprehends four
competition years, while all the other youth road cycling categories include only two competition years. Hence,
while U19year1 compete against cyclists no more than one year older, U23year1 race against other competitors who
could be even two or three years older, thus increasing the probability to be outperformed even by less talented
cyclists.
Relative age effect
In contrast to findings in some other sports (Hancock et al., 2013; Mann & van Ginneken, 2017), relatively
older cyclists do not achieve more success compared to other cyclists across all the competition years, except for
U17year1 in NON-PRO. This is in line with Mostaert and colleagues (2021b), who also found no RAE on success in U19,
while it was still present in U15. The age category is known to be a moderator of risk of RAE in sports, with the risk
that gradually declines while progressing towards adulthood (Cobley et al., 2009). Accordingly, this study
underlines the findings of Mostaert and colleagues (2021b) that in road cycling the RAE tends to disappear after
puberty. Possibly, the RAE does not play a role in the chances of becoming a pro-cyclist because important
regional/national selections, and the extra training opportunities that go with a selection, take place later
compared to other sports (usually only from U19 on)(Webdale et al., 2020).
129
Probability of becoming professional cyclists
Success in youth competitions provides an indication of probability of becoming professional cyclists from
the U17 onwards, with U23year1 presenting the best prediction value (see also Table 2.16.). This agrees with Svendsen
and colleagues (2018), who found that places at U19 national Championships have good accuracy in predicting
whether a cyclist would later reach World Tour level. Also, the study of Mostaert and colleagues (2021b) reported
that U17 and U19 competition results have predictive value on whether cyclists become professional or not. On the
other hand, they found that U15 competition success did not have a potential prediction on future successes. In our
study, the higher predictive value was found for U23 category. Therefore, from these findings, it seems that the
prediction power of youth success for being a professional cyclist increases as the categories level-up. When
considering single competition years, the highest prediction value was found for U23 year1. For example, according to
the current sample, obtaining 40 points in the first year of U23 would result in a probability of 55% to reach the
pro-level as an adult, while the same score in U17 and U19 would only result in a probability of 5-12%. The U23
category includes a higher number of competition years compared to U17 and U19. Thus, as discussed above, U23year1
is the only youth competition year in which cyclists have to face competitors who are two or three years older.
Hence, it is not surprising that both NON-PRO and PRO achieve their lower scores in this competition year.
Nevertheless, it seems that due this greater difficulty of competing against more experienced cyclists, a higher
success score in this competition year provides the best prediction for turning professional as an adult. In other
words, it seems that the ability to adjust to a different competition level, with more experienced opponents, is what
could discriminate best between PRO and NON-PRO.
Practical applications
Overall, these findings give useful insights for sports stakeholders and practitioners in evaluating youth
road cycling race results for talent identification and development. Professional teams should particularly consider
hiring the most successful cyclists in the U23year1, as this is the competition year in which success score presented
the greatest predictive value on whether cyclists become professional or not. From another point of view, the
decrease in success score when transitioning from a new category in both future professional and non-professional
cyclists could induce national cycling federations to split the youth categories on annual basis. This could help
youth ego-oriented athletes not to lose motivation during the first year in the age category.
130
Conclusions
This study confirmed some of the previous findings on a Belgium cohort, and extended to an Italian road cyclists’
cohort: (i) youth competition success predicts the odds to become a professional cyclist with predictive value
increasing with age-category; (ii) no RAE from the U17 category onwards, which means older cyclists within the
same annual age group are not advantaged compared to their younger counterparts. A novel and interesting
finding of this study is that the most predictive value was reported for the U23 category, particularly the U23year1
competition year. The transition presenting the highest decrease in success score was U19 year2-U23year1. This indicates
that handling the most difficult transition well, and therefore having success competing against older and more
experienced cyclists, is an important factor for talent identification in youth cycling.
Acknowledgements
The results of the study are presented clearly, honestly, and without fabrication, falsification, or
inappropriate data manipulation. The authors thank all the athletes and the clubs involved in the study for their
contribution. Authors declare no conflict of interests.
131
References
Abbott, A., & Collins, D. (2002). A theoretical and empirical analysis of a'state of the art'talent identification model.
High ability studies, 13(2), 157-178.
Cervelló, E. M., Escartí, A., & Guzmán, J. F. (2007). Youth sport dropout from the achievement goal theory. Psicothema,
65-71.
39(3), 235-256.
1049-1051.
Hancock, D. J., Ste-Marie, D. M., & Young, B. W. (2013). Coach selections and the relative age effect in male youth ice
hockey. Research quarterly for exercise and sport, 84(1), 126-130.
Hosmer Jr, D. W., Lemeshow, S., & Sturdivant, R. X. (2013). Applied Logistic Regression (Vol. 398): John Wiley & Sons.
Mann, D. L., & van Ginneken, P. J. (2017). Age-ordered shirt numbering reduces the selection bias associated with the
relative age effect. Journal of sports sciences, 35(8), 784-790.
Mostaert, M., Vansteenkiste, P., Pion, J., Deconinck, F. J., & Lenoir, M. (2021). The importance of performance in
competitions as an indicator of future success in cycling. European Journal of Sport Science, 1-10.
doi:10.1080/17461391.2021.1877359
Pizzuto, F., Bonato, M., Vernillo, G., La Torre, A., & Piacentini, M. F. (2017). Are the world junior championship finalists
for middle-and long-distance events currently competing at international level? International journal of
sports physiology and performance, 12(3), 316-321.
Webdale, K., Baker, J., Schorer, J., & Wattie, N. (2020). Solving sport’s ‘relative age’problem: A systematic review of
proposed solutions. International Review of Sport and Exercise Psychology, 13(1), 187-204.
132
PART 3 : GENERAL DISCUSSION
‘Measuring performance to predict performance – the development of the Cycling Compass’ refers to talent
research within cycling. Test performances measured with a test battery were examined within the different cycling
disciplines (road, track, cyclo-cross and MTB), aiming to get an understanding into the similarities and differences
of the athletic profiles to facilitate participation in all the cycling disciplines. The test performances and current
cycling performances were investigated within road cycling to predict future cycling performance and identify high
potentials. The main findings of these studies can be found in section 1 of the general discussion. Next, the findings
are discussed within a broader framework and supplemented with results from two pilot studies (section 2). In the
third section, the findings have been combined into a practical tool, the ‘Cycling Compass’, to support daily practice
for federations, coaches, professional cycling teams, athletes, and other stakeholders. The research presented in
the previous chapters, and the pilot investigations that are part of the general discussion have sought to answer
some of the questions on talent identification in cycling, but at the same time have also raised several other
questions. Therefore, we conclude the general discussion with strengths, limitations, and suggestions for future
research (section 4).
1. Summary of the main findings
1.1. Investigating the athletic profiles of cycling athletes
The first objective of this dissertation was to investigate the athletic profiles of cyclists from four different cycling
disciplines. For this purpose, the similarities and differences between the road, track, cyclo-cross, and MTB cycling
have been investigated. In line with recent talent research where a (non-)sport-specific test battery was applied in
various sports (Pion et al., 2014; Opstoel et al., 2015; Pion et al., 2015; Pion et al., 2021), a similar test battery was
used for the first time within cycling (study 1). The athletic profiles of road, track, cyclo-cross and MTB were
compared using a test battery consisting of anthropometric, physical, motor coordination and sport-specific tests
(study 1). The results showed that the athletic profiles for the four cycling disciplines could only be distinguished
when athletes reached a young adult level (≥ 16 years), with significant differences for anthropometry, physical,
motor coordination and cycling-specific test(s). Using a discriminant analysis, the young adult athletes could be
correctly assigned to their discipline for 80.7%. Track cyclists distinguished themselves from the other cycling
disciplines by a higher sprint speed and better motor coordination. MTB and cyclo-cross athletes outperformed the
other disciplines on the shuttle bike agility test. Road cyclists, however, showed significant overlap in performance
characteristics with the other disciplines. Despite the less pronounced discriminative character in the adolescent
cycling population (12.0 – 15.9 years; 51.0%), the findings are in line with the results of the young adults. These
findings indicate that test performances measured with a test battery can provide valuable information about
137
talent orientation and transfer to coaches and federations in late specialisation sports such as cycling with all its
specific disciplines.
1.2. Prediction of future performance in road cycling
1.2.1. Predictions based on a (non-)sport-specific test battery
The second objective of this dissertation was to gain insight into the extent to which test performance obtained
through a test battery can predict future cycling performance. The extent to which a (non-)sport-specific test battery
can predict performance and identify talent within road cycling has not been investigated before. In other sports,
it has been shown that this test battery can be used as a talent identification tool (Vandorpe et al., 2012; Matthys
et al., 2013; Deprez, 2015; Mostaert et al., 2016; Norjali Wazir et al., 2018; Mostaert et al., 2020). Study 1 has already
shown that the test performances could differentiate the athletic profiles of adolescent riders and could be applied
as a talent orientation tool. Additionally, research has shown that physiological parameters such as aerobic and
anaerobic performance levels, power profiles, cycling economy are helpful to identify the better performers within
the junior and senior categories. However, research has shown that these characteristics alone are not sensitive
enough to predict future competitive success at adult age (Menaspà et al., 2010; Sanders & van Erp, 2020). These
findings suggest that alternative methods of identifying and predicting talent in road cycling should be examined
(study 2). A (non-)sport-specific test battery was evaluated in U15 and U17 male road cyclists at baseline in study
2. The results showed that the test performances in the U15 accounted for 22.6% of the variance in future cycling
success two to three years later. Maturity and general motor coordination were the only variables that uniquely
explained +/- 5% future success. The results emphasise that having a high level of motor coordination in road
cycling may be promising for future success. In addition, the possible influence of maturity on future competition
outcomes should be taken into account in youth cycling. This could be explained by the fact that early maturing
athletes seem to have a better chance of achieving better race results two to three years later. From the U17
category onwards, neither maturity, relative age, motor coordination, physical performance, nor cycling-specific
performance could predict race outcomes in the U19year2 category. Given the reduced predictive value from the U17
category onwards, these tests alone do not seem sufficient to predict future success reliably. In summary, the
findings provide a better understanding of the usefulness of a (non-)sport-specific test battery for young cyclists,
to predict future cycling performance two to three years later. These findings can aid coaches and federations when
they consider (de)selection in function of talent identification.
138
1.2.2. Predictions based on competition results
The third aim of this dissertation was to gain an understanding into the extent to which current cycling performances
during competitions and relative age can predict future cycling performance. Coaches and federations regularly use
competition results as an evaluation tool. However, research is needed to examine the predictive value of cycling
performance in road cycling. The results of study 3 showed that estimating potential based on competition results
was unpredictable in the U15 category, but the importance of cycling performance increased with increasing age,
while the influence of the relative age effect (RAE) decreased. It was found that only in the U15 category and to a
lesser extent in the U17 category, the relatively older road cyclists performed significantly better than their younger
peers. For U17 and U19 cyclists, each additional top 10 result in national and provincial championships seemed to
be associated respectively with 3-5% and 6% more chance of reaching the elite level in adulthood. The most
accurate predictions for performance at the senior elite level were found in the U19 category and for World Tour
riders. Furthermore, the findings showed significant variability in success rate in the different race categories
indicating that the career trajectories are rather nonlinear (Güllich & Emrich, 2014). Therefore, it is suggested that
the cycling performance should be interpreted with caution.
Additionally, the third study's findings were verified with research results from abroad, e.g., Italy (study 4). A
detailed ranking system was used to examine the career trajectories and the predictability of cycling performance,
with the difference that all races in the youth years (U17, U19) and U23 were included. The results showed that the
PRO cyclists performed significantly better in all (youth) categories (U17, U19, U23) compared to the NON-PRO
cyclists. From the U17 category onwards, the road cycling competition results showed an indication of the chance
of becoming a PRO cyclist. With each extra point in the success score, cyclists have a 4 to 10 % higher chance of
becoming a PRO, with U23year1 having the best predictive value. In accordance with the findings of studies 3 and 4,
athletes experienced a decrease in performance when transitioning to a new category (i.e. U17year2 to U19year1 and
U19year2 to U23year1) and an increase in performance after one year of experience within the same category (i.e. U17year1
to U17year2, U19year1 to U19year2 and, in PRO-only, U23year1 to U23year2). These insights into the career development of
young cyclists up to (elite) adult level can be used as a reference for coaches, professional cycling teams, and
federations when evaluating cycling performance for talent identification and development.
139
2. Overall discussion
2.1. Motor coordination
2.1.1. The role of motor coordination in the talent process
One of the main findings of this dissertation was that, similar to other sports, motor coordination seems to play an
essential role in the development of cycling athletes (studies 1 and 2). Motor coordination, which also refers to
fundamental movement skills, motor competence, or motor performance (Robinson et al., 2015), can be defined as
the ability to perform a wide range of motor skills proficiently (Burton & Miller, 1998; Gabbard, 2015; Haga, 2008).
These fundamental movement skills provide the basis for performing more advanced and specific movement skills
useful in cycling. Around the age of five years, children start to learn to ride a bicycle (Figure 3.1.). The motor
component for cycling includes skills such as steering, pedalling, balancing, and braking. Also, the perceptual-motor
component involves perception, attention, concentration, planning and decision making (Briem et al., 2004; Corden
et al., 2005; Zeuwts, 2016). Cycling skills can be defined as the ability to control the bicycle in such a way that the
bicycle becomes part of the body, enabling the cyclist to perform all required tasks (Briem et al., 2004; Van Houcke
et al., 2009). Cycling skills are subject to age-related improvements. Until the age of 14 years, the control over the
bicycle and the quality of movement execution increases. The most significant improvements are observed between
five and ten years of age. More experienced cyclists are suggested to have better control over their bicycles
(Schepers, 2013). The earlier a child learns to cycle, the greater his or her experience and the better his or her
performance will be. Hansen et al. (2005) found that the age at onset of cycling is a stronger predictor of cycling
skills than the amount of time spent on the bike. Consequently, having a well-developed motor coordination seems
also crucial in cycling.
Figure 3.1.: Learning how to bike (Ayelet-Keshet).

140
2.1.2. Additional analysis towards detecting motor coordination levels at a young age
Although the main part of this dissertation did not focus on motor coordination of younger (< 12 years) cyclists, it
is hypothesised that motor coordination is crucial for development in sports, even in sports like cycling where the
role of coordination is not very pronounced at first glance. However, data are available to explore motor
coordination of younger cyclists in this overall discussion. During the extensive data collection for this dissertation,
between 2014 and 2019, data was also collected from cyclists under the age of 12 years. A post-hoc analysis was
conducted to compare the performances on the motor coordination tests between (young) cyclists, BMX athletes,
and a healthy reference population that is (not) formally engaged in organised sport (detailed information in
Addendum A). BMX athletes were considered as a separate group because a) they have not been studied explicitly
in the previous chapters and b) we hypothesise that, given the nature of the discipline, they might outperform
other cyclists in general motor coordination. Based on findings by Opstoel et al. (2015), it is expected to find higher
levels for motor coordination in children who train the most hours per week. The differences and similarities in
athletic profiles on the motor coordination, but also on the anthropometric and physical tests are displayed in the
spider plot (Figure 3.2.), discussed and displayed in detail in the addenda (Addendum A: Tables A.1. – A.2.; Figures
A.1. – A.3.). For motor coordination, it was found that in the youngest age group (7.0 – 11.9 years), cyclists and BMX
cyclists performed significantly better on the jumping sideways and moving sideways tests compared to the
(non-)sporting reference population. In addition, the BMX athletes seem to outperform the cyclists on the motor
coordination tests (significant differences found on the jumping sideways test).
At a young age, cyclists and BMX cyclists already possess significantly better motor coordination skills compared
to the healthy (non-)sporting reference population. These better performances can be linked to the motor skills
required for cycling. For example, the cyclists are used to maintain their balance on a bike (possible link with the
performances on the balance beam), perform an optimal cadence with their legs (possible link with the
performances on jumping sideways), while controlling the direction with the handlebars and speed with the breaks
(possible link with the performances on moving sideways). In addition, the higher levels of motor coordination
performance may also be due to the cyclists and BMX riders practising sports regularly. Similar results were found
by Opstoel et al. (2015) and De Meester (2017), indicating that regardless of the type of sport, sports participation
(hours spent within sport(s); organised sports) at a young age positively contributes to the development of the
child’s motor coordination. These types of activities provide more opportunities to learn and refine motor skill
executions (Fisher et al., 2005; De Meester, 2017).
141
The critical period in development ('window of opportunity') for learning motor coordination is before the age of 9
in boys (Ford et al., 2011). However, Lloyd and Oliver (2012) found that the development of motor coordination
should not be limited to this age range. Throughout development, motor coordination and other physical
performance variables should be trained, but the response can differ depending on the stage of growth (Lloyd &
Oliver, 2012). In order to optimally develop motor coordination, it is recommended to challenge children’s motor
skills from an early age through various forms of movement, practising different sports and disciplines. Fransen et
al. (2012) also suggested that boys aged 10–12 years should participate in various sports. Boys participating in
different sports outperform the others specializing in a single sport on motor coordination. Being engaged in
diversified training and free play may also positively influence their fitness level, motor coordination and sports
development (Sieghartsleitner et al., 2018).
Body height
BB 1,2 Sitting height
JS 0,8 Body weight
0,4
MS 0,0 BMI
-0,4
CMJarms -0,8 SBJ
-1,2
CMJHips SAR
Knee push ups ESHR
Curl ups 30m sprint at 5m

30m sprint at 30m 30m sprint at 10m
30m sprint at 20m
Reference population (no sport) Reference population (sport)

Cycling BMX
Figure 3.2.: The spider plot displays the z-scores of the different (non-)sport-specific tests for the (non-)sporting reference
population, cyclists and BMX cyclists. A total sample of 348 children between 7.0 – 11.9 years of age (110 reference
population no sport; 105 reference population sport; 86 cyclists, 47 BMX cyclists) were included in spider plot of pilot study 1
142
2.2. Talent orientation and transfer
2.2.1. Road, track, cyclo-cross, and MTB
Novak and Dascombe (2014) examined the similarities and differences between the athletic profiles of elite adult
road, BMX and MTB cyclists based on physiological tests. Mujika and Padilla (2001), Peinado et al. (2011), Sanders &
van Erp (2020), and Lucia et al. (2000) instead focused on the differences and similarities between the
specialisations within elite adult road cycling. Menaspà et al. (2012) conducted a similar study based on a
physiological test battery, comparing the athletic profiles of junior road cyclists. These studies showed that
different physiological profiles could be distinguished from the junior and elite adult level onwards. Nevertheless,
these studies lack uniformity in the physiological test procedures, which complicates the comparison between
studies. Moreover, these comparative studies are limited to the U19, U23, and elite adult categories. In contrast,
study 1 compared the athletic profiles of four different cycling disciplines (road, track, cyclo-cross, and MTB), using
a standardised (non-)sport-specific test battery in a population of adolescent (12.0 – 15.9 years) and young adult
(≥ 16 years) cyclists. Consequently, the athletic profiles could be studied in the younger age groups and could be
easily compared with other talent research.
The results from study 1 showed that there was an overlap of roughly 50% between the four disciplines in the
adolescent age group (12.0 – 15.9 years), while the differences were more pronounced in the young adult group
(≥ 16 years). Similar results were found in studies investigating the similarities and differences between different
sports disciplines and age groups (Leone & Larivière, 1998; Pion et al., 2014; Pion et al., 2015; Pion et al., 2021). The
study of Pion et al. (2014) indicated that combat sports could be 100% correctly discriminated from each other in
the U13 and U18 categories, however, the differences in performance on the different tests of the test battery were
only significant in the U18 category (Pion et al., 2014). A second study by Pion et al. (2015) found that a non-sport-
specific test battery could be used to distinguish young elite male athletes (U18; 16.1 ± 0.8 years) from different
sports (badminton, basketball, gymnastics, handball, judo, soccer, table tennis, triathlon, and volleyball).
Obviously the discrimination between different sports is more straightforward than identifying differences and
similarities within the group of cycling athletes of different cycling disciplines. Study 1 also indicated that the profile
of road cyclists overlapped significantly with the other three disciplines, both for adolescents and young adults.
This might be due to the seasonal schedule of road cycling, which is compatible with the competition season for
the three other disciplines. This results in many cyclo-cross, track and MTB cyclists combining their discipline with
road cycling. Apart from this, the high variability within road cycling might be explained by the emerging
143
specialisations (sprinting, climbing, etc.) from the U19 category onwards (Mujika & Padilla, 2001; Peinado et al.,
2011). The cycling specialties are characterised by differences in anthropometric and physiological characteristics
(Padilla et al., 1999; Lucia et al., 2000; Impellizzeri et al., 2008; Menaspà et al., 2012). Similar results were found by
Menaspà et al. (2012) indicating that junior road cyclists could also be classified into different cycling specialities.
These findings could also be applied to other sports, for example, team sports such as basketball, soccer, and
volleyball which may be characterised by an athletic profile that is to some extent overlapping, while the different
playing positions are still associated with different performance characteristics. For example, Pion et al. (2018)
showed that the different playing positions of professional basketball players can be distinguished by means of
the anthropometrical, physiological and physical tests. Milić et al. (2017) found that young female volleyball players
(aged 13 to 15 years) in various playing positions differed significantly in height and all three somatotype
components, whereas no significant differences were found on the physical performance measurements. Our
research has shown that there is no such thing as ‘the profile’ of the young cyclist, but that the different sub-
disciplines are characterised by similarities and differences, which opens the door for reorientation, transfer, or
combing different sub-disciplines in cycling.
2.2.2. Additional analysis towards the profile of BMX
As indicated before, study 1 examined the differences between the cycling disciplines road, track, cyclo-cross and
MTB, except for BMX. BMX was not included in study 1 as it was considered a different discipline within cycling since
its technical character, unique competition format (+/-30 sec effort on a short but challenging track), and the use
of a different type of bicycle (Novak & Dascombe, 2014). However, BMX is an Olympic discipline, and further research
is needed in the field of talent orientation and transfer. For this purpose, a second pilot study was carried out. The
athletic profiles of BMX, road, cyclo-cross, track, and MTB cyclists were examined in a sample of young adolescents
between 12.0 – 15.9 years of age. The spider plot below visualises the BMX’ athletic profile compared to the other
athletic cycling profiles, using z-scores (Figure 3.3.).
144
Shuttle bike
Body height 1,0 SBJ
ESHR 0,8 CMJarms
0,6
Body weight 0,4 CMJHips
0,2
0,0
SAR JS
-0,2
-0,4
Curl ups -0,6 Maximal cadence
30m sprint at 5m Knee push ups
30m sprint at 10m BB
MS 30m sprint at 30m

