Development and validation of a novel movement screen to predict lower extremity injury in male youth soccer players

Transcription

1 Development and validation of a novel movement screen to predict lower extremity injury in male youth soccer players Paul James Read MSc, BSc (hons), ASCC, CSCS*D A thesis submitted to the School of Sport at Cardiff Metropolitan University in fulfillment of the requirements for the degree of Doctor of Philosophy in January 2016 Director of Studies: Dr. Rhodri S. Lloyd Supervisors: Dr. Jon L. Oliver and Prof. Mark B. De Ste Croix Academic Advisor: Dr. Greg D. Myer 1

2 PRIFYSGOL CYMRU : UNIVERSITY OF WALES SUBMISSION FOR THE DEGREE OF DOCTOR OF PHILOSOPHY Declaration and statements of student in respect of their submitted work This work is being submitted in fulfillment of the requirements for the degree of Doctor of Philosophy within the field of paediatric exercise science and injury prevention. I certify that this work has not been previously submitted or accepted for the fulfillment of any degree, and is not being concurrently submitted in candidature for any other academic programmeme. I further certify that the whole of this work is the result of my individual efforts, except where otherwise stated. All quotations from books and journals have been acknowledged and a bibliography is appended. I hereby give consent for my thesis, if accepted, to be available for photocopying, and for inter-library loan, and for the title and summary to be made available to outside organisations. I hereby give consent for the University to electronically store, copy or translate the thesis to any approved medium or format for the purpose of future preservation and accessibility. I hereby give consent that upon deposition in the digital repository, for the thesis to be made accessible to a wide variety of people and institutions, including automated agents and search engines via the World Wide Web. I hereby give consent for the thesis to be incorporated into public access catalogues or services, such as national databases of electronic theses. Signed: (Candidate) P. Read Date: Wednesday 27th January, 2016 Certificate of supervising tutor in respect of the student s submitted work I am satisfied that this work is the result of the above-named student s own efforts. Signed: (Director of Studies) Rhodri S. Lloyd Date: Wednesday, 27th January,

3 Abstract Lower extremity non-contact injuries are common in male youth soccer players. Altered neuromuscular control defined as muscle strength, power or activation patterns that lead to increased joint loads has been suggested as a mechanism that underpins the occurrence of these injuries during rapid deceleration tasks. However, data pertaining to neuromuscular risk factors and screening in male youth soccer players is sparse. The purpose of this thesis was to investigate the validity of a novel movement screen to predict injury risk in elite male youth soccer players. Study 1 examined the test re-test reliability of a range of field-based neuromuscular control tests reporting acceptable values for measures of single leg dynamic balance, landing force, maximal hop distance and tuck jump knee kinematics; however, other commonly used assessments were more variable. The effects of chronological age on the measures deemed reliable in study 1 were then analysed in study 2. A number of between group differences were evident but this pattern was variable across the different constructs of neuromuscular control, thus an age-specific training emphasis may be required at different stages of a young player s development. Also, the normative data included for a range of chronological age groups in this study may be useful for practitioners, from which fluctuations in performances can be identified. The results of study 3 showed acceptable within subject variation on the majority of the tests measured at three time points across a soccer season. However, percentage change scores in neuromuscular control were more variable and differences between test sessions were often considerably lower than the random variation, thus observed changes may not be meaningful. However, single leg countermovement jump forces increased considerably throughout the season indicating that a real change occurred and this may be associated with greater injury risk. Seasonal variation in injury occurrence was also identified in study 4, but a major finding was a three-fold 3

4 increase in player incidence rate since the introduction of an early soccer specialisation model in the United Kingdom. This indicates that elite male youth players are now at a greater risk of injury. The final study examined the ability of a novel movement screen to predict noncontact lower extremity injury risk in male youth soccer players. A combination of anthropometric and neuromuscular risk factors were shown as predictors, but there was variability across the different chronological age groups. Single leg countermovement jump landing force asymmetry was the most frequently reported risk factor and univariate analysis also identified a number of significant predictors in respective chronological age groups. The results of this study provide an evidenced-based diagnostic assessment tool from which at risk players could be identified; the injury prevention asymmetry soccer screen (i-pass). However, greater weightings of specific assessments may be required at different stages of a child s development due to variability across age groups. This thesis has made an original and significant contribution to the existing paediatric injury risk screening literature for soccer players. Furthermore, these findings can easily be applied by practitioners to more accurately screen their players and develop targeted prevention strategies to reduce injury risk. 4

5 Journal publications and conference presentations from this thesis Peer-reviewed journal articles 1. Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. Injury risk factors in male youth soccer players. Strength Cond J, 37(5): 1-7, Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. Assessment of injury risk factors in male youth soccer players. Strength Cond J, (In press) 3. Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. Reliability of the tuck jump injury risk screening assessment in elite male youth soccer players. J Strength Cond Res, (In Press). 4. Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. Neuromuscular risk factors for knee and ankle injuries in male youth soccer players. Sports Med. (In Press). 5. Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. The Elite Player Performance Plan: Injury Risks Associated with this Modern Day example of early sport specialisation. J Sports Sci. (Under Review). 6. Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. Consistency of field-based neuromuscular screening tests using force plate diagnostics in elite male youth soccer players. J Strength Cond Res. (Under Review) Oral Presentations 1. Read, PJ. Identifying developmental players at risk of injury and methods of preventing injury. Proceedings of the Learn2Perform Sport Science Paediatric Development Conference, Stoke City Football Stadium, England. May Read, PJ. Injury Risk Factors and Screening in elite male youth soccer players. Proceedings of the ACPESM Young Athlete Conference, Brighton, England. Oct

6 Poster Presentations 1. Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. Reliability of the tuck jump injury risk screening assessment in elite male youth soccer players. Proceedings of the United Kingdom Strength and Conditioning Association (UKSCA) National Conference. Warwick, England, September

7 Acknowledgements The production of this thesis has been an extremely challenging and at times humbling process but at the same time also hugely rewarding. I owe an enormous debt of gratitude to a number of people, who have helped me through this journey, I must first thank my academic supervisors who have been pivotal to my development. As my director of studies, Dr. Rhodri Lloyd has been so dedicated in his role, offering continued guidance and support, for which I will be forever thankful. I am very lucky to have such a great supervisor, mentor and friend and I am extremely grateful for everything you have done and continue to do for me. I must also thank my additional supervisors and advisors, Dr. Jon Oliver, Prof. Mark De Ste Croix and Dr. Greg Myer. I could not have asked for a better team, and they have all been so generous and willing with their time. I believe the combined expertise of my supervisors make them undoubtedly the world leaders in paediatric exercise science. I truly hope to continue working with you all for many years to come. I would also like to thank my family for their continued support throughout my life. They are always there when I need them and I hope you know how much that means to me. In particular my parents, who have instilled in me a set of core values that help me strive to work hard and constantly improve. Finally, I would like to thank Pixie and Phillippa. For Pixie, she has been a new ignition to my life and spending time with her has made even the toughest days worthwhile. Phillippa, it has not always been easy, but I appreciate the sacrifices you have made and I am so grateful to you for standing by me through this journey. I will never forget your love and support and I hope now I can repay you as we move into the next chapter. 7

8 Table of contents DECLERATION STATEMENT 2 ABSTRACT 3 PUBLICATIONS AND PRESENTATIONS FROM THE THESIS 5 ACKNOWLEDGEMENTS 7 TABLE OF CONTENTS 8 LIST OF TABLES 14 LIFT OF FIGURES 16 CHAPTER 1 PREFACE INTRODUCTION AIMS AND OBJECTIVES OF THE THESIS THESIS ORGANISATION CHAPTER 2 LITERATURE REVIEW 2.1 INJURY OCCURRENCE IN ELITE MALE YOUTH SOCCER PLAYER Injury incidence in elite male youth soccer players Injury risk in different chronological age groups Location and type of injuries Injury mechanisms Severity Seasonal variation Summary of injury occurrence in male youth soccer players INJURY RISK FACTORS IN MALE YOUTH SOCCER PLAYERS Operational terms 36 8

9 2.2.2 The role of growth ad maturation Movement skill Fatigue Previous injury Summary of injury risk factors in male youth soccer players ASSESSMENT OF INJURY RISK FACTORS Monitoring growth and maturation Assessment of movement skill Fatigue monitoring - methods for assessing training load Monitoring neuromuscular fatigue to determine state of recovery and readiness Previous injury - assessment and identification Developing a return to play criteria Summary of assessing injury risk factors in male youth soccer players NEUROMUSCULAR RISK FACTORS FOR LOWER EXTREMITY INJURY IN MALE YOUTH SOCCER PLAYERS Quadriceps dominance Leg dominance Ligament dominance Neuromuscular feed-forward strategies Dynamic balance Trunk dominance Injury risk factor hierarchical model Summary of neuromuscular risk factors in male youth soccer players

10 2.5 FIELD-BASED ASSESSMENTS OF NEUROMUSCULAR CONTROL IN MALE YOUTH SOCCER PLAYERS Assessments of quadriceps dominance Hand held dynamometry Isometric hamstring tests Field-based hamstring strength tests Assessment of leg dominance (asymmetry) Assessment of ligament dominance (knee valgus) Clinic based landing tool assessment Landing Error Scoring System (L.E.S.S) Box height considerations for drop vertical jump tasks Type of jumping tasks Tuck jump assessment Assessment of trunk dominance Assessments of dynamic stability Time to stabilisation Star excursion or y-balance test Summary of field-based assessments of neuromuscular control in male youth soccer LITERATURE REVIEW SUMMARY 110 CHAPTER 3 - STUDY 1 THE RELIABILITY OF FIELD-BASED MEASURES OF NEUROMUSCULAR CONTROL IN ELITE MALE YOUTH SOCCER PLAYERS INTRODUCTION METHODS

11 3.2.1 Participants Experimental design Procedures Statistical analysis RESULTS DISCUSSION SUMMARY AND PRACTICAL APPLICATIONS. 144 CHAPTER 4 - STUDY 2 THE EFFECTS OF CHRONOLOGICAL AGE ON FIELD-BASED MEASURES OF NEUROMUSCULAR CONTROL IN ELITE MALE YOUTH SOCCER PLAYERS INTRODUCTION METHODS Participants Sample size estimation Experimental design Procedures Statistical analysis RESULTS DISCUSSION SUMMARY AND PRACTICAL APPLICATIONS. 183 CHAPTER 5 - STUDY 3 SEASONAL VARIATION IN FIELD BASED MEASURES OF NEUROMUSCULAR CONTROL IN ELITE MALE YOUTH SOCCER PLAYERS INTRODUCTION

12 5.2 METHODS Participants Experimental design Procedures Statistical analysis RESULTS DISCUSSION SUMMARY AND PRACTICAL APPLICATIONS. 203 CHAPTER 6 - STUDY 4 INJURY OCCURRENCE IN ELITE ENGLISH MALE YOUTH SOCCER PLAYERS INTRODUCTION METHODS Experimental design Procedures Statistical analysis RESULTS DISCUSSION SUMMARY AND PRACTICAL APPLICATIONS CHAPTER 7 - STUDY 5 THE ASSESSMENT OF LOWER EXTRMEITY INJURY RISK FACTORS IN ELITE MALE YOUTH SOCCER PLAYERS INTRODUCTION METHODS Participants

13 7.2.2 Experimental design Procedures Statistical analysis RESULTS DISCUSSION SUMMARY AND PRACTICAL APPLICATIONS. 256 CHAPTER 8 SUMMARY AND PRACTICAL APPLICATIONS, LIMITATIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH SUMMARY AND PRACTICAL APPLICATIONS LIMITATIONS OF THE RESEARCH DIRECTIONS FOR FUTURE RESEARCH. 268 CHAPTER 9 REFERENCES 272 APPENDICES 296 APPENDIX A - EXAMPLE FORMS FOR ETHICS APPROVAL PARENTAL INFORMATION, PARENTAL CONSENT, PARTICIPANT INFORMATION, PARTICIPANT ASSENT AND PHYSICAL ACTIVITY READINESS QUESTIONNAIRE APPENDIX B STUDY 2 CROSS SECTIONAL ANALYSIS BETWEEN GROUP STATISTICS. 306 APPENDIX C - PUBLISHED RESEARCH ARTICLES EMANATING FROM THE THESIS 316 APPENDIX D STANDARDISED PROTOCOL INSTRUCTIONS FOR TESTERS

14 List of Tables Table 2.1 Suggested test battery to evaluate the movement skill of a male youth soccer player. 51 Table 2.2 Borg Scale and session rate of perceived exertion (RPE) scale. 55 Table 2.3 Schematic evaluation of the load associated with a training programme in an elite male youth soccer player. 56 Table 2.4 Sample perceptual monitoring questionnaire.. 58 Table 2.5 Sample prospective injury reporting questionnaire. 60 Table 2.6 Return to play criteria physical performance checklist Table 2.7 Assessments of quadriceps dominance in male youth athletes Table 2.8 Assessments of leg dominance in male youth athletes 89 Table 2.9 Assessments of knee valgus in male youth athletes 100 Table 3.1 Field-based neuromuscular control test descriptive statistics and reliability values for pre peak height velocity players. Table 3.2 Field-based neuromuscular control test descriptive statistics and reliability values for pre peak height velocity players. Table 3.3 Asymmetry descriptive statistics and reliability values for pre peak Height velocity players Table 3.4 Asymmetry descriptive statistics and reliability values for post peak height velocity players. Table 3.5 Kappa ratings of agreement for tuck jump plyometric technique errors Table 4.1 Cross sectional study participant characteristics. 153 Table 4.2 Tuck jump mode scores per chronological age group. 167 Table 4.3 Tuck jump frequency scores for each grade classification per chronological age group... Table 5.1 Descriptive statistics and percentage changes in test scores across the season Table 5.2 Seasonal reliability statistics for all test variables

15 Table 6.1 Number of injuries, player incidence ratio and mean time loss per age group 210 Table 6.2 Anatomical location of injuries sustained Table 6.3 Anatomical location of injuries per age group 211 Table 6.4 Overall types of injury sustained. 213 Table 6.5 Types of injury sustained per age group. 213 Table 7.1 Mean (s) values for participant details for all players combined and sub-chronological groups. 229 Table 7.2 Anthropometrics of all injured and non-injured players. 234 Table 7.3 Descriptive statistics and univariate odds ratios for all injured and noninjured players Table 7.4 Multivariate analysis of risk factors for all players. 236 Table 7.5 Anthropometrics of U11 & U12 injured and non-injured players Table 7.6 Descriptive statistics and univariate odds ratios for injured and noninjured U11 & U12 players Table 7.7 Multivariate analysis of risk factors for U11 & U12 age groups 239 Table 7.8 Anthropometrics of U13 & U14 injured and non-injured players. 240 Table 7.9 Descriptive statistics and univariate odds ratio s for injured and noninjured U13 & U14 players Table 7.10 Multivariate analysis of risk factors for U13 & U14 age groups Table 7.11 Anthropometrics of U15 & U16 injured and non-injured players 242 Table 7.12 Descriptive statistics and univariate odds ratio s for injured and noninjured U15 & U16 players Table 7.13 Multivariate analysis of risk factors for U15 & U16 age groups. 244 Table 7.14 Anthropometrics of U18 injured and non-injured players 245 Table 7.15 Descriptive statistics and univariate odds ratio s for injured and non-injured U18 players Table 7.16 Multivariate analysis of risk factors for U18 age group

16 List of Figures Figure 2.1 Sample Child Friendly Retrospective Injury Reporting Questionnaire. 60 Figure 2.2 Knee and ankle ligament injury neuromuscular risk factor hierarchical model 77 Figure 2.3 Clinic based nomogram to predict high knee abduction loads.. 91 Figure 2.4 Landing Error Scoring System (L.E.S.S) real-time scoring criteria.. 93 Figure 2.5 Tuck jump assessment plyometric technique error grading criteria.. 98 Figure 2.6 Third-order polynomial time to stabilisation range variation example. 104 Figure 2.7 Vertical time to stabilisation example based. 105 Figure 4.1 Absolute y-balance scores per chronological age group Figure 4.2 Relative y-balance scores per chronological age group 159 Figure 4.3 Absolute single leg hop for distance scores per chronological age group. Figure 4.4 Relative single leg hop for distance scores per chronological age group.. Figure 4.5 Absolute 75% hop peak landing scores per chronological age group.. Figure 4.6 Absolute single leg countermovement jump peak landing scores per chronological age group Figure 4.7 Relative 75% hop peak landing scores per chronological age group 165 Figure 4.8 Relative single leg countermovement jump peak landing scores per chronological age group. 166 Figure 4.9 Asymmetry scores for all tests per chronological age group 169 Figure 6.1 Anatomical location of muscle strain injuries Figure 6.2 Growth overuse injuries per chronological age group Figure 6.3 Player incidence rate for severe injuries in each chronological age group Figure 6.4 Seasonal variation of injuries sustained

17 Figure 8.1 A novel field-based movement screen to assess neuromuscular control, the i-pass

18 Chapter 1 Preface 1.1 Introduction The sport of soccer imposes high physiological demands with repeated maximal effort running bouts and rapid high intensity movements such as jumping, cutting and contact with other players (Bangsbo, 1994; Daniel et al., 1994). Due to the nature of the sport, injury risk is an accepted aspect of participation where these high intensity actions predispose players to positions in which suboptimal movement mechanics and contact with opposition are common. The predominance of injuries that occur are to the lower extremity in male youth soccer and a high proportion of these injuries are non-contact in nature (Le Gall et al., 2006; Price et al., 2004; Rumpf and Cronin, 2012). Despite the commonality and anticipated risk of injury for youth players, it is reasonable to suggest that the degree of injury risk could be reduced with appropriate screening techniques and targeted prevention strategies. Injury risk is multifactorial (Meeuwise, 1994), and with youth players frequently exposed to high levels of competition and training loads, practitioners must be cognisant of a number of risk factors. These include growth and maturation, movement skill, fatigue, and previous injury (Read et al., 2015). While developing a clear understanding of such factors and how they affect relative risk of injury is essential, definitive evidence confirming that anatomical and genetic risk factors increase injury rates remains sparse (Griffin et al., 2000). Altered neuromuscular control has been suggested as a mechanism that underpins the occurrence of noncontact lower extremity injuries that occur during running, cutting, landing and side stepping movements (Hewett et al., 2004). While growth and maturation and previous 18