Planking BMI
30m sprint at 20m
BMX Road Cyclo-cross Track MTB
Figure 3.3.: The spider plot displays the z-scores of the different (non-)sport-specific tests for athletes between
12.0 – 15.9 years of age from five different cycling disciplines (BMX, road, cyclo-cross, track, and MTB cycling).
The spider plot showed that BMX athletes were smaller and had higher body weight and BMI compared to the other
cycling disciplines. Additionally, the plot indicated that the BMX cyclists outperform the other cycling disciplines on
the explosive tests (SBJ, CMJ, knee-push ups, 30m sprint tests) and skill-based tests (BB, MS, JS, shuttle bike and
maximal cadence). However, the MANCOVA results revealed that the significant differences on the jumping
sideways and maximal cadence tests disappeared when controlling for the covariates maturity, RAE, and training
history (Addendum B - Table B.1.). Furthermore, it was found that the cycling disciplines scored significantly better
on the ESHR compared to the BMX cyclists. The results are discussed and displayed in detail in the addenda
(Addendum B – Table B.1.; Figure B.1.).
A possible explanation for the better performance on motor coordination tests by BMX cyclists is that they are
familiar with performing complex skills such as jumping, “pumping”, keeping balance, and avoiding obstacles at
high speed in a short time. BMX seems to be more challenging with respect to motor coordination compared to the
other cycling disciplines. The better results found on the explosive tests (SBJ, CMJ) for the BMX cyclists can be
associated with the starting position in BMX races, which is considered to be a performance determining factor for
the final position of the race. In the first 8-10 seconds of a race, which requires well-developed lower limb muscle
strength (Rylands & Roberts, 2014). The differences found on the physical tests are supported by the study of Novak
145
and Dascombe (2014). They found that adult BMX cyclists could be differentiated from road and MTB cyclists in
terms of physiological parameters. Higher explosive power output values were registered for BMX athletes
compared to the others, while road and cross-country MTB cyclists achieved higher aerobic fitness values. The study
of Lievens et al. (2021) showed that adult elite BMX athletes, with the exception of track sprinters, differ from the
other cycling disciplines in muscle typology. BMX and track sprint cyclists have a fast muscle typology compared to
the other cycling disciplines, with a slow to intermediate muscle typology. In BMX and track sprint cycling, the
ability to generate high speeds and power outputs in a short time is an essential performance indicator (Janssen
& Cornelissen, 2017).
Further research is needed to identify potential transfer (cross-over) between BMX and the other cycling disciplines.
Successful cross-overs between sports or disciplines are possible when athletes are competent in comparable
(transferable) characteristics within different sports or disciplines (Teunissen et al., 2021). Early involvement in BMX
(from the age of 5) may be beneficial for mastering complex technical skills. BMX cyclists reach their peak
performance at a relatively young age (23.2 - 25.0 years), at least in comparison to the other cycling disciplines
(Longo et al., 2016). In other words, it can be assumed that the transition to BMX at a later age (> 16 years) can
cause difficulties in learning the technical complexity specific to the sport, which may prevent these athletes from
reaching the top. Therefore, talent orientation or transfer towards BMX is suggested at an age younger than 16
years. However, athletes may still benefit from combining different sports and disciplines. Further research is
needed to investigate the maximum age to enter the discipline in order to become successful. Similar research
within gymnastics (early specialisation sport) showed that it is advisable to start at a young age (4-5 years), but
also confirmed that for extreme talents, it is still possible to excel in gymnastics after the age of 10 (Zurc, 2017).
Contrarily, it can be beneficial for BMX cyclists to transfer towards another discipline or sport, or some athletes will
be forced to do so. Possible reasons for this kind of talent transfer are (a) the difficulty of the BMX trail increases
as the athletes get older. In addition, not every BMX cyclist is able to perform the track with an 8 m high start ramp;
(b) BMX performance may drop-out due to injuries or falls; (c) specialisation starts at a young age. Young athletes
who specialise too early can lose their motivation or drop-out of the sport (Baker, 2003). Talent transfer has a
chance of succeeding when the athletes have well developed fundamental movement skills and when transitioning
to a complementary discipline (Pion et al., 2021; Teunissen et al., 2021). For example, the study of Rylands and
Roberts (2014) found similarities in the athletic profiles of BMX and track sprint cyclists (Janssen & Cornelissen,
2017). From the age of 12, BMX cyclists could start participating or transferring to the indoor track sprint discipline.
The British Cycling Talent development pathway is a practical example of stimulating the transfer from and to BMX
in the U15 category (Staff et al., 2021; British Cycling, 2021-2022).
146
2.3. Talent identification, selection and performance prediction
2.3.1. Longitudinal research in this dissertation
Talent identification research aims to identify the key characteristics for success, but more importantly, future
success. Studies limited to cross-sectional research have already attempted to distinguish the elite from sub-elite
athletes within the same category and sport, using either a sport-specific, or non-sport-specific test battery. The
results showed that such test batteries could distinguish athletes of different levels within several sports such as
figure skating, badminton, football, handball, gymnastics, and combat sports (Mostaert et al., 2016; Norjali Wazir et
al., 2018; Robertson et al., 2021). However, cross-sectional research can only partially consider the talent's dynamic
nature and development (Vaeyens et al., 2008). Longitudinal research instead measures the progression
possibilities and potential of the athletes (Vaeyens et al., 2008; Baker et al., 2018). Therefore, a longitudinal design
was preferred in the studies 2 – 4, to investigate whether test and current cycling performances could predict
future cycling performance. Based on the findings, it is suggested that a (non-)sport-specific test battery could be
applied in the U15 category for talent identification (study 2). From U17 onwards, both the test battery and the
competition results could be interpreted, while in the U19 category, when the athletes reach their APHV, only
competition results could be used for talent identification (studies 2- 4). These findings are similar to the ‘sliding
population approach’ of Régnier et al. (1993). This stepwise approach suggests that different test batteries should
be applied for different age groups.
The strength of studies 2 – 4 was that they value more long-term outcomes than short-term rewards (Baker et al.,
2018). The results of study 2 underlined the importance of motor coordination tests, while the physical and cycling-
specific tests did not significantly contribute to future success. Similar results were found for elite gymnasts. A non-
sport-specific motor test battery explained more than 40% of the variation in competition performance two years
later (Vandorpe et al., 2012). Neither the coaches’ judgement, the anthropometric, nor the physical characteristics
were sensitive enough to predict performance (Vandorpe et al., 2012). The study of Mostaert et al. (2020) analysed
the predictive value of the test battery of 10- to 13-year old volleyball players for reaching the international youth
competitive level 4–8 years later. The findings showed that being taller and jumping higher is related to higher
chances of being selected for the national team (female and male) (Mostaert et al., 2020). In addition, for female
players, excelling on jumping sideways also increased chances for future selection. These studies have in common
that motor coordination tests might be valuable in the early identification phase of athletes. The discriminative
and predictive qualities of motor coordination might be sufficiently powerful for selection of potential athletes
within a relatively homogeneous population exhibiting similar anthropometric and physical profiles (Vandorpe et
147
al., 2012). However, talent (de)selection in youth athletes is challenging and comes with several contemporary
concerns. Additional future research is needed to provide insight into the performance determinants predicting
future success within cycling to reduce the risk of deselecting athletes with potential (Baker et al., 2018).
Given the current uncertainty on the use of a (non-)sport-specific test battery in cycling, it was a strength of study
2 to investigate the prognostic validity of each dimension of the test battery (motor coordination, physical, and
cycling-specific performance) in terms of talent prediction. Schorer et al. (2017) conducted a similar study in a
sample of female handball players and compared the extent to which motor coordination and coach assessment
best predicted future performance. This study showed that motor coordination provided the best predictive value.
Sieghartsleitner et al. (2019) indicated that each of the classification models for talent prediction (coach
assessment, motor performance tests, and multidimensional date) contributed significantly to the prediction of
performance in elite youth football. However, a multidimensional approach was preferred rather than motor
performance only. It can be noted that these studies did not include competition performance in the analyses.
Nevertheless, studies 3 and 4 showed that competitive performance could partially predict future performance in
cycling. Therefore, it is suggested that future research should also investigate the effectiveness of a combined
approach, including test and cycling performances in the analysis, within cycling.
Competitions are mainly about winning and achieving a gold, silver, or bronze medal. On the one hand, athletes
experience pressure to improve their personal times or distances, and on the other hand, they want to win the
competition. However, it is challenging to balance short-term and long-term goals. Whether or not they experience
success could encourage continued sports participation or drop-out. The conflict between short-term and long-
term goals also potentially decreases the efficiency of the entire “talent system” (Baker et al., 2018). Therefore, the
added value of studies 3 and 4 was that the competitive career trajectories were described within road cycling,
from youth to adult elite level in Belgian and Italian riders. An interesting finding was that predicting future
performance in the U15 competition category is problematic (study 3), which is consistent with other research,
although it might vary from sport to sport (Li et al., 2018). In sports and disciplines such as cycling (studies 3, 4),
tennis (Brouwers et al., 2012; Li et al., 2020), long and high jump (Boccia et al., 2021), swimming (Brustio et al., 2021),
etc. in which competition performance at the age of 16 is not a good predictor of adult performance, it is suggested
to keep the children involved in the sport, well-coached, and motivated to discover their own limits.
Additionally, the findings of studies 3 and 4 have shown that the predictive value of cycling performance gradually
increases from the U17 onwards (3 -10% higher chance of reaching adult elite level) in road cycling, with the best
148
predictive value in the U23year1 category. However, it should be noted that even if the average results clearly
indicated that the future achievers significantly outperformed the future non-achievers, a large inter-individual
variability was present. This means that career pathways are highly variable across different athletes with complex
oscillations featuring highly varied transitions (Gulbin et al., 2013). In this perspective, some very successful
individuals in youth competitions could not become professional and, conversely, there are some with only a few
youth successes who reach professional elite level later. In line with this, a success score indicating a probability
of 90% to obtain a PRO contract would mean that the young cyclist is very promising, but still means that 1 out of
10 with this success score will not reach the PRO level. Similar results were found in the literature (Abbott et al.,
2005; Gulbin et al., 2013). Therefore, it is suggested that other potential performance determining factors should
also be taken into account such as: the (non-)sport-specific test battery, the coaches-eye, other individual
dimensions (e.g., cognitive, psychological, physiological features), tactical, strategic, contextual, and global features
and organisational constraints (Phillips & Hopkins, 2020).
Furthermore, studies 3 and 4, as well as Schumacher et al. (2006), and Svendsen et al. (2018) investigated the
between-group differences of road cyclists who reached the elite adult level or those who did not. As a result, these
studies could determine which characteristics might be relevant for talent identification, but further research is
needed to identify talent on an individual basis. The study of Van Bulck et al. (2021) is one of the first to develop a
computer-aided system to assist scouts in professional road cycling to support the selection of potential riders
who are likely to become top professional riders. The model is currently limited to the prediction of the first three
years of professional performance of riders that participated in their last U23 race. Therefore, further research is
recommended.
2.3.2. Confounding factors
In talent research of youth athletes, growth, tempo, status, and relative age were shown as confounding variables
in predicting future success (Hill et al., 2020). With increasing age, the youth athletes become taller and gain more
muscle mass due to the body's neurological, hormonal, and musculoskeletal changes during puberty (Malina et al.,
2004; Nimmerichter, 2018). Growth velocity curves show considerable inter-individual variation during adolescence
(Pearson et al., 2006; Teunissen et al., 2020). Consequently, study 2 found that athletes in the U15 category who
matured earlier according to the non-invasive estimate are more likely to achieve better competitive results two
to three years later. The results indicating that maturity status was a significant predictor of performance, are also
observed in other sports (Pearson et al., 2006; López-Plaza et al., 2017).
149
Furthermore, given the fact that maturation and relative age should be recognised as independent constructs that
exist and operate independently (Hill et al., 2020), it is suggested that future research should benefit from including
maturation as well as relative age in the analyses. Nevertheless, studies 3 and 4 investigated only the possible
influence of relative age on the performance results. An over-representation of those born in the beginning of the
competition year was found for the youngest competition categories of road cycling (U15, to a lesser extent in the
U17 category). However, this effect disappeared from the U19 category onwards (study 3). Similar results were
found in swimming, indicating that performance advantages associated with relative age are transient (Cobley et
al., 2018). In other sports such as soccer, a RAE was found in all youth categories, in which players born in the first
quarter of the selection year (from January to March) were overrepresented in the competitions. They were more
likely to be considered “talented” because they had physical advantages compared to their “younger” peers (Helsen
et al., 2005). A possible explanation for the differences found in RAE for road cycling compared to, e.g., soccer, might
be due to less (de)selection at a young age in cycling and the discipline is an individual sport. The selections for the
professional cycling teams mainly take place from the U19 or U23 category at the earliest. These athletes are
already past puberty, and the advantages they experienced by being relatively older and/or early mature may have
disappeared (Malina et al., 2013). These results suggest that it is not advisable to start (de)selecting at an
increasingly younger age in road cycling to limit the possible impact of relative age.
Moreover, the results showed a decrease in performance when transitioning to a new competition category and an
increase after one year of experience within the same category (U17, U19, U23). Study 4 observed the highest
decrease in performance transitioning from the U19year2 to the U23year1 category. Similar results were found in the
literature, indicating that the transition from junior to senior appears to be one of the most critical and complex
transitions in the athletic career (Stambulova et al., 2009). There are several possible underlying causes: (a) the
transition comes with a set of demands/challenges for practice (e.g., balance sports goals with life goals),
competitions (e.g., coping with the pressure of selections) and life in general (e.g., from school to college or
university) (Stambulova et al., 2009); (b) specific for road cycling, when progressing to a new age category the
riders have to compete against, e.g., three-year older riders who are more experienced and more developed.
Therefore, Stambulova et al. (2009) emphasise the importance of adequate psychological support (coping
resources, psychological skills training) during the various transitions in the athletes' careers in order to avoid
drop-out (Morris, 2013). How athletes judge their success could improve or decrease their well being, and could
positively contribute to more adaptive or maladaptive goal motives (Ruiz et al. 2021). In particular, athletes benefit
from judging their performances based on self-referenced criteria (task orientation) and should avoid social
comparison (ego orientation) to feel more competent (Martinez-Gonzalez et al. 2022). Also, coaches can promote
150
greater task orientation in their athletes by recognising and emphasising exerted effort, cooperation, the mistakes
as part of the learning process, and taking into account all the contributions that each athlete makes independently
of their ability (Ruiz et al., 2021). In addition, by offering athletes adapted physical training sessions and
competitions in which athletes of the same biological age or level compete together, could also positively influence
athletic development (discussed in section 4). Further research is needed to examine how factors such as the type
of the race, training history, maturity, weather conditions, etc. might affect the cycling performances and determine
whether or not an athlete achieves excellent results.
Furthermore, drop-out is a prevalent complication in the analysis of data from longitudinal studies (studies 2 – 4)
(Hogan et al., 2004). Due to the COVID-19 pandemic and drop-out, examining more extended periods of longitudinal
follow-up with the group of cycling athletes was impossible. However, it could have provided an additional
understanding of the development and the predictive value of the test battery on performance. In addition, due to
the structure of our dataset in studies 3 and 4, for the competition years in which zero points were gathered, we
were not able to distinguish cyclists who did not gather points from cyclists who dropped out from competitive
road cycling. As a result, there is a possibility for selection bias, as the riders who experienced less success during
competitions may have dropped out from cycling, and the analyses may have consisted of only the ‘better’ riders.
Nevertheless, it would have been interesting to investigate the precise mechanisms for dropping out within cycling.
For example, a follow-up study on drop-out rates in Norwegian track and field athletes between 14 and 18 years
found that dropping out is multifactorial, and 66.4% of the reasons for dropping out from the competition was
related to injuries (24.3%), school priority (21.4%) and lack of motivation (20.7%) (Enoksen, 2011). Therefore,
identifying when and why cyclists dropped out could have provided further insights into youth road cycling
dynamics.
2.4. Talent development
The present dissertation (studies 1 – 4) indicated that test performances measured with a (non-)sport-specific test
battery could be applied for monitoring athletes in all the youth competition categories (U12 – U19). On the other
hand, longitudinal follow-up of cycling performances could also be valuable, however, it is suggested only from
the U17 category onwards within road cycling. As talent development is a process which manifests itself throughout
the athletic career, benchmarks have been developed to help coaches, federations, and athletes compare
performances, identify strengths and weaknesses, and monitor the athletes' progress over time. Due to the
extensive data collection of the past six years (2014-2019) in cooperation with Cycling Vlaanderen, benchmarks
151
could be established (Addendum C ; Tables C.1. – C.5.). The data consisted of male athletes who were delegated by
the qualified club trainers to the test days based on their potential to excel compared to an average cycling athlete
of the same age, and male high performance athletes (active in the national cycling academy, national selection,
and professional riders) who volunteered to participate. Both cross-sectional and longitudinal data were included
in the benchmarks, in which duplicates within the same age were excluded. The cyclists were classified according
to their self-reported main or secondary discipline. The number of participants, mean, standard deviation, and
percentile scores (P25, P50, P75) were reported for each test (anthropometric, physical, motor coordination, and
cycling-specific tests), age category (7 – 8 years, 9 – 10 years, 11 – 12 years, 13 – 14 years, 15 – 16 years, 17 – 18 years),
and discipline (track, road, cyclo-cross, MTB, and BMX), for which sufficient data were available (Addendum C ; Tables
C.1. – C.5.).
These benchmark data should be considered as dynamic data, which should be regularly updated due to the
evolution of the sport and the changes in athletic characteristics over time (Pion, 2015; Lath et al., 2020). Coaches
should consider the biological age (i.e. stage of maturation) and relative age of the athletes when interpreting
benchmarks and test results (studies 1 – 4). They have to be aware of the characteristics of the growth spurt and
recognise that changes in growth and performance are highly individually determined (Deprez, 2015). Additionally,
the athletic profiles of youth athletes may differ between high potential athletes. The high potential athletes might
lack various skills and abilities but ‘compensate’ for them in other areas (Robertson-Martens, 2021). The so-called
‘compensation phenomenon’ allows the athletes to be selected by coaches and/or excel at the highest level of
performance (Vaeyens et al., 2008). Therefore, the coaches-eye should also be involved in interpreting the test and
competition results. Experienced coaches use an intuitive approach, and evaluations are based on the athlete as a
whole (holistic character), taking into account information from different dimensions (Buekers et al., 2015;
Sieghartsleitner et al., 2019).
152
3. Practical implications
3.1. Cycling compass
The findings of pilot studies 1-2, studies 1-4, and the age- and discipline-specific benchmark data are summarised
into a practical tool for talent development in cycling, ‘the Cycling Compass’, which can be considered a new applied
and cycling-specific version of the previously developed Flemish Sports Compass (Pion, 2015). It contains
information about the applicability of test performances measured with a (non-)sport-specific test battery and
cycling performance based on competition results, specified per competition category (U12, U15, U17, U19, U23,
professionals or non-professionals). The Cycling Compass could be applied by coaches, federations, athletes, and
other stakeholders in addition to other performance determining factors (e.g. physiological, psychological, and
environmental characteristics), which were beyond the scope of this dissertation. It is suggested to be a dynamic
and supportive tool, which needs to be updated regularly according to the needs and wishes of the coaches, and
other stakeholders. Four different levels could be differentiated and are displayed in the Cycling Compass
(Figure 3.4.):
 Talent detection
 Talent orientation
 Talent identification
 Talent development
Below we elaborate on each level of the Cycling Compass and the practical implications are discussed.
153
Figure 3.4.: Graphical representation of the Cycling Compass. Four different levels could be distinguished: talent detection, orientation, identification, and development.
Note: the symbols represent: = (non-)sport-specific test battery*, with emphasis on motor coordination; = competition results*;
* larger size = more important
= road cycling; = track cycling; = cyclo-cross; = MTB; = BMX
154
3.1.1. Talent detection
The first level of the Cycling Compass focuses on talent detection (Figure 3.4.), ‘the discovery of potential performers
who are currently not involved in the sport’ (Vaeyens et al., 2008). It is suggested that a non-sport-specific test
battery can be applied to detect the better movers at a young age (< 12 years). Children with good motor
coordination skills, seem to benefit when starting with cycling and/or BMX (pilot study 1; Addendum A). BMX athletes
require even higher levels of motor coordination compared to the cyclists. It seems appropriate that the federation
and coaches encourage young athletes to participate in various cycling disciplines and/or sports, aiming to develop
broadly. Stimulating diversification of skills at a young age could positively affect skill acquisition later in the
development.
The talent detection phase is currently organised by the Flemish Sports Compass in primary schools (3 rd and 4th
grade), assessing a non-sport-specific test battery to detect and orient children towards the sports and disciplines
that best suit their motivation, preferences, and physical/motor abilities. The strength of the Flemish Sports
Compass is that a comprehensive list of different sports is advised to the children. Separating the disciplines of
BMX and cycling in the Flemish Sports Compass can result in more diverse advice (pilot study 1; Addendum A).
3.1.2. Talent orientation and transfer
The second level of the Cycling Compass includes talent orientation and transfer advise (Figure 3.4.), to support
athletes (with potential) in the search for the cycling discipline(s) that best fits their athletic profile (road, track,
cyclo-cross, MTB, and/or BMX). It seems appropriate to promote participation in the various cycling disciplines in
youth (12 – 15 years), since mainly similarities in performance characteristics were found between the different
cycling disciplines (road, track, cyclo-cross and MTB; study 1), except for BMX (pilot study 2; Addendum B). Based on
the findings of study 1, it is suggested that coaches should focus on broad development by challenging athletes
through diversified training, free play, and encouraging them to participate in different disciplines and sports which
may positively impact athletic development. It could support athletes in their search for the discipline they enjoy
and feel most competent (De Meester et al., 2018). Based on the results of pilot study 2 (Addendum B), talent
orientation or transfer to or from BMX should occur earlier in the development compared to the other cycling
disciplines. Explosive and skilled riders (i.e., those with significantly better performances on the explosive and
motor coordination tests) could be encouraged to start or remain active in BMX. It seems appropriate to promote
talent transfer and orientation at a younger age for BMX (< 16 years) (unless they are already practising BMX in
combination with other disciplines). Additionally, it is suggested that BMX athletes may benefit from transferring
to other discipline(s) with complementary characteristics, for example, track sprint, or from combining their
155
discipline with other cycling disciplines at an early age. Depending on individual progress, goals and motivation,
the transfer from, to, and within cycling disciplines can occur at different ages. Federations could consider the
organisation of Pentathlons for youth. Through these Pentathlons, children could be encouraged to participate in
competitions of different cycling disciplines (e.g., BMX, road, track, cyclo-cross, and MTB). Consequently, they could
be challenged mentally, physically, and motorically, which could positively affect their athletic development.
From 16 years onwards, coaches and federations could support athletes searching for their main discipline(s) by
investigating the athletic profiles using the (non-)sport-specific test battery. The riders who perform well on the
explosive and motor coordination tests could be advised to start or remain active in track cycling, while the riders
with excellent bike skills could be encouraged to start with or stay active in cyclo-cross and/or MTB. The other riders
could possess potential for road cycling, however, they need to discover which speciality fits their abilities the most
in the long-term. Based on previous research, additional procedures such as physiological tests (e.g. aerobic fitness,
power-profiling, etc.) and the Muscle Talent Scan could also support the talent transfer (Menaspà et al., 2012;
Lievens et al., 2021).
3.1.3. Talent identification and performance prediction
The third level of the Cycling Compass provides advice about talent identification and future cycling performance
prediction in road cycling (Figure 3.4.), to support coaches and federations in the search for potential athletes, who
are already active in the sport. Based on the findings, it is suggested that coaches could use a (non-)sport-specific
test battery to partially identify potential talented athletes in the U15 category (study2). However, it does not
already seem appropriate to interpret the cycling performances to predict future performance of road cyclists
(study 3). As mentioned earlier, coaches should pay particular attention to the athletes' motor development and
encourage free play, participation in various sports and/or disciplines. When interpreting the performances in the
U15 category (physical, sport specific tests) the biological age (maturity status) of the athletes should be taken into
account. For example, early maturing athletes may experience an advantage because they are physically more
developed than their average or late-maturing peers, but these differences may disappear after puberty.
In the U17 category, it is suggested that coaches should not only rely on the test battery but also on provincial and
national competition results or ranking scores to partially predict future success in competitions (study 2). The test
performances are useful to evaluate the athletes' strengths and weaknesses and measure if the athletes' minimum
basic level is reached to be able to experience success and enjoyment in their sport. In order to predict future
cycling performance, competition results could be applied to analyse the current cycling performances and to
monitor the career development of the road cyclists. The predictive value of competition results gradually increases
(U17, U19, U23), with the highest predictions found in the U23year1 category. Given that the most significant decrease
156
in performance results was observed from U19year2 to U23year1, it seems crucial to provide additional psychological
support and guidance for young cyclists in this transition. This could prevent athletes from early drop-out from
competitive cycling. The coaches and athletes should be made aware that a lower level of competitive success is a
natural consequence of racing against opponents with superior training and competition experience. Additionally,
the federation could also consider to reorganise the competition system, by creating age-based competition
categories and/or rankings. Nevertheless, it is suggested that coaches and federations should not only rely on
competition results to identify talent, as a high inter-individual variability was found in the analyses. Other
potential performance determining factors should be taken into account such as: the coaches-eye, other individual
dimensions (e.g., cognitive, emotional, physiological features), tactical, strategic, contextual, and global features
and organisational constraints (Phillips & Hopkins, 2020).
3.1.4. Talent development
The fourth level of the Cycling Compass includes talent development (Figure 3.4.), a process which manifests itself
throughout the athletic career. On the one hand, the federation and coaches could apply a (non-)sport-specific test
battery for a longitudinal follow-up of athletes in the U12, U15, U17, and U19 categories. On the other hand,
longitudinal follow-up of cycling performances could also be valuable, however, it is suggested only from the U17
category onwards within road cycling. From the moment the federation and coaches decide to apply the test
battery, we suggest to provide the athletes with an overview of their test results, displaying their performances
against a benchmark (Addendum C, Tables C.1. – C.5.). This allows a comparison of test results against peers,
performance levels, and disciplines. These benchmarks should be regularly updated as the sport, the needs, and
characteristics of the sport continue to change. When interpreting the test results, coaches and the athletes should
be aware of the possible influence of maturity status, training history, relative age effect, the compensation
phenomenon, etc. on the results. Therefore, it seems appropriate to involve (national, provincial, regional, local)
coaches in the talent program and educate them how to interpret the results, and how to support the athletes in
their further development.
157
4. Strengths, limitations, and suggestions for future research
Throughout this dissertation, some pertinent questions with regard to talent research in youth cycling have been
(partly) answered. This research highlighted some essential components of the talent development process, such
as natural abilities - physical domain, development, competencies, and talent. Given the complexity of talent, there
is still room for additional research (e.g. mental domain, intra-personal, environmental catalysts, etc.). This was
illustrated as an example in the Differentiated Model of Giftedness and Talent developed by Gagné et al. (2010)
(Figure 3.5.).
Figure 3.5.: Gagné's Differentiated Model of Giftedness and Talent (DMGT 2.0; 2008 update) (Gagné, 2010). Green: investigated
within this dissertation; Blue: suggestions for future research.
In the following section, the strengths and limitations will be discussed, and suggestions for future research will
also be provided.
4.1. Strengths
4.1.1. Adaptations of the test battery for future research
The assessment of a (non-)sport-specific test battery within a cycling population seems valuable to detect, orient,
identify, select, monitor, develop athletes with potential, and even partially predict future performance (studies 1,
158
2). The test battery consists of a large set of tests, which were all applied to the different age groups. According to
'the sliding population approach' of Régnier et al. (1993) and the results in studies 1 and 2, the content and focus
of a test battery can differ per age group as a result of the athletes' development in adolescence. An overview of
the tests per age group is provided in Table 3.1.. No distinction was made between the different talent phases and
cycling disciplines (talent detection, orientation, identification, and development). The tests are discussed more in
detail in the following paragraphs, and recommendations for future research are considered. The recommendations
take into account the practical feasibility and relevance of the tests. Based on future research and in collaboration
with the federation, coaches, and researchers, these findings should be evaluated regularly and adjusted if
necessary.
Table 3.1.: An overview of the tests that could be included in the test batteries per age group (X = included; (X) = optional).
U12 U15 U17 U19
(< 12 years) (12-14 years) (15-16 years) (17-18 years)
Anthropometry
Body height X X X X
Sitting height X X X (X)
Body weight X X X X
Body mass index X X X X
Physical tests
Endurance plank test X X X
Standing broad jump (SBJ) X
Counter movement jump (CMJ) X X X
Sit and reach (SAR) X X X X
Curl-ups X
Knee push-ups X
30m sprint X X X X
Endurance shuttle run (ESHR) X X X (X)
Motor coordination
Moving sideways X X X X
Jumping sideways X X X X
Balance beam X X X X
Cycling specific tests
Shuttle bike X X X
Maximal cadence X X X
30" all-out cycling test (X) X X
159
4.1.1.1. Anthropometry
Anthropometric measurements are an essential element of talent identification also within cycling (Tables 3.1., 3.2.).
It gives an indication of the athletes' body composition and can be used for talent orientation (study 1). However,
caution is suggested when orienting, identifying, and selecting athletes solely on anthropometric measurements
since the timing and tempo of growth (maturity) and relative age differ inter-individually (Mostaert et al., 2016;
Teunissen et al., 2020). Furthermore, the study of Santos et al. (2014) showed that information on body composition
could be used to optimise competitive performance, monitor training and detect medical problems (Ackland et al.,
2012; Sundgot-Borgen et al., 2013; Santos et al., 2014).
Table 3.2.: A critical review on the anthropometric measurements used in this dissertation. Suggestions for future research.
Measurements Retain or not Comments
(yes or no)
Anthropometry
Body height (cm) Yes U12, U15, U17, U19
Sitting height (cm) Yes U12, U15, U17, and optional in the U19
Body weight (kg) Yes U12, U15, U17, U19
Fat percentage (%) No Fat percentage was measured using a foot-to-foot bioelectrical impedance system (Tanita,
BC420SMA, Weda B.V., Holland). Fat percentage was not reported in the current studies due
to inaccuracy (estimation of body fat). Alternative methods such as skinfolds could be
considered; however, this approach is time-consuming and has low accuracy (Wells &
Fewtrell, 2006). More accurate body fat measurements could be considered once the
athletes are selected and have reached APHV (e.g., DEXA scan, underwater weighing)(Wells
& Fewtrell, 2006).
Calculations
BMI (kg/m²) Yes U12, U15, U17, U19
BMI could differentiate track from cyclo-cross cyclists, whereby the track cyclists possessed
a higher BMI. BMI is considered a (rough) estimate or body composition, and correlates well
with body fat. However, it does not, for example, take age, sex, muscle mass, fat distribution
or bone structure into consideration (Rothman, 2008). Consequently, BMI sometimes gives
an overestimation of body fat. For example, lean athletes (BMX) with high muscle mass
(Rothman, 2008) may have a high BMI score. Therefore, BMI should be interpreted with
caution for these reasons and others, especially within a sporting population (Rothman,
2008).
Maturity Yes U12, U15, U17, and optional in the U19
(APHV or PAH) The formula of, e.g., Fransen et al. (2018), Mirwald et al. (2002) could be used to control for
maturity status (age of peak height velocity = APHV). Moreover, the Kamish-Roche
procedure could be applied for bio-banding (predicted adult height = PAH)(Cumming et al.,
2017). Nevertheless, any non-invasive method that estimates the biological age (e.g.,
Khamis-Roche, Mirwald, Fransen) should account for a certain degree of error (Teunissen et
al., 2020; Laureys et al., 2021b).
Relative age effect Yes U12, U15, U17, and optional in the U19
(RAE) The RAE was present in the youngest categories (U15, lesser U17 road cycling) but
disappeared from U19 onwards. Information about age should be presented appropriately
to reduce selection bias associated with the RAE (Mann & van Ginneken, 2017).
160
4.1.1.2. Physical tests
The physical tests applied in this dissertation are based on the generic tests of the Flemish Sports Compass and
provide valuable information about the athletes' general physical development and physical characteristics (Pion
et al., 2015) (Tables 3.1., 3.3.). Study 1 and pilot study 2 indicated that the physical tests could be used to compare
the athletic profiles of different cycling disciplines (talent orientation and transfer) (Addendum B). In terms of
talent identification, it was shown that these physical tests are less suitable for predicting future performance. This
finding is in agreement with previous research reporting that physical performance could determine current ability
but is less suitable as an indication for the potential to excel in the future (Vandorpe et al., 2012). This might be due
to the fact that training, growth, RAE, and maturation have an impact on these physical characteristics. For these
reasons, the test battery results should be interpreted with caution, especially for early and late mature athletes,
which requires knowledge and insight from the coaches and athletes. The CMJ, curl-up and knee push-up were
added to the test battery at a later stage (2018; pilot studies 1 and 2). The added value of a CMJ was shown in pilot
studies 1 and 2 (Addenda A, B).
161
Table 3.3.: A critical review on the physical tests used in this dissertation. Suggestions for future research.
Retain or
Measurements not retain Comments
(yes or no)
PHYSICAL CHARACTERISTICS
Endurance plank Yes U15, U17, U19
test (s) The endurance plank test was added to the test battery to measure static strength in cycling. The test was able to differentiate road cyclists from the other disciplines. The test could
be used for talent research in cycling; however, it is suggested to adapt the protocol. A time limit should be applied to ensure that static strength is measured rather than perseverance.
SBJ (cm) Yes U12 and optional in the U15, U17, U19
The SBJ test measures explosive leg power and could distinguish BMX cyclists from the other cycling disciplines. The test is easy to apply in children and adolescents, and can be
generalised across different sports. Significantly large correlations were found between CMJ and SBJ (r = 0.57, p <0.001) (Popowczak et al., 2019). Since the athletes have to be tested
in a relatively short time-span with a minimum of equipment, the assessment of the CMJ is preferred over the SBJ.
CMJ (cm) Yes U15, U17, U19
The CMJ test measures also explosive leg power and could be used to differentiate BMX athletes from the other disciplines. The test can be recorded with an OptoJump device (MicroGate,
Bolzano, Italy). Although the purchase for this equipment is more expensive compared to the SBJ, the CMJ is preferred due to the more accurate measurements and the equipment can
be used to measure, e.g., different types of jumps (CMJ hips, CMJ with arms, SJ). If there are limited financial resources in the federation, the SBJ is a valuable alternative for the CMJ.
SJ (squat jump) Yes Future research should consider incorporating the squat jump test in the talent test battery (Bertucci et al., 2007; Robert et al., 2020). The study of Bertucci et al. (2007) found a
(cm) significant correlation between the height reached in the squat jump, counter movement jump, and the BMX performance.
SAR (cm) Yes U12, U15, U17, U19
Since the sit and reach test is a health-related fitness test, it is suggested to keep this test also in the test battery of cycling (Council of Europe, 1988).
Curl-up (N) No U12
From 2018 onwards, the curl-up and knee-push up tests were added to the test battery of cycling to compare with the Flemish Sports Compass administered in schools (Council of
Knee push-up (N) No Europe, 1988). No significant differences were found between the cycling disciplines and the healthy reference population (pilot study 1). Due to the test protocol, which encourages
the athletes to perform as many curl-ups or knee-push ups in 30 seconds, the quality of the exercises could not always be guaranteed. Therefore, it is suggested to not include these
tests in further cycling research.
30m sprint (s) Yes U12, U15, U17, U19
The 30m sprint test with split times at 5m, 10m, 20m, and 30m measures speed and could distinguish track and BMX cyclists from the other cycling disciplines. The test can be recorded
with a MicroGate Racetime2 chronometry and Polifemo Light Photocells (MicroGate, Italy). This equipment is rather expensive, however, it can also be assessed for other tests (e.g.,
shuttle bike test).
ESHR (mins) Yes U12, U15, U17, and optional in the U19
The endurance shuttle run measures endurance, which is considered an essential characteristic in cycling. Differences in performance were found between BMX cyclists and the other
disciplines. The test is easy to apply in children and adolescents. The results can be generalised across different sports. The endurance shuttle run is less time consuming and less
expensive compared to an incremental cycling test. However, when working with a select group of athletes, more cycling-specific endurance tests could be preferred.
Physical quotient Yes U15, U17, U19
Recalculating the test scores into overall quotient scores, displaying the population as a normal distribution, enables to control for the compensation phenomenon and compare
between disciplines, ages and sex (Vaeyens et al., 2008). These quotient scores, expressed in positive scores (average is equal to 100), provide an overall impression of the athlete's
abilities independently of the individual test results. The quotients are calculated using the following formulas:
z-score = (raw scoretest - meantest) / standard deviationtest
Q-score = 100 + (z-score * 15)
PQ= Average Q-scores of the different tests
162
4.1.1.3. Motor coordination
A KTK3 test battery was used to measure motor coordination in cycling (Tables 3.1., 3.4.). This test battery measures
a broad spectrum of motor coordination skills, from locomotion to balance, and relies on components of physical
fitness and motor coordination (Mostaert et al., 2021). Study 2 showed that having good motor coordination in the
U15 category can lead to better performances 2-3 years later. Furthermore, previous studies found that diversified
training, free play, participation in different disciplines could positively influence the motor development of
athletes (Sieghartsleitner et al., 2018). Study 1 indicated that talent orientation should not occur early in the
development of cyclists. Instead, the athletes could use the opportunity to develop broadly. As a result, athletes
possessing excellent motor coordination skills are more likely to easily and quickly learn new skills and sports,
which facilitates the transition to their main discipline(s). Anecdotally, the findings could be confirmed with
practical examples, since cyclists such as Primož Roglič (ski jump), Wout Van Aert (cyclo-cross), Mathieu Van der
Poel (soccer, MTB, cyclo-cross), Mark Cavendish (track) practised multiple disciplines during their youth sports
career and matured into champions. Consequently, the KTK3 test battery should be used for talent research in
cycling.
Table 3.4.: A critical review on the motor coordination tests used in this dissertation. Suggestions for future research.
Retain or
Measurements not retain Comments
(yes or no)
MOTOR COORDINATION
Moving sideways (N) Yes
U12, U15, U17, U19
The KTK3 test battery is a reliable and valid test battery to measure general motor
Jumping sideways (N) Yes coordination in children and adolescents (Vandorpe et al., 2011; Kiphard & Schilling,
2017; Mostaert et al., 2021). The KTK3 test battery can be used to predict cycling
performance two to three years later and distinguish athletic cycling profiles. All three
Balance beam (N) Yes tests should be administered.
Motor quotient Yes U15, U17, U19

Recalculating the test scores into overall quotient scores, displaying the population as
a normal distribution, enables to control for the compensation phenomenon and
compare between disciplines, ages and sex (Vaeyens et al., 2008). These quotient scores,
expressed in positive scores (average is equal to 100), provide an overall impression of
the athlete's abilities independently of the individual test results. The quotients are
calculated using the following formulas:
MQ= Average Q-scores of the different tests
163
4.1.1.4. Cycling-specific tests
The shuttle bike and maximal cadence tests used in this dissertation were chosen because they have a link with
the sports disciplines, include motor coordination skills, technical skills, and leg speed, and are easy and time-
efficient to administer (Tables 3.1., 3.5.). Study 1 and pilot study 2 showed that the tests could distinguish the more
technical disciplines from the others (Addendum B). These cycling-specific tests were not able to predict future
performance(s), which was in line with the findings of the physical tests within cycling (study 2). The results might
be influenced by training history, biological, and/or relative age. However, these tests should remain part of the
test battery, and could be used by coaches and federations to evaluate the athletes' strengths and weaknesses in
order to adjust their training to it.
The 30" all-out cycling test was added to the test battery at a later stage (2018). It was mainly assessed for riders
older than 15 years of age, and rather exceptional in the younger age groups (12.0 – 14.9 years). Due to the limited
data gathered, the test was not investigated in studies 1 and 2. However, benchmarks are provided in the Addenda
(Addendum C; Tables C.1. – C.5.). This test procedure correlates with the heights registered in an SBJ and CMJ.
Therefore, it is expected that the 30” all-out cycling test should be able to identify the explosive athletic profiles.
Table 3.5.: A critical review on the cycling-specific measurements used in this dissertation. Suggestions for future research.
Measurements Retain or not Comments
(yes or no)
CYCLING-SPECIFIC TESTS
Shuttle bike (s) Yes U15, U17, U19
The shuttle bike test evaluates agility, speed and cycling skills (bike handling), in which riders have to
complete an eight-shaped circuit as quickly as possible. The test was able to differentiate cyclo-cross,
MTB, and BMX cyclists from the other disciplines.
Maximal cadence (N) Yes U15, U17, U19
The maximal cadence test measures agility. The cyclist needs to perform their maximal cadence within
ten seconds, on a standardized cycling ergometer without resistance. Pilot study 2 indicated that the
BMX cyclists outperform the other cycling disciplines on maximal cadence test, however the
differences disappeared when controlling for the covariates maturity, RAE, and training history.
30” all-out cycling Yes U17, U19, and optional in the U15
test (W, W/kg) A 30 seconds all-out cycling test evaluates the neuromuscular power, sprint ability, anaerobic and
aerobic performance levels of the cyclists. The athletes need to perform an all-out effort for 30
seconds against a constant braking force on a standardized ergometer (Wattbike ergometer). However,
growth and maturation have been associated with profound changes in markers of aerobic and
anaerobic muscle metabolism. As children mature there is a greater increase in performances
supported by the anaerobic metabolism than in performances supported by the oxidative metabolism
(Armstrong et al., 2015). Therefore, it is suggested to apply the 30” all-out cycling test in athletes post
puberty.
Sport-specific Yes U15, U17, U19
quotient Recalculating the test scores into overall quotient scores, displaying the population as a normal
distribution, enables to control for the compensation phenomenon and compare between disciplines,
ages and sex (Vaeyens et al., 2008). These quotient scores, expressed in positive scores (average is
equal to 100), provide an overall impression of the athlete's abilities independently of the individual
test results. The quotients are calculated using the following formulas:
SQ= Average Q-scores of the different tests
164
4.1.2. Suggestions for measuring cycling performance for future research
Within this dissertation different methodologies were used to measure and quantify cycling performance. On the
one hand, study 3 used top 10 results from national and provincial competitions. The success rate was computed
for each competition category and/or competition year, taking into account the number of top 10 results achieved
(yes = top 10 result achieved; no = did not achieve a top 10 result) and the number of competitions listed (U15total:
N=9; U17 total: N=12; U19 total: N=12). On the other hand, study 4 determined the success in the youth categories as
the number of points gathered in the annual Italian national ranking of the Italian Cycling Federation. In this scoring
system a number of points are awarded for the top 5 athletes of each race in the national calendar. Both studies
have in common that high standards were imposed (top 10 or top 5) and multiple competitions were included with
different difficulty levels (minimum 9). The results of both studies showed that cycling performance starts to be
indicative for future success only from the U17 category onwards (Figure 3.6.). Furthermore, a similar competitive
trajectory was found for the achievers/professionals and non-achievers/non-professionals. Therefore, we believe
that both methodologies could be used by coaches, the federation, and other stakeholders to evaluate cycling
performance from the U17 category onwards. The use of the Italian ranking system (study 4) or alternative rankings
(study 2, UCI rankings, etc.) might be the most preferred methods. Rankings might be better at taking into account
the fact that riders do not participate in all races and not all races may be equally difficult to win, as more
prestigious races that typically attract the best riders receive more points (Van Bulck et al., 2021). As stated by Van
Bulck et al. (2021) it should be acknowledged that even 'more detailed data on races (e.g. level of competition) and
race events (e.g. position data for each rider on short time frames), as well as physiological data on the riders (e.g.
weight, maximal oxygen uptake) could be of great value, however, currently no such data is (publicly) available,
and certainly not on the scale required for talent identification' (p. 2).
Table 3.6.: A critical review on cycling performance per age group.
U12 U15 U17 U19 U23