19 injury are unaffected by targeted interventions, previous data show that a key construct of movement skill (neuromuscular control) may be highly modifiable (Williams et al., 2001). Neuromuscular control has been defined as activation of the dynamic restraints which occur in preparation for, and in response to, joint movements and forces which provides functional joint stability during dynamic tasks (Riemann and Lephart, 2002). Sub-optimal neuromuscular control has been defined as muscle strength, power or activation patterns that lead to increased joint loads (Myer et al., 2004). Within the available literature, specific neuromuscular imbalances have been identified that proliferate the risk of ligamentous injury including: ligament dominance, indicating abnormalities in neuromuscular control when ligaments, rather than the lower extremity musculature provide the majority of the required dynamic stabilisation (Hewett et al., 2002); quadriceps dominance, defined as an imbalance between quadriceps and hamstring recruitment during rapid deceleration tasks; leg dominance, classified as an asymmetry of the contralateral lower extremity in joint kinematics, force production and / or landing force control (Myer et al., 2004); and trunk dominance, defined as an imbalance between the inertial demands of the trunk and the ability of the core musculature to control and resist it (Myer et al., 2011). Additional neuromuscular risk factors also include aberrant neuromuscular activation patterns (Kubo et al., 2007) and reduced dynamic stability (Ross and Guskiwicz, 2008). Further research is required to establish if these risk factors are present in elite male youth soccer players. In recent years, an emerging body of literature has focused on the assessment and trainability of different constructs of neuromuscular control (Myer et al., 2004; Hewett et al., 2005; Myer et al., 2006; Myer et al., 2008; Myer et al., 2011) to aid in 19

20 the identification of high risk athletes and subsequent development of targeted prevention strategies to reduce this risk. Neuromuscular injury risk factors have been proposed in female soccer players (Alernton-Geli et al., 2009) and more recently in adult male soccer players (Alernton-Geli et al., 2014). However, data pertaining to neuromuscular risk factors in male youth players remains sparse. Importantly, the validity and reliability of various neuromuscular screening techniques has yet to be examined in elite male youth soccer players (Read et al., in press; Read et al., in press). Despite the prophylactic benefits of neuromuscular training as measured through relative risk reduction reports, the development of a valid and reliable screening system is needed to identify players at high risk, thus allowing more targeted preventative strategies (Sugimoto et al., 2014). This is especially important for male youth soccer players who have previously been identified as a target group for injury prevention (Schmikli et al., 2011). This is confounded by recent concerns following the implementation of an early sport specialisation model in professional soccer clubs within the United Kingdom (Read et al., in press c). The potential impact of significant increases in training volume for young male soccer players that are experiencing a range of growth and maturational processes is currently unknown and requires further investigation to determine pertinent injury risk factors (Read et al., in press c). Traditionally, assessments of neuromuscular control have often required the use of specialised and expensive equipment, such as isokinetic dynamometers, electromyography and three-dimensional video analysis. These approaches are not cost-effective and involve labour intensive data collection. This limits the ability to screen large numbers of athletes, as is the case in youth soccer academies where budgets and time available are often comparatively less than in full time professional 20

21 programmemes. Consequently, a range of field based methods have been developed to screen large numbers of athletes which are cost and time effective (Padua et al., 2009; Myer et al., 2008; Myer et al., 2011). These measures have been reported to correlate significantly with more sophisticated laboratory techniques and will assist in the early identification of high risk athletes and the development of targeted prehabilitation programmemes (Hewett et al., 2005; Myer et al., 2010; Myer et al., 2011). However, available literature to support their use in paediatric athletes, and in particular elite male youth soccer players, is sparse. Therefore, the development of a valid and reliable field-based screening battery to assess different constructs of neuromuscular control is required for elite male youth soccer players. 1.2 Aims and objectives of the thesis The current thesis aims to develop a valid and reliable battery of field-tests to screen elite male youth soccer players and identify the injury risk characteristics associated with the different developmental phases within a professional soccer academy. The use of field-based measures, which are time-efficient and cost-effective have been included to widen participation and increase the sample size, providing a more reflective account of the players that are represented in this cohort. The information derived from this thesis will be used to determine if specific areas of focus are required based on the associated mechanisms of injury at different stages of growth and maturation. This will allow the development of more targeted prevention strategies in comparison to the currently proposed and commercially available programmemes such as the F-Marc F.I.F.A 11+ (Impellizzeri et al., 2013). Furthermore, a key aim of the thesis is to ensure that the results will directly inform practice in the field, and that the methods proposed are practically viable within the 21

22 constraints of a soccer academy environment. The specific aims of the thesis are as follows: 1. Synthesise existing literature to determine the occurrence, type and severity of injury in elite male youth soccer players and report the current data to examine the effects of a recently implemented early soccer specialisation programmeme 2. Develop a valid and reliable battery of field-based tests to assess lower extremity neuromuscular control in elite male youth soccer players 3. Determine the effects of chronological age on a range of lower extremity fieldbased neuromuscular control assessments in this cohort 4. Examine the seasonal variation in field-based measures of neuromuscular control to determine if periods of heightened risk are present during an academy soccer season 5. Identify relationships between measures of neuromuscular control and injury risk in elite male youth soccer players 1.3 Thesis organization The main focus of this thesis is to further our understanding of field-based neuromuscular control screening modalities and to identify relevant risk factors for injury in elite male youth soccer players. In order to do this, the thesis has been structured to provide a clear narrative for the reader whereby the existing literature is first presented and the subsequent original investigations are designed to investigate specific areas for which there is currently a paucity of data, using a systematic approach. Chapter 2 provides a review of the existing literature and critical analysis of a range of topics pertinent to the thesis including: injury prevalence, injury risk factors, and assessment 22

23 considerations for the use of field-based techniques in this cohort. This section has been formatted as a series of independent reviews which include published material by the author as indicated by the journal titles included under each sub-heading. Chapters 3 7 then consist of the original investigations. Chapter 3 establishes the reliability of a range of field-based neuromuscular control assessments using a mixed sample of elite male youth players who were pre or post peak height velocity (PHV). Using this information, Chapter 4 used tests that were deemed reliable to examine the effects of chronological age on different constructs of neuromuscular control in male academy players aged years of age. To investigate seasonal changes in neuromuscular control and the stability of these measures across a soccer season, Chapter 5 used a repeated measures design to track an academy soccer club through a competitive season with test sessions conducted at the pre, mid and end of season stages. Chapter 6 was included to track injury prevalence across a soccer season in order to establish the most frequently occurring injuries, player incidence ratios and severity in this cohort. Finally, Chapter 7 examined the relationships between performance during the reliable neuromuscular control screening battery and non-contact lower extremity injury risk in elite male youth soccer players to elucidate risk factors for injury. Chapter 8 provides a summary of the main research findings from this thesis and associated implications for practitioners. In addition, limitations are identified and future directions for continued research are proposed which are based on a critical interpretation of the data collected and presented throughout the thesis. Chapter 9 provides an overall reference list. At the end of the thesis, a number of appendices are provided, which include: ethics information; sample participant information forms; and publications from this thesis. All chapters have been presented in accordance with the guidelines set out by Cardiff Metropolitan University. 23

24 Chapter 2 LITERATURE REVIEW 2.1 Injury Incidence in Elite Male Youth Soccer Players A range of studies have reported the injury incidence data of amateur male youth soccer players (Schmikili et al., 2011; Timpka et al., 2007; Brito et al., 2012). The results of these studies have shown that the majority of injuries occur in the lower limb, predominantly at the ankle, knee and thigh, and the number of injuries sustained generally increases with age. Overall injury incidence has been reported to range from 1.2 to 6.8 injuries per 1,000 hours of exposure (Schmikili et al., 2011; Timpka et al., 2007; Brito et al., 2012). These values could be considered relatively low in comparison to those of adult players and are likely due to the lower levels of competition (Hagguland et al., 2009; Ekstrand et al., 2011). It has been shown that as the playing level increases so does the injury incidence (Timpka et al., 2007). Thus, caution should be applied when extrapolating these findings to elite youth players, in which playing exposure and speed of play is increased. A recent review of the available literature by Rumpf and Cronin (2012) examined the injury incidence in youth soccer. The authors proposed a number of key findings: 1) the frequency of injury increases with age, especially after age 14; 2) injuries occur predominantly in the lower extremities (80%), mainly at the ankle and knee; 3) more injuries occur during match play than training; 4) defence and midfield are the positions with the highest injury rate; and 5) higher injury incidence occurs in females. Furthermore, incidence data was reported per thousand hours of exposure by sub-division respective chronological age groups. The results showed 8/1,000h in the 9-12 year olds, 65.8/1,000h for year 24

25 olds and 8.4/1,000h in the year olds. This clearly indicates that those children aged participating in organised football are at a significantly greater risk of injury. This period coincides with the onset of peak height velocity (PHV), defined as the maximal accelerated growth spurt (Malina et al., 2004a). However, caution should be applied when interpreting these findings as in the review by Rumpf and Cronin (2012), both sexes were included in the data set, and methodological differences were present between studies. Specifically, a variety of performance levels were included and in some studies injury classification was not clearly defined. Subsequently, the data reported per thousand hours of exposure may not provide an accurate representation of the injury incidence in elite male youth soccer players Injury incidence in elite male youth soccer players A growing body of data are available to examine the injury incidence in elite male youth soccer players (Junge et al., 2000; Volpi et al., 2003; Price et al., 2004; Le Gall et al., 2006; Cloke et al., 2009; Brink et al., 2010; Cloke et al., 2011; Moore et al., 2011). However, a number of these studies have restricted their analysis to single injuries (i.e. knee or ankle) (Cloke et al., 2009; Cloke et al., 2011; Moore et al., 2011) or severe injuries only (Volpi et al., 2003). Overall injury incidence has ranged from injuries per 1,000h (Junge et al., 2000; Le Gall et al., 2006; Brink et al., 2010). The largest study in this population used a prospective cohort design to audit the frequency, type and severity of injuries that occurred over a two year period in 4,773 elite academy soccer players aged 9 to 19 years of age in the United Kingdom (Price et al., 2004). During the study period, 3,805 injuries occurred which equated to a player incidence rate of 0.40 injuries per player, per season, and a mean absence of 21.9 days missed per injury. In French academy players, the overall injury incidence rate has been reported as 4.8/1,000h, although some variation was evident across age groups (4.9, 25

26 4.6 and 5.2/1,000h) in the U14s, U15s and U16s respectively (Le Gall et al., 2006). A mean injury rate of 2.2 injuries per player, per season was also shown for all age groups. Disparity between the number of injuries sustained during training and competitions has also been reported, with significantly higher incidence in matches (11.2 vs. 3.9/1000h respectively) (Le Gall et al., 2006). This discrepancy was further magnified in the findings of Brink et al. (2010), who reported 26.65/1000h time-loss injuries in competition, with much fewer occurring in training (6.74/1000h). Conversely, Price et al. (2004) reported a more even distribution with 50.4% and 48.7% of injuries sustained during competition and training respectively, with no activity specified to account for the remaining 0.9% of injuries (Price et al., 2004). The timing of injuries that occur during match play should also be considered, with a higher frequency towards the end of each half (41 and 50% respectively) (Price et al., 2004). Specifically, the greatest percentage of injuries occurred in the final 15 minutes of the game and this could be as a result of fatigue, resulting in altered patterns of neuromuscular control (Padua et al. 2006). However, limited data are available in youth populations to examine the effects of fatigue on neuromuscular control, thus further investigation is warranted Injury risk in different chronological age groups Few studies have reported injury incidence across a wide range of chronological age groups, which is reflective of the organizational structure within a professional soccer academy. Prospective analysis has shown a trend of a linear increase in injury with age (Price et al., 2004), a non-significant reduction at U16, followed by a subsequent increase when players commence full time academy scholarship programmemes after this point (Price et al., 2004). 26

27 This is likely due to increased exposure levels, heightened intensities of play and the greater physicality of players in the later stages of adolescence and young adulthood. Recent data has shown periods of heightened risk around the time of peak height velocity (PHV) (Rumpf and Cronin, 2012; van der Sluis et al., 2014; van der Sluis et al., 2015). Specifically, in elite male youth soccer players, the frequency of traumatic injuries was significantly greater during the year of PHV (van der Sluis et al., 2014). The number of days absent from training and competition was also highest during this period but these differences were not statistically significant. The authors stated that heightened injury risk around this period followed by a reduction post-phv, identifies this as a time of increased vulnerability. Players in this age group could be considered at greater risk of injury due to rapid limb length changes effecting force dissipation (Hewett et al., 2005b; Hewett et al., 2004) and changes in biomechanical characteristics (Shea et al., 2004). While the number of traumatic injuries appears to reduce post-phv, a trend depicting a linear increase in overuse injuries through maturation was also shown (van der Sluis et al., 2014). This could be attributed to developmental growth lags, whereby, a delay between growth in muscle length and cross-sectional area has been reported (Xu et al., 2009). Changes in muscle and tendon length in the absence of hypertrophy, which can occur sometime after the peak growth spurt, requires the muscles to function at a greater percentage of their maximum capacity to normalise the increases in limb length and subsequent moment of inertia (Hawkins and Metheny, 2001). This results in larger forces and strain on the tendinous structures, which can lead to injuries such as patella and/or achilles tendinopathy. 27

28 2.1.3 Location and type of injuries Injuries occur mainly in the lower extremities (71-80%), predominantly to the upper thigh, knee and ankle (Price et al., 2004; Le Gall et al., 2006). The preponderance of thigh injuries have been reported as strains to the quadriceps (43%) and hamstrings (57%) (Price et al., 2004), while the highest proportion of injuries to the knee were ligament injuries (28%) largely occurring in the medial collateral ligament (85%) (Price et al., 2004; Moore et al., 2011). Ankle injuries were mostly ligament sprains (72%), predominantly to the anterior talofibular ligament (Price et al., 2004; Cloke et al., 2009). A high proportion of ligament sprains to the knee and ankle have also been reported in other studies using elite male youth soccer players, with this injury type the second most frequently occurring injury behind muscle contusions and hematomas (Le Gall et al., 2006). An eight year audit of the ankle injuries sustained in male youth soccer academies in the United Kingdom reported an incidence rate of 1.0 ankle injuries per player, per year with an average time loss per injury of 20.4 days (Cloke et al., 2009). A linear increase in the number of injuries was shown from the U9s to the U16s, with a high frequency of anterior talofibular ligament sprains far exceeding the next highest recorded injury (Cloke et al., 2009). The incidence of knee injuries has also been reported, with a rate of 0.71 knee injuries per player per year, equating to 17 training days absent and 2 matches missed per knee injury (Moore et al., 2011). The U14 to U16 chronological age groups appear to be at a greater risk of knee injury, and the incidence was highest in the U14s (Moore et al., 2011) which may be due to the aforementioned period of rapid growth. However, no data were provided for players in older age groups. The high proportion of medial collateral ligament injuries should also be considered in the assessment and training of elite male youth soccer players. Two primary injury mechanisms have been identified for this structure, specifically a contact mechanism in which a forceful valgus load is applied to the knee joint and secondly, a non- 28

29 contact mechanism with the knee positioned in valgus during cutting and deceleration movements (Wijdicks et al., 2009). Identification of players who demonstrate these latter high-risk movement patterns using appropriate screening protocols may aid in the prevention of such injuries (Read et al., in press). Based on the available data, growth related conditions appear to account for a much lower proportion of injuries in this cohort (Price et al., 2004; Le Gall et al., 2006). Osgood- Schlatter s disease is consistently the most prevalent injury in this category, followed by Sever s disease (Price et al., 2004; Le Gall et al., 2006). Peak incidence of Osgood Schlatter disease has been observed between the ages of 12 and 14 years (Moore et al., 2011), while other investigations have shown the highest numbers of these injuries are sustained in the U13s and U14s (Price et al., 2004). The greatest prevalence of Sever s disease has been reported to occur in the U11s (Price et al., 2004). These time points (particularly the U13s and U14s) correspond with the onset of the pubertal growth spurt and a period of maximal accelerated growth (Malina, 2004a). Thus, regular monitoring of growth rates in male youth soccer players is suggested to identify players who may be at a greater risk of injury during these sensitive periods, with adjustments made to prescribed training loads and avoidance of repetitive type activities Injury mechanisms Within the current literature, a clear delineation of the number of contact versus non-contact injuries has not been reported; however, available research does indicate a high frequency of non-contact injuries in this population (Price et al., 2004; Cloke et al., 2009; Moore et al., 2011). Data showing the mechanism of specific injuries has been reported in elite male youth players, with non-contact mechanisms implicated in 55% of all knee injuries sustained 29