(< 12 years) (12-14 years) (15-16 years) (17-18 years) (19-22 years)
Cycling performance X X X
4.2. Limitations
4.2.1. Main focus on male cycling
The present dissertation has only focused on male cycling. Nevertheless, future talent research should investigate
the athletic profiles of women cyclists. In the last decades, women’s road cycling gained increasing attention from
the public, UCI, sponsors and research (Martin et al., 2001; Van Erp, 2019). Women differ from men in terms of
165
biological development. They reach their APHV around the age of 12, while men reach their APHV on average around
the age of 14 (Mirwald et al., 2002). Since puberty on average occurs at a younger age in women, maturity and/or
RAE seems to start and disappear earlier compared to men (Cobley et al., 2018). Therefore, peak performance could
be reached two to three years earlier. Besides the anthropometric differences, it can be noted that professional
female cycling races are shorter in duration and distance, lower in total work and absolute power output (Lamberts,
2014; Sanders et al., 2019; Van Erp et al., 2019). Therefore, they spend a significantly greater proportion in the higher
intensity zones during the competitions compared to men. Due to the differences in race regulations, different race
tactics are needed. If the race regulations of women’s cycling could change, the gap with men could be reduced as
more similarities in the race and training characteristics would occur between women and men (Van Erp, 2019).
Currently, limited research is available on the career trajectories of female cyclists. Brustio et al. (2021) investigated
the junior-senior transition rate in female swimmers and found that more young women reached the highest
performance level in adulthood compared to men. This may be explained by the fact that women performance
could be predicted two to three years earlier, since they reach their peak performance on average earlier.
Furthermore, Lepers (2019) indicated that sex-related differences might disappear over the years between men
and women in endurance events. Given the differences and similarities described in the literature, it is suggested
that future research is required within this area.
4.2.2. Main focus on road cycling
Predictions based on cycling and test performances were examined only within road cycling (studies 2, 3, and 4).
Given the differences in race duration, terrain characteristics, equipment, number of participants (UCI Regulations,
2018; Lucia et al., 2001; Beedie and Foad, 2009; Spindler et al., 2018) between the different cycling disciplines, it is
expected that the results from road cycling could not unambiguously be generalised to the other disciplines. For
example, it is suggested that within track cycling, relative season-best scores could be considered to evaluate the
race outcomes of these events with fixed distances in sprint and endurance track events (team pursuit, team sprint,
individual sprint) (Post et al., 2020a). Relative season-best scores, also used in sport such as swimming, speed
skating, and track and field, allow a more reliable comparison between athletes, correcting for the changes in the
sport over time (faster times, new equipment) by relating the season-best times with the prevailing world record
(de Koning, 2010; Post et al., 2020b). Future research is needed in track cycling to examine the usefulness of these
relative season-best scores.
In contrast, given some similar features in the competition formats of MTB, cyclo-cross, and some track events
(scratch, points race, keirin) with road cycling, it is expected that top 10 results from national and provincial
166
competitions, and/or points achieved in the annual national ranking seem appropriate methods to examine and
visualise the cyclists’s career trajectories. However, future research is needed and should also incorporate domestic
and foreign riders in their analyses to expand the database (Van Bulck et al., 2021).
Concerning the predictive value of test performances, similar results are expected in the different cycling
disciplines, and categories since the same category divisions are used (U12, U15, U17, U19, U23). It is expected that
motor coordination will be a performance determinant within all disciplines, but especially within track cycling
(study 1). It is also expected that cycling-specific tests can predict performance 2 to 3 years later within MTB and
cyclo-cross (study 1). Further research is needed to confirm these hypotheses.
4.2.3. Limited multidimensional approach
This dissertation is limited to the investigation of the athletic profile of cyclists based on test performances
assessed with a (non-)sport specific test battery and cycling performances. However, as previously mentioned, the
development of athletes is much more complex. It is influenced by different internal (intra-personal, chance) and
external (environment, chances) factors (Gagné, 2004; Gulbin et al., 2013). There is a need for a more ecological
design, ‘which places an emphasis on the interactions of the individual in his/her environment where intentions,
perceptions, and actions are interconnected rather than treated as separate entities’ (Davids et al., 2013; p.106).
Phillips and Hopkins (2020) emphasised specifically for cycling that the determinants of performance in cycling
are a complex interplay of different dimensions and features (e.g., individual, tactical, strategic and global features)
regulating performance in elite cycling competitions. Therefore, future research should consider a broader
spectrum of characteristics to identify talent. For example, intrapersonal factors such as psychological
characteristics could be tested, using a PCDEQ-C questionnaire to assess and monitor the developmental
psychological characteristics of young athletes (Laureys et al., 2021a). Additional research is required to determine
whether cognitive characteristics, such as executive functions (mental domain), linked to inhibitory control or
tactical features can distinguish athletes of different levels within cycling. In practice, there is a trend for
professional teams to select riders with tactical insight (those who can read the race) in addition to the
physiological and physical profile. Phillips and Hopkins (2020) indicated that tactical features, the quality of an
individual athlete to perform the right action at het right moment (Kannekens et al., 2011), could be significant in
cycling since opponents interact with each other during the competitions. Furthermore, riders have to constantly
adapt to the actions of others in a continuous changing racing environment.
167
Winning a race might be dependent on the tactical insight of riders:
 Drafting: riding in a slipstream of others, the energy cost of maintaining speed can be reduced. Riding in
a peloton makes this possible.
 Positioning: positioning in the beginning, middle and end of the race. It could be partly related to drafting.
 Pacing: dividing the race according to the individual characteristics of the athlete. Pacing also depends on
the characteristics of the opponent, actions and contextual features.
 Breakaway: attacking cyclists forming a smaller group. Moreover, the timing and cooperation dilemma are
crucial factors
 Assessing contextual features: for example, weather and road conditions.
 Team dynamics: cooperation between riders within of between teams.
Future research is needed to investigate to what extent the tactical insights are innate or could be trained.
Moreover, it should be examined if testing the tactical domain could be used to identify high potentials at a young
age.
4.2.4. Limited information on training history
In the current research project, the cyclists participating in the talent test days were asked to report about their
training history through a demographic questionnaire (from 2016 onwards, no data available of 2014 and 2015).
The cyclists retrospectively reported about the number of years competing in cycling (years), the number of hours
per week spent in cycling practise (h/week) and non-sport specific sports practise (h/week). The results did not
show an influence of training history on the results in the young adult and adolescent group (study 1). However, it
should be noted that a self-reported questionnaire comes with certain constraints: e.g., inferior level of evidence
compared to prospective studies and a risk of selection, recall and misclassification bias. Therefore, it is suggested
that future research designs could be optimised by implementing more detailed information on volume and
intensity of sport-specific practice and training (Rees et al. 2016). Research could benefit from using more
sophisticated methods to register training history such as training logs, which could provide information about
training volume and intensity. Unfortunately, young cyclists are not yet familiar with training logs. An experimental
or longitudinal study could prevent this shortcoming.
168
4.3. Suggestions for future research
4.3.1. Evolution in cycling
The current dissertation was the first to examine and compare the athletic profiles of five different cycling
disciplines (road, cyclo-cross, track, MTB, and BMX) (Figure 3.6.). However, as previously mentioned, the sport
continues to change (environmental factor) (Lath et al., 2020). The evolutions and changes within sports depend
on, e.g., media interest, culture (Lath et al., 2020). For example, the interest in sports may increase significantly
when the sport is included in the Olympic Games, when important international competitions are organised or
when media attention increases. Federations, as well as scientific research, must remain attentive and responsive
to these new trends. Disciplines such as BMX freestyle, trial, gravel racing, artistic cycling and cyclo-ball are not
discussed in this dissertation. In the future, these discipline(s) could become a part of the Cycling Compass. In order
to anticipate these changes and evolutions, athletes should be encouraged to develop broadly (studies 1 and 4).
New scientific insights will be needed to determine how the new discipline(s) relate to the disciplines previously
researched. For example, individual technical sports with a high-risk factor (e.g., trial) are more likely to correspond
to MTB and BMX. Strategic endurance disciplines, which rely on technique and the maintenance of balance due to
the cluttered terrain (e.g., gravel racing), will be more closely related to road and cyclo-cross or MTB. A (non-)sport-
specific test battery can be used as a tool to stimulate and guide athletes to a new discipline that suits their athletic
profile. Athletes with good motor coordination will potentially have fewer problems learning new skills, which
increases the chances of success. As athletes achieve more success in the new discipline(s) and a sufficiently large
population participates in the discipline(s), the athletic profile can be more closely examined and refined using the
(non-)sport-specific test battery. Furthermore, competition results and other performance determinants should be
investigated more closely in the future. In other words, the process described in this dissertation will be repeated.
Figure 3.6.: A visual representation of the interrelationship of the different cycling disciplines (LTAD-VOLUME, 2008).
169
4.3.2. Bio-banding
As indicated before, the strength of this dissertation was that benchmarks were provided (studies 1, 2; Addendum
C). Previous research showed that benchmarks could also be constructed and interpreted through the concept of
'bio-banding' (Cumming et al., 2017). In this approach, athletes are grouped according to their biological age rather
than their chronological age, which limits the misinterpretation of the performance results due to inter-individual
differences in biological maturation. Nevertheless, any non-invasive method that estimates the biological age (e.g.,
method of Khamis-Roche estimating the percentage of predicted adult height (%PAH), Mirwald calculating maturity
offset, Fransen calculating maturity ratio, etc.) should account for a certain degree of error (Khamis & Roche, 1994;
Mirwald et al., 2002; Cumming et al., 2017; Fransen et al., 2018; Teunissen et al., 2020; Laureys et al., 2021b). Since
growth varies considerably from individual to individual during adolescence, it is recommended to monitor the
growth status to adjust training goals accordingly. The study by Teunissen et al. (2020) assessed the athletes' body
measurements every three months. These more frequent growth assessments provide the opportunity to identify
'mini growth spurts’. Further research should explore the 'mini-growth spurts' and their potential additional value
in the context of bio-banding within cycling (Cumming et al., 2017).
Bio-banding is also applied in training. Scientific research indicated that adjustments in the athletic profile resulting
from a training stimulus depend on the athlete's growth phase and maturity (Smith, 2003). According to the Youth
Physical Development model (YPD model), periodisation implies that all physical attributes can be trained during
each phase of development. However, how the various attributes are trained should be adapted to the athletes'
growth phase (Lloyd & Oliver, 2012). The YPD model emphasises technical competence and learning experience.
This approach could avoid short- and long-term disadvantageous effects, such as chronic injuries, but might slow
down the fast gains in performance (Faigenbaum et al., 2019). In addition, some sports and countries are recently
investing in bio-banding training groups and competition categories. By offering adapted physical training and
competitions, athletes of the same biological age can compete together. This could reduce the injury risk in
(pre)adolescent athletes and possibly also positively influence athletic development (Rommers, 2020). For
example, soccer clubs in the United Kingdom and the United States work with bio-banding categories. Further
research is suggested to investigate the potential benefits of bio-banding categories within cycling.
4.3.3. Early specialisation or diversification/sampling
Based on the results of study 1, it can be suggested that early specialisation is not necessary for cycling. However,
future research is required to investigate which pathways of specialisation or diversification/sampling are the most
170
effective in cycling. Bompa and Carrera (2015) found similar results indicating that cyclists could consider
specialising in one or more disciplines from the ages of 16 and 18 and hopefully reach peak performance between
22 and 28 years of the age. In the literature, different pathways of specialisation or diversification/sampling are as
well debated as the implementation of practice- or play-oriented activities (Sieghartsleitner et al., 2018). These
pathways aim to optimise the learning environment for developing the athletes, to identify the athletes with the
potential to become elite athletes. Early specialisation leads to faster target sports performance improvement and
peak performance at 15–16 years of age (Bompa & Carrera, 2015). However, it could cause social isolation, over-
dependence, burn-out, and an increased risk of overuse injuries (Malina, 2010; Fransen et al., 2012). In contrast,
Sieghartsleitner et al. (2018) introduced “Specialised sampling", focusing on the primary sport domain and the
positive effect of participation within diversified forms of domain-specific free play. Côté and Hancock (2016) and
Fransen et al. (2012), on the other hand, suggests that athletes, especially in sports where peak performance is
reached after maturation, should focus on early diversification, deliberate play and fun to build a foundation for
elite performance, personal development, and lifelong sport participation. Specialised sampling and early
diversification have in common that they establish a range of motor and cognitive experiences that children can
ultimately bring to their principal sport of interest (Côté & Hancock, 2016). Furthermore, many international athletes
participated in various sports or movement activities before being successful in one specific sport (Seifert et al.,
2013; Teunissen et al., 2021). Therefore, we could expect that stimulating free play and sport participation in the
different cycling disciplines is promising for later success in cycling (studies 1 and 4). Nevertheless, further research
is required within this field of area.
4.3.4. Globalisation of the findings
Study 3 has shown that the findings on competition results can be generalised to other countries with a similar
culture. Particularly, the results found for the Belgian cyclists were similar to the findings of the Italian cyclists
(studies 3 and 4). It is suggested to be cautious when comparing the test performances from the (non-)sport-
specific test battery with other cultures, contexts, and countries. Future studies should be extended upon this to
interpret the results on a global scale (Mostaert et al., 2021).
171
5. Conclusion
The current dissertation provided an understanding of the similarities and differences in athletic profiles between
the different cycling disciplines and examined the predictive value of test and cycling performances within road
cycling. Based on the current results, a Cycling Compass was proposed, including both the (non-)sport-specific test
battery and competition results. Our findings of the (non-)sport-specific test battery showed that the athletic
profiles for the different cycling disciplines (road, track, cyclo-cross, and MTB) could only be discriminated from the
young adult level (≥ 16 years) on, except for BMX. This finding supports the idea that within cycling, it is not
necessary to promote early specialisation for one specific discipline, and consequently time can be spent on a
broader development. In a similar line of thinking, within road cycling, general motor coordination seems to be an
essential factor in the U15 category, while cycling performance starts to be indicative for future success only from
the U17 category onwards. This again supports the importance of a broad development in the early years of sport
participation. Finally, benchmarks on the characteristics evaluated in the test battery could, in conjunction with
cycling performance from U17 on, be used for longitudinal follow-up of cycling athletes. Nevertheless, it should be
noted that this work, except for the talent orientation study, focussed on male road cyclists, (non-)sport-specific
tests, and cycling performances. Therefore, future research should study other cycling disciplines, female cyclists,
and consider including additional features (physiological, tactical, psychological, cognitive, etc.) to better
understand the performance determinants within cycling.
172
6. References
Ackland, T. R., Lohman, T. G., Sundgot-Borgen, J., Maughan, R. J., Meyer, N. L., Stewart, A. D., & Müller, W. (2012). Current
status of body composition assessment in sport. Sports Medicine, 42(3), 227-249.
Armstrong, N., Barker, A. R., & McManus, A. M. (2015). Muscle metabolism changes with age and maturation: How do
they relate to youth sport performance? British Journal of Sports Medicine, 49(13), 860-864.
Ayelet-Keshet. In ayelet-keshet/shutterstock.com (Ed.), Shutterstock.
85-94.
Baker, J., Schorer, J., & Wattie, N. (2018). Compromising talent: Issues in identifying and selecting talent in sport.
Quest, 70(1), 48-63.
Bertucci, W., Hourde, C., Manolova, A., & Vettoretti, F. (2007). Facteurs mécaniques de la performance lors de la phase
d'accélération en BMX chez des pilotes entraînés. Science & sports, 22(3-4), 179-181.
Boccia, G., Cardinale, M., & Brustio, P. R. (2021). Performance progression of elite jumpers: Early performances do not
predict later success. Scandinavian Journal of Medicine & Science in Sports, 31(1), 132-139.
Bompa, T. O., & Carrera, M. (2015). Conditioning young athletes: Human Kinetics.
Briem, V., Radeborg, K., Salo, I., & Bengtsson, H. (2004). Developmental aspects of children's behavior and safety
while cycling. Journal of pediatric psychology, 29(5), 369-377.
British Cycling, n. d. (2021-2022). Rider route - performance pathway.
475.
Brustio, P., Cardinale, M., Lupo, C., Varalda, M., De Pasquale, P., & Boccia, G. (2021). Being a top swimmer during the
early career is not a prerequisite for success: a study on sprinter strokes. Journal of Science and Medicine
in Sport, 24(12), 1272-1277. doi:10.1016/j.jsams.2021.05.015
Buekers, M., Borry, P., & Rowe, P. (2015). Talent in sports. Some reflections about the search for future champions.
Movement & Sport Sciences-Science & Motricité(88), 3-12.
Cobley, S., Abbott, S., Dogramaci, S., Kable, A., Salter, J., Hintermann, M., & Romann, M. (2018). Transient relative age
effects across annual age groups in national level Australian swimming. Journal of Science and Medicine
in Sport, 21(8), 839-845.
Corden, T. E., Tripathy, N., Pierce, S. E., & Katcher, M. L. (2005). The role of the health care professional in bicycle safety.
Wisconsin medical journal, 104(2), 35-38.
Côté, J., & Hancock, D. J. (2016). Evidence-based policies for youth sport programmes. International Journal of Sport
Policy and Politics, 8(1), 51-65.
Cumming, S. P., Lloyd, R. S., Oliver, J. L., Eisenmann, J. C., & Malina, R. M. (2017). Bio-banding in sport: applications to
competition, talent identification, and strength and conditioning of youth athletes. Strength & Conditioning
Journal, 39(2), 34-47.
Davids, K., Araújo, D., Vilar, L., Renshaw, I., & Pinder, R. (2013). An ecological dynamics approach to skill acquisition:
Implications for development of talent in sport. Talent Development and Excellence, 5(1), 21-34.
de Koning, J. J. (2010). World records: how much athlete? How much technology? International journal of sports
physiology and performance, 5(2), 262-267.
De Meester, A. (2017). Motivating children and adolescents to develop a physically active lifestyle: The role of
extracurricular school-based sports and motor competence. Ghent University,
173
De Meester, A., Stodden, D., Goodway, J., True, L., Brian, A., Ferkel, R., & Haerens, L. (2018). Identifying a motor
proficiency barrier for meeting physical activity guidelines in children. Journal of Science and Medicine in
Sport, 21(1), 58-62.
Deprez, D. (2015). Anthropometrical, physical fitness and maturational characteristics in youth soccer:
methodological issues and a longitudinal approach to talent identification and development. Ghent
Enoksen, E. (2011). Drop-out rate and drop-out reasons among promising Norwegian track and field athletes: A 25
year study.
Faigenbaum, A. D., Lloyd, R. S., & Oliver, J. L. (2019). Essentials of youth fitness: Human Kinetics Publishers.
Fisher, A., Reilly, J. J., Kelly, L. A., Montgomery, C., Williamson, A., Paton, J. Y., & Grant, S. (2005). Fundamental movement
skills and habitual physical activity in young children. Med Sci Sports Exerc, 37(4), 684-688.
Ford, P., De Ste Croix, M., Lloyd, R., Meyers, R., Moosavi, M., Oliver, J., Till, K., & Williams, C. (2011). The Long-Term Athlete
Development model: Physiological evidence and application. Journal of sports sciences, 29(4), 389-402.
doi:10.1080/02640414.2010.536849
30(2), 296-307. doi:10.1123/pes.2017-0009
Fransen, J., Pion, J., Vandendriessche, J., Vandorpe, B., Vaeyens, R., Lenoir, M., & Philippaerts, R. M. (2012). Differences
in physical fitness and gross motor coordination in boys aged 6–12 years specializing in one versus
sampling more than one sport. Journal of sports sciences, 30(4), 379-386.
119-147. doi:10.1080/1359813042000314682
Gagné, F. (2010). Motivation within the DMGT 2.0 framework. High ability studies, 21(2), 81-99.
Hansen, K. S., Eide, G. E., Omenaas, E., Engesæter, L. B., & Viste, A. (2005). Bicycle-related injuries among young
children related to age at debut of cycling. Accident Analysis & Prevention, 37(1), 71-75.
Helsen, W. F., Van Winckel, J., & Williams, A. M. (2005). The relative age effect in youth soccer across Europe. Journal
of sports sciences, 23(6), 629-636.
Hill, M., Scott, S., Malina, R. M., McGee, D., & Cumming, S. P. (2020). Relative age and maturation selection biases in
academy football. Journal of sports sciences, 38(11-12), 1359-1367.
Hogan, J. W., Roy, J., & Korkontzelou, C. (2004). Handling drop‐out in longitudinal studies. Statistics in medicine,
23(9), 1455-1497.
Impellizzeri, F., Ebert, T., Sassi, A., Menaspa, P., Rampinini, E., & Martin, D. (2008). Level ground and uphill cycling
335-341.
Janssen, I., & Cornelissen, J. (2017). Pedal forces during the BMX and track sprint cycling start. ISBS Proceedings
Archive, 35(1), 277.
Johnston, K., Wattie, N., Schorer, J., & Baker, J. (2018). Talent identification in sport: a systematic review. Sports
Medicine, 48(1), 97-109.
Kannekens, R., Elferink‐ Gemser, M., & Visscher, C. (2011). Positioning and deciding: key factors for talent
development in soccer. Scandinavian Journal of Medicine & Science in Sports, 21(6), 846-852.
Khamis, H. J., & Roche, A. F. (1994). Predicting adult stature without using skeletal age: the Khamis-Roche method.
Pediatrics, 94(4), 504-507.
174
Lamberts, R. P. (2014). Predicting cycling performance in trained to elite male and female cyclists. International
journal of sports physiology and performance, 9(4), 610-614.
Lath, F., den Hartigh, R., Wattie, N., & Schorer, J. (2020). Talent Selection: Making Decisions and Prognoses about
Athletes. In Talent Identification and Development in Sport (pp. 50-65): Routledge.
Laureys, F., Collins, D., Deconinck, F. J., & Lenoir, M. (2021a). Exploring the use of the Psychological Characteristics of
Developing Excellence (PCDEs) in younger age groups: First steps in the validation process of the PCDE
Questionnaire for Children (PCDEQ-C). PloS one, 16(11), e0259396.
Laureys, F., Middelbos, L., Rommers, N., De Waelle, S., Coppens, E., Mostaert, M., Deconinck, F. J., & Lenoir, M. (2021b).
The effects of age, biological maturation and sex on the development of executive functions in
adolescents. Frontiers in physiology, 12, 703312. doi:Laureys, F., Middelbos, L., Rommers, N., De Waelle, S.,
Coppens, E., Mostaert, M., Deconinck, F., & Lenoir, M. (2021). The Effects of Age, Biological Maturation and
Sex on the Development of Executive Functions in Adolescents. Frontiers in physiology, 12, 703312.
https://doi.org/10.3389/fphys.2021.703312
Leone, M., & Larivière, G. (1998). Anthropometric and biomotor characteristics of elite adolescent male athletes
competing in four different sports. Science & sports, 1(13), 26-33.
Lepers, R. (2019). Sex difference in triathlon performance. Frontiers in physiology, 10, 973.
Li, P., De Bosscher, V., Pion, J., Weissensteiner, J. R., & Vertonghen, J. (2018). Is international junior success a reliable
550-559.
Li, P., Weissensteiner, J. R., Pion, J., & Bosscher, V. D. (2020). Predicting elite success: Evidence comparing the career
pathways of top 10 to 300 professional tennis players. International Journal of Sports Science & Coaching,
15(5-6), 793-802. doi:10.1177/1747954120935828
Lievens, E., Bellinger, P., Van Vossel, K., Vancompernolle, J., Bex, T., Minahan, C., & Derave, W. (2021). Muscle typology
of world-class cyclists across various disciplines and events. Medicine and Science in Sports and Exercise,
53(4), 816-824.
Lloyd, R. S., & Oliver, J. L. (2012). The youth physical development model: A new approach to long-term athletic
development. Strength & Conditioning Journal, 34(3), 61-72.
doi:10.14198/jhse.2016.111.03
López-Plaza, D., Alacid, F., Muyor, J. M., & López-Miñarro, P. Á. (2017). Sprint kayaking and canoeing performance
prediction based on the relationship between maturity status, anthropometry and physical fitness in
young elite paddlers. Journal of sports sciences, 35(11), 1083-1090.
LTAD-VOLUME, C. (2008). Long-Term Athlete Development - Cycling. 1.
Lucia, A., Joyos, H., & Chicharro, J. (2000). Physiological response to professional road cycling: climbers vs. time
trialists. International journal of sports medicine, 21(07), 505-512.
Malina, R., e Silva, M. C., & Figueiredo, A. (2013). Growth and maturity status of youth players. In Science and soccer
(pp. 319-344): Routledge.
Malina, R. M. (2010). Early sport specialization: roots, effectiveness, risks. Current sports medicine reports, 9(6), 364-
371.
Mann, D. L., & van Ginneken, P. J. (2017). Age-ordered shirt numbering reduces the selection bias associated with the
relative age effect. Journal of sports sciences, 35(8), 784-790.
Martin, D. T., McLean, B., Trewin, C., Lee, H., Victor, J., & Hahn, A. G. (2001). Physiological characteristics of nationally
competitive female road cyclists and demands of competition. Sports Medicine, 31(7), 469-477.
Martínez-González, N., Atienza, F. L., Duda, J. L., & Balaguer, I. (2022). The Role of Dispositional Orientations and Goal
Motives on Athletes’ Well-and Ill-Being. International Journal of Environmental Research and Public
Health, 19(1), 289.
175
Menaspà, P., Rampinini, E., Bosio, A., Carlomagno, D., Riggio, M., & Sassi, A. (2012). Physiological and anthropometric
Menaspà, P., Sassi, A., & Impellizzeri, F. M. (2010). Aerobic fitness variables do not predict the professional career of
Milić, M., Grgantov, Z., Chamari, K., Ardigò, L., Bianco, A., & Padulo, J. (2017). Anthropometric and physical
characteristics allow differentiation of young female volleyball players according to playing position and
level of expertise. Biology of sport, 34(1), 19.
Mirwald, R. L., Baxter-Jones, A. D., Bailey, D. A., & Beunen, G. P. (2002). An assessment of maturity from
anthropometric measurements. Medicine & Science in Sports & Exercise, 34(4), 689-694.
Morris, R. (2013). Investigating the youth to senior transition in sport: From theory to practice. Aberystwyth
University.
Mostaert, M., Coppens, E., Laureys, F., D'Hondt, E., Deconinck, F. J., & Lenoir, M. (2021). Validation of a motor competence
assessment tool for children and adolescents (KTK3+) with normative values for 6-to 19-year-olds.
Frontiers in physiology, 12, 652952. doi:10.3389/fphys.2021.652952
Mostaert, M., Deconinck, F., Pion, J., & Lenoir, M. (2016). Anthropometry, physical fitness and coordination of young
figure skaters of different levels. International journal of sports medicine, 37(07), 531-538. doi:10.1055/s-
0042-100280
Mostaert, M., Pion, J., Lenoir, M., & Vansteenkiste, P. (2020). A Retrospective Analysis of the National Youth Teams in
Volleyball: Were They Always Faster, Taller, and Stronger? Journal of Strength and Conditioning Research,
1-7. doi:10.1519/jsc.0000000000003847
Norjali Wazir, M. R. W., Mostaert, M., Pion, J., & Lenoir, M. (2018). Anthropometry, physical performance, and motor
coordination of medallist and non-medallist young fencers. Archives of Budo, 14, 33-40. doi:1854/LU-
8599260
Opstoel, K., Pion, J., Elferink-Gemser, M., Hartman, E., Willemse, B., Philippaerts, R., Visscher, C., & Lenoir, M. (2015).
Anthropometric characteristics, physical fitness and motor coordination of 9 to 11 year old children
participating in a wide range of sports. PloS one, 10(5), e0126282.
Padilla, S., Mujika, I., Cuesta, G., & Goiriena, J. J. (1999). Level ground and uphill cycling ability in professional road
cycling. Medicine and Science in Sports and Exercise, 31(6), 878-885.
489.
176
Pion, J., Fransen, J., Lenoir, M., & Segers, V. (2014). The value of non-sport-specific characteristics for talent
orientation in young male judo, karate and taekwondo athletes. Archives of Budo, 10(1), 147-154.
doi:10453/94313
Pion, J., Segers, V., Fransen, J., Debuyck, G., Deprez, D., Haerens, L., Vaeyens, R., Philippaerts, R. M., & Lenoir, M. (2015).
Pion, J., Segers, V., Stautemas, J., Boone, J., Lenoir, M., & Bourgois, J. G. (2018). Position-specific performance profiles,
using predictive classification models in senior basketball. International Journal of Sports Science &
Coaching, 13(6), 1072-1080.
Popowczak, M., Rokita, A., Świerzko, K., Szczepan, S., Michalski, R., & Maćkała, K. (2019). Are Linear Speed and Jumping
Ability Determinants of Change of Direction Movements in Young Male Soccer Players? Journal of sports
science & medicine, 18(1), 109.
Post, A. K., Koning, R. H., Stoter, I. K., Visscher, C., & Elferink-Gemser, M. T. (2020a). Interim Performance Progression
(IPP) During Consecutive Season Best Performances of Talented Swimmers. Frontiers in sports and active
living, 2, 579008. doi:10.3389/fspor.2020.579008
Post, A. K., Koning, R. H., Visscher, C., & Elferink‐Gemser, M. T. (2020b). Multigenerational performance development
of male and female top‐elite swimmers–A global study of the 100 m freestyle event. Scandinavian Journal
of Medicine & Science in Sports, 30(3), 564-571.
Robert, P., Cirer-Sastre, R., Matas-Garcia, S., Corbi, F., Julia-Sanchez, S., Alvarez-Herms, J., & Lopez-Laval, I. (2020).
Relationship Between Jump Capacity and Performance in BMX Cycling. 140, 37-43.
Robertson-Martens, K. (2021). Multidisciplinary contributions to talent identification in young elite badminton
players. Ghent University, Ghent, Belgium.
Robertson, K., Laureys, F., Mostaert, M., Pion, J., Deconinck, F. J., & Lenoir, M. (2021). Mind, body, and shuttle:
multidimensional benchmarks for talent identi, cation in male youth badminton. Biology of sport, 38(4),
79-94.
Rommers, N. (2020). Injury incidence, patterns, and risk factors in youth elite-level football players. Ghent
University,
Rothman, K. J. (2008). BMI-related errors in the measurement of obesity. International journal of obesity, 32(3),
S56-S59.
Ruiz, M. C., Appleton, P. R., Duda, J. L., Bortoli, L., & Robazza, C. (2021). Social environmental antecedents of athletes’
emotions. International Journal of Environmental Research and Public Health, 18(9), 4997.
Rylands, L., & Roberts, S. J. (2014). Relationship between starting and finishing position in World Cup BMX racing.
International Journal of Performance Analysis in Sport, 14(1), 14-23.
Sanders, D., & van Erp, T. (2020). The Physical Demands and Power Profile of Professional Men’s Cycling Races: An
Updated Review. International journal of sports physiology and performance, 1(aop), 1-10.
Sanders, D., van Erp, T., & de Koning, J. J. (2019). Intensity and load characteristics of professional road cycling:
differences between men’s and women’s races. International journal of sports physiology and
performance, 14(3), 296-302.
177
Santos, D. A., Dawson, J. A., Matias, C. N., Rocha, P. M., Minderico, C. S., Allison, D. B., Sardinha, L. B., & Silva, A. M. (2014).
Reference values for body composition and anthropometric measurements in athletes. PloS one, 9(5),
e97846.
Schorer, J., Rienhoff, R., Fischer, L., & Baker, J. (2017). Long-term prognostic validity of talent selections: comparing
national and regional coaches, laypersons and novices. Frontiers in Psychology, 8, 1146.
Smith, D. J. (2003). A framework for understanding the training process leading to elite performance. Sports
Medicine, 33(15), 1103-1126.
Stambulova, N., Alfermann, D., Statler, T., & CôTé, J. (2009). ISSP position stand: Career development and transitions
of athletes. International Journal of Sport and Exercise Psychology, 7(4), 395-412.
Sundgot-Borgen, J., Meyer, N. L., Lohman, T. G., Ackland, T. R., Maughan, R. J., Stewart, A. D., & Müller, W. (2013). How to
minimise the health risks to athletes who compete in weight-sensitive sports review and position
statement on behalf of the Ad Hoc Research Working Group on Body Composition, Health and Performance,
under the auspices of the IOC Medical Commission. British Journal of Sports Medicine, 47(16), 1012-1022.
Teunissen, J. W., Rommers, N., Pion, J., Cumming, S. P., Rössler, R., D’Hondt, E., Lenoir, M., Savelsbergh, G. J., & Malina,
R. M. (2020). Accuracy of maturity prediction equations in individual elite male football players. Annals of
Human Biology, 47(4), 409-416.
Teunissen, J. W. A., Ter Welle, S. S., Platvoet, S. S., Faber, I., Pion, J., & Lenoir, M. (2021). Similarities and differences
between sports subserving systematic talent transfer and development: The case of paddle sports. Journal
of Science and Medicine in Sport, 24(2), 200-205.
Van Bulck, D., Vande Weghe, A., & Goossens, D. (2021). Result-based talent identification in road cycling: discovering
the next Eddy Merckx. Annals of Operations Research, 1-18. doi:10.1007/s10479-021-04280-0
Van Erp, T. (2019). The Development of Women's Professional Cycling. Journal of Science and Cycling, 8(3), 1-2.
Van Erp, T., Sanders, D., & De Koning, J. J. (2019). Training characteristics of male and female professional road
cyclists: a 4-year retrospective analysis. International journal of sports physiology and performance, 15(4),
534-540. doi:10.1123/ijspp.2019-0320
Van Houcke, H., Gabriels, D., & De Smedt, T. (2009). FietsIEfieTS. Mechelen: Jan Peumans.
Vandorpe, B., Vandendriessche, J., Lefèvre, J., Pion, J., Vaeyens, R., Matthys, S., Philippaerts, R., & Lenoir, M. (2011). The
0838.2009.01067.x
Vandorpe, B., Vandendriessche, J. B., Vaeyens, R., Pion, J., Lefevre, J., Philippaerts, R. M., & Lenoir, M. (2012). The value
178
Wells, J., & Fewtrell, M. (2006). Measuring body composition. Archives of disease in childhood, 91(7), 612-617.
Zeuwts, L. (2016). Understanding and stimulating the development of perceptual-motor skills in child bicyclists.
Ghent University, Ghent, Belgium.
Zurc, J. (2017). It was worth it - I would do it again!: phenomenological perspectives on life in the elite women's
artistic gymnastics. Science of Gymnastics Journal, 9(1), 41-59.
179
ADDENDA
Addendum A: pilot study 1
Pilot study 1 compared the performances on a (non-)sport-specific test battery between (young) cyclists, BMX
athletes and a healthy (non-)sporting reference population. BMX and the cycling discipline were separately
included in the analyses, as the Olympic discipline BMX can be considered a different discipline within cycling since
its technical character, unique competition format (+/-30 sec effort on a short but challenging track), and the use
of a different type of bicycle.
Data on the healthy reference population were collected between 2018 and 2019 using a large-sampled study in
which Flemish children and adolescents between 6 and 19 years old participated. Convenience sampling resulted
in seven primary schools, seven secondary schools, two summer camp organisations and one university. This
approach resulted in a total sample of 2271 participants (1112 boys and 1159 girls). Of this sample, 411 male
participants between 7.0 – 15.99 years of age were included in the study. Two hundred and five participants
registered that they were not involved in organised sport, while 206 participants were involved in organised sport
for at least 1-hour/week.
Between 2014 and 2019, 328 male BMX athletes and 1325 male cyclists (active in road, track, cyclo-cross, or MTB)
between the ages of 7 and 18 were tested. 114 BMX cyclists and 380 cyclists between 7.0 – 15.99 years of age were
included in the pilot study. For the current pilot study, the data were divided into two age groups, 7.0 – 11.99 years
of age and 12.0 – 15.99 years of age (similar to study 1). The data consisted only of athletes who all performed four
anthropometric, eight physical and three motor coordination tests and registered training history using a
demographic questionnaire. Maturity ratio (Fransen et al., 2018), age, and training history were entered as
covariates in the MANCOVA analyses. Discriminant analyses were used to discriminate the profiles of the four
groups (for both young adults and adolescents). The differences and similarities in athletic profiles on the
anthropometric, physical, and motor coordination tests for both age groups are displayed in the spider plots
(Figure A1.., Figure 3.3. in the general discussion).
183
Body height
BB 1,2 Sitting height
JS 0,8 Body weight
0,4
MS 0,0 BMI
-0,4
CMJarms -0,8 SBJ
-1,2
CMJHips SAR
Knee push ups ESHR
Curl ups 30m sprint at 5m

30m sprint at 30m 30m sprint at 10m
30m sprint at 20m
Reference population (no sport) Reference population (sport)

Cycling BMX
Figure A.1.: The spider plot displays the z-scores of the different (non-)sport-specific tests for the (non-) sporting reference
population, cyclists and BMX. A total sample of 557 adolescents between 12.0 – 15.9 years of age (95 reference population
no sport; 101 reference population sport; 294 cyclists, 67 BMX cyclists) were included in pilot study 1.
The MANCOVA results revealed a multivariate effect for both age groups on anthropometry, physical characteristics
and motor coordination (Table A.1., A.2.). In other words, cyclists, BMX athletes, and the healthy (non-)sporting
reference population could be distinguished based on their performances on the test battery.
In the youngest age group (7.0-11.99 years), cyclists possessed a higher BMI, and performed significantly better on
the explosive (SBJ, 30m sprint, CMJhips), upper body strength (curl-up), endurance test (ESHR), and motor
coordination tests (JS, MS) compared to the (non-)sporting reference population. BMX athletes possessed a higher
BMI, and outperformed the (non-)sporting reference population on all physical and motor coordination tests except
for the ESHR, knee push up, 30m sprint at 5m and 10m tests, balance beam test. BMX cyclists could be differentiated
from cycling on the curl-ups and jumping sideways test.
In the older age group (12.0-15.99 years), cyclists possessed a higher BMI, and performed significantly better on the
explosive (SBJ, 30m sprint), upper body strength (curl-up), endurance test (ESHR), and motor coordination tests
(JS, MS, BB) compared to the (non-)sporting reference population. BMX athletes were smaller, possessed a higher
BMI, and performed significantly better on all physical and motor coordination tests with exception of the ESHR,
knee push up, 30m sprint at 5m and 10m tests compared to the (non-)sporting reference population. The BMX
184
cyclists possessed a higher BMI and outperformed the cyclists on the explosive tests (SBJ, CMJhips, CMJarms), upper
body strength (knee push-ups) and on the jumping sideways test.
The discriminant analyses revealed that in the 7.0-11.99 year old age group, only 55.7% of the participants could be
correctly assigned to the reference population, BMX or cycling discipline (50.3% cross-validated) (Figure B). In this
discriminant model of original grouped cases 64.5% of the non-sporting reference population, 43.8% of the sporting
reference population, 67.4% of the cyclists, and 40.4% of the BMX cyclists were correctly assigned to their group.
Figure A.2.: Graphical representation of the discriminant analysis of cyclists, BMX and a (non-)sporting reference population
between 7.0 and 11.99 years of age based on performance results on a non-sport-specific test battery. Note: the scatter plot
has the canonical discriminant function coefficients as its axes, with Function 1 on x-axis and Function 2 on the y-axis.
In the 12.0-15.99 year old age group, the discriminant analyses revealed that 83.1% of the cyclists could be correctly
classified (80.3% cross-validated) using the generic test battery (Figure C). 71.6% of the non-sporting reference
population, 77.2% of the sporting reference population, 96.6% of the cyclists, and 49.3% of the BMX cyclists were
correctly assigned to their discipline.
185
Figure A3.: Discrimination of cyclists, BMX cyclists and the (non-)sporting reference population between 12.0 and 15.9 years of
age based on performance results on a non-sport-specific test battery (71.6% correctly classified; 80.4% cross-validated). Note:
the scatter plot has the canonical discriminant function coefficients as its axes, with Function 1 on x-axis and Function 2 on the
y-axis.
186
Table A.1.: Average performance on the anthropometric, physical, and motor coordination tests (± SD) of the (non-)sporting reference population, cyclists, and BMX cyclists between the age of 7.0 and 11.99 years
Reference population Reference population