30 (Moore et al., 2011). A more even distribution has been shown for ankle injuries, with just over half of all ankle injuries associated with a contact mechanism (51.7%) (Cloke et al., 2009). Price et al. (2004) showed that the highest percentage of non-contact injuries occurred while running (19%), although the stage of the gait cycle, or a description if these were experienced during accelerations or decelerations was not reported. Other non-contact injuries were sustained during twisting/turning (7%), over-stretching (4%), and landing movements (Price et al., 2004). In the analysis of ankle injuries, non-contact mechanisms included running (10.6%), twisting/turning (10%) and landing (7.7%). A further description of the specific mechanisms of injuries occurring during training and competitions that lead to injury is challenging due to the paucity of data. To the knowledge of the author, only one study has reported this information (Price et al., 2004). A similar trend was shown between the mechanisms of injury in these two environments; however, a higher frequency of injuries were due to tackling and running in competition and training respectively. Cumulatively, the data indicates that the mechanisms of injury are similar in training and competition. Also, non-contact injuries occur during high speed movements and those that involve rapid decelerations, indicating that screening approaches to identify risk should consider the assessment of neuromuscular control using movements that produce high speeds and ground reaction forces Severity Moderate injuries (classified as 1-4 weeks absent from training and competition) have been reported as the most frequent (44%) injury classification in elite male youth soccer players in the United Kingdom (UK), with severe injuries (> 4 weeks absence) accounting for nearly a quarter (22%) of all injuries (Price et al., 2004). Conversely, Le Gall et al. (2006) showed a 30

31 near equal distribution (approx. 30%) of slight, minor and moderate injuries in all age groups (U14-U16) and less than 10% severe injuries in French academies (Le Gall et al., 2006). Also, Le Gall et al. (2006) only provided data for players in the U14-U16 age groups, whereas, Price et al., (2004) including players throughout each academy chronological age group (U9-U19). Based on the available data, it could be indicated from these studies that the risk of serious injury is greater for youth male players participating at the elite level in the UK. While purely speculative, this may be due to heightened intensities of play and/or differences in training approaches that exist in the respective countries where the players are registered. Sites of severe injury vary based on chronological age but most commonly include the knee, followed by the ankle (Volpi et al., 2003). Osteochondrosis and skeletal fractures also occur regularly with muscle lesions and tendinopathies experienced less frequently (Volpi et al., 2003). The most frequent severe non-contact knee injuries include ligament sprains (23%) and Osgood Schlatter s (21%), followed by meniscal tears (14%) (Moore et al., 2011). Severe ankle injuries commonly include fractures, dislocations and ligament ruptures; however no incidence data was reported to confirm their incidence in this cohort (Cloke et al., 2009). Currently, it is not possible to differentiate the type and frequency of severe injuries that occur between age groups. For example, Volpi et al. (2003) classified participants from their study as either over or under 15 years of age, which does not permit accurate interpretation of the variations in growth and maturation that may be present in different chronological age groups. Thus, further research is required to determine injury severity and types across a range of age groups in elite male youth soccer players. 31

32 2.1.6 Seasonal variation In addition to recording the total incidence of injury, it may also be useful to examine the seasonal variation in incidence to elucidate if injury risk is greater at certain time points during the competitive season. Price et al. (2004) reported two peaks in training injuries (August and January), which coincide with the end of pre-season (August) and the return to training after the mid-winter break (January) in soccer academies in the United Kingdom. A similar pattern was observed in other studies that focused on injuries to the ankle (Cloke et al., 2009) and knee (Moore et al., 2011). Price et al. (2004) reported that the incidence of injuries sustained during competition revealed a linear increase from July, with a peak in October (p <.05), followed by a decline leading up to the Christmas break. In elite academy French male youth players, the highest number of injuries for all age groups (U14-U16) were reported in September (7.4/1000h). Conversely, the number of training injuries peaked in November, however when analysed per chronological age group, peak season incidence occurred earlier in the U14s vs. the older groups (September and November respectively) (Le Gall et al., 2006). In addition, most of the major competition and training injuries for all age groups occurred in August. Cumulatively, the high frequency of injuries towards the end of the preparation periods and early season could be linked to a lack of preparatory conditioning, inappropriate management of training volume, the accumulation of fatigue from high density of training, and/or environmental factors such as harder pitch surfaces. Spikes in incidence rates following the mid-winter break could also be as a result of deconditioning and rapid increases in training volume. In support of this, no significant differences in the frequency, severity, and type of injuries sustained have been shown immediately after vacation periods in French Academy players (Le Gall et al., 2006). Players attending the French National Academy receive shorter, but more frequent vacation breaks, which may provide greater overall periods of regeneration and avoid pronounced detraining effects (Le 32

33 Gall et al., 2006). Therefore, governing bodies may wish to adopt a similar approach to aid in the prevention of soccer-related injuries Summary of injury occurrence in male youth soccer players The available data examining the injury epidemiology in elite male youth soccer players shows an approximate equal distribution between contact and non-contact injuries. These injuries predominantly occur at the lower limb, with the thigh, knee and ankle the most frequently reported anatomical sites. Variation is evident between studies with regards to injury incidence but this may be due to different geographical locations and methods of reporting. However, it can be expected that on average, players will miss up to one-month of the competitive season through injury, with the resulting time loss much greater in the case of severe injuries. A general linear increase in injury rates has been observed with advances in chronological age; however, periods of heightened risk may be present during stages of rapid growth. Seasonal variation in injury incidence also appears to be evident, with a spike in injury rates following extensive preparation periods and the commencement of competitive periods. A secondary spike has also been observed following the mid-winter break which may be associated with rapid increases in training volume. When interpreting the available injury epidemiological data, it is important to note that each of these studies was conducted prior to the implementation of the Elite Player Performance Plan (EPPP). The inception of this scheme has resulted in a substantial increase in the volume of soccer specific training and requires a linear increase in exposure as players progress through each developmental phase. The impact of such a significant increase in training volume for these young athletes who are experiencing a range of growth and maturational processes is currently unknown. However, current concerns amongst researchers 33

34 and practitioners alike, suggests that research is urgently required to examine the potential increased risks of injury associated with the adoption of an early specialisation approach to the development of young soccer players (Read et al., in press c). 34

35 2.2 Injury Risk Factors in Male Youth Soccer Players Paul J. Read, Jon L. Oliver, Mark B. A. De Ste Croix, Gregory D. Myer, Rhodri S. Lloyd. Injury risk factors in male youth soccer players. Strength and Conditioning Journal 37(5), 1-7, For young athletes participating in high intensity sport, an inherent risk of sports-related injury exists, and this is heightened at various stages of growth and maturation (Rumpf and Cronin, 2012). Specifically, with an increase in a child s age, there is greater exposure to training and competition, which involves high levels of repetitive loading that can increase injury risk (Hogan and Gross, 2003). Further, a linear increase in injury rates has been reported from 9 to 15 years of age in male players (Price et al., 2004), with a marked increase around the age of 13 years (Emery, 2003; Rumpf and Cronin, 2012). Chronologically, these ages coincide with rapid changes in stature and mass as a result of maturational processes. During adolescence, males will experience peak height velocity (PHV) at around age 14, which refers to the time at maximal rate of growth during the adolescent growth spurt (Malina et al., 2004). Significantly, recent research shows that elite male youth soccer players experience more traumatic injuries in the year of peak height velocity (van der Sluis et al., 2014), which underlines the greater occurrence of sports injuries with later stages of maturation (Michaud et al., 2001). Recent trends have highlighted a range of injury risk factors and the importance of injury prevention strategies within female soccer players (Alternton-Geli et al., 2009). However, there is a paucity of information on male youth players. Due to the physical demands of soccer, the associated injury risk and the number of children and adolescents who participate in the sport, there is a clear need for increased research within male soccer players to identify age and sex specific injury risk factors (Alternton-Geli et al., 2014). Specifically, 35

36 practitioners working with youth male players must be cognizant of a range of modifiable and non-modifiable risk factors that are specific to paediatric populations which may increase injury risk. Hence, the focus of this review is to outline a range of considerations pertaining to male youth soccer players which may contribute to their relative risk of injury Operational terms For the purposes of this review, growth refers to quantifiable change in anthropometrics, body composition, body size, or the size of specific regions of the body (Beunen and Malina, 2005). Maturation refers to qualitative system changes, both structural and functional in nature in the organism s progress towards a mature state. The timing and tempo of maturation is variable among bodily systems (Beunen and Malina, 2005) and while growth and maturation are often used interchangeably, growth should be viewed as a constantly evolving process, whereas, maturation has a definitive end point (i.e. when an adolescent becomes fully mature. Childhood is the period of pre-pubescence and extends from the end of infancy to the start of adolescence (Malina et al., 2004). Adolescence is more difficult to define via chronological age due to large variation in maturation rates, but can be referred to as the period between childhood and adulthood (Malina et al., 2004) The Role of growth and maturation One factor which cannot be modified but should be monitored regularly as part of a screening approach is individual growth and maturation. This is highlighted by Michaud et al. (2001) who showed that children display an increase in sports related injury occurrence as they mature. Recent data show that there is a heightened risk of injury for youth male soccer players in the year of PHV (van der Sluis et al., 2014) and that with maturation, there is an increased risk of ligament sprain and a concomitant decrease in bone fractures which are 36

37 likely due to increased body mass, altered bony lever lengths influencing increased joint loads and greater intensities of play (Adrim and Cheng, 2003; Ford et al., 2010a; Ford et al., 2010b). Thus, an awareness of growth and maturational processes is essential for developing an understanding of changes in performance and alterations in motor control at various stages throughout childhood and adolescence (Lloyd et al., 2012; Malina et al., 2004). Variability in the growth and development of various physiological systems can be considered a key risk factor for injury, specifically around periods of accelerated growth. For example, with rapid growth in skeletal structures, the muscular system must simultaneously develop both in length (to normalize tension from bone growth), and also in size, so that greater levels of force production are possible to support and move the larger and heavier skeleton (Williams et al., 2012). However, it should be noted that the preceding growth in skeletal structures provides a stimulus for morphological adaptation of muscle tissue, thus an inherent time lag is present between the rate of bone growth and subsequent muscle lengthening. This has connotations for the incidence of traction apophyseal injuries in youth athletes, particularly prevalent in soccer between the ages of 11-14, with peak incidence occurring in males for the under 13 and under 14 age groups (Price et al., 2004). The disproportionate growth rates of bone and the muscle tendon complex result in greater forces experienced by the involved tissues when they are in a relaxed state (previously referred to as tissue preload (Hawkins and Metheny, 2001)), and this has been suggested as a contributing factor in the occurrence of traction apophyseal injuries (Mountjoy et al., 2008). Further, a delay between growth in muscle length and cross-sectional area has also been reported (Xu et al., 2009). This development lag in cross-sectional area may result in altered neuromuscular control strategies making dynamic stabilisation more challenging (Hewett et al., 2004; Ford et al., 2010a; Ford et al., 2010b). A reason for this may be the subsequent change in lever 37

38 lengths, which result in a higher center of mass and concomitant increases in joint torques in order to attenuate forces in the absence of adequate hypertrophy and strength (Myer et al., 2008; Williams et al., 2012). Furthermore, it has been acknowledged that during the peak growth spurt, a differential growth rate exists between the legs and the trunk, whereby, the long bones (limbs) experience peak growth prior to the short bones (trunk) (Mirwald et al., 2002). Therefore, the presence of musculoskeletal growth lags following the onset of a growth spurt up to, and around the period of peak height velocity, need to be considered in male youth soccer players to ensure the risk of overuse and apophyseal injuries is reduced during these key periods of growth. As a consequence of rapid increases in limb length, young soccer players may experience temporary decrements in motor skill performance, which has commonly been referred to as a period of adolescent awkwardness (Philippaerts et al., 2006). While this period of awkwardness does not necessarily affect all youth, adolescents may experience disruption in motor control due to the continual growth of anatomical structures, disproportionate growth of skeletal and muscle tissue and changes in neuromuscular functioning (De Ste Croix and Deighan, 2012). An awareness of the adolescent awkwardness phenomenon is important as it has previously been suggested that acquired skills and movement patterns may need to be re-perfected during this period (Drabik, 1996). Cumulatively, literature indicates that periods of accelerated growth may be a key injury risk mechanism and therefore practitioners should adopt an appropriate system of monitoring growth in young athletes. 38

39 2.2.3 Movement skill Precise neural regulation of muscle optimizes human movement and the execution of finely tuned motor skills, in addition to increasing joint stability by dynamic restraint (defined as the role muscles play in joint stability) (Hewett et al., 2002). Frequent stimulation of neural pathways will subsequently enhance motor programmeming, preparatory muscle activity and reflexive neuromuscular responses, which will contribute to greater levels of dynamic joint stabilisation and skill (Chimera et al., 2004). Individuals who lack fundamental movement skill development during the prepubescent period may compromise dynamic stability as they enter puberty and adolescence (Beck and Wildermuth, 1985; Hewett et al., 2004). Therefore, targeted interventions to develop fundamental movement skills during the prepubertal years are deemed critical due to the accelerated periods of neural plasticity associated with prepubescence, resulting from the natural development of the neuromuscular system (Borms, 1986). This is further highlighted by suggestions that the optimal age for movement skill development is during the prepubertal period (Myer et al., 2011), with the ages of 7-11, determined as a sensitive period for sequential development of gross motor skill (Gallahue, 1982) and movement coordination (Hirtz and Starosta, 2002). The relationship between movement skill competency and injury is not consistent within the available literature. Harmon and Randall (1998) showed that skill level does not relate to Anterior Cruciate Ligament (ACL) injury risk across a period of seven years in male and female basketball and soccer players. Further, in female athletes, higher skill levels have been reported as a risk factor for lower extremity and back injury (Hopper et al., 1995). Conversely, in male soccer players, an association between skill level and injury has been reported with low skill players demonstrating a two-fold increased incidence of lower extremity injuries, particularly in the knee and ankle (Chomiak et al., 2000; Peterson et al., 39

40 2000). However, it should be considered that in the aforementioned research studies, their classification of skill was determined by sport playing level (i.e. division 1, 2 or 3) and not a reflection of an individual s ability to perform fundamental movement skills. According to Blume (1981), seven coordinative abilities provide the key components of skill, including: motor differentiation; motor connection; balance preservation; spatial orientation; motor rhythmitization; speed reaction; and motor transformation. Thus, it should be considered that the ability to compete at a high level in sport is an over simplistic classification of one s skill. This is highlighted by findings that in male subjects, altered movement patterns and deficits in neuromuscular control were able to predict injury (Gomes et al., 2008; Sheehan et al., 2012). Therefore, the ability to perform movements safely in desirable patterns associated with successful performance would be a more appropriate means of evaluation. Movements that lead to injury in male youth soccer include running, twisting, turning, over-stretching and landing (Price et al., 2004), with altered neuromuscular control during such actions a suggested mechanism (Alerton-Geli et al., 2009). Neuromuscular control has been defined as the activation of dynamic restraints which occur in preparation for, and in response to, joint movements and forces to provide functional joint stability (Riemann and Lephart, 2002). Deficits in neuromuscular control direct excessive stress to the passive ligamentous structures, exceeding their strength limit, and result in mechanical failure (Li et al., 1998). Specific imbalances which have been identified in female athletes include: quadriceps dominance; leg dominance or asymmetry; ligament dominance, and trunk dominance or core dysfunction, (Hewett et al., 2002; Myer et al., 2004; Myer et al., 2011). Further, reductions in fundamental movement skills of injured professional male athletes (Kiesel et al., 2007), lower scores on a single leg jump and balance assessment (Goosens et al., 2014), and greater leg asymmetry in dynamic balance tasks (Plisky et al., 2006) have been 40

41 identified in male and female students as positive predictors of injury. Therefore, it could be argued that the assessment of an athlete s neuromuscular control provides a better indication of their movement skill, and if a number of deficits are identified, a greater risk of injury is present. However, despite the growing body of evidence in females, there is a lack of available literature to confirm pertinent injury risk factors for male youth soccer players at different stages of growth and maturation, and thus requires further investigation Fatigue Heightened fatigue following an acute bout of exercise has been reported to increase known markers of injury risk, which may subsequently effect dynamic joint stabilisation (McLean et al., 2007; Padua et al., 2006; Small et al., 2010). In soccer, greater levels of fatigue have been reported to increase injury incidence in both adult male professionals (Ekstrand et al., 2011) and elite male youth players (Cloke et al., 2009; Price et al., 2004), with injuries occurring more frequently towards the end of the first and second half respectively (Price et al., 2004). It is suggested that this timeframe may be indicative of reduced neuromuscular function and control, as evidenced by recent data showing that electromechanical delay increases (De Ste Croix et al., in press) and feed-forward reflex activity decreases (Oliver et al., 2014) in females and males respectively following exposure to acute soccer-specific fatigue protocols. Furthermore, in a group of male youth soccer player s, fatigue induced changes have been reported during a drop jump task (Oliver et al., 2008). Specifically, the subjects increased landing forces, with a reduction in muscle activity of the tibialis anterior, knee flexors and extensors. Conversely, aemg of the soleus increased, suggesting that in a fatigued state, young athletes are less able to tolerate ground reaction forces, and due to lower muscle activation, experience greater skeletal loading. Also, youth players may utilise a more ankle dominant landing strategy with reductions in neuromuscular control around the knee. Finally, 41