(no sport) (sport)
Cycling BMX MANOVA Effect size MANCOVA Effect size
Mean ± SD Mean ± SD Mean ± SD Mean ± SD F p F p
N 110 105 86 47
Anthropometry 8.627 <0.001 0.091 4.280 <0.001 0.048
Body height (cm) 137.1 ± 10.4 137.6 ± 10.6 144.9 ± 8 140 ± 8.3 12.489 <0.001 0.098 1.912 0.127 0.017
Sitting height (cm) 71.7 ± 5.1 b 72.3 ± 5.5 a 75.4 ± 3.9 73.4 ± 3.5 11.018 <0.001 0.088 2.706 0.045 0.023
Body weight (kg) 31.2 ± 8.2 30.2 ± 6.2 34.9 ± 6.4 33.2 ± 6 8.208 <0.001 0.067 1.435 0.232 0.012
Body mass index (kg/m²) 12.2 ± 10.6 c.d 10.9 ± 1.6 c.d 16.5 ± 1.7 a.b 16.8 ± 1.4 a.b 19.522 <0.001 0.145 11.299 0.000 0.090
Physical characteristics 5.406 <0.001 0.151 2.776 <0.001 0.084
Standing broad jump (cm) 143.4 ± 18 c.d 146.7 ± 19.1 d 162.7 ± 17.7 a 164.7 ± 16.3 a.b 29.203 <0.001 0.203 9.559 <0.001 0.078
Sit and reach (cm) 17.7 ± 6.1 18.3 ± 6.6 d 19 ± 5.5 21.8 ± 5 b 5.260 0.001 0.044 4.663 0.003 0.039
Endurance shuttle run (min) 4.7 ± 2 c 5.6 ± 1.9 c 7.7 ± 1.8 a.b 6.5 ± 1.9 38.853 <0.001 0.253 10.558 <0.001 0.085
30m sprint - 5m (s) 1.38 ± 0.141 1.366 ± 0.167 1.313 ± 0.12 1.329 ± 0.119 4.325 0.005 0.036 0.118 0.950 0.001
30m sprint - 10m (s) 2.392 ± 0.237 2.371 ± 0.27 2.234 ± 0.158 2.276 ± 0.159 10.337 <0.001 0.083 1.275 0.283 0.011
30m sprint - 20m (s) 4.246 ± 0.369 c.d 4.153 ± 0.306 3.902 ± 0.257 a 3.963 ± 0.27 a 23.433 <0.001 0.170 4.697 0.003 0.040
30m sprint - 30m (s) 6.106 ± 0.668 c.d 5.984 ± 0.548 5.544 ± 0.387 a 5.659 ± 0.374 a 21.676 <0.001 0.159 4.453 0.004 0.038
Curl ups (N) 25 ± 7 d 25 ± 7 d 30 ± 7 d 32 ± 8 a.b.c 17.714 <0.001 0.134 8.550 <0.001 0.070
Knee push ups (N) 28 ± 8 28 ± 7 68 ± 378 28 ± 7 0.969 0.408 0.008 0.823 0.482 0.007
Counter movement jump hips (cm) 19.2 ± 4.2 c.d 20 ± 4.2 d 23.2 ± 4 a 22.8 ± 4.5 a.b 19.613 <0.001 0.146 5.758 0.001 0.048
Counter movement jump armswing (cm) 22.8 ± 4.9 d 23.7 ± 5 d 27.6 ± 4.8 27.8 ± 5.2 a.b 22.807 <0.001 0.166 6.230 <0.001 0.052
Motor coordination 15.918 <0.001 0.121 7.885 <0.001 0.065
Moving sideways (N) 44 ± 12 c.d 42 ± 13 d 55 ± 8 a 54 ± 8 a.b 33.416 <0.001 0.226 15.192 <0.001 0.118
Jumping sideways (N) 60 ± 13 c.d 61 ± 14 c.d 76 ± 12 a.b.d 77 ± 12 a.b.c 42.157 <0.001 0.269 16.342 <0.001 0.126
Balance beam (N) 45 ± 13 45 ± 14 54 ± 12 53 ± 13 12.423 <0.001 0.098 2.365 0.071 0.020
Note: Non-sporting reference population = a, Sporting reference population = b, Cycling= c, BMX= d, which indicates significant differences (p<0.05) between the different groups based on the MANCOVA (post hoc
analyses).
187
Table A.2.: Average performance on the anthropometric, physical, and motor coordination tests (± SD) of the (non-)sporting reference population, cyclists, and BMX cyclists between the age of 12.0 and 15.99 years
Reference population Reference population
(no sport) (sport)
Cycling BMX MANOVA Effect size MANCOVA Effect size
Mean ± SD Mean ± SD Mean ± SD Mean ± SD F p F p
N 95 101 294 67
Anthropometry 182.564 <0.001 0.555 167.944 <0.001 0.537
Body height (cm) 166.8 ± 10.2 167 ± 10 d 164.8 ± 10.6 161.1 ± 9.4 b 5.456 0.001 0.029 5.316 0.001 0.028
Sitting height (cm) 85.5 ± 5.8 85.2 ± 5.2 84.3 ± 5.9 83.1 ± 5 2.969 0.031 0.016 1.797 0.147 0.010
Body weight (kg) 53.7 ± 11.9 53.4 ± 10.3 50.6 ± 10.3 50.2 ± 10.1 3.411 0.017 0.018 2.522 0.057 0.014
Body mass index (kg/m²) 16 ± 3 c.d 15.9 ± 2.4 c.d 18.4 ± 2 a.b.d 19.1 ± 2.2 a.b.c 57.137 <0.001 0.237 93.220 <0.001 0.337
Standing broad jump (cm) 182.3 ± 23.6 c.d 183.4 ± 24.1 c.d 187.9 ± 21.6 a.b.d 207.6 ± 22 a.b.c 20.363 <0.001 0.099 34.196 <0.001 0.157
Sit and reach (cm) 15.9 ± 7.4 c.d 17.6 ± 8 c.d 20.4 ± 7.1 a.b 21.5 ± 6.2 a.b 13.279 <0.001 0.067 14.821 <0.001 0.075
Endurance shuttle run (min) 7.3 ± 2.4 c 8.6 ± 2.5 c 9.4 ± 1.9 a.b.d 8.7 ± 1.8 c 25.600 <0.001 0.122 29.693 <0.001 0.139
30m sprint - 5m (s) 1.256 ± 0.11 1.218 ± 0.135 1.235 ± 0.125 1.218 ± 0.143 1.825 0.142 0.010 1.751 0.155 0.009
30m sprint - 10m (s) 2.104 ± 0.162 2.047 ± 0.207 2.084 ± 0.175 2.057 ± 0.177 2.078 0.102 0.011 1.427 0.234 0.008
30m sprint - 20m (s) 3.645 ± 0.31 b.c.d 3.517 ± 0.291 a 3.606 ± 0.261 a 3.547 ± 0.259 a 4.520 0.004 0.024 4.087 0.007 0.022
30m sprint - 30m (s) 5.217 ± 0.486 b.c.d 4.959 ± 0.444 a 5.09 ± 0.377 a.d 4.982 ± 0.377 a.c 7.770 <0.001 0.040 8.296 <0.001 0.043
Curl ups (N) 32 ± 7 31 ± 7 c.d 33 ± 7 b 35 ± 6 b 4.453 0.004 0.024 6.790 <0.001 0.036
Knee push ups (N) 30 ± 9 32 ± 7 30 ± 6 d 33 ± 5 c 4.731 0.003 0.025 2.911 0.034 0.016
Counter movement jump hips (cm) 27.4 ± 6.2 d 28.5 ± 6.3 d 27.6 ± 4.9 d 31 ± 5.9 a.b.c 7.789 <0.001 0.041 9.693 <0.001 0.050
Counter movement jump armswing (cm) 31.9 ± 6.3 d 34.3 ± 7 d 32.8 ± 5.6 d 36.8 ± 6.5 a.b.c 10.773 <0.001 0.055 11.425 <0.001 0.059
Motor coordination 27.468 0.000 0.128 27.456 0.000 0.129
Moving sideways (N) 55 ± 9 b.c.d 53 ± 8 a.c.d 65 ± 10 a.b 67 ± 9 a.b 64.770 0.000 0.260 84.212 0.000 0.315
Jumping sideways (N) 78 ± 11 c.d 84 ± 13 c.d 88 ± 13 a.b.d 96 ± 12 a.b.c 31.793 0.000 0.147 24.516 0.000 0.118
Balance beam (N) 49 ± 12 c.d 54 ± 11 c.d 56 ± 12 a.b 62 ± 9 a.b 19.206 0.000 0.094 9.159 0.000 0.048
Note: Non-sporting reference population = a, Sporting reference population = b, Cycling= c, BMX= d, which indicates significant differences (p<0.05) between the different groups based on the MANCOVA (post hoc
analyses).
188
Addendum B: pilot study 2
Between 2014 and 2019, 328 male BMX athletes and 1325 male cyclists between the ages of 7.0 and 18.99 were
tested. The cyclists were classified according to their self-reported main discipline. In total, 66 BMX, 62 road, 107
cyclo-cross, 72 track, and 16 MTB cyclists between 12.0 – 15.99 years of age were included in the pilot study. The
data consisted only of athletes who all performed four anthropometric, nine physical, three motor coordination
and two cycling-specific tests and registered their training history using a demographic questionnaire. Maturity
ratio (Fransen et al., 2018), age, and training history were included as covariates in the MANCOVA analyses.
Discriminant analyses were used to discriminate the profiles of the five cycling disciplines.
A multivariate effect on anthropometry, physical characteristics, motor coordination and the cycling-specific
characteristics was found for the MANCOVA analyses. The athletic profiles of BMX athletes and the different cycling
disciplines could be distinguished based on their performances on the test battery.
The BMX athletes were smaller, they have a higher body weight and BMI compared to other cycling disciplines. With
exception of the planking, curl-up test, ESHR, moving sideways, jumping sideways, and maximal cadence tests, BMX
cyclists significantly outperformed the other disciplines on the physical, motor coordination, and cycling-specific
tests. On the ESHR, the other disciplines performed significantly better compared to the BMX cyclists. The results
are displayed in Table B.1..
The discriminant analyses revealed that 61.6% of the cyclists could be correctly classified (54.2% cross-validated)
using the generic test battery (Figure B.1.). 78.8% of the BMX, 41.9% of the road, 71.0% of the cyclo-cross, 62.5% of
the track, and 0% of the MTB cyclist could be correctly assigned to their discipline.
189
Figure B.1.: Graphical representation of the discriminant analysis of the different cycling disciplines (BMX, road, cyclo-cross,
track and MTB) between 12.0 and 15.99 years of age based on performance results on a non-sport-specific test battery.
Note: the scatter plot has the canonical discriminant function coefficients as its axes, with Function 1 on x-axis and Function
2 on the y-axis.
190
Table B.1.: Average performance on the anthropometric, physical, motor coordination, and cycling-specific tests (± SD) of the cycling athletes between 12.0 - 15.99 years of age from the four different disciplines and
BMX Road Cyclo-cross Track cycling MTB MANOVA Effect size MANCOVA Effect size
Mean ± SD Mean ± SD Mean ± SD Mean ± SD Mean ± SD F p F p
N 66 62 107 72 16
Anthropometry 3.904 <0.001 0.047 2.570 0.001 0.032
Body height (cm) 161.1 ± 9.5 b 166 ± 10.3 a 162.5 ± 10.2 168.9 ± 10.4 162 ± 12.7 6.664 <0.001 0.077 2.560 0.039 0.031
Sitting height (cm) 83.1 ± 5 84.4 ± 5.8 83.1 ± 5.6 86.8 ± 5.6 83.1 ± 7.1 5.624 <0.001 0.066 0.903 0.463 0.011
Body weight (kg) 50.3 ± 10.2 c 50.8 ± 10.4 48.1 ± 9.7 a 55.9 ± 10.1 48.2 ± 10.1 6.919 <0.001 0.080 3.498 0.008 0.043
Body mass index (kg/m²) 19.2 ± 2.2 b.c 18.2 ± 2.3 c 18 ± 1.9 a 19.4 ± 1.9 18.2 ± 1.4 7.252 <0.001 0.084 6.061 <0.001 0.071
Planking (s) 182.5 ± 148.2 158.6 ± 83.2 149.8 ± 73.5 149.1 ± 71.2 166.9 ± 83.8 1.479 0.208 0.018 0.622 0.647 0.008
Standing broad jump (cm) 208 ± 22 b.c.d.e 189 ± 23 a 186 ± 21 a 189 ± 22 a 189 ± 23 a 11.551 <0.001 0.127 14.438 <0.001 0.155
Sit and reach (cm) 21.6 ± 6.1 19.3 ± 6.9 20.8 ± 7.1 20.5 ± 7.4 17.8 ± 7.5 1.561 0.184 0.019 2.669 0.032 0.033
Endurance shuttle run (min) 8.7 ± 1.9 c.e 9.2 ± 2 9.4 ± 1.9 a 9.4 ± 1.9 9.9 ± 2.2 a 2.123 0.078 0.026 3.095 0.016 0.038
30m sprint - 5m (s) 1.217 ± 0.144 1.228 ± 0.128 1.271 ± 0.12 d 1.195 ± 0.128 c 1.228 ± 0.104 4.250 0.002 0.051 3.562 0.007 0.043
30m sprint - 10m (s) 2.056 ± 0.178 c 2.066 ± 0.195 2.133 ± 0.164 a 2.031 ± 0.173 2.077 ± 0.133 4.297 0.002 0.051 3.118 0.015 0.038
30m sprint - 20m (s) 3.546 ± 0.26 c 3.597 ± 0.255 3.673 ± 0.254 a 3.52 ± 0.259 3.596 ± 0.2 4.674 0.001 0.056 4.018 0.003 0.049
30m sprint - 30m (s) 4.982 ± 0.379 c 5.071 ± 0.373 5.176 ± 0.366 a 4.969 ± 0.369 5.071 ± 0.299 4.536 0.001 0.054 4.038 0.003 0.049
Curl ups (N) 35 ± 6 32 ± 7 33 ± 7 36 ± 7 35 ± 7 3.062 0.017 0.037 2.219 0.067 0.027
Knee push ups (N) 33 ± 5 d 30 ± 6 31 ± 6 30 ± 6 a 32 ± 5 3.940 0.004 0.047 2.984 0.019 0.037
Counter movement jump hips (cm) 31 ± 5.9 b.c.d 27.1 ± 4.6 a 26.8 ± 4.8 a 28.7 ± 5 a 28.7 ± 3.8 8.234 <0.001 0.094 7.497 <0.001 0.087
Counter movement jump armswing (cm) 36.8 ± 6.6 b.c.d 32.2 ± 5.4 a 32.3 ± 5.5 a 34 ± 5.1 a 34.1 ± 4.8 7.898 <0.001 0.090 8.759 <0.001 0.100
Motor coordination 6.743 <0.001 0.078 5.449 0.001 0.050
Moving sideways (N) 67 ± 9 65 ± 11 65 ± 10 67 ± 9 65 ± 10 1.057 0.378 0.013 0.261 0.903 0.003
Jumping sideways (N) 96 ± 12 86 ± 15 89 ± 12 89 ± 12 88 ± 11 5.694 <0.001 0.067 2.021 0.091 0.025
Balance beam (N) 62 ± 9 b.d 53 ± 12 a 59 ± 10 52 ± 13 a 61 ± 8 10.687 <0.001 0.118 7.192 <0.001 0.084
Cycling specific tests 14.533 <0.001 0.155 9.171 <0.001 0.105
Shuttle bike (s) 10.27 ± 0.779 b.c.d.e 12.154 ± 1.359 a 11.677 ± 1.091 a 12.248 ± 1.484 a 11.542 ± 1.263 a 28.903 <0.001 0.267 18.702 <0.001 0.192
Maximal cadence (N) 214 ± 16 205 ± 15 207 ± 15 210 ± 13 204 ± 16 3.701 0.006 0.044 1.075 0.369 0.013
Note: BMX = a, Road = b, Cyclo-cross = c, Track = d, MTB= e, which indicates significant differences (p<0.05) between the different groups based on the MANCOVA (post hoc analyses).
191
Addendum C: Benchmarks
Table C.1.: Benchmarks (number of participants (N), mean, standard deviation (SD), percentile 25, percentile 50, and percentile 75) of male track cyclists for the different age groups and tests of the (non-)sport-specific test battery.
30m 30m 30m 30m Knee 30" ABS 30" ABS 30" REL
Training Body Body Sprint Sprint Sprint sprint Curl Push CMJ CMJ Shuttle Maximal Average Peak Average
history APHV Height Weight BMI Planking SBJ SAR ESHR 5m 10m 20m 30m Ups Ups Hips Arms MS JS BB Bike cadence Power Power power
Track (years) (years) (cm) (kg) (kg/m²) (s) (cm) (cm) (min) (s) (s) (s) (s) (N) (N) (cm) (cm) (N) (N) (N) (s) (N) (W) (W) (W/kg)
9 - 10 years N 6 11 11 11 11 8 11 11 8 11 11 11 11 11 11 11 11 8
Mean 2.5 12.9 144.8 40.2 19.19 91.4 143 16.8 6.2 1.335 2.289 3.978 5.787 50 71 42 13.976 173
SD 0.8 0.5 4.6 8.0 3.93 36.7 20 4.4 2.1 0.098 0.141 0.286 0.380 6 8 10 1.697 14
P25 1.8 12.7 141.4 33.6 15.76 64.8 130 12.5 4.3 1.267 2.218 3.802 5.500 46 65 33 12.590 168
P50 3.0 13.0 144.3 40.1 18.40 79.5 150 18.0 5.8 1.360 2.283 3.904 5.683 50 69 41 13.960 173
P75 3.0 13.3 146.0 44.1 21.23 131.3 160 20.0 8.4 1.406 2.385 4.196 6.190 52 79 53 15.800 183
11 - 12 years N 44 62 62 62 62 59 62 62 59 62 62 62 62 16 16 16 16 62 62 62 62 55 8 8 8
Mean 3.3 13.6 153.3 42.3 17.93 119.6 164 19.9 7.7 1.263 2.157 3.765 5.337 31 27 26.1 30.6 58 80 51 13.131 190 363.5 544.3 8.0
SD 1.7 0.6 7.8 6.8 2.31 74.5 19 6.8 1.8 0.094 0.140 0.246 0.352 5 5 3.5 3.8 8 10 14 1.634 17 76.8 160.4 1.1
P25 2.0 13.2 147.8 37.7 16.43 65.0 150 14.0 6.0 1.203 2.046 3.588 5.051 29 22 22.7 27.5 53 75 41 11.967 180 284.0 422.3 7.3
P50 4.0 13.7 150.9 40.9 17.71 101.0 165 19.5 8.0 1.254 2.130 3.710 5.259 32 28 26.5 30.7 58 80 52 13.146 189 378.0 532.5 8.3
P75 5.0 14.0 159.4 47.1 18.81 174.0 180 25.0 9.0 1.333 2.265 3.902 5.566 37 32 28.6 34.3 63 89 63 13.880 201 410.5 595.3 8.9
13 - 14 years N 83 113 113 113 113 102 113 113 102 113 113 113 113 43 42 43 43 113 113 113 113 103 24 24 25
Mean 4.4 13.6 168.4 54.7 19.15 133.6 187 22.7 9.3 1.198 2.035 3.512 4.944 37 29 28.1 33.3 66 89 53 12.349 208 516.0 724.5 8.4
SD 2.0 0.7 9.8 9.8 1.96 51.9 21 7.5 1.8 0.098 0.135 0.209 0.317 8 6 5.0 4.9 9 11 11 1.322 12 133.7 203.6 2.3
P25 3.0 13.2 160.8 47.8 17.90 92.0 176 18.0 8.5 1.124 1.942 3.375 4.729 31 24 24.2 29.6 60 82 46 11.364 200 406.5 578.5 7.4
P50 4.0 13.5 168.3 53.9 19.00 124.1 190 23.0 9.5 1.198 2.024 3.465 4.874 35 29 27.7 32.6 66 88 55 12.173 207 499.5 685.0 8.5
P75 6.0 14.1 175.2 61.7 20.45 164.3 202 28.0 10.5 1.261 2.123 3.657 5.154 42 33 32.1 36.9 73 98 61 13.211 217 628.8 847.0 9.5
15 - 16 years N 81 96 96 96 96 92 96 96 88 94 94 94 94 30 30 30 30 96 96 96 96 90 65 65 65
Mean 4.9 13.9 175.5 64.0 20.71 154.4 206 25.3 10.8 1.151 1.944 3.351 4.692 36 33 31.9 38.2 72 97 57 11.904 219 626.5 944.9 9.7
SD 2.4 0.4 7.5 8.4 1.75 58.7 21 7.8 1.7 0.113 0.143 0.207 0.293 7 5 4.3 4.6 11 11 12 1.282 14 108.5 180.5 0.8
P25 3.0 13.6 170.8 58.8 19.62 115.3 190 20.0 10.0 1.082 1.867 3.230 4.521 29 30 30.1 35.7 65 89 51 10.972 211 555.0 789.5 9.3
P50 5.0 13.9 175.8 63.3 20.46 146.0 209 26.0 11.0 1.131 1.932 3.325 4.678 34 33 31.5 38.1 72 96 59 11.614 220 620.0 944.0 9.7
P75 7.0 14.2 180.6 69.1 21.77 179.3 224 31.0 12.0 1.221 2.016 3.427 4.795 41 36 34.1 40.9 79 104 67 12.636 228 709.5 1.085.5 10.2
17 - 18 years N 17 24 24 25 24 23 24 24 16 24 24 24 24 7 8 8 8 24 24 24 23 22 16 16 16
Mean 6.4 14.6 179.8 70.7 21.90 194.9 222 31.7 11.7 1.128 1.890 3.220 4.488 37 37 36.4 43.8 79 107 58 11.775 230 772.6 1.184.6 10.7
SD 1.9 0.3 5.4 5.5 1.11 67.1 24 7.7 1.5 0.068 0.081 0.120 0.177 5 7 5.0 5.4 10 10 12 1.375 12 106.2 175.8 1.0
P25 5.0 14.3 176.0 66.4 21.07 147.0 213 27.6 10.6 1.082 1.829 3.135 4.362 31 30 31.9 40.8 70 102 48 10.509 221 681.3 1.029.3 10.1
P50 6.0 14.5 179.1 70.3 21.95 192.0 225 32.0 11.1 1.121 1.903 3.219 4.472 38 41 36.5 43.8 76 108 61 12.021 231 777.0 1.181.0 10.4
P75 7.5 14.9 183.4 74.4 22.68 260.0 235 38.0 13.4 1.180 1.941 3.307 4.584 42 42 39.1 48.6 88 114 68 12.560 241 856.3 1.352.8 11.7
192
Table C.2.: Benchmarks (number of participants (N), mean, standard deviation (SD), percentile 25, percentile 50, and percentile 75) of male road cyclists for the different age groups and tests of the (non-)sport-specific test battery.
Training Body Body 30m 30m 30m 30m Curl Knee CMJ CMJ Shuttle Maximal 30" ABS 30" ABS 30" REL
history APHV Height Weight BMI Planking SBJ SAR ESHR Sprint Sprint Sprint sprint Ups Push Hips Arms MS JS BB Bike cadence Average Peak Average
Road cycling (years) (years) (cm) (kg) (kg/m²) (s) (cm) (cm) (min) 5m (s) 10m (s) 20m (s) 30m (s) (N) Ups (N) (cm) (cm) (N) (N) (N) (s) (N) Power (W) Power (W) power (W/kg)
7 - 8 years N 6 6 6 6 6 6 6 6 6 6 6 6 6 6
Mean 12.5 135.0 28.6 15.74 151 24.3 1.342 2.290 4.037 5.800 44 65 48 14.013
SD 0.3 7.1 1.7 1.30 14 6.1 0.100 0.122 0.179 0.255 6 16 12 1.785
P25 12.3 129.3 26.8 14.53 140 18.8 1.242 2.188 3.951 5.607 38 52 35 12.638
P50 12.6 132.1 29.0 15.85 146 24.5 1.357 2.297 4.007 5.746 46 60 52 13.705
P75 12.8 143.0 30.0 16.51 167 27.5 1.404 2.353 4.107 5.987 49 85 57 15.265
9 - 10 years N 15 37 37 37 37 25 37 37 25 37 37 37 37 7 7 7 7 37 37 37 37 25
Mean 1.9 13.2 142.8 33.4 16.37 118.3 154 21.9 7.2 1.307 2.242 3.921 5.589 28 29 22.6 26.7 50 73 48 14.084 176
SD 1.1 0.5 6.3 4.4 1.86 63.8 16 6.4 1.7 0.076 0.129 0.218 0.322 7 3 4.1 4.1 7 14 12 2.066 16
P25 1.0 12.9 137.6 30.6 15.05 71.6 145 17.0 6.2 1.250 2.168 3.806 5.388 23 26 18.9 23.9 45 59 40 12.801 165
P50 2.0 13.3 142.2 32.8 16.20 120.0 155 22.0 7.5 1.312 2.240 3.894 5.566 27 30 22.1 25.6 49 72 51 13.690 178
P75 3.0 13.6 147.4 35.3 17.70 151.5 165 28.0 8.5 1.357 2.297 4.040 5.750 31 31 27.4 30.5 56 83 57 14.976 192
11 - 12 years N 36 75 75 75 75 51 75 74 51 75 75 75 75 20 20 20 20 75 75 75 75 49 8 8 8
Mean 3.0 13.7 154.0 40.9 17.09 116.0 164 20.4 8.2 1.273 2.163 3.758 5.334 32 25 25.2 29.5 56 79 49 13.274 191 310.8 448.4 7.5
SD 1.7 0.6 9.3 8.6 2.05 60.9 20 6.9 1.5 0.090 0.117 0.203 0.297 7 7 2.8 3.8 7 12 13 1.691 16 77.6 101.5 0.8
P25 2.0 13.3 146.7 35.0 15.59 75.0 150 15.8 7.0 1.205 2.062 3.611 5.123 27 21 23.1 27.7 50 70 41 12.030 182 248.0 377.8 7.2
P50 3.0 13.7 153.5 39.2 16.76 116.0 168 20.0 8.0 1.274 2.170 3.724 5.320 31 24 25.1 29.9 56 80 49 13.060 193 301.0 441.5 7.5
P75 4.0 14.1 159.6 43.4 18.29 148.0 179 25.0 9.0 1.344 2.244 3.917 5.559 34 32 27.6 31.4 61 86 58 14.217 204 379.3 528.0 8.2
13 - 14 years N 75 137 137 137 137 100 137 137 99 136 136 136 136 37 37 37 37 135 137 136 137 99 22 22 23
Mean 3.9 13.7 167.3 52.5 18.64 146.6 187 22.5 9.8 1.219 2.050 3.542 4.990 31 30 26.9 31.7 63 89 53 12.684 204 446.5 670.0 7.8
SD 2.2 0.6 9.0 9.5 2.24 67.3 20 8.2 1.9 0.102 0.157 0.226 0.344 6 5 4.7 5.3 7 12 12 1.558 14 114.0 162.6 2.0
P25 2.0 13.3 161.3 45.6 17.37 104.3 172 16.5 8.5 1.147 1.941 3.336 4.706 29 27 23.1 27.7 58 80 46 11.725 196 361.5 508.0 7.2
P50 4.0 13.6 167.7 52.3 18.40 134.0 185 22.0 10.0 1.207 2.051 3.545 4.972 32 31 27.2 31.9 63 90 55 12.481 203 431.5 645.5 8.2
P75 6.0 14.1 173.4 58.7 19.83 180.8 200 28.0 11.0 1.304 2.160 3.696 5.240 34 33 29.0 34.5 68 97 63 13.485 214 513.5 777.0 9.0
15 - 16 years N 84 108 108 108 108 107 108 108 99 108 108 108 108 42 41 42 42 108 108 108 108 103 68 68 68
Mean 4.3 14.0 176.1 61.3 19.74 164.1 201 26.6 11.0 1.178 1.986 3.396 4.749 36 34 30.9 37.1 69 93 55 12.250 214 574.5 842.5 9.4
SD 2.6 0.5 6.5 7.6 2.12 59.8 20 6.9 1.4 0.104 0.129 0.183 0.257 6 5 4.4 4.8 10 11 12 1.328 12 96.2 150.2 0.9
P25 2.0 13.6 170.9 55.9 18.60 122.0 190 22.0 10.0 1.097 1.898 3.263 4.580 33 31 27.3 33.2 62 86 47 11.299 205 516.5 741.5 8.9
P50 4.0 14.1 175.6 62.3 19.60 158.0 204 26.0 11.0 1.174 1.976 3.377 4.705 36 34 31.1 37.6 68 92 57 12.161 217 570.5 812.5 9.5
P75 6.0 14.3 181.2 66.6 20.67 195.0 215 31.0 12.0 1.226 2.068 3.500 4.886 39 38 35.1 41.1 75 101 65 13.093 223 650.8 958.0 10.0
17 - 18 years N 15 17 17 17 17 16 16 16 6 16 16 16 16 9 9 9 9 16 16 16 15 18 15 15 14
Mean 7.3 14.8 178.0 64.4 20.29 172.3 207 25.9 10.6 1.175 1.978 3.396 4.762 34 29 29.0 35.6 71 91 54 11.499 223 623.7 924.5 9.7
SD 2.5 0.4 6.9 6.5 1.34 63.6 16 8.9 2.1 0.103 0.123 0.173 0.246 6 5 3.2 1.8 13 11 14 1.167 16 96.5 144.2 0.9
P25 5.0 14.4 172.2 58.5 19.55 126.3 195 18.3 8.0 1.089 1.901 3.286 4.615 29 26 26.4 34.4 62 83 40 10.800 215 552.0 833.0 9.2
P50 7.0 14.9 178.2 63.3 20.00 182.5 200 26.5 11.3 1.195 1.978 3.395 4.761 34 30 28.8 35.1 69 92 58 11.221 223 604.0 905.0 9.9
P75 9.0 15.2 182.6 70.3 21.00 210.8 223 33.0 12.3 1.273 2.076 3.468 4.880 37 33 30.8 37.3 78 98 64 12.004 229 709.0 1.048.0 10.5
193
Table C.3.: Benchmarks (number of participants (N), mean, standard deviation (SD), percentile 25, percentile 50, and percentile 75) of male cyclo-cross cyclists for the different age groups and tests of the (non-)sport-specific test
battery.
Cyclo-cross (years) (years) (cm) (kg) (kg/m²) (s) (cm) (cm) (min) (s) (s) (s) (s) (N) (N) (cm) (cm) (N) (N) (N) (s) (N) (W) (W) (W/kg)
7 - 8 years N 10 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13
Mean 1.2 12.4 137.0 30.3 16.15 77.8 142 19.9 5.6 1.365 2.334 4.101 5.891 47 69 39 15.694 162
SD 0.8 0.5 5.0 2.9 0.97 55.7 18 6.5 1.9 0.080 0.086 0.183 0.317 7 11 15 2.771 17
P25 0.8 12.1 133.0 27.9 15.48 45.4 133 14.3 4.5 1.312 2.267 3.924 5.588 41 61 26 14.591 148
P50 1.0 12.4 135.6 29.5 16.01 57.0 142 20.0 5.5 1.367 2.333 4.074 5.808 46 67 37 15.460 163
P75 2.0 12.9 140.7 33.4 16.76 96.5 155 25.5 6.5 1.428 2.406 4.310 6.258 52 78 52 16.645 177
9 - 10 years N 59 87 87 87 87 75 87 87 75 87 87 87 87 23 23 24 24 87 87 87 87 73 4 4 4
Mean 2.3 13.2 142.4 33.5 16.44 119.7 155 21.0 6.8 1.308 2.242 3.928 5.591 31 25 23.1 27.6 52 73 49 13.263 178 271.5 413.5 7.7
SD 1.1 0.6 6.8 5.8 2.07 79.0 16 6.1 1.9 0.089 0.121 0.213 0.305 7 7 3.0 3.1 8 11 11 1.798 17 46.1 57.2 0.6
P25 1.0 12.8 137.2 29.3 15.20 68.0 145 17.0 6.0 1.251 2.168 3.796 5.390 23 20 21.5 25.3 47 65 42 11.985 169 232.8 353.3 7.2
P50 2.0 13.3 142.2 32.6 16.25 96.0 155 21.0 7.0 1.316 2.248 3.920 5.560 30 26 22.4 27.4 51 72 51 12.773 180 264.5 428.5 7.7
P75 3.0 13.6 146.5 36.7 17.32 148.0 165 26.0 8.0 1.355 2.309 4.031 5.783 34 30 25.2 29.9 56 82 58 14.240 190 317.3 458.8 8.3
11 - 12 years N 114 159 159 159 159 139 159 159 140 159 159 159 159 67 67 67 67 159 158 159 159 137 32 32 32
Mean 3.5 13.8 152.3 40.2 17.18 130.5 170 20.2 8.2 1.288 2.186 3.777 5.348 31 82 24.3 29.4 57 82 56 12.188 193 315.7 463.4 7.4
SD 1.6 0.6 8.6 8.1 1.91 77.4 18 6.5 1.7 0.110 0.147 0.221 0.313 7 428 3.7 4.8 8 11 11 1.156 14 71.7 113.7 1.0
P25 2.8 13.5 146.3 34.5 15.91 76.0 156 16.0 7.0 1.218 2.094 3.640 5.160 26 26 21.4 26.2 52 75 50 11.340 185 268.8 385.0 6.6
P50 3.0 13.9 151.6 39.3 17.00 115.0 170 20.0 8.2 1.276 2.170 3.751 5.316 30 29 23.7 28.6 56 82 57 12.080 195 310.5 435.5 7.4
P75 5.0 14.2 156.7 44.0 18.10 161.0 180 25.0 9.5 1.352 2.267 3.889 5.524 34 33 26.9 32.1 63 90 64 12.929 203 339.5 537.0 8.1
13 - 14 years N 128 167 167 167 167 147 167 166 148 167 167 167 167 71 71 70 70 167 167 167 167 147 51 51 52
Mean 4.4 14.0 163.4 49.2 18.28 145.5 185 22.8 9.8 1.236 2.085 3.596 5.062 33 30 27.5 32.8 64 90 58 11.640 203 414.7 613.5 8.0
SD 2.0 0.7 9.3 8.8 1.94 69.5 18 7.2 1.6 0.096 0.143 0.210 0.332 7 6 4.3 5.1 9 13 11 1.259 17 104.5 166.0 1.6
P25 3.0 13.5 157.0 42.1 16.87 90.0 175 18.0 8.5 1.169 1.997 3.448 4.858 28 27 25.1 29.1 58 81 52 10.814 195 337.0 498.0 7.3
P50 4.0 14.0 162.8 48.8 18.10 128.0 186 22.3 10.0 1.225 2.082 3.588 5.063 34 31 27.3 32.3 65 89 60 11.465 204 389.0 570.0 8.1
P75 6.0 14.5 169.8 55.3 19.43 180.0 195 28.0 11.0 1.300 2.166 3.728 5.257 38 34 31.0 35.8 70 99 67 12.312 212 493.0 730.0 8.9
15 - 16 years N 49 53 53 53 53 53 53 53 48 53 53 53 53 27 27 27 27 53 53 53 53 52 42 42 42
Mean 5.2 14.2 173.1 58.1 19.29 151.2 208 26.6 10.8 1.198 2.001 3.434 4.811 32 31 30.8 36.6 71 94 60 11.168 219 524.8 790.0 9.2
SD 3.1 0.6 8.3 8.1 1.58 49.2 16 6.0 1.9 0.101 0.129 0.195 0.280 6 6 4.1 4.7 10 10 10 1.069 13 91.8 147.1 0.7
P25 2.5 13.7 166.6 52.4 18.10 120.0 196 22.5 10.0 1.125 1.906 3.280 4.586 27 27 28.6 33.4 64 87 55 10.502 209 454.8 677.3 8.8
P50 6.0 14.2 174.5 58.7 19.20 146.0 210 27.0 11.0 1.178 2.008 3.401 4.775 31 32 30.4 36.1 69 93 61 11.032 219 521.5 795.0 9.3
P75 7.5 14.6 180.0 63.2 20.22 180.0 220 31.5 12.0 1.271 2.065 3.567 4.959 36 36 34.3 39.5 82 103 69 11.787 226 594.0 915.8 9.6
194
Table C.4.: Benchmarks (number of participants (N), mean, standard deviation (SD), percentile 25, percentile 50, and percentile 75) of male MTB cyclists for the different age groups and tests of the (non-)sport-specific test battery.
MTB (years) (years) (cm) (kg) (kg/m²) (s) (cm) (cm) (min) (s) (s) (s) (s) (N) (N) (cm) (cm) (N) (N) (N) (s) (N) (W) (W) (W/kg)
7 - 8 years N 6 10 10 10 10 9 10 10 9 10 10 10 10 3 3 3 3 10 10 10 10 9
Mean 1.8 12.3 132.4 28.5 16.05 84.0 139 23.7 5.4 1.409 2.408 4.268 6.134 32 30 17.5 21.3 47 67 45 14.266 159
SD 0.4 0.4 8.6 6.7 1.86 20.3 14 4.1 1.1 0.060 0.110 0.234 0.399 2 10 3.5 5.6 5 11 11 1.463 18
P25 1.8 11.9 124.8 23.0 14.60 68.5 134 20.5 4.8 1.360 2.319 4.087 5.819 30 19 14.4 16.2 45 58 37 13.321 147
P50 2.0 12.3 133.6 27.5 15.49 84.0 140 23.0 5.0 1.410 2.402 4.242 6.085 32 32 16.9 20.3 47 66 47 14.354 159
P75 2.0 12.6 139.0 32.6 17.02 102.0 143 26.8 6.3 1.457 2.517 4.500 6.478 50 73 55 15.578 168
9 - 10 years N 19 23 23 23 23 22 23 23 22 23 23 23 23 7 7 7 7 23 23 23 23 22
Mean 3.2 13.2 139.0 32.3 16.60 121.6 153 20.2 7.1 1.320 2.282 4.047 5.712 28 25 18.8 22.2 53 73 53 12.382 172
SD 0.8 0.4 6.0 5.5 1.72 61.3 16 5.2 1.6 0.101 0.134 0.287 0.332 5 6 5.0 3.8 5 12 8 2.022 18
P25 3.0 13.0 135.0 28.2 15.25 71.0 141 17.0 6.0 1.271 2.189 3.902 5.591 23 18 14.3 18.9 50 66 50 11.064 161
P50 3.0 13.1 137.5 31.1 15.92 102.0 150 20.0 7.3 1.327 2.250 3.975 5.677 28 26 17.1 21.7 52 68 53 11.670 175
P75 4.0 13.6 141.7 35.4 17.20 170.5 166 25.0 8.0 1.371 2.411 4.194 5.954 32 30 24.3 25.9 55 80 58 13.606 185
11 - 12 years N 30 36 36 36 36 36 36 36 36 36 36 36 36 13 13 13 13 36 36 36 36 36 8 8 8
Mean 3.8 13.9 149.7 39.5 17.54 121.9 169 19.9 8.4 1.267 2.175 3.796 5.386 29 29 22.6 26.9 57 79 57 11.653 190 279.9 418.1 7.0
SD 1.7 0.4 8.8 6.3 1.50 52.7 20 6.2 2.0 0.107 0.130 0.229 0.341 8 6 3.8 4.8 8 10 8 1.112 14 28.7 60.3 0.6
P25 3.0 13.8 145.8 36.1 16.55 89.5 151 15.3 6.6 1.185 2.082 3.604 5.159 19 24 19.2 21.7 51 74 51 10.847 182 264.8 369.5 6.4
P50 4.0 14.0 149.7 39.0 17.50 119.0 169 21.0 8.5 1.259 2.140 3.720 5.280 32 26 24.2 27.5 58 82 58 11.485 191 278.5 410.5 7.0
P75 5.0 14.2 155.7 43.0 18.68 144.3 185 24.5 9.9 1.332 2.290 3.982 5.616 35 35 25.6 31.4 61 85 63 12.105 199 302.0 481.0 7.5
13 - 14 years N 27 37 37 37 37 34 37 37 34 37 37 37 37 12 11 12 12 37 37 37 37 33 6 6 6
Mean 5.2 14.1 161.0 47.1 18.03 154.5 190 23.0 9.3 1.195 2.040 3.533 5.003 35 30 28.9 34.0 64 88 58 11.628 204 500.8 722.2 9.4
SD 2.1 0.7 8.7 8.2 1.78 71.3 20 7.2 2.0 0.094 0.127 0.234 0.357 7 3 5.4 7.1 8 11 8 1.423 16 117.6 198.5 1.2
P25 4.0 13.7 155.8 41.3 16.60 106.3 177 16.5 8.0 1.108 1.950 3.392 4.771 28 28 24.6 27.9 59 80 54 10.563 188 397.0 539.3 8.3
P50 5.0 14.1 160.6 44.9 17.71 138.0 190 22.0 10.0 1.206 2.041 3.487 4.945 35 29 27.4 32.7 63 89 58 11.300 204 503.5 685.5 9.2
P75 7.0 14.8 167.6 54.5 19.30 175.0 203 29.8 11.0 1.269 2.142 3.682 5.238 41 34 31.6 39.2 68 97 64 12.600 215 584.5 899.3 10.4
15 - 16 years N 13 18 18 18 18 18 17 17 17 18 18 18 18 7 7 7 7 18 18 18 18 16 12 12 12
Mean 6.7 14.4 172.1 58.8 19.75 153.3 205 25.2 10.6 1.168 1.973 3.382 4.781 36 32 32.4 38.7 70 98 62 10.805 213 578.6 929.8 9.6
SD 1.6 0.4 7.5 8.8 1.67 66.6 21 6.4 1.5 0.065 0.109 0.158 0.256 6 4 5.5 5.4 11 13 6 0.723 15 101.8 215.2 0.9
P25 5.0 14.2 166.3 50.4 18.15 105.5 190 22.0 9.8 1.120 1.899 3.305 4.651 30 28 27.8 34.8 62 86 59 10.194 201 497.3 788.3 8.8
P50 7.0 14.4 171.8 59.4 20.16 127.5 203 25.0 10.5 1.156 1.978 3.371 4.750 38 32 31.6 38.4 71 100 63 10.910 213 590.0 883.5 9.6
P75 8.0 14.7 177.8 64.6 21.22 194.0 215 27.5 11.8 1.228 2.051 3.477 4.918 41 36 38.7 45.3 77 107 66 11.298 227 670.3 1.086.0 10.3
195
Table C.5.: Benchmarks (number of participants (N), mean, standard deviation (SD), percentile 25, percentile 50, and percentile 75) of male BMX cyclists for the different age groups and tests of the (non-)sport-specific test battery.
30" ABS 30" ABS 30" REL
Training Body Body 30m 30m 30m Curl Knee CMJ CMJ Shuttle Maximal Average Peak Average
history APHV Height Weight BMI Planking SBJ SAR ESHR Sprint Sprint 30m Sprint sprint Ups Push Hips Arms MS JS BB Bike cadence Power Power power
BMX (years) (years) (cm) (kg) (kg/m²) (s) (cm) (cm) (min) 5m (s) 10m (s) 20m (s) 30m (s) (N) Ups (N) (cm) (cm) (N) (N) (N) (s) (N) (W) (W) (W/kg)
7 - 8 years N 30 45 45 45 45 37 45 45 37 45 45 45 45 16 16 16 16 45 45 45 45 36
Mean 3.0 12.4 133.1 28.9 16.26 106.3 150 23.5 5.5 1.362 2.335 4.098 5.881 29 25 21.3 25.3 48 69 47 12.704 168
SD 0.8 0.4 5.6 3.8 1.14 58.8 14 6.1 1.4 0.100 0.146 0.258 0.351 7 8 4.3 4.3 6 9 12 1.397 13
P25 2.0 12.1 128.2 25.7 15.60 69.0 140 20.0 4.5 1.290 2.227 3.940 5.646 22 21 17.0 22.0 44 63 39 11.682 158
P50 3.0 12.4 132.9 28.9 16.10 95.0 150 23.0 5.5 1.349 2.317 4.083 5.852 29 26 21.9 25.3 48 69 45 12.650 170
P75 3.3 12.7 138.2 31.6 16.75 118.0 156 27.0 6.3 1.413 2.403 4.258 6.165 34 33 24.1 27.8 52 75 58 13.765 177
9 - 10 years N 48 70 70 70 70 64 70 70 64 69 69 69 69 19 19 19 19 70 70 70 70 64 4 4 4
Mean 4.0 13.1 141.9 33.7 16.67 112.6 168 22.3 7.2 1.299 2.212 3.848 5.494 34 30 23.5 29.2 54 78 52 11.291 182 226.8 348.3 6.6
SD 0.9 0.4 6.0 4.8 1.37 56.6 12 5.8 1.7 0.083 0.113 0.210 0.287 10 6 4.9 5.9 5 10 11 1.243 16 59.9 104.6 1.4
P25 4.0 12.8 137.6 29.7 15.46 75.3 160 17.0 6.0 1.242 2.146 3.704 5.256 27 24 20.5 24.6 50 74 44 10.454 173 172.5 257.0 5.1
P50 4.0 13.1 143.0 34.2 16.75 104.5 169 22.8 7.5 1.304 2.211 3.825 5.471 32 30 21.9 27.5 55 79 54 11.065 182 221.0 332.0 7.0
P75 5.0 13.5 146.3 37.0 17.60 134.8 180 26.1 8.0 1.360 2.293 3.983 5.708 44 33 29.6 34.7 57 84 60 11.914 194 286.8 455.8 7.7
11 - 12 years N 70 85 85 85 85 79 85 85 78 85 85 85 85 40 39 40 40 85 85 85 85 79 10 10 10
Mean 5.5 13.8 150.5 40.2 17.65 133.3 185 22.2 8.0 1.252 2.137 3.691 5.216 34 31 26.1 31.2 61 89 60 10.654 199 325.7 510.1 7.1
SD 1.4 0.5 7.3 6.4 1.74 75.7 17 5.6 1.7 0.095 0.124 0.189 0.282 6 6 4.4 4.5 7 10 9 0.845 16 33.2 72.5 0.4
P25 5.0 13.5 144.5 35.9 16.05 86.0 174 19.0 7.0 1.182 2.055 3.548 4.987 30 27 22.1 28.3 57 83 53 10.026 190 301.0 475.3 6.8
P50 6.0 13.8 151.1 40.1 17.53 121.0 182 23.0 8.0 1.244 2.127 3.698 5.187 34 31 25.0 30.6 61 88 62 10.529 199 336.5 527.5 7.3
P75 6.0 14.1 155.3 44.9 19.20 170.0 200 26.0 9.5 1.317 2.228 3.844 5.382 39 36 29.9 33.8 67 95 67 11.220 210 350.5 548.3 7.5
13 - 14 years N 62 78 78 78 78 72 78 78 73 78 78 78 78 44 44 43 43 78 78 78 78 73 24 24 24
Mean 6.8 14.0 164.4 53.1 19.51 163.6 208 22.4 8.9 1.181 2.006 3.469 4.880 35 34 32.3 38.1 66 96 63 10.327 214 453.5 687.8 8.4
SD 1.9 0.7 8.5 8.9 2.02 134.3 21 6.6 1.6 0.119 0.149 0.220 0.312 6 5 4.7 5.4 10 11 9 1.030 16 110.3 191.1 1.1
P25 5.8 13.6 159.3 46.5 17.79 105.0 199 17.0 8.0 1.107 1.904 3.312 4.650 30 31 30.0 34.6 60 88 58 9.712 204 349.5 577.3 7.4
P50 7.0 13.9 164.0 52.6 19.57 125.0 210 23.0 9.0 1.179 2.014 3.467 4.885 34 34 33.5 39.0 66 96 65 10.182 214 463.5 738.5 8.7
P75 8.0 14.3 169.4 58.9 20.75 189.3 223 26.6 10.0 1.271 2.119 3.622 5.080 38 36 36.4 41.4 72 104 69 10.854 227 518.8 778.8 9.2
15 - 16 years N 27 30 30 30 30 30 30 30 29 30 30 30 30 9 9 9 9 30 30 30 30 30 25 25 25
Mean 7.9 14.1 173.6 65.5 21.66 144.3 230 26.0 9.6 1.145 1.925 3.284 4.576 35 33 38.1 44.6 70 102 60 9.932 229 663.5 1.044.4 10.0
SD 1.6 0.5 7.2 8.4 1.73 46.0 25 5.0 2.4 0.091 0.125 0.210 0.313 5 4 4.6 3.9 11 14 11 0.876 15 138.2 216.9 1.2
P25 7.0 14.0 168.2 59.9 20.40 118.3 215 22.0 7.8 1.080 1.836 3.124 4.357 30 30 33.8 40.6 59 95 55 9.483 220 579.5 964.5 9.3
P50 8.0 14.2 174.7 65.4 21.72 136.0 233 25.5 10.5 1.120 1.878 3.219 4.499 35 33 37.5 44.0 68 105 64 9.836 230 639.0 1.016.0 10.0
P75 9.0 14.4 177.0 72.0 23.11 166.5 246 28.9 11.0 1.223 2.049 3.446 4.787 41 37 42.4 48.6 77 113 69 10.365 238 744.0 1.184.0 10.4
17 - 18 years N 8 9 9 9 9 9 9 9 4 9 9 9 9 9 9 9 9 9 9 9 9
Mean 8.5 14.7 177.2 74.7 23.78 171.3 267 25.4 10.4 1.130 1.870 3.139 4.348 71 108 65 9.201 246 835.0 1.396.8 11.2
SD 2.2 0.3 5.1 7.1 1.79 67.1 18 6.1 0.9 0.044 0.059 0.073 0.110 9 13 7 0.576 6 115.5 248.0 1.0
P25 6.3 14.5 172.2 70.1 23.29 111.5 251 20.3 9.6 1.092 1.828 3.084 4.232 66 104 60 8.842 240 733.0 1.137.0 10.5
P50 8.5 14.6 179.4 74.5 23.90 159.0 270 25.0 10.3 1.104 1.860 3.108 4.369 69 108 67 9.187 246 842.0 1.433.0 11.0
P75 10.5 14.9 181.3 79.3 24.83 222.5 280 30.0 11.3 1.177 1.913 3.200 4.423 75 115 71 9.283 252 923.5 1.599.5 11.8
196
CURRICULUM VITAE
Personalia
Name: Mireille Mostaert
Email (professional): mireille.mostaert@ugent.be
Email (personal): mireillemostaert@hotmail.com
Educations and certifications