42 the measurement of landing kinetics and kinematics in a fatigued state has identified both male and female subjects significantly increased peak proximal tibial anterior shear forces, increased valgus moments, and decreased knee flexion angles (Chappell et al., 2005). Combined, these fatigue-induced neuromuscular alterations lead to an overall reduction in dynamic stabilisation upon ground contact (Fuksuhi Yamada et al., 2012), thus placing the lower limb at increased risk of injury. Recent evidence shows that paediatric subjects respond to fatigue differently based on age and maturation (De Ste Croix, 2012). In this report submitted to U.E.F.A, data showed that pre, circa and post pubertal females experience changes in leg stiffness, electromechanical delay and functional quadriceps: hamstring ratio following an acute simulated soccer fatigue protocol. These changes differed according to age and maturation, with prepubertal and circa pubertal youth showing the greatest decrement in electromechanical delay or functional hamstring: quadriceps ratio respectively, which may negatively impact dynamic joint stabilisation. Thus, whilst these responses were measured in female subjects, recent evidence suggests in a fatigued state, altered patterns of neuromuscular control are evident in male youth soccer players (Oliver et al., 2008; Oliver et al., 2014), and therefore, it is reasonable to assume that differential responses based on age, growth and maturation are also likely in males. Consequently, practitioners should be aware of individualized responses to fatigue and consider implementing more targeted intervention strategies to enhance neuromuscular capacities in both rested and fatigued states. However, due to the paucity of current data available, specific responses to fatigue and the subsequent effects on movement mechanics in male youth soccer players requires investigation. 42

43 A further point of consideration is the effects of chronic fatigue on the level of afferent feedback. Specifically in males, measured responses to a simulated eccentric fatigue protocol have demonstrated that although muscle function was restored close to baseline post 96 hours, electromechanical delay was significantly greater for all of the reported contraction conditions (Howatson, 2010). The author proposed that this may be due to changes in postsynaptic events, specifically exciting-contraction coupling (Warren, 2001; Howatson, 2010). This has implications for monitoring neuromuscular readiness to re-perform following periods of heavily fatiguing exercise which include of a high number of eccentric muscle actions, such as the repeated decelerations and changes of direction occurring in soccer. Therefore, despite the aforementioned study utilizing adult male subjects, it is reasonable to assume that neuromuscular performance will be altered in youth males when preceded by fatiguing conditions. Subsequently, assessments of neuromuscular function performed in a pre-fatigued state may enhance ecological validity. However, there is currently a paucity of data to describe the baseline characteristics of male youth soccer players, and is thus a logical start point. Following this, the effects of acute fatigue on movement performance can be more accurately identified Previous injury Previous injury has been reported as a significant risk factor for future injury occurrence (Dvorak and Junge, 2000; Hagguland et al., 2006; Kucera et al., 2005). For example, male soccer players with a history of ankle or knee sprain were at a greater risk of re-injury to the same sites (odds ratio (OR) = 4.6 and 5.3 respectively) (Arnason et al., 2004). Furthermore, Ekstrand and Tropp (1990) followed 639 male soccer players over a period of one season, identifying that those who had previously experienced an ankle sprain were at a 2.3 times greater risk of injury. Evidence of this risk factor in youth players has also been reported, 43

44 with males and females (age range U12-U18) that experienced a previous injury presenting a two-fold greater risk of a secondary occurrence (Kucera et al., 2005). Also, there was a threefold risk for subjects who encountered two or more injuries suggesting that this is a pertinent risk factor for youth soccer players as the risk of re-injury seems to increase exponentially with the number of injuries occurred (Kucera et al., 2005). When interpreting the available research for previous injury as a risk factor for reinjury, practitioners should be cognizant of the fact that often the relationship between previous injury and subsequent re-injury is measured via a retrospective analysis, which relies on the individual s ability to recall their own injury history. Such methodologies may lead to recall bias which can occur in both long and short term retrospective reporting (Junge and Dvorak, 2000). Thus, a greater body of research is required in paediatric male populations that utilise prospectively recorded injury data and subsequent tracking of players over a longitudinal period. An example of this type of research design, albeit in adult populations, was used to determine previous injury as a risk factor in elite male soccer (Hagguland et al., 2006), confirming that players who experienced an injury in the first season were at a greater risk of sustaining any type of injury in the following season than previously non-injured players (hazard ratio 2.7; 95% confidence interval 1.7 to 4.3). Also, players who encountered a hamstring, groin, or knee joint injury were 2-3 times more likely to experience an identical injury in the following season. This highlights the lack of prospective screening and associated surveillance data and the need for detailed reporting of injury history during the initial athlete screening process, and subsequent monitoring of changes in movement patterns following the occurrence of an injury. Further, appropriate rehabilitation and training interventions are required to ensure that suitable neuromuscular 44

45 control is displayed, and athletes are able to demonstrate low risk movement patterns prior to returning to play. The mechanisms associated with a high level of injury recurrence are not clearly understood. However, it has been suggested that neuromuscular inhibition may lead to altered movement and stabilisation patterns (Opar et al., 2013). An example of this has been identified following injury to the ACL, whereby maximal voluntary quadriceps activation is significantly reduced following injury (Hurley, 1997). Furthermore, neuromuscular inhibition patterns effecting knee joint stabilisation have been reported following injury, with the hamstrings demonstrating greater deficits in eccentric rather than concentric strength (Crosier and Creelard, 2000; Lee et al., 2009). A key role of the hamstring musculature during landing and deceleration tasks is to counteract the anterior shear of the tibia relative to the femur, providing control and joint support through eccentric actions (Osternig, 2000). Altered activation patterns of the hamstrings will likely increase the risk of knee joint injuries due to a reduced ability to attenuate forces. Moreover, the gluteal musculature, which provide key synergistic actions to assist with knee joint stabilisation, subtaler joint positioning, and resultant center of mass control have demonstrated inhibition following the occurrence of an ankle injury (Bullock-Saxton et al., 1994; Friel et al., 2006). Thus, practitioners involved in the assessment and prevention of injury should be cognizant of altered movement patterns and muscular activation sequencing which may occur following an injury, and their relative effect on injury reoccurrence. Specifically in male youth athletes, there is a paucity of information confirming the mechanisms associated with injury reoccurrence, and therefore further investigation is warranted. 45

46 2.2.6 Summary of injury risk factors in male youth soccer players For individuals working with youth athletes, and in particular, male youth soccer players, a range of factors need to be considered to aid in the prevention of injury. Some of the key factors have been presented in the manuscript and are summarized below: 1. Injury risk may be higher in youth male soccer players, particularly during critical periods of accelerated growth and development 2. A clearer definition of the term skill is required to differentiate between technical sport proficiency and movement competency. In addition, a validated assessment to outline the movement proficiency of youth athletes is required 3. Individuals should also be aware that confounding factors including fatigue and previous injury heighten injury risk, highlighting the importance of appropriate screening protocols to identify alterations in neuromuscular control and the implementation of monitoring approaches to manage training volume 46

47 2.3 Assessment of Injury Risk Factors in Elite Male Youth Soccer Players Paul J. Read, Jon L. Oliver, Mark B. A. De Ste Croix, Gregory D. Myer, Rhodri S. Lloyd. Assessment of injury risk factors in male youth soccer players. Strength and Conditioning Journal. 38(1): 12-21, While an understanding of the risk factors outlined in the previous section is important, practitioners should also utilise suitable assessment and monitoring tools that can be used to systematically screen athletes for these risk factors. Therefore, this section will provide an evidenced based discussion of suitable assessment and monitoring strategies that are practically viable and effective in reducing injury risk Monitoring Growth and Maturation Youth male soccer players display a heightened risk of sport related injuries during the year of peak height velocity (van der Sluis et al., 2014); implementing methods to monitor the rate of growth and maturation is deemed necessary to identify this at risk population. Lloyd et al. (2014) offer an extensive commentary of the range of methods available to determine maturity status and their implications for programmeming. It is beyond the scope of this review to discuss in great detail; however, the key points are highlighted below: 1. Skeletal age assessment may be considered the preferable method but are costly and require specialised expertise; 2. Tanner staging only provides categorical data and could be considered an unnecessarily invasive method of assessment; 3. Longitudinal tracking of growth rates provide a simple approach which may be possible in the context of a soccer academy and could be considered the most important aspect to monitor for identifying growth spurts associated with periods of increased injury risk; 47

48 4. In the absence of longitudinal data, predictive equations using age and anthropometric variables can provide an estimate of maturity. The predictive equations of Mirwald et al. (2002) have recently been adapted to provide a simpler and more accurate equation, which only requires measures of age and standing height (Moore et al., in press). Thus, longitudinal tracking of growth rates and periodic estimations of somatic maturity (approx. every three months) allow practitioners to identify periods of accelerated growth indicative of increased injury risk in male youth soccer players. Also, the above measures should be considered alongside an assessment of training age and technical competency to ensure appropriate adjustments are made to individual training prescription throughout each stage of growth and maturation (Lloyd et al., 2014) Assessment of movement skill Within the available literature, a range of movement skill assessments have been proposed for children, in particular, the Movement-ABC 2 (Henderson et al., 2007), Bruininks-Oseretsky Test of Motor Proficiency (BOTMP-BOT-2) (Bruininks and Bruininks, 2005) and more recently the coordinative motor skills scale (CMSS) (Bardaglio et al., 2012) (see Cools et al. (2009) for a detailed review). It should be noted that these assessments have been primarily used with young children to identify developmental coordination disorders (Hartman et al., 2010; Hung and Pang, 2010). These methods may not be reflective of youth males engaged in competitive sport, in particular those involved in elite level youth soccer. Therefore, it would appear that the availability of valid and reliable assessments of movement skill for this population is somewhat limited. 48

49 An alternative movement assessment is the functional movement screen (FMS) (Cook et al., 2006). Utilised frequently in professional male soccer (McCall et al. in press), this seven station assessment includes a range of mobility and stability based tasks. This modality has shown acceptable validity and reliability (Onate et al., 2012) and was originally intended for determining a foundational movement baseline (Cook et al., 2006). However more recently, available literature has assessed the relationship between movement performance and injury identifying a total reporting score of <14 is a positive predictor of injury (Kiesel et al., 2007; Chorba et al., 2010; O Connor et al., 2011; Lisman et al., 2013). However, none of these studies assessed male youth soccer players. Also, recent reports have highlighted limitations in solely utilizing the total score due to low internal consistency in FMS sum scores (Kraus et al., 2014; Kazmen et al., 2014). Practitioners should also be cautious when interpreting scores on mobility based tests without analyzing the quality of the movement, as high scores may be achieved in the presence of hypermobility which may be a risk factor for certain injuries (Pacey et al., 2010). Both inter and intra-rater reliability has been established in youth populations (Wright et al., 2015), but to the knowledge of the authors no research has been conducted which a) validates FMS total score as an injury predictor or b) reports inter-session reliability. Therefore, practitioners should consider the above limitations when screening their young athletes. A final point of consideration for practitioners using the FMS (and those of similar screening protocols) is that the tests are primarily comprised of static tasks unlike the dynamic nature of soccer. This is confounded by reports that highlight poor relationships with dynamic movements of athletic performance (Parchmann and McBride, 2011). However, in youth male soccer players conflicting evidence is present. Significant correlations (r = ) are reported between the FMS total score, deep overhead squat, in-line lunge, active 49

50 straight leg raise and rotary stability tests and various measures of jump and agility performance (Lloyd et al., 2015). Additionally, in-line lunge performance explained the greatest variance in reactive strength index (R² = 47%) and agility cut (R² = 38%) performances, which are both dynamic measures in nature. Therefore, in youth soccer players, variation in physical performance could partly be explained by functional movement quality; however, this requires further investigation due to the paucity of empirical research. Skill proficiency is often based on the determination of sports specific constructs involving a range of ball skills (Figueiredo et al., 2009). This determination may not be appropriate for injury risk identification since these skills are not reflective of positions that perform high risk maneuvers involving rapid decelerations and high eccentric forces (e.g. jumping and landing tasks). To screen an athlete s level of skill and injury risk during jumping and landing tasks, a range of time efficient, non-invasive methods have recently been developed (Padua et al., 2009; Myer et al., 2008; Myer et al., 2011). Such assessments provide coach friendly diagnostic tools which have been reported to correlate significantly with more sophisticated laboratory techniques (Padua et al., 2009; Myer et al., 2011). Examples include a tuck jump assessment (Myer et al., 2008), the Landing Error Scoring System (L.E.S.S) (Padua et al., 2009), and a predictive nomogram (Myer et al., 2011). These methods focus on mechanisms of injury which relate to knee injuries (more specifically the anterior cruciate ligament [ACL]) and can assist in the early identification of high risk athletes (Myer et al., 2011). Such injuries occur frequently in youth soccer, in addition to other common knee injuries including medial collateral ligament sprains (Price et al., 2004; Le Gall et al., 2006). Practitioners should also consider other diagnostic tools to assess a range of movement deficiencies for other prevalent injuries including ankle sprains and hamstring strains (Price et al., 2004). Examples of assessments which have been able to 50

51 predict such injuries include single leg jumping and landing (Goosens et al., 2015), dynamic balance tasks (Plisky et al., 2006), and eccentric hamstring strength (Lehance et al., 2009; Goosen et al., 2015). Practitioners should be encouraged to investigate the mechanisms of frequently occurring injuries and utilise assessments to detect functional deficits which may increase injury risk. Availability of literature to confirm the validity and reliability of assessments to quantify neuromuscular control in youth male soccer players is insufficient. Using available literature, Table 2.1 provides a sample movement skills test that could be used to assess lower extremity neuromuscular control for a youth male soccer player. 51

52 Table 2.1 A suggested test battery to evaluate the movement skill of a male youth soccer player Test Risk Factor Test Variables Application Tuck Jump (using video analysis) Observed landing mechanics Knee Valgus, Trunk Dominance, Leg Dominance Inability to control frontal plane motion in landing increases risk of ACL, MCL and medial meniscus injury (Myer et al., 2011). Single Leg CMJ (force plate and video analysis optional) Single Hop For Distance (Force Plate and video analysis optional) Y-Balance Bodyweight Back Squat Screen Asymmetry and Vertical Single Leg Landing Forces Asymmetry and Horizontal Single Leg Landing Forces Single Leg Balance and Stability Ankle, Hip and Thoracic Mobility, Knee, hip, and lumbar stability Peak vertical landing ground reaction forces, Single Leg Frontal Plane Projection Angle for each limb and determine asymmetry index (ASI) Hop Distance on each leg and determine asymmetry index (ASI). Peak vertical landing ground reaction forces. Reach distance in 3 directions (anterior, posterior medial and posterior lateral) Scoring criteria for upper and lower quarter and movement mechanics Asymmetry deemed an important injury predictor (14). SLCMJ was the only test to identify an asymmetry < 85% when comparing isokinetic and a range of jump tests (Petschnig et al., 1998) Assesses the ability to attenuate force during a single limb landing and horizontal force generation comparison between limbs. Reductions in performance have been reported as a predictor of hamstring injury (Goosens et al., 2015) Assessment of dynamic balance in a single limb stance demonstrating relationships to hip extension / abduction strength (Hubbert et al., 2007). Youth male athletes with an anterior right/left reach distance difference 2.5 x more likely to sustain a lower extremity injury (Plisky et al., 2006). Assessment of functional movement to determine deficits in a range of factors including: neuromuscular control, strength and mobility. A higher number of deficits may indicate an increased risk of injury and loss of performance (Myer et al., 2014) 52

53 2.3.3 Fatigue monitoring - methods for assessing training load Biochemical monitoring of training stress and recovery often involves the use of invasive protocols and costly equipment to measure changes in markers such as salivery testosterone, cortisol and IgA (Neville et al., 2008; Mclellan and Lovell, 2010). Monitoring time motion data and the residual fatigue of training (for example using heart rate telemetry) have distinct limitations too, such as day to day variation (Waldeck and Lambert, 2003) and inaccuracies due to varied exercise duration, hydration and training status, and competition anxiety (Lambert et al., 1998). In the assessment of youth male soccer players, heart rate measures used to determine player recovery status were not strongly associated with performance decrements. This leads to questioning their use for the determination of functional overeaching in youth soccer players (Buccheit et al., 2012). When considering which monitoring approach to implement, an awareness of the validity and variation present within each measure is essential to ensure accurate and reliable data is obtained. Hill-Hass et al. (2008) investigated the variability in a range of physiological parameters, perceptual responses and time-motion profiles during small sided games played by adolescent male soccer players. Typical error of estimates (%) were reported for a range of variables, with heart rate responses (<5%) and session rate of perceived exertion (between 1 and 2 units) demonstrating low variability, while blood lactate variability was high (range 16-34%). Although total distance covered at lower movement speeds displayed acceptable values (<5%), higher movement speeds indiciative of the intensities of match play (>8 km/h) had greater variability. Also, recent reports have confirmed high variance between GPS units when monitoring time-motion analysis, with some units measuring two to six times more acceleration / deceleration occurences than others (Buchheit et al., 2014). Cumulatively, due to feasibility, cost effectiveness and time efficiency, it is recommended that approaches for 53

54 monitoring training load should focus on easy to implement tools which are valid and reliable. In addition to the measurement error associated with monitoring time-motion data, of particular importance for practitioners when considering associations with injury is the differential between the quantification of external load (distance covered, number of accelerations / decelerations) and internal load (physiological load experienced by the player). For example, a squad of players may be exposed to the same external load, however, individual physiological responses may be markedly different (Bouchard and Rankinen, 2001) potentially increasing a player s injury risk. Available data show greater volumes of high velocity running are associated with soft tissue injuries in adult males (Gabbett and Shahid, 2012), and in team sport athletes cumulative weekly loads (specifically three-weekly loads) are indicative of greater injury risk (Colby et al., 2014). Therefore, while monitoring such markers may be useful, purely analyzing game related data may be over-simplistic and coaches are advised to consider the internal load experienced by each player. These recommendations are supported by Hill-Hass et al. (2009) who analyzed the acute physiological and time motion characteristics of three different small sided game formats (2 vs. 2, 4 vs. 4, and 6 vs. 6) using the same pitch size. The authors reported that the physiological load (measured via blood lactate, HR, and RPE) was greatest in the 2 vs. 2 player format, despite GPS data demonstrating that players completed lower total distances at the variety of movement speeds. Thus, approaches which monitor internal training load maybe more appropriate. Traditional field-based approaches that monitor internal training load have utilised training impulse (TRIMP) measured from the average heart rate multiplied by the session duration. However, this method does not reflect the intermittent nature of soccer match play 54

55 due to the averaging of heart rate and is thus inappropriate (Padilla et al., 2000). An alternative approach is to use session rate of perceived exertion. The athletes rate of perceived exertion (using an adapted Borg Scale see table 2.2) is assessed 30 minutes following the completion of the training session and mutliplied by the value of the session duration (in minutes) (Foster, 1998). This method can be adapted for resistance training by multiplying the number of repetitions performed by the session rate of perceived exertion. An example to outline training load monitoring of a standard soccer week using these methods has been provided in table 2.3. Such approaches have been used in elite male youth soccer players with multinominal regression demonstrating that physical stress was related to both injury and illness (OR ) (Brink et al., 2010). Specifically, for young players it is important to be educated on how to use the rating scale and how to interpret questions such as "how hard was the session? to ensure the accuracy of the data collected. Also, ensuring a consistent time frame between the end of a session (or game) and the time of rating is essential. Table 2.2 Borg Scale (Borg, 1982) and session rate of perceived exertion (RPE) scale. Adapted from (Foster et al., 1998) Borg Scale Category Ratio Scale Session RPE 0 Nothing at all Rest 1 Very weak Really easy 2 Weak Easy 3 Moderate Moderate 4 Somewhat strong Sort of hard 5 Strong Hard 6 7 Very strong Really hard 8 9 Really, really hard 10 Very, very strong Just like my hardest game 55

56 Table 2.3 Schematic evaluation of the load associated with a training programme in an elite male youth soccer player. Adapted from Turner et al. (2013) Day Training Type Duration/Reps RPE Session TL Daily TL Mon (am) Coaching Mon (pm) S&C Tue (am) Coaching Tue (pm) Rest Wed (am) Active Rec Wed (pm) Rest Thu (am) Coaching Thu (pm) S&C Fri (am) Coaching Fri (pm) Rest Sat (am) Match Sat (pm) Rest Sun (am) Rest Sun (pm) Rest Total TL 3, 835 Daily mean TL (=total TL/No. of training days) Daily SD of TL S&C = Strength & Conditioning Active Rec = Active Recovery Session Monitoring neuromuscular fatigue to determine state of recovery and readiness to play Neuromuscular fatigue is commonly measured via acute and chronic reductions in performance following a bout of exercise (Coutts et al., 2002). Specifically, the inclusion of functionally relevant, dynamic movements with an inherent stretch shortening cycle (SSC) component may provide a suitable assessment strategy (Komi, 2000; Nicol, Avela and Komi, 2006), as these will incorporate the mechanical, metabolic and neural elements of fatigue (Nicol et al., 2006). A measure commonly used is the assessment of jump height during a countermovement jump (CMJ) (Robson et al., 2009). It has been suggested that reductions in CMJ may provide an early indication of overreaching (Welsh et al., 2008); however, the 56

57 sensitivity of this approach may not be sufficient to identify deficits (Cormack et al., 2008). This is supported by the reported effects of soccer-specific fatigue on jump performance (measured during a squat jump, CMJ and drop jump) in male youth soccer players (Oliver et al., 2008). Decrements in jump height were present during all tasks, however, impact forces during the drop jump were the only landing force variables to show a significant change in response to fatigue. Thus, while reductions in CMJ height likely resulting from muscular fatigue were present, increased impact forces and skeletal loading associated with decrements in DJ performance might be of greater relevance for identifying injury risk. This highlights that practitioners should apply caution when just measuring isometric force production ability for monitoring purposes, assuming that neuromuscular capability has returned to pre-fatigue levels. Confounding this, following a fatiguing eccentric protocol, force production returned to baseline post 48 hours, whereas, electromechanical delay (EMD) was still comprised post 96 hours (Howatson, 2010). Changes in leg stiffness during a maximal hopping task have also been reported in youth male players in response to fatigue due to reductions in preplanned muscle feed-forward activity (Oliver et al., 2008). Therefore, muscular force production and neuromuscular feed-forward and feedback mechanisms exhibit differential responses to fatigue and are separate risk factors. Such decrements have been associated with reductions in joint stability and increased injury risk due to greater stress placed on soft tissue structures (Padua et al., 2005; Hughes and Watkins, 2008). However, individualized responses were noted (Oliver et al., in press); practitioners are encouraged to monitor the training response of each player to accurately quantify their relative risk of injury. While these methods provide valuable information regarding acute responses to fatigue, practitioners should also consider the effects of accumulated fatigue that will occur throughout a soccer season, of which there is currently a paucity of research available. 57

58 Aside from neuromuscular responses, the athlete s perception of fatigue is also an important aspect to consider, since perceptual measures are sensitive in detecting fatigue (McClean et al., 2010). Using a psychological questionnaire (table 2.4), players rate their state of well-being, with significant reductions in total score following competition indicative of greater levels of perceptual fatigue. Due to the fact that injury and illness appear to be related to the balance between stress and recovery (Brink et al., 2010), monitoring of psychosocial stress-recovery balance may be warranted. A method to quantify this is the Recovery Stress Questionnaire for Athletes (RESTQ-Sport). This questionnaire has recently been used with elite male youth soccer players to determine psychosocial stress across a period of two-seasons, and was associated with the occurrence of illness (Brink et al., 2010). Table 2.4 Perceptual monitoring questionnaire adapted from McClean et al. (2010) Wellness marker Record score here Fatigue Very fresh Fresh Normal More tired than normal Always tired Sleep quality Very restful Good Difficulty falling asleep Restless Sleep Insomnia General muscle soreness Feeling great Feeling good Normal Increased soreness / tightness Sore Stress levels Very relaxed Relaxed Normal Feeling stressed Highly stressed Mood Very positive Generally good Less interested in others / and or activities than normal Snappiness at team mates / family members Highly annoyed / irritable / down 58

59 2.3.5 Previous injury - assessment and identification A common method for investigating previous injuries experienced by a player is through the use of retrospective analysis. This approach relies on the individual s ability to recall their own injury history and may lead to recall bias which can occur with both long and short term retrospective reporting (Junge and Dvorak, 2000). Recall accuracy declines as the level of detail requested increases, i.e. exact number of injuries, body region and diagnosis of each injury sustained is reduced, whereas, the injury occurrence and location is more accurate (Gabbe et al., 2003). Due to their stage of learning, this factor may be further confounded in youth athletes. An alternative approach is a prospective reporting system, and an example form has been provided in table 4. Further guidelines for incidence reporting have been provided by the FIFA Medical Assessment and Research Centre (F-Marc) (Fuller et al., 2006). Specific recommendations for practitioners indicate that the number of games and training sessions should be documented for each player, and caution should be applied in reporting injuries per position due to frequent rotations (Junge and Dvorak, 2000). This approach may be useful if used in conjunction with tests to measure functional deficits and identify players who may be at a heightened risk of injury. However, in the practical setting of a football academy, the use of prospective analysis may not always be possible due to the frequency of registering new players and transfers between clubs. Thus, alternative systems are needed which are child friendly, time and cost efficient, and provide suitable information as to a players injury history. To date, a validated questionnaire that can accurately identify previous injury history is currently unavailable within the literature. Therefore, practitioners are encouraged to develop a simple document that is child appropriate (see figure 1.1 for an example) and used in conjunction with a professionally led medical assessment of musculoskeletal function and movement quality. 59

60 Table 2.5 A sample prospective injury reporting questionnaire Complete when a player sustains an injury and misses either a training session or a match DOB Date of Injury Training or match? Contact, noncontact or overuse? Time in activity when injury occurred (min) Injury location Comments / further description Time absent (days ) 13/09/ /04/2014 Match Non-Contact Knee During landing 3 On the diagram circle in red pen areas which currently hurt. In blue pen, circle areas which you have hurt previously. If you can remember accurately, try and number each injury in order of the time when it happened (list the first injury you had with the number 1). For each injury, list as much detail as you can remember below. This could include: type of injury (sprain / strain / fracture etc.), how long you were injured for, did you injure it playing sport, and if you had surgery. Injury 1: Injury 2: Injury 3: Injury 4: Injury 5: Figure 2.1 Sample child friendly retrospective injury reporting questionnaire 60

61 2.3.6 Developing a return to play criteria Classifying an appropriate level of function assessed via rehabilitation constructs (pain, muscle strength, joint stability), movement competency and performance capacity is a critical component in the determination of an appropriate time-point for an athlete to return to play. Players that return to competitive activities too soon after injury will subsequently heighten their risk of re-injury (Hagglund et al., 2005). In an attempt to provide clear return to play guidelines a recent investigation by Haines et al. (2013) utilised the Delphi method to identify a set of criteria from which to evaluate the suitability of assessment tools for previously injured athletes. This method has been commonly applied within available literature to validate the use of assessment tools (Stheeman et al., 1995; Grahm et al., 2003). Following three rounds of questionnaires, a consensus decision was made confirming a physical performance assessment checklist for prospective test batteries to be used in preparation for returning to sport following lower extremity injury (see table 5). The agreed constructs include a range of assessments to measure: stability, power, neuromuscular control, coordination, pain, movement, balance, and psychological factors (Haines et al., 2013). For further examples and guidelines of targeting deficits before return to sport see (Myer et al., 2008a). 61

62 Table 2.6 Return to play criteria physical performance checklist. Adapted from Haines et al. (2013) Construct 1) Evaluation of multiple planes of lower extremity movement 2) Tests are specific to the movement demands of the target sport 3) Evaluation of the athlete landing after explosive movements, unilaterally or bilaterally 4) Evaluation of appropriate neuromuscular control of bilateral lower extremities during deceleration activities or sports specific movement 5) Identification of compensatory and or/ acquired dysfunctional movement patterns 6) Identification if pain during or after the movement modifies the movement during the assessment 7) Includes a validated and reliable threshold which determines a specific return to sport 8) Evaluates efficiency of changing direction during sport specific movements 9) Tests for symmetrical motor control (stability) of active lower extremity movement bilaterally and asymmetrically 10) Evaluation of trunk control/strength required for sport specific activity 11) Tests are specific to the movement demands of the involved joint for a particular sport 12) Includes the fear avoidance questionnaire Summary of assessing injury risk factors in male youth soccer players Individuals working with youth athletes, and in particular, male youth soccer players, need to consider effective strategies to implement a range of assessment and monitoring tools which assist in the identification of a players relative level of injury risk. The key factors for effective implementation have been reviewed in this manuscript and are summarized below: 1. When monitoring growth and maturation of players, practitioners are advised to include a range of assessment tools (tracking growth curves and predictions of APHV) to increase their accuracy 62

63 2. Qualifying a player s level of movement skill should involve tasks which challenge their level of neuromuscular control in a range of tasks which identify deficits associated to the mechanisms of frequently occurring injuries 3. Due to the associated risk of injury following acute and cumulative fatigue, methods to monitor training load, and a players state of neuromuscular readiness should be implemented to ensure they are able to meet the demands of competition 4. Previous injury should be recorded using methods that are child friendly, time and cost efficient, however, practitioners should be aware of recall inaccuracies 5. Developing an appropriate return to play criteria with specified rehabilitation constructs and physical performance capacities is recommended to reduce the risk of re-injury 63

64 2.4 Neuromuscular Risk Factors for Lower Extremity Injury in Male Youth Soccer Players Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD and Lloyd, RS. Neuromuscular Risk Factors for Knee and Ankle Ligament Injuries in Male Youth Soccer Players. Sports Med. In Press The sport of soccer imposes high physiological demands and an inherent risk of injury due to the repeated high intensity movements such as jumping, cutting and contact with opposition players when fatigued (16). Injuries in male youth soccer players occur mainly in the lower extremities (71-80%), with a low proportion (5%) of overuse injuries and a high proportion (20%) of acute traumatic ligament sprains at the ankle and knee (Le Gall et al., 2006; Price et al., 2004). More specifically, the medial collateral ligament (MCL) and anterior talofibular ligament are the most commonly reported injuries (Cloke et al., 2009; Price et al., 2004). Altered neuromuscular control during dynamic activities (e.g. running, cutting and landing) is indicated as a key mechanism for lower limb ligament injuries (Hewett et al., 2004; Hewett et al., 2005). Deficits in neuromuscular control direct excessive stress to the passive ligamentous structures, exceeding their tensile threshold, resulting in mechanical failure (Li et al., 1998). Specific neuromuscular imbalances have been identified for female athletes and adult males including: quadriceps dominance (Myer et al., 2004), leg dominance (Myer et al., 2011), ligament dominance (Hewett et al., 2002); trunk dominance, (Myer et al., 2011), neuromuscular activation patterns (Kubo et al., 2007) and dynamic stability (Ross et al., 2008). The presence of these deficiencies does not indicate an explicit causative factor for injury per se, however, it should be noted that ligamentous injuries likely occur when active muscular restraints are unable to adequately reduce joint torques during dynamic movements involving deceleration and high forces (Beynnon and Flemming, 1998; Powell and Barber- 64

65 Foss, 2000). In-spite of the growing body evidence in adults and young female athletes, there is a paucity of literature available examining injury risk factors in male youth soccer players. Due to the physical demands of youth soccer, the associated injury risk, and the number of children and adolescents who participate in the sport with an age related increase in incidence (Michaud et al., 2001; Price et al., 2004), there is a clear need for increased research in this cohort to identify age and sex-specific injury risk factors (Alernton-Geli et al., 2014). This article will review the available literature and investigate key neuromuscular risk factors that may contribute to their relative risk of injury Quadriceps dominance Disproportionate knee moments between recruitment of the quadriceps and hamstrings may be reflective of an imbalance in force absorption (Myer et al., 2011). Actions in soccer which require rapid decelerations (such as landing, pivoting and cutting) involve substantial eccentric muscle force contributions from the knee extensors, increasing the risk of noncontact ligamentous injuries (Simonsen et al., 2000). The considerable anterior shear of the tibia relative to the femur is counteracted by the anterior cruciate ligament (ACL), MCL and co-activation of the knee flexors (Beynnon et al., 1992; Dragenich and Vahey, 1990). The functional hamstrings: quadriceps ratio has been shown to increase in post-pubertal children and adolescents as a requirement for co-activation of the hamstrings to reduce anterior tibial translation and high shear forces of the quadriceps during high velocity movements as a form of counterbalance (De Ste Croix, 2007). In male youth soccer players, alterations in the functional hamstring: quadriceps ratio (H: Q) may be present due to muscle loading patterns that asymmetrically strengthen the quadriceps during repetitive training and competitions (Iga et al., 2009). This alters the reciprocal balance of strength and dynamic stabilisation around 65

66 the knee as indicated by compromised function of the hamstrings during high velocity actions (Iga et al., 2009). Confounding factors such as fatigue and previous injury also need to be considered within the remit of injury risk factors as the final stages of a game appear to be a particular time of risk in male youth players (Price et al., 2004). In adults, reports of a more quadriceps dominant landing strategy and reduced peak muscle activation of the tibialis anterior and hamstrings (Padua et al., 2006), in addition to changes in the functional H/ Q ratio (Small et al., 2010), have been cited as plausible explanations. Neuromuscular inhibition of the hamstrings may also be present following the occurrence of a hamstring injury (Opar et al., 2013), and such injuries occur at regular frequencies in elite male youth soccer (Le Gall et al., 2006; Price et al., 2004). The knee flexors are key antagonist muscles, are heavily involved with knee joint stabilisation, and demonstrate greater deficits in eccentric rather than concentric strength (Crosier and Creelard, 2000; Lee et al., 1994). Cumulatively, it could be conceived that reduced activation of the hamstrings relative to the quadriceps increases injury risk in male youth soccer players. Therefore, practitioners should devise appropriate injury prevention strategies that target deficits in knee flexor strength. Limited data is available in youth populations, however, in collegiate female athletes a 6-week programme of emphasized hamstring resistance training significantly increased the functional H: Q ratio to acceptable levels (>1) for the reduction of ACL injuries (Holcomb et al., 2007). Significant improvements in sagittal plane knee kinematics indicative of a reduced quadriceps dominant landing strategy (greater initial contact knee flexion) have also been observed following a short duration (7 weeks) neuromuscular training programme that included both dynamic balance and plyometric exercises (Myer et al., 2006). The authors suggested that the prescribed interventions could have increased knee flexor co-contraction, however; muscle 66

67 activation patterns were not measured; further research is warranted to examine the effects of such activities on measures of quadriceps dominance in male youth soccer players Leg dominance Leg dominance has been defined as an imbalance in strength, coordination and control between the two lower extremities (Myer et al., 2004). A discrepancy in excess of 15% has been deemed a key predictor of injury (Crosier and Creelard, 2000). Asymmetry places additional stress on the weaker leg, compromising performance and predisposing athletes to various injuries during cutting / landing activities (Hewit et al., 2012). This risk factor is inherent to soccer, where preferred limb dominance is evident. For example, Zebis et al. (2009) identified that during a cutting maneuver, all subjects who subsequently experienced an ACL rupture in the following 2 competitive seasons injured their preferred push off leg. Additionally, analysis of the distribution of non-contact ACL injuries in male players confirmed 74.1% injured their dominant (kicking) leg (Brophy et al., 2010). There is a paucity of data in youth players confirming the presence of asymmetry and associated injury risk. One available study used the star excursion balance test (Plisky et al., 2006). Logistic regression indicated that high school athletes with an anterior right-left reach difference >4 cm represented a 2.5 times greater risk of lower extremity injury. To the knowledge of the authors, no studies have been conducted with male youth soccer players to examine the relationship between limb dominance and prevalence of injury. In spite of the lack of available evidence to report the relationship with injury, asymmetry has been reported in male youth soccer players (Atkins et al., 2015; Daneshjoo et al., 2013). Daneshjoo et al. (2013) measured isokinetic hamstring and quadriceps strength and hip joint flexibility in young male professional soccer players. Of the 36 players tested, all but one 67