Educations:
Voskenslaan Ghent: Sciences Mathematics (7 h)
Ghent University: 2009-2013: Master of Science in Physical Education and Movement Sciences — Main
Subject Sports Training - magna cum laude
2013-2014: Master of Science in Physical Education and Movement Sciences — Main
Subject Sports Management – magna cum laude
2014-2017: Academic Teacher Education in Physical Education and Movement Sciences –
cum laude
Certifications:
Highest coaching degree in Flanders: Trainer A single skating
Initiator team sports, artistic sports, athletics, swimming, and climbing certificate
Additional educations:
2017: Advanced Academic English
2018: Statistics - Introduction to R
2019: Power analyses
Work experiences
2013 – 2022: Doctoral assistant (AAP) Ghent University, Department of Movement and Sports Sciences,
Research unit Biomechanics and Motor Control of Human Movement
2020 – … : Sport technical coordinator Flemish Skating Union
2010 – … : National synchronized skating coach
Occassionaly: VTS instructor
Publications as part of this thesis

Mostaert, M., Laureys, F., Vansteenkiste, P., Pion, J., Deconinck, F., & Lenoir, M. (2020). Discriminating
performance profiles of cycling disciplines. International Journal of Sports Science & Coaching, 16(1), 110–122.
doi: 10.1177/1747954120948146
Mostaert, M., Vansteenkiste, P., Pion, J., Deconinck, F., & Lenoir, M. (2021). The importance of performance in
youth competitions as an indicator of future success in cycling. European Journal of Sport Science, 1-10. doi:
10.1080/17461391.2021.1877359
201
Mostaert, M*., Gallo, G*., Faelli, E., Ruggeri, P., Delbarba, S., Codella, R., Vansteenkiste, P., Filipas, L. (2022). Do
race results in youth competitions predict future success as a road cyclist? A retrospective study in the Italian
Cycling Federation. International Journal of Sports Physiology and Performance, 1(aop), 1-6.
(*Mostaert, M, Gallo, G share first authorship)
Mostaert, M., Vansteenkiste, P., Laureys, L., Rommers, N., Pion, J., Deconinck, F., Lenoir, M. Is motor coordination
the key to success in youth cycling? (2021 – submitted to the International Journal of Sports Physiology and
Performance).
Other publications
Coppens, E.*, Laureys, F*., Mostaert, M*., D'Hondt, E., Deconinck, F., & Matthieu, L. (2021). Validation of a motor
competence assessment tool for children and adolescents (KTK3+) with normative values for 6- to 19-year-
olds. Frontiers in Physiology, 12. doi: 0.3389/fphys.2021.703312
(*Coppens, E., Laureys, F., Mostaert, M. share first authorship)
Laureys, F., Middelbos, L., Rommers, N., De Waelle, S., Coppens, E., Mostaert, M., Deconinck, F, Lenoir, M. (2021).
The effects of age, biological maturation and sex on the development of executive functions in adolescents.
Frontiers in Physiology, 12, 1453. doi: 10.3389/fphys.2021.703312
Mostaert, M., Pion, J., Lenoir, M., & Vansteenkiste, P. (2020). A retrospective analysis of the national youth teams
in volleyball : were they always faster, taller, and stronger? Journal of Strength and Conditioning Research, 1-
7. doi: 10.1519/JSC.0000000000003847
Robertson, K., Laureys, F., Mostaert, M., Pion, J., Deconinck, F., & Lenoir, M. (2020). Mind, body, and shuttle:
multidimensional benchmarks for talent identification in male youth badminton. Biology of Sport, 39(1), 79–
94. doi: 10.5114/biolsport.2021.101603
Norjali Wazir, Van Hiel, M., Mostaert, M., Deconinck, F.J.A., Pion, J., Lenoir, M. (2019). Identification of elite
performance characteristics in a small sample of taekwondo athletes, PLOS ONE, 14(5), e0217358.
Rommers, N., Mostaert, M., Goossens, L., Vaeyens, R., Witvrouw, E., Lenoir, M., D’Hondt, E. (2018). Age and maturity
related differences in motor coordination among male elite youth soccer players. Journal of Sports Sciences,
37(2), 196-203. doi: 10.1080/02640414.2018.1488454
Norjali Wazir, M.R.W., Torfs, M., Mostaert, M., Pion, J., Lenoir, M. (2018). Anthropometry, physical performance,
and motor coordination of medallist and non-medallist young fencers. Archives of Budo,14, 33-40.
Robertson K, Pion J, Mostaert M, Norjali Wazir MRW, Kramer T, Faber IR, Lenoir M. (2018). Coaches’ perspective
on the contribution of anthropometry, physical performance, and motor coordination in racquet sports;
Journal of Sport Sciences, 36(23), 2706-2715.
Norjali Wazir, M.R.W., Torfs, M., Mostaert, M., Pion, J., Lenoir, M. (2017). Predicting judo champions and medallists
using statistical modelling. Archives of Budo; 13: 161-168.
202
Mostaert, M.*, Deconinck, F.*, Pion, J., Lenoir, M. (2016). Anthropometry, physical fitness and coordination of
young figure skaters of different levels. International Journal of Sports Medicine; 37: 531-538.
(*Mostaert, M., and Deconinck, F. share first authorship)
Presentation at national and international conferences

Oral presentation at the Science & Cycling conference, Leuven, Belgium (22-23 September 2021):
Mostaert, M., Vansteenkiste, V., Laureys, F., Rommers, R., Pion, J., Deconinck, F., Lenoir, M. (23/09/2021). Motor
coordination may be the key to success in youth cycling.
Toronto Talent Conference, Toronto, Ontario - Canada (13-15 May 2020) - CANCELLED due to COVID-19:
Mostaert, M., Vansteenkiste, P., Pion, J., Lenoir, M. (2020). A retrospective analysis of the national youth teams
in volleyball -Were they always faster, taller, and stronger?
Oral presentations at the ECSS conference, Prague, Czech Republic (3-6 July 2019):
Mostaert, M., Laureys, F., Pion, J., Deconinck, F., Lenoir, M. (3/7/2019). The importance of performance in
competitions as an indicator of future success in cycling (finalist of the Young Investigator Award).
Norjali Wazir, M.R.W., Van Hiel, M., Mostaert, M., Deconinck, F., Pion, J., Lenoir, M. (4/7/2019). Identification of elite
performance characteristics in small sample of taekwondo athletes.
Oral presentation at the Science & Cycling conference, Cane, France (4-5 July 2018):
Mostaert, M., Laureys, F., Pion, J., Deconinck, F., Lenoir, M. (5/7/2018). Anthropometry, physical fitness, and
coordination in young cyclists of different disciplines.
Poster presentation at the NASPSPA conference, San Diego, California - USA (4 – 7 June 2017):
Mostaert, M., Deconinck, F., Norjali Wazir, M.R.W., Robertson, K., Pion, J., Lenoir, M. (6/6/2017). Identifying the
differences of anthropometric, physical, motor coordination and cycling specific characteristics of young
cyclists from five different cycling disciplines (BMX, road cycling, track cycling, MTB, cyclo-cross).
Oral presentation at the Youth Sport Conference, Ljubljana, Slovenia (9 - 10 December 2016):
Mostaert, M., Deconinck, F., Pion, J., Lenoir, M., (10/12/2016). Anthropometry, physical fitness and coordination of
young figure skaters of different levels.
Oral presentation at VK symposium, Leuven, Belgium (15 December 2015):

Mostaert M, Deconinck F, Pion J, Lenoir M. (15/12/2015). Anthropometry, physical fitness and coordination of
young figure skaters of different levels (won the VTS-award).
203
ACKNOWLEDGEMENTS – DANKWOORD
Vandaag, 31 maart 2022, is een emotionele dag voor mij. Een dag waarop ik een heel groot hoofdstuk uit mijn leven
moet afsluiten en dit zowel letterlijk als figuurlijk. Zo een 13 jaar geleden begon het allemaal voor mij hier aan het
HILO. Studeren werd ingeruild voor een job als curriculummanager en vervolgens een assistentschap met een
doctoraat als kers op de taart. In mijn gehele traject heb ik heel veel fantastische (co)promotoren, begeleiders,
collega’s, vrienden, studenten en zakenpartners leren kennen, die ik dan ook vandaag graag in de bloemetjes wil
zetten.
Mijn eerste woorden van dank gaan uit naar mijn promotor, copromotor en de leden van de begeleidingscommissie.
Zonder jullie hulp stond ik hier niet vandaag en had ik de vele opportuniteiten die jullie mij gegeven hebben, niet
ten volle kunnen benutten.
Beste Prof. dr. Lenoir, beste Matthieu, heel erg bedankt voor je steun en begeleiding de afgelopen zeven jaar. Uw
passie voor sport en talent werkte aanstekelijk. Dankzij jou ben ik mezelf blijven ontwikkelen als onderzoeker,
sporter, trainer, persoon. Je vond steeds de gepaste woorden om mijn werk naar een hoger niveau te tillen. Ik ben
je ontzettend dankbaar voor alle kansen die je mij hebt gegeven.
Beste Prof. dr. Deconinck, beste Frederik, ik vond het heel fijn om met je samen te werken. Dank je wel voor de vele
gedetailleerde feedback en kritische blik op mijn werk.
Beste dr. Pion, beste Johan, jij was mijn allergrootste schoolmeester op het vlak van ‘talent’. Niet alleen hebben we
samen vele talenten ontdekt, maar heb jij ook het talent in mezelf naar boven gebracht. Dank je wel voor de vele
steun, begeleiding en vriendschap!
Beste dr. Vansteenkiste, beste Pieter, je kwam in mijn traject op een moment waarop ik het moeilijk had. Door jouw
aanmoedigende woorden en geloof in mij, stimuleerde je mij om verder te doen. Samen hebben we uiteindelijk
veel artikels geschreven. Je hebt mij veel kennis en inzichten bijgebracht. Ik ben je ontzettend dankbaar daarvoor!
I would like to express my sincere gratitude to the members of the examination board: Prof. dr. Jan Bourgois
(Chairman), Prof. dr. Roberto Codella, Prof. dr. Marije Elferink-Gemser, Prof. dr. Reinout Van Schuylenbergh, Prof. dr.
Dirk De Clercq, Prof. dr. Jan Boone, dr. Bram Constandt. Thank you for your time and effort to read my dissertation
and to provide valuable feedback. Your constructive comments and interesting questions were very helpful to
improve the manuscript.
Een speciale dankjewel gaat uit naar de samenwerking met Cycling Vlaanderen en Belgian Cycling, met name Koen
Beeckman, Frederik Broché, Frank Glorieux, Jan Vancompernolle, Erwin Konincks, topsport trainers (Kieran en
Kristof), Massimo Van Lancker, Anne-Laure Gheerardyn, Chris Jacobs, het personeel van Cycling Vlaanderen, Belgian
207
Cycling en Sport Vlaanderen. Het vertrouwen en de mogelijkheden die jullie mij en mijn onderzoek hebben geboden
is van onschatbare waarde. Ook Felien, de studenten en trainers die hebben meegeholpen met de organisatie van
de vele testdagen en de wielrenners die hebben deelgenomen aan de testdagen wil ik enorm bedanken. Ik hoop
dat ik met mijn proefschrift jullie toch een stap vooruit heb kunnen brengen.
Niet alleen ben ik werkzaam geweest bij Cycling Vlaanderen, ook heb ik vele uren gedraaid tijdens testdagen in
scholen en bij diverse sportfederaties. Dank je wel voor jullie vertrouwen en aan iedereen die geholpen heeft en
heeft deelgenomen aan de testdagen.
Mijn tijd aan het HILO was onvergetelijk dankzij mijn fantastische collega’s. Samen hebben we naast het harde
werken en inhoudelijke discussies zoveel plezier kunnen beleven aan het HILO, tijdens de testdagen, middagpauzes
(frietjes eten), koffiepauzes, uitstapjes, drinks,… Jullie steunden mij door dik en dun, ook als het af en toe eens wat
minder ging. COVID-19 en het naderen van de deadline van mijn proefschrift gooide jammer wat roet in het eten
en creëerde wat meer afstand. Maar dankzij jullie kon ik mijn batterijen telkens terug opladen toen ik naar het HILO
kwam. In het bijzonder wil ik graag Felien, Nikki, Eline, Mireille, Kamasha, Griet, Silke, Lee, Pieter, Johan, Jade, Linus,
Jolien, Sien, An, Jozefien,… bedanken voor de fijne tijd. Ik ga jullie heel hard missen!
Ook mijn collega’s van het eerste verdiep wil ik graag apart bedanken. Jullie ontvingen mij met open armen als
kersvers afgestudeerde en vertrouwden mij de job van curriculummanger toe. Ik weet nog dat ik het ontzettend
moeilijk had om jullie bureau te verlaten. Jullie zijn altijd mijn veilige thuishaven geweest op het HILO en maakten
deel uit van mijn dagelijkse routine. Wat zal ik de fijne babbeltjes missen met jullie, met name Petra, Erwin, Vera
en Lore. Ook bij het bovenste verdiep kon ik altijd terecht bij Nadia, Isabelle, Rudi, Tine. Dank je wel daarvoor!
In het voorjaar 2020 kreeg ik de kans om halftijds aan de slag te gaan bij de Vlaamse Schaatsunie, naast het
halftijds afronden van mijn doctoraatsverhandeling. Een droom die uitkwam en perfect aansloot bij mijn ambities
en passie. Bedankt Prof. Lenoir om mij deze kans te geven. Het was uiteraard niet altijd even gemakkelijk om alles
te combineren. Ik wil dan ook mijn collega’s aan de Vlaamse Schaatsunie hartelijk bedanken voor jullie flexibiliteit
en geloof in mij. Dank je wel Jurgen Schroyen, Gaby Deckmyn, Corné Lepoeter, Annelies Dom, Jeroen Straathof,
Gunter Werner, Ludo Daemen, Maria Bouwens en Vera Huys.
Als coach ben ik vele uren naast mijn doctoraat bezig geweest met synchro les geven. Toen het einde van mijn
proefschrift in zicht kwam, werd deze combinatie steeds lastiger en ook mede door mijn dikker wordende buik. Met
pijn in het hart, heb ik mijn schaatsers minder frequent gezien, maar dankzij de fantastische hulp van voornamelijk
208
Maxime, Yanna, Loïs en Els werd dit zeer goed opgevangen. Heel erg bedankt daarvoor, want op deze manier kon
ik met een gerust hart verder schrijven en hebben de schaatsers steeds progressie blijven maken, hun talenten
verder blijven ontwikkelen.
Ook mijn lieve vrienden horen thuis in dit lijstje. Een ongelofelijk dikke dankjewel aan Eva, mijn naamgenootje
Mireille, Jolien, Lynn, Yanna, Yoline, Maxime, Kamasha, Yannick en Barbara, … om mij zo hard te steunen, af te leiden,
aan te moedigen, mijn gezaag te aanhoren. Bij jullie kon ik altijd terecht! Jullie zijn vriendinnen uit de duizend! Ik
kijk er naar uit om terug samen veel tijd met elkaar door te brengen en bij te kletsen.
Tot slot wil ik graag de mensen bedanken die al (bijna) mijn hele leven voor mij klaar staan. Mijn familie en
schoonfamilie Lieve, Gilbert, Sofie, Karel, Peter en Sara. Wat fijn om jullie aan mijn zijde te hebben en jullie steun te
voelen. Mijn allerliefste ouders, mama, papa, zonder jullie had ik hier niet gestaan! Jullie blijven altijd
onvoorwaardelijk in mij geloven, steunen, helpen mij, hebben mij lief en geven mij ten gepaste tijde een duwtje in
de rug. Papa, voilà hier sta ik dan en ook al geef ik het niet graag toe, ik ben blij dat ik in je voetsporen ben getreden.
Ik ben jullie onbeschrijfelijk dankbaar!
Mijn allerliefste man, mijn Jan, toekomstige papa. Wat heb ik je het leven lastig gemaakt het afgelopen jaar tijdens
het afronden van mijn doctoraatsverhandeling. Een oprechte sorry is hier op zijn plaats, maar ook een
onbeschrijfelijke dank je wel! Ik kijk ernaar uit om eindelijk terug het normale leven op te pikken en samen met
jou ons voor te bereiden op een nieuw hoofdstuk met onze kleine spruit.
209
View publication stats

Measuring Performance To Predict Performance - The Development of The Cycling Compass

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Measuring Performance To Predict Performance - The Development of The Cycling Compass

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Measuring performance to predict performance - the development of the

Thesis · January 2022

Talent identification within cycling View project

The user has requested enhancement of the downloaded file.

Thesis submitted in fulfilment of the requirements for the degree of

Printed by University Press, Wachtebeke

List of abbreviations .....................................................................................................................................................................................................................................I

PART 1 : GENERAL INTRODUCTION ................................................................................................................................................................................................ I

3.1. The Sports Compass .......................................................................................................................................................................................... 5

4. Potential components of the Cycling Compass................................................................................................................................... 9

4.1. Lessons from other sports .......................................................................................................................................................................... 9

4.1.1. Growth, maturation and age .................................................................................................................................................................... 9

4.2. Focus on cycling ................................................................................................................................................................................................. 14

4.2.1. Features and athletic profile of road cycling ........................................................................................................................... 16

4.3. Talent identification based on competition performance ........................................................................................... 23

4.3.1. Competition results ....................................................................................................................................................................................... 23

5. Hiatuses for the development of a Cycling Compass ................................................................................................................... 27

5.2.1. Predictions based on a (non-)sport-specific test battery............................................................................................28

5.3. Talent development....................................................................................................................................................................................... 29

6. General overview and aims of this research ..................................................................................................................................... 30

PART 2 : ORIGINAL RESEARCH .....................................................................................................................................................................................................45

Study 1 - Discriminating performance profiles of cycling disciplines ...........................................................................................................49

Chapter 2 : Prediction of future performance in road cycling ............................................................................................................................. 75

Study 2 : Is motor coordination the key to success in youth cycling? .......................................................................................................... 77

PART 3 : GENERAL DISCUSSION ................................................................................................................................................................................................ 135

1. Summary of the main findings................................................................................................................................................................... 137

1.1. Investigating the athletic profiles of cycling athletes .................................................................................................. 137

1.2.1. Predictions based on a (non-)sport-specific test battery..........................................................................................138

2. Overall discussion .............................................................................................................................................................................................. 140

2.1. Motor coordination ...................................................................................................................................................................................... 140

2.2. Talent orientation and transfer ......................................................................................................................................................... 143

2.3. Talent identification, selection and performance prediction .................................................................................. 147

2.3.1. Longitudinal research in this dissertation............................................................................................................................... 147

2.4. Talent development...................................................................................................................................................................................... 151

3. Practical implications ....................................................................................................................................................................................... 153

3.1. Cycling compass .............................................................................................................................................................................................. 153

3.1.1. Talent detection.............................................................................................................................................................................................. 155

4. Strengths, limitations, and suggestions for future research ................................................................................................ 158

4.1. Strengths ...............................................................................................................................................................................................................158

4.1.1. Adaptations of the test battery for future research .......................................................................................................158

4.2. Limitations ...........................................................................................................................................................................................................165

4.2.1. Main focus on male cycling ..................................................................................................................................................................165

4.3. Suggestions for future research ...................................................................................................................................................... 169

4.3.1. Evolution in cycling..................................................................................................................................................................................... 169

ADDENDA ............................................................................................................................................................................................................................................... 181

CURRICULUM VITAE ........................................................................................................................................................................................................................ 199

30” ABS = 30 seconds all-out cycling test absolute score

3.1. The Sports Compass

3.2. Current approaches in cycling

3.3. The Cycling Compass

4.1. Lessons from other sports

4.1.1. Growth, maturation and age

4.1.2. Anthropometric tests

4.1.3. Physical tests

4.1.5. Sport-specific tests

4.1.6. Additional constraints

4.2. Focus on cycling

4.2.2. Features and athletic profile of track cycling

4.2.3. Features and athletic profile of cyclo-cross

4.2.4. Features and athletic profile of mountain bike

4.2.5. Features and athletic profile of BMX

Stage 2 0.205 9.204 * 0.128 17.172 *

Note: p<0.05; p<0.01; p<0.001

Note: p<0.05; p<0.01; p<0.001