68 reported musculoskeletal imbalances >10% and heightened levels of dominant leg hip joint range of motion. In addition, Atkins et al. (2015) investigated limb asymmetry by examining contralateral differences in peak ground reaction forces during the deep squat exercise in elite male youth soccer players. Based on chronological age, significant differences were identified between limbs (p 0.05) in all age groups except for the U13s and U17s (the youngest and oldest groups respectively), indicating that asymmetry increases during the period of peak height velocity (PHV) and the early stages of adolescence. This time point corresponds with a stage of adolescent awkwardness, in which decrements in motor skill performance are often evident resulting from rapid gains in limb length (Lloyd et al., 2014; Philippaerts et al., 2006). Further, elite male youth soccer players experience more traumatic injuries in the year of PHV (van der Sluis et al., 2014) and total injury incidence is highest in this period (Price et al., 2004; Rumpf and Cronin, 2012). An awareness of leg dominance and the development of appropriate screening protocols will aid practitioners to identify youth players who may be at a greater risk of injury. Utilising single limb tasks is preferred to bilateral variations due to their enhanced ability and sensitivity for determining deficits in neuromuscular control (Myer et al., 2011b). Furthermore, assessments of leg power across multiple directions (vertical, horizontal, and lateral) have reported insignificant relationships between tests in the various movement planes (Hewit et al., 2012; Maulder and Cronin, 2005; Meylan et al., 2005). Testing athletes across different directions may measure relatively independent leg-power qualities, thus utilising a range of assessments targeting multi-planar actions is recommended. When interpreting athlete test scores, practitioners should also be cognizant that accurately determining an asymmetry threshold to predict injury is challenging. Adult data indicate a 15% difference between limbs increases injury risk, however; currently no research is available to examine the relationship between asymmetry and injury risk in male youth soccer players and this requires further investigation. 68

69 2.4.3 Ligament dominance Dynamic valgus alignment has been defined as a medial or valgus collapse of the knee during tasks which involve hip and knee flexion (Hewett et al., 2005a; Myer et al., 2004). Displacement primarily occurs in the frontal plane from a combination of hip internal rotation, knee valgus and tibial external rotation (Krosshaug et al., 2007). This is coupled with decreased knee and hip flexion angles and pronation at the subtalar joint (Brophy et al., 2010). Such knee positions indicative of reduced frontal plane control on ground contact have been reported in male subjects who subsequently experienced an ACL injury (Padua et al., 2009). Due to the increased load on the MCL and ACL in this position, greater valgus angles have been indicated as a predisposing injury risk factor for athletes (Hewett et al., 2005). In male youth soccer, incidence reports have confirmed that 17.1% of all injuries occur at the knee (37, 49), and the MCL accounts for between 77-83% of all knee ligament injuries (Moore et al., 2011; Price et al., 2004). The knee is also the most frequent site of major injuries (classified as >4 weeks absence from match play) in this population (Volpi et al., 2003), most commonly ACL lesions (all of which required surgery), followed by meniscal and MCL sprains. Knee valgus has been a commonly associated risk factor for non-contact ACL injury in female athletes (Hewett et al., 2005a). During the prepubescent period, similar valgus alignment has been reported between boys and girls (Barber-westin et al., 2005; Ford et al., 2010a; Quatman et al., 2005). In a large cohort study, Barber-Westin et al. (2005) identified that during a drop jump test, although boys (aged between 9-17) had greater mean ankle and knee separation distances than aged matched girls on take-off, a large proportion of the boys displayed distinct lower limb valgus alignment upon landing. No significant sex differences for mean total medial knee displacement were present and normalized medial knee displacement difference between boys and girls aged 15 was minimal. Therefore, this risk 69

70 factor may be present in both youth males and females. However, caution should be applied to the findings of Barber-Westin et al. (2005) study, as sizeable variability were present in the data; the study only used a single plane camera set up which may have limited the accuracy of identifying potential differences. These findings also conflict with previous investigations that have reported that after the onset of maturation, girls display greater medial knee displacement than boys (Hewett et al., 2004; Hewett et al., 2005b). During the prepubescent period, ACL injury occurrence is less frequent (Tursz and Crost, 1986), however, the risk of ligament sprain increases as youth approach adolescence (Adrim and Cheng, 2003). Smaller stature, lower body mass and relatively slower velocities of play involved in children s sporting activities may help reduce the risk of injury in pre-pubertal athletes (Barber-westin et al., 2005). Cumulatively, the available data show that younger male players will display greater levels of knee valgus than older youths; however, these findings are not consistent across the all investigations (Noyes et al., 2005). A lack of evidence currently exists to describe knee angles during landing and cutting-based tasks for male youth soccer players at different stages of growth and maturation and their association with injury risk. This requires further investigation to validate its presence as a relevant risk factor for knee injury in this cohort. However, due to the potential for increased injury risk, young male soccer players, and specifically those who demonstrate valgus alignment upon landing, should be targeted to undertake a progressive and periodized integrative neuromuscular training (INT) programme comprised of general and specific activities including strengthbuilding, plyometrics, balance and agility (Faigenbaum et al., 2011). This should be completed in addition to their regular soccer practice due to the beneficial effects of INT on lower extremity kinematics (Myer et al., 2006; Myer et al., 2011c; Myer et al., 2011d). 70

71 2.4.4 Neuromuscular feed-forward strategies During actions such as landing from a jump or side cutting in competitive match play, the time available for decision making and postural repositioning is limited. These reactive tasks provide insufficient time to make the necessary postural adjustments, resulting in compromised leg positioning and significantly greater loads on the knee joint (Besier et al., 2001). Moreover, Krosshaug et al. (2007) using in-vivo video analysis reported that the timing of non-contact ACL injury ranged between milliseconds following initial ground contact. This short timeframe does not provide a suitable period to utilise reflexive neuromuscular feedback mechanisms, but rather relies on feed-forward muscle activation. This serves to maintain joint integrity by providing early recruitment of the involved musculature prior to loading to better enable force absorption and, in doing so, reducing joint torques and ligamentous loadings (Beard et al., 1993; Besier et al., 2001; Hewett et al., 2005a). Therefore, altered or imbalanced sequences of neuromuscular firing during dynamic actions occurring in soccer such as landing or cutting may increase the risk of injury. There is a paucity of research that has examined the relationship between muscle preactivation and injury risk in pediatric populations. Available literature has compared muscle pre-activation in pre-pubescent boys (aged 9-11 years) and post-pubescent males (aged years) during a vertical jumping task (Croce et al., 2004). Post-pubescent youth displayed greater levels of hamstring activity and co-contraction ratio prior to landing. Conversely, greater hamstring activity post-landing and during initial contact to maximum knee flexion was present in pre-pubertal subjects, indicating greater co-contraction ratios post landing than older subjects. Intuitively, this suggests that a more efficient neuromuscular feed-forward strategy is developed prior to landing as males mature to control ground reaction forces and regulate anterior displacement of the knee. Confirming this, research shows that preparatory co-contraction ratios were 2 times higher in adults compared to children in landing from a 71

72 vertical jump task (Russell et al., 2007). Data also show that as children mature they become more reliant on supra-spinal feed-forward input and short latency stretch reflexes (Lloyd et al., 2012), suggesting that pre-activation strategies are a learnt skill that develop with maturation. During a period of adolescent awkwardness, disrupted sequences of neuromuscular feed-forward muscle recruitment may increase injury risk; thus re-establishing the correct performance of fundamental motor skills and landing mechanics is an important consideration for practitioners responsible for youth-based exercise prescription. Furthermore, practitioners should consider the inclusion of plyometric exercises to reduce the risk of injury. Data are available in young female athletes that show 6-weeks of plyometric training increased preparatory muscle activation of the adductors and abductor-adductor coactivation ratio, which was suggestive of a training-induced pre-active motor strategy (Guskiewicz et al., 1996). Thus, while this study utilised female subjects, it could be inferred that such activities would also be beneficial for male youth soccer players, however this requires further investigation Dynamic balance Effective performance of both static and dynamic stability tasks requires the integration of visual, vestibular and proprioceptive inputs which provide an efferent response to control the body s center of mass within its base of support (Guskiewicz et al., 1996). Deficits in postural control and reflex stabilisation have been reported in the assessment of subjects with functional ankle instability via the calculation of time to stabilisation (Ross and Guskiewicz, 2008). A delay in neural feedback responses, which contribute to lower limb stability, may increase injury risk highlighted by the timing of ACL injuries 17-50ms after initial ground contact (Krosshaug et al., 2007). In male soccer players, the physiological training effects of 72

73 a warm up injury prevention programme have been reported and time to stabilisation in the training group was 90ms faster than the soccer training only matched controls (Imprezzilini et al., 2013). Therefore, it can be implied that shorter muscle feedback responses will enhance neuromuscular control via active restraint, thus lowering injury risk. Available data to confirm relationships between impaired dynamic stability and injury risk in youth populations is sparse. Previous literature has suggested that maturation of the neurological, visual, vestibular and proprioceptive systems may lead to enhanced performance during single leg balancing tasks (Mickle et al., 2011). Younger subjects demonstrate greater postural sway during single leg balance maneuvers which may compromise stability (Mickle et al., 2011). In male high school basketball players, higher postural sway during unilateral balancing was also associated with increased risk of ankle sprain (McGuine et al., 2000). Subjects who demonstrated greater sway experienced nearly seven times as many ankle sprains as those with good balance. In male youth soccer, ankle injuries account for 19% of total injury incidence (Cloke et al., 2009) and the most common diagnosis is anterior talofibular ligament sprain (Price et al., 2004). Improving dynamic balance has significantly reduced the risk of ankle sprains in high school soccer and basketball players who performed a series of single leg balance and squat exercises in both stable and unstable conditions (risk ratio, 0.56; 95% CI, ; p = 0.033) (McGuine and Keeme, 2006). Moreover, youth male soccer players undertaking a proprioception training intervention enhanced postural stability indices in both anterior-posterior and medial-lateral directions on the star excursion balance test (Malliou et al., 2004). Additionally, a significant reduction in the number of knee and ankle sprains was reported across the course of a soccer season in these players in comparison to those who only completed their normal soccer training (Malliou et al., 2004). Enhanced trunk stabilisation may also improve dynamic stability and balance due to improved trunk motion control (Hewett and Myer, 2011). This 73

74 has been confirmed in male youth soccer players, whereby an intervention consisting of trunk stabilisation exercises including quadruped contralateral raises, front planks, back and side bridges enhanced performance during specified reach directions of the star excursion balanced test and single leg static balance tasks (Imai et al., 2014). Cumulatively, the available literature suggests that deficiencies in dynamic stability may increase the risk of lower limb injuries in male youth soccer players. Targeting these deficits using appropriately prescribed balance and trunk stabilisation exercises may subsequently reduce this risk Trunk dominance Trunk dominance has been defined as an imbalance between the inertial demands of the trunk and the ability of the core to resist perturbations to the center of mass (Hewett et al., 2005a; Myer et al., 2011b). This inability to dissipate force effectively results in excessive trunk motion primarily in the frontal plane and increased ground reaction forces and knee joint torques (Hewett and Johnson, 2010). When distal segments are fixed during closed chain sporting actions, motion at more proximal segments will influence the kinetics and kinematics of other segments in the chain (Leetun et al., 2004). A key action of the abdominal musculature is to provide adequate control of the pelvis due to increases in femoral internal rotation and adduction which may be coupled with increased anterior pelvic tilt (Ireland, 2002). Active proprioceptive repositioning of the trunk has also predicted knee injury status with 90% sensitivity and 56% specificity in female athletes (Zazulak et al., 2007). Thus, reduced pre-activation of the trunk may result in a loss of control of the body s center of mass and can be considered essential for controlling excessive spinal motions which may contribute to altered biomechanics of the lower limbs during dynamic movements in soccer. 74

75 There is currently a paucity of literature pertaining to measures of core stability, trunk dominance and injury incidence in youth male soccer players. While the aforementioned assessments (Zazulak et al., 2007) have demonstrated favorable results, it should be considered that such measures were derived during artificial conditions and postures in which the pelvis is immobilized, limiting contributions from more distal musculature. Ecological validity may be questioned and the highly specialised and costly equipment will likely limit application to larger scale youth athlete screening programmes. Alternative field-based trunk muscle endurance assessments also contain ecological validity concerns due to their prolonged isometric actions and non-functionality. This is confounded by data reporting weak-moderate relationships (Intra-class correlation coefficient range = ) between performance on isometric trunk endurance tests and a range of athletic measures (Nesser et al., 2008). Moreover, Leetun et al. (2004) analyzed isometric trunk endurance tests and additional measures of hip abduction and external rotation strength, identifying that hip external rotation strength was the only useful predictor of injury status (Odds Ratio = 0.86). Therefore, further investigation is required to elucidate if trunk dysfunction is a prevalent risk factor in youth male soccer players. In addition, appropriate trunk dominant neuromuscular control tests are needed to accurately measure potential deficits that may predispose athletes to a greater risk of injury Injury risk factor hierarchical model Following the identification of prevalent injury risk factors, the authors propose a systematic model to screen male youth players and subsequently develop individualized programmes to reduce their relative risk of knee and ankle injuries (Figure 1.2). Each risk factor is linked to a neuromuscular screening assessment and target exercises are then selected to improve 75

76 relevant neuromuscular control deficits. It is beyond the scope of this review to discuss each assessment listed here; however, practitioners are encouraged to examine comprehensive reviews that discuss their application (Myer et al., 2004; Myer et al., 2011b) and to utilise assessments that: 1) focus on mechanisms and associated risk factors of injury, 2) are able to detect functional deficits assisting in the early identification of players at high risk, and 3) demonstrate suitable validity and reliability. Also, practitioners may wish to assess their players movement abilities under conditions of fatigue to determine changes in neuromuscular control. In male youth soccer, injuries occur more frequently towards the end of the first and second half respectively (Price et al., 2004). Solely screening players in a nonfatigued state may not accurately identify those individuals whose movement mechanics deteriorate towards the end of a match, affecting their relative risk of injury. The final step requires the selection of appropriate exercises that are associated with each test. It is proposed that following an appropriate training intervention, neuromuscular deficits can be reduced, lowering injury risk. In selecting exercises to reduce injury risk, practitioners are advised to consider INT (Myer et al., 2011c). The integration of such activities that develop fundamental movement skills should be initiated during preadolescence and maintained through adolescence to enhance skill related fitness and reduce the risk of sports related injury (Faigenbaum et al., 2011; Myer et al., 2011d). Periodized INT programmes included during the pre- and off-season periods are also critical, especially for youth soccer players who engage in specialised sports practice where exposure to a wide range of developmental motor skill activities is limited (Myer et al., 2011c; Myer et al., 2011d). 76

77 Injury risk factor knee and ankle ligament hierarchical model Quadriceps dominance Leg dominance Ligament dominance Muscle feedforward stabilisation Dynamic balance Trunk dominance Knee flexor/extensor strength assessment Single leg hop assessments Kinematic jumplanding assessment Drop jump leg stiffness assessment Y-balance assessment Trunk stability push up and rotary stability assessments Posterior chain and eccentric hamstring strengthening Single leg strength training, landing and hopping tasks Jump-landing exercises Plyometrics (including jumplanding technical exercises) Balance and proprioception training Ant-flexion/extension /rotation training and trunk perturbation drills Figure 2.2 Knee and ankle ligament injury neuromuscular risk factor hierarchical model The top tier of the model includes identification of the associated neuromuscular risk factors for knee and ankle ligament injuries. Practitioners are then required to select appropriate assessments that are able to detect functional deficits assisting in the early identification of players at high risk (tier 2 of the model). The final step involves the selection of appropriate exercises that target each of the respective neuromuscular risk factors. It is proposed that following an appropriate training intervention, test performance will improve and subsequent neuromuscular deficits can be reduced, lowering injury risk 77

78 2.4.8 Summary of neuromuscular risk factors in male youth soccer players The occurrence of lower limb injuries in male youth soccer players is highly prevalent with risk increasing as players grow and mature. This review has identified neuromuscular imbalances for common injuries in male youth soccer providing an overview of the available pediatric literature. Existing research suggests that heightening neuromuscular control in the aforementioned areas may reduce the risk of injury in male youth soccer players; however practitioners should be cognizant that available data in this cohort are sparse. A hierarchical injury risk factor model has also been proposed to provide practitioners with a systematic model to screen youth players and determine their level of risk. Hypothetically, this information can then be used to develop individualized programmes that target deficits in neuromuscular control. 78

79 2.5 Field-Based Assessments of Neuromuscular Control in Male Youth Soccer Players Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD, Lloyd, RS. A review of field-based assessments of neuromuscular control and their applicability for youth soccer players. Sports Medicine. Under review. Neuromuscular imbalances during the performance of high impact force and velocity tasks such as jumping and landing have been cited as suggested mechanisms for sports-related injury (Myer et al., 2004; Myer et al., 2011a; Sugimoto et al., 2015). Appropriate screening methods to assess deficits in neuromuscular control are essential to identify youth athletes who may be at a greater risk of injury. Also, the practical application of such measurements has to be considered due to the cost implications and time inefficiency of screening a large number of athletes; thus, in the context of a soccer academy, field-based assessments are likely more appropriate. There is currently a paucity of literature available to examine injury risk factors and reliable screening techniques in male youth soccer players (Alernton et al., 2014; Read et al. in press); therefore more data are required to examine the validity and reliability of field-based assessments of neuromuscular control in this cohort. This section will review the available literature and describe a range of field-based neuromuscular screening tests, critically analysing their suitability for use with male youth soccer players Assessment of quadriceps dominance The gold standard measurement for determining the ratio between hamstring and quadriceps (H/Q) strength involves the use of an isokinetic dynamometer to assess muscular torque in both concentric and eccentric actions. Advantages of this method include; the ability to control range of motion and movement velocities, accommodating resistance, and strong 79

80 reliability (Davies and Heiderscheit, 1997). However, the feasibility of using such measurements should be considered due to the cost implications and time inefficiency of screening a large number of athletes. More efficient, cost effective field-based methods of assessment are likely more appropriate in academy soccer programmemes due to the high number of players, congested training and competition scheduling, and financial resources available Hand-held dynamometry Handheld dynamometry determines the greatest force an individual can resist during an isometric contraction, with the value reported in kilograms. This method is efficient and cost effective and can be easily administered to assess strength imbalances of the knee flexors and extensors. A systematic review to examine the relationship between hand-held and isokinetic dynamometry reported moderate-strong relationships between these two methods (Stark et al., 2011). The authors concluded that the hand-held mode of assessment can be considered valid and reliable, which benefits from ease of use, portability and cost effectiveness. Also, a previous investigation of test re-test reliability reported moderate to strong inter-instrument relationships (ICC = 0.79) (Bohannon, 1990). Despite the hand-held method providing a practical alternative for the clinical assessment of muscle strength, distinct limitations are present. These include the isometric nature of assessment, which may not be indicative of dynamic sporting actions; no control or variation in movement speed which limits the interpretation of muscle torque relationships during high speed manoeuvres; and performance may be affected by previous injury. This has been shown by low to moderate correlations (r = ) when comparing hand held and isokinetic dynamometry in subjects with knee dysfunction (Reinking et al., 1996). Practitioners must also be highly skilled, experienced, 80

81 and display sufficient strength to resist the individual. This may be suitable for younger players, but with advanced maturity and strength players may be able to overcome the resistance (Bohannon et al., 1990). Straps may be used in this case but this presents further limitations for practitioners in establishing the correct set up and accuracy of interpretation (Hsieh, 1990). There is also a paucity of literature in paediatric populations, with a majority of studies focusing on subjects with neuromuscular disorders Isometric hamstring strength tests A simple and practically viable technique that may be used to assess isometric hamstring strength has been proposed recently in male professional soccer players (McCall et al., 2015). In this test, the peak force exerted is measured in a supine position with the heel of the testing leg placed on a raised force platform with the knee flexed at either 90 or 30 to preferentially recruit either the semimembranosus, semitendinosus or biceps femoris respectively (Onishi et al., 2002). Good to strong reliability of this assessment has been reported in elite male soccer players (CV = %) (McCall et al., 2015). Additionally, the test was sensitive to changes in performance following match induced fatigue at both angles recorded, indicating its practical usefulness in the assessment of isolated hamstring strength (McCall et al., 2015). A limitation of the work of McCall et al. (2015) is the test position, which is not reflective of positions of high risk during sprinting (Chumanov et al., 2011) and their role in resisting anterior tibial translation of the knee joint during cutting and landing manoeuvres (Hewett et al., 2005a). A recent investigation examined the reliability of an isometric hamstring test, whereby, the subjects were positioned with their foot locked in a load cell secured to the floor, lying prone on a plinth, underneath a portable 45 wedge board in order to standardise the hip position and replicate a more functional position of the hamstrings 81

82 during the terminal swing phase of the running action (Wollin et al., in press). Strong reliability was reported for measures of peak torque (ICC range = ; SEM% range %) and a minimal detectable change of between 11.8 and 15.9% (Wollin et al., in press). This range has previously been associated with an increased risk of hamstring injury (Crosier and Creelard, 2000). However, caution should be applied as there is currently limited data to indicate the predictive validity of isometric tests in soccer players who may be at a greater risk of experiencing hamstring strain injury and whether there is transference to paediatric populations. Also, the test set up described is both time consuming and requires the use of specialised equipment Field-based hamstring strength tests There is currently a paucity of literature to report the validity and reliability of field-based hamstring strength assessments in male youth athletes. A previous investigation of risk factors for hamstring injury in adult players assessed injury history and included a Nordic hamstring strength assessment within a field-based screening battery (Engebretsen et al., 2010). The test was scored as either weak or strong based on the player s ability to hold the required body position during a Nordic hamstring curl beyond a 30 angle for 10 seconds (Engebretsen et al., 2010). However, no association with increased risk of hamstring injury was identified and inter-rater reliability was weak (k = 0.24) (Engebretsen et al., 2010). More recently, the Nordic hamstring curl has been used to measure knee angular displacement via two-dimensional analysis, whereby, a greater knee angle prior to the moment where the athlete loses eccentric control may be indicative of heightened eccentric hamstring strength (Sconce et al., 2015). The relationship between the break point angle (the angle at which the subject is unable to resist the gravitational moment) and isokinetic hamstring peak torque 82

83 showed a significant correlation (r = -0.80, R² = 65%) in male and female adult soccer players, but not the angle of peak torque (Sconce et al., 2015). A limitation of this assessment is the requirement for testers to hold the athletes feet during the movement. Issues regarding standardisation of pressure and inter-tester differences are present which may affect testretest reliability. A more sophisticated methodology has been proposed that assesses eccentric peak force and bilateral muscle balance during the Nordic hamstring curl exercise on an instrumented device with uniaxial load cells (Opar et al., 2013). Test-retest reliability of this device has been reported with recreationally active males displaying high to moderate reliability (ICC = ; TE = 21.7N N; CV% = ) (Opar et al., 2013). In adult male Australian Rules football players, eccentric hamstring strength below 256N (risk ratio = 2.7; p = 0.006) and 279N (risk ratio = 4.3; p = 0.002) at the start and end of the preseason respectively increased the risk of future hamstring injury (Opar et al., 2015), while asymmetrical limb differences of > 10% did not significantly increase injury risk (Opar et al., 2015). It should be noted that for all of the aforementioned Nordic hamstring curl assessments, in contrast to isokinetic measures, movement speed cannot be controlled. It is also not possible to determine the angle at which peak torque of the knee flexors occurs. In addition, comparative assessments between knee extensor and flexor strength to assess hamstring to quadriceps ratios cannot be easily administered, thus limiting the information available to identify this injury risk factor. An alternative measure is the single leg hamstring bridge test that requires the athlete to position themselves in a supine position and place one foot on top of a box with the aim of performing as many repetitions as possible using a straight leg hip extension motion. A recent prospective study showed that young male Australian Rules players who experienced a hamstring strain injury during the course of a season performed a significantly lower number 83

84 of repetitions than the non-injured control group (Freckleton, Cook, Pizzari, 2014). However, there was a low overall injury rate and confounding factors were reported including, age and previous injuries, which are known risk factors for hamstring strain injury (Engebretsen et al., 2010; Opar et al., 2013). This assessment could also be considered a test of muscular endurance as opposed to strength, and places a greater emphasis on the concentric function of the hamstrings. Based on the current body of evidence, there is a paucity of valid and reliable fieldbased assessments to accurately measure quadriceps and hamstring strength and H/Q ratios in male youth soccer players. The majority of investigations to date have either utilised adult subjects or expensive and time-consuming laboratory equipment. Also, the predictive validity of these assessments remains unclear and requires further investigation. An overview of the available research using paediatric populations to measure quadriceps and / or hamstring strength is summarised in table

85 Table 2.7 Assessments of quadriceps dominance in male youth athletes Reference Subjects Measurement Tool Summary of findings Herbert et al. 74 school-age boys and 2011 girls (age 4-17) Hill et al Mercer Stemmons al Freckleton al et et Wollin et al. in press 18 boys and 7 girls (aged 9-11) 17 healthy children (aged 11.1 ±2.4) 482 semi elite Australian Rules players (age range years) 16 elite male youth soccer players (age ± 0.54 years) Assessment of reliability and concurrent validity between the Isokinetic and hand held dynamometer Relationships between isokinetic and hand held dynamometry measured at different joint angles and movement speeds Test re-test reliability measurements of hand held dynamometry in healthy children and those with down syndrome Single leg Hamstring Bridge Test (SLHB). Pre-season screen and season monitoring of Hamstring Injuries. Intra and Inter-day reliability of a prone single leg isometric hamstring test using a calibrated load cell Mean intra and inter-rater reliability (ICC range = ). SEM varied from 0.5 to 6.0 Nm and was highest for the hip extensors and ankle plantarflexion. Mean concurrent validity (ICC) varied from 0.48 to 0.93 Isometric strength was able to predict low-velocity dynamic strength with moderate-high reliability (ICC range r = ). Greater speeds displayed lower relationships (120 s-¹, ICC range r = ; 180 s-¹ ) Reliability for normal children (ICC = , SEM = N). Lower ICC and higher SEM were reported for children with down syndrome Reliability of SLHB (ICC = ; inter-tester ICC = ). Players sustaining a right hamstring strain during the season had a significantly lower mean right SLHB score ( p=0.029), were older ( p=0.002) and more likely to have sustained a past right hamstring injury (p=0.02) Good to strong inter-test reliability was reported (ICC range = ; SEM% range = %). A minimal detectable change of % was reported to accurately determine a clinical outcome following an intervention. 85

86 2.5.4 Assessment of leg dominance (asymmetry) To confirm the presence of asymmetry, single limb tasks are preferred to bilateral variations due to their enhanced ability and sensitivity for determining deficits in neuromuscular control (Myer et al., 2011a). Also, a variety of assessments may be warranted due to different tasks demands (vertical vs. horizontal) and increased sensitivity (Barber et al., 1990). Using a range of single leg jumping tests in previously injured ACL patients, 60% of participants demonstrated abnormal levels of asymmetry (>15%) on at least one out of two tests. However, when only one test was evaluated, the range of abnormality decreased to 42-50% depending on the test used (Barber et al., 1990). A discrepancy in excess of 15% has been deemed a key predictor of injury in adults (Crosier and Crielaard, 2000). Conversely, interlimb asymmetries in non-injured youth athletes measured during sprinting have been reported to range from 15-20% (Rumpf et al., 2014). This indicates that greater movement variability is evident in youth populations, which is in accordance with previous paediatric motor control literature (Gerodomis et al., 2008). Furthermore, assessments of leg power across 3 directions (vertical, horizontal, and lateral) have reported insignificant relationships between tests in the various movement planes (Maulder and Cronin, 2005; Meylan et al., 2009; Hewit et al., 2012), thus, utilising a range of assessments targeting multi-planar actions is recommended. Commonly used single leg hop tests have reported strong reliability (Bandy et al., 1994; Paterno and Greenberger, 1996; Bolgla and Kesula, 1997; Brosky et al., 1999; Ross et al., 2002; Maulder and Cronin, 2005; Meylan et al., 2009; Munro et al., 2011) (ICC range = ), including single leg vertical jump, single hop for distance, triple hop for distance, crossover hop for distance, and a six-meter hop for time. Of all the horizontal hop tests, standard error of measurement is consistently lowest in the single hop for distance (Ross et al., 2002; Reid et al., 2007; Munro et al., 2011) but the repeated hopping tests may display greater ecological validity for soccer players. For example, the triple hop comprises a 86

87 deceleration component followed by the application of concentric force and use of the stretch-shortening cycle (SSC). The ability to attenuate force during a single limb stance and subsequently regenerate and direct motion may be a key factor for reducing injury risk (Lephart et al., 2002). This test has also been established as a strong predictor of vertical jump height (R² = 69.5%) and isokinetic measures of hamstring and quadriceps peak torque (R² = 49% %) (Hamilton et al., 2008). However, a rebound task performed in a unilateral stance is highly demanding with significant eccentric loading present and may not be suitable for youth subjects with limited exposure to plyometric training. The single hop for distance may be more appropriate as part of an initial screen in younger athletes, and once subjects have developed an appropriate training age and requisite technical competency, the triple hop can be introduced. The single hop for distance has also been used recently to identify young students who possess a greater risk of hamstring injury (Goosens et al., 2015), a frequently occurring injury in male youth soccer (Price et al., 2004). In addition to horizontal jumping, single leg countermovement jumps (SLCMJ) should also be considered due to the frequency of such tasks during game play and previously reported strong reliability in youth athletes for peak force and peak power variables (ICC range = ) (Ceroni et al., 2012). The validity and sensitivity of this measure have also been confirmed when analysing a range of single leg jumping assessments, with authors reporting that the SLCMJ was the only test to identify an asymmetry >15% (Petschnig et al., 1998). Also, statistically significant differences for peak force and peak power were demonstrated on the dominant leg (Petschnig et al., 1998), which previous literature has suggested may be indicative of an increased risk of soccer injury in male players (Brophy et al., 2010). In-spite of the above literature indicating suitable validity and reliability of single leg jumping assessments, less information is readily available to confirm the relationship between asymmetrical landing forces and injury risk. Previous suggestions have indicated 87

88 that normalised landing forces should be < 3x body weight (Myer et al., 2006b). However, this requires further investigation in male youth soccer players due to the lack of available data. Within the existing literature, leg dominance has also been identified in male youth soccer players using alternative tasks including an overhead squat screen (Atkins et al., in press) and range of motion assessments (Daneshjoo et al., 2013) but reliability data has not been consistently reported. In high school basketball players, asymmetrical reach scores > 4cm in the anterior reach direction of the Y-balance test have identified athletes at a 2.5 times greater risk of injury (Plisky et al., 2006). Further research is required to examine the withinsubject variation of selected test measures and their relationship with injury risk in male youth players. An overview of the available research using assessments to measure leg dominance in paediatric populations is summarised in table

89 Table 2.8 Assessments of leg dominance in male youth athletes Ref Subjects Assessment Summary of findings Atkins et al. (in 74 Youth males, Overhead Deep Squat on a twin press) assigned to Force Plate system measuring performance groups peak ground reaction force according to (PGRF) chronological age (Under 13-17). Ceroni et al. (2012) Plisky et al. (2006) Reid et al. (2007) Daneshjoo et al. (2013) Noyes et al. (1991) Youth males (n=117 age ± 1.93) & females (n=106 age ± 1.87) 235 high school basketball players 42 patients aged 15 to 45 years of age, who had undergone ACL reconstruction 36 male professional soccer players (age 18.9 ± 1.4 years 40 male and 27 female recreational subjects (age range 16-48) with a history of ACL injury Single Leg Vertical Jump on a force plate without arm swing measuring peak vertical force (PVF) & power (PW) Pre-season Star Excursion Balance Test measures and daily injury report to document time loss injuries over a season 6m timed hop, single leg hop, cross over hop and triple hop for distance 16 weeks after surgery and a further follow up session 6 weeks later Biodex Isokinetic Dynamometer measures of Peak Torque (PT) for the hamstrings and quadriceps KT-1000 arthrometer and biodex lower 60 &300 d/s and 4 hop tests single leg hop for distance, timed hop, triple hop for distance, crossover hop for distance. Significant differences (p 0.05) were identified between right and left side PGRF for all groups except the youngest (U13) and oldest (U17). Non-dominant 'sides' showed the highest levels of PGRF across all groups. The magnitude of PGRF was not significantly different both within and between groups, except for the left side in the U13 to U15 groups (p = 0.04). ICC of test measures (range = ) with 20 30% showing a difference of >15% between limbs. Between group asymmetry differences (>15%) were evident: females PVF = 25.5%; PW = 32.7%; Males PVF = 21.4%; PW = 21.4%. Statistically significant differences for peak force and power on the dominant leg were reported in boys only. Logistic regression models indicated that players with an anterior right/left reach distance difference >4 cm were 2.5 times more likely to sustain a lower extremity injury (P<.05). Girls with a composite reach distance <94.0% of their limb length were 6.5 times more likely to have a lower extremity injury (P<.05). ICC of all hop tests (range = ), SEM ( %), MDC ( %). Statistically greater changes in hop scores were reported on the operative vs. the nonoperative leg. PT of both hams & quads in the non dominant leg at all angular velocities showed nonsignificant higher tendencies than the dominant leg. Asymmetry deficits were abnormal (>10%) at all angular velocities, with 97.2% reported to have at least one musculoskeletal abnormality >10%. Also flexibility in the non dominant leg was lower than the dominant leg. 50% of subjects had limb asymmetry >15% on one of the single hop tests. When the results of two hop tests were combined, number of subjects with asymmetry >15% increased to 62%. Statistical trends were also noted between limb asymmetry on the hop tests and low velocity quadriceps isokinetic test results but not fast velocity. 89

90 2.5.5 Assessment of ligament dominance (knee valgus) While the gold standard for kinematic assessment for knee valgus is via three-dimensional motion analysis, this approach requires specialised equipment and labour intensive data collection, thus limiting the ability to effectively utilise this technique when screening large numbers of athletes. Subsequently, alternative time-efficient and non-invasive clinic-based methods have been proposed (Padua et al., 2009; Myer et al., 2008a; Myer et al., 2011). These measures use two-dimensional (2D) video analysis, which have shown good reliability (Munro et al., 2012), correlating significantly with more sophisticated laboratory techniques (Padua et al., 2009; Myer et al., 2010; Myer et al., 2011b). The most prominent methods from the available literature have been summarised below and an overview of assessments to measure ligament dominance in paediatric populations is provided in table Clinic based landing assessment tool A recently developed nomogram has been proposed using 5 laboratory-based measurements, contributing to 78% of the explained variance in knee abduction moments (Myer et al., 2011b). The measures were derived from the landing phase of a drop vertical jump using three-dimensional motion analysis and strength diagnostics using an isokinetic dynamometer. Variables within the nomogram include: 1) knee abduction angle; 2) relative quadriceps recruitment; 3) knee flexion range of motion; 4) tibia length; and 5) mass normalized to body height that accompanies growth. A subsequent logistic regression model was applied predicting high knee abduction status in adolescent female athletes with 85% sensitivity and 93% specificity (Myer et al., 2010). However, the authors stated that this technique requires expensive equipment and labour intensive measurements, limiting the potential to screen large numbers of athletes simultaneously (Myer et al., 2010). Hence, the development of a 90

91 validated assessment tool which is simpler to perform and can be administered in a clinic environment was recommended (Myer et al., 2011) (see figure 2.3). Figure 2. Clinic based nomogram to predict high knee abduction loads. Reprinted with permission from Myer et al. (2011b) This clinician friendly method requires two standard video cameras positioned in the saggital and frontal planes and has reported moderate-high agreement between the validated nomogram and the clinic based tool (ICC range ) (Myer et al., 2010; Myer et al., 2011b). Actual versus modelled predictions of high knee abduction moments were also reported with 79% sensitivity and 83% specificity, and 77% sensitivity and 71% specificity for the laboratory and clinic based assessments respectively. However, this method was validated using female subjects and no data is available for male youth soccer players. A 91

92 criteria within the nomogram suggests the use of isokinetic measures of knee extension/flexion (concentric/concentric) H/Q ratios and if this equipment is not available, a surrogate measure of Quad:Ham ratio based on the athletes body mass can be used (Myer et al., 2010). Caution should be applied using this approach with male youth soccer players, as while this increases efficiency the use of the surrogate calculation with males, and in particular male youth during different stages of growth and maturation may not be suitable. Furthermore, the conventional H/Q ratio (CON:CON) may not be reflective of true knee joint movement that only allows eccentric muscle action to be combined with concentric quadriceps muscle action during extension and flexion respectively (Aagaard et al., 1998). Therefore, further research is required in order to validate an appropriate surrogate measure for use in male youth. In addition, the use of the functional H/Q ratio (ECC Ham: CON Quad) may be more ecologically valid (Aagaard et al., 1998; Coombs and Garbutt, 2002) Landing error scoring system (L.E.S.S) The L.E.S.S is a clinical assessment tool of an individual s jump-landing biomechanics using 2D analysis with cameras positioned in the frontal and sagittal planes. This method was validated against three-dimensional motion analysis and force plate diagnostics (Padua et al., 2009). Subjects were required to perform a drop jump task from a 30 cm box landing on a force plate 50% of the maximum vertical jump distance away. The L.E.S.S score is determined using a count of technique errors based on a standardised checklist (see figure 2.4), which is calculated retrospectively. Higher scores are indicative of poor landing mechanics and the authors reported that a score > 6 was poor and < 4 was excellent. 92

93 Figure 2.4 L.E.S.S real-time scoring criteria, adapted from Padua et al. (2011) In adult subjects, inter-rater and intra-rater reliability of this method has been confirmed as strong to very strong respectively (ICC = 0.84; SEM = 0.71; ICC = 0.91; SEM = 0.42) (Padua et al., 2009). Statistically significant relationships have also been reported between expert raters and three-dimensional motion analysis of knee valgus range of motion (p <.05) (Onate et al., 2011). Cumulatively, the L.E.S.S can be considered a valid and reliable tool to identify subjects with altered landing mechanics reflective of high injury risk. However, no data are available in male youth athletes and relationships between raters on some of the 93

94 marking criteria (symmetrical landing and lateral trunk flexion) were deemed poor (range 10-42%) (Onate et al., 2011). Also, the use of the aforementioned scoring classification system (i.e. < 4 = excellent, versus > 6 = poor) in clinical settings may not be appropriate for all male youth soccer players as their results were based on quartiles from military participants, including both male and female adults (Padua et al., 2009). Thus, determining normative data for different groups including male youth soccer players is required for accurate interpretation Box height considerations for drop vertical jumping tasks In selecting which box height to use during a drop vertical jump assessment with youth subjects, practitioners must consider what an appropriate drop height is for their athletes due to high variation in their stage of growth and maturation. Within the available literature, the most commonly selected box heights are approximately 30 cm (Noyes et al., 2005; Barber- Westin et al., 2006; Quatman et al., 2006; Padua et al., 2009; Schmitz et al., 2009; Ford et al., 2010; Myer et al., 2011b). An understanding of the kinetics and kinematics (namely; muscle activation, ground reaction forces, ground contact times and knee flexion angles) during landing and take-off from different heights is important to establish in the assessment and training process. In a recent investigation, Peng et al. (2011) reported a greater magnitude of ground reaction forces with increases in box height. However, activation of the biceps femoris remained largely unchanged irrespective of drop height during both the landing and take-off phases. Conversely, greater drop heights increased quadriceps activation, with the highest values reported from the uppermost condition (60cm box). Conversely, Ruan and Li (2010) showed that increases in drop height (progressing from a 15-60cm box) did not augment 94

95 greater muscle activation of the rectus femoris or biceps femoris during landing. However, it should be considered that Ruan and Li (2010) included a week of pre-conditioning prior to testing and muscle activation sequences following a period of training may elicit different responses. Due to the paucity of available literature in youth populations and inconsistencies reported from the data in adult populations, intuitively practitioners may wish to standardise the box height at 30 cm to allow comparisons with previous research. Different heights may provide either insufficient or excessive forces from which to elicit an appropriate response and this may be magnified when working with youth athletes. Alternatively, analysis of the second landing could be performed, providing a height reflective of their individual neuromuscular capabilities and a more pertubated landing position (Bates et al., 2013). Kinetic and kinematic measures during the first and second landing of a drop vertical jump from a 31 cm box have recently been examined in adolescent female basketball players (Bates et al., 2013). No significant differences were shown in peak vertical ground reaction forces between landings (p = 0.445) but greater asymmetry was present in the second landing and this was combined with a higher centre of mass position. The authors suggested that these factors are more reflective of sporting activities and heightened injury risk (Bates et al., 2013). However, the altered kinetics and kinematics displayed may be due to the constraints of the task, whereby, athletes are required to land and stick on a force plate following the vertical elevation of the initial drop jump Type of jumping task In addition to the inherent difficulty associated with selecting an appropriate box height during a DVJ, practitioners should also consider the ecological validity of such assessments. 95

96 An alternative approach involves the measurement of jump-landing variables during traditional vertical jumps (Lazaradis et al., 2010) or repeated jumping tasks such as the aforementioned tuck jump assessment (Myer et al., 2008a). This movement pattern could be considered more innate and may therefore require less familiarisation for young athletes. Also, during drop and stop, or drop and go landing tasks (akin to the DVJ), impact forces at the point of ground contact are controlled eccentrically by muscles that were previously at rest. Conversely, in more functional tasks that reflect the athlete s neuromuscular ability, such as countermovement or repeated jumping tests, landing heights are equivalent to those regularly demonstrated in match play and forces are controlled via a preceded shortening of the involved musculature which are required to perform propulsive motions (i.e. the initial jumping action). This type of assessment may better represent the ability of the neuromuscular system to provide adequate stabilisation. It could be inferred that drop jumping tasks may artificially induce feed-forward stabilisation mechanisms which is a learnt skill, developed through maturation (Lazaridis et al., 2010). Also, the pre-planned nature of these assessments that do not require a stimulus-response component, are not reflective of the open environments in which many sporting situations and subsequent injuries occur. For example, actions such as rapidly changing direction or landing and jumping in response to movements from an opponent are characterised by perturbations to the body s centre of mass and stress perception and decision making skills (Young and Farrow, 2006). These situations do not allow sufficient time for the neuromuscular system to adequately adjust landing postures, which in turn increases landing forces and compromises the integrity of joints and soft tissue (Besier et al., 2001). However, standardisation of tasks in which athletes are reacting to external stimuli such as a light system does not truly replicate the player s ability to react to the postural cues or the kinematics of an opponent, subsequently reducing ecological validity (Abernethy et al., 1998). Also, the use of other players in the testing 96

97 environment to mirror the role of an opponent will subsequently increase human error and affect the reliability of the measures used. Hence, based on the current evidence, pre-planned assessments of landing mechanics are suggested. Specifically, the repeated nature of the tuck jump assessment discussed below (Myer et al., 2008a) provides some inherent perturbation and may more accurately reflect the movement demands and high risk mechanics involved in competition Tuck jump assessment The repeated tuck jump assessment has been proposed to identify neuromuscular deficits and plyometric technique flaws indicative of high injury risk (Myer et al., 2008a; Myer et al., 2013). Performance on this test has been suggested to provide an indication of the following risk factors: quadriceps dominance; ligament dominance; leg dominance; and trunk dominance (Myer et al., 2008b; Myer et al., 2011a). Repeated tuck jumps are performed in place for a period of 10 seconds and subjects are assessed using a ten point rating scale (see figure 2.5) with a greater a number of deficits indicating increased injury risk (Myer et al., 2008a). To increase accuracy, 2D video cameras can be used to capture the test and grade each player s technique retrospectively. 97

98 Figure 2.5 Tuck jump screening criteria. Reprinted, by permission, from G.D. Myer, K.R. Ford, and T.E. Hewett, 2008, Tuck jump assessment for reducing anterior cruciate ligament injury risk, Athletic Therapy Today 13(5): Human Kinetics, Inc. This assessment has been used previously to quantify the effectiveness of in-season neuromuscular training as opposed to a control group that only followed a soccer training programmeme (Klugman et al., 2011). Also, more recently tuck jump performance was measured before and after task specific feedback interventions (Stroube et al., 2013; Myer et al., 2013). No data are currently available in boys that determine the validity and reliability of the tuck jump assessment. In adults, initial pilot studies have indicated moderate-strong interrater reliability (ICC = ) (Myer et al., 2008). More recently, intra and inter-tester reliability have been reported with average percentage of exact agreement of 93% (kappa 98

99 measurement (k = 0.88)) (Herrington et al., 2012). However, no data are available to confirm within-subject variation and this requires further investigation. 99

100 Table 2.9 Assessments of knee valgus in male youth athletes Reference Subjects Measurement Tool Summary of findings Paterno et al. (2010) Quatman et al. (2005) Ford et al. (2003) Noyes et al. (2005) 56 subjects (n =35 female: 25 male; age ± 2.97) 5 pubertal and prepubertal subjects. (No specific data given on these subjects) 81 High school subjects (males age, 16 ± 0.2; Females age, 16 ± 0.2) 325 females (age; 14.1 ± 1.7; range years) and 130 male athletes (age; 14.6 ± 2.0; range years) Drop Vertical Jump (3DMA) and force plate measures of ground reaction force; Postural Stability using a Biodex Balance system and Anterior/Posterior Knee Laxity using a CompuKT Drop Vertical Jump (3DMA) and vertical ground reaction force (VGRF) using 2 force plates. Drop Vertical Jump (3DMA) Integrated landing prediction model for ACL injury reported high sensitivity (0.92) and specificity (0.88). Subjects who sustained a second ACL injury had altered landing mechanics and deficits in postural stability. Reliability statistics for repeated measures across three sessions included: maximum VGRF at landing (ICC = 0.89), maximum VGRF at takeoff (ICC = 0.98) and maximum vertical jump height (ICC= 0.98) Strong reliability was reported for both knee separation distance at maximum valgus angle (ICC = 0.92) and the difference between knee valgus angle at initial contact and maximum valgus angle (ICC = 0.84). Drop Vertical Jump 2D Analysis Test-retest reliability for hip separation distance was strong (ICC = 0.96 preland; land, 0.94; takeoff, 0.94). Following a 3 days p/wk. 6 week neuromuscular training programmeme. 100

101 2.5.6 Assessment of trunk dominance The assessment of core proprioception has commonly involved the use of specialised equipment to isolate motion of the lumbar spine (Zazulak et al., 2007a; Zazulak et al., 2007b). One such example utilised a seated apparatus, where subjects were tested both passively (with the seat driven by a step motor) and actively where the subjects rotated themselves 20 away from the neutral position (Zazulak et al., 2007a). In both tests, participants were required to stop the apparatus when they perceived they had returned to the neutral position. Moderate reliability was reported for this test (ICC = ) for active and passive repositioning respectively. A modified version of this assessment has also been utilised, in which multi-directional forces applied through a cable pulley system are resisted during testing (Zazulak et al., 2007b). No reliability data was reported for this assessment, although it was identified that trunk displacement was greater in athletes with knee injuries than un-injured athletes (p < 0.05). Lateral trunk displacement was also the strongest predictor of ligament injury (p < 0.05), however; these measures were derived during artificial conditions and postures in which the pelvis is immobilised, thus reducing ecological validity. Furthermore, highly specialised and costly equipment is required, limiting their application to larger scale youth athlete screening programmemes. Field based measures to assess core stability have also been devised due to their ease of administration and cost effectiveness, but currently limited data is available that reports their reliability in male youth athletes. In adults, a number of core exercises and standing based tasks including; prone bridge, single leg squat, and lateral step down have shown poor intraobserver reliability (ICC range ) (Weir et al., 2010). Trunk muscle endurance assessments such as isometric holds in a variety of positions have displayed stronger reliability (ICC range ) (McGill et al., 1999); however, the content validity of such measures may be questioned based on their prolonged isometric actions and non- 101

102 functionality. This is confounded by reports of weak to moderate relationships (ICC range = ) between performance on the aforementioned core tests and a range of athletic measures (Nesser et al., 2008). In addition, Leetun et al. (2004) used a modification of this protocol with additional measures of hip abduction and external rotation strength. Regression analysis demonstrated that hip external rotation strength was the only predictor of injury status (OR = 0.86, 95% CI = 0.77, 0.097), therefore, the assessment of isolated measures of core stability to infer lower limb injury risk and performance measures provokes questionable validity. Alternatively, movement abnormalities such as trunk dominance and a loss of core control may be detectable using more dynamic approaches, for example during the tuck jump assessment (Myer et al., 2008a) or the L.E.S.S test (Padua et al., 2009) Assessments of dynamic stability Within the available literature, the majority of investigations that have examined balance abilities in youth populations have predominantly utilised static tasks (Riach and Starkes, 1994; Kirsenbaum et al., 2001; Nolan et al., 2005; Steindl et al., 2006; Cumberwoth et al., 2007; Bieć and Kuczyński, 2010; Pau et al., 2014). Static balance postures are not reflective of the dynamic nature of soccer activities during which injuries occur. This is supported by previous data that identified weak relationships between static and dynamic tasks used to assess balance performance in male youth soccer players (Pau et al. in press). Thus, assessment of dynamic balance and stability should be comprised of more functionally relevant tasks indicative of the dynamic actions regularly involved in soccer training and competition. Two common methods are time to stabilisation (Ross et al., 2008; Ross et al., 2009; Ebben et al., 2010) and the Y-balance assessment (Plisky et al., 2006). 102

103 Time to stabilisation Measurement of time to stabilisation (TTS) involves the use of a force plate to quantify the speed in which individuals stabilise after a jump and/or landing task (Ross et al., 2008; Ebben et al., 2010). Although both drop jumps (Flanagen, Ebben and Jensen, 2008) and single leg drop landings (DiStefano et al., 2010) have been used, the most common form of assessment is a horizontal single leg jump task (Myer et al., 2006b; Ross and Guskiewicz, 2008; Gribble et al., 2012). Single leg landing assessments may be more ecologically valid due to nature of the movements in soccer competitions and are also indicative of greater injury risk (Lephart et al., 2002: Sugiomoto et al., 2015). A comparison of double versus single leg landing tasks in adult males identified that single leg energy dissipation reduced in the sagittal plane but increased in the frontal plane (Yeow et al., 2010). The sagittal plane is the primary motion of the knee joint and its associated musculature. In the frontal plane, range of motion and knee joint muscular stability is limited which increases the loading of the passive structures. Therefore, assessing single leg landing kinetics may be a more appropriate measure of injury risk. Two prominent methods of analysis have been applied to quantify TTS. The first involves scanning the components of ground reaction force from the last two windows of the last 10 seconds during a 20 second static hold following landing (Ross and Guskiewicz, 2008). The window with the smallest ground reaction force range is accepted as the optimal range variation (Ross and Guskiewicz, 2008). The data is then rectified and from the moment of peak ground reaction force an unbounded third order polynomial is fitted, with the TTS determined as the point in which this polynomial transects the horizontal range variation line (Ross and Guskiewicz, 2008) (see figure 2.6). The second method quantifies the time taken for an athlete upon landing to reach and stabilise within a GRF range representative of 5% of the athlete s bodyweight for a period of one second (see figure 5) (Flanagen et al, 2008; 103

104 Ebben et al, 2010). For younger athletes, the requirement to spend long periods standing still on the force plate, as in the method proposed by Ross and Guskiewicz (2008), will likely demonstrate greater postural sway, thus affecting the ground reaction force range. Consequently, the method of Flanagen et al. (2008) may be more suitable for younger populations. Also, as a measure to detect injury risk, the requirement to maintain stability for a period of one-second is likely more reflective of match play situations, as opposed to longer periods of static stability. Furthermore, the shorter recording period (7 seconds as used by Flangen et al., 2008) has implications for testing a large number of athletes, particularly youth athletes who may demonstrate lower levels of concentration. Figure 2.6 Third-order polynomial anterior-posterior ground-reaction-force time to stabilisation, adapted from Brown et al. (2004) 104

105 Figure 2.7 Time to stabilisation example of vertical ground reaction force during a countermovement jump, adapted from Ebben et al. (2010). A useful feature of this assessment is that it involves both vertical and horizontal displacement, and stabilisation mechanisms inherent to soccer, such as landing from a jump in a single leg stance (Brown et al., 2004). However, controlling for jump height and distance is an important factor due to the variation in acceleration upon landing. Previous recommendations have been established which involve normalising jump height to 50% of the individual s maximum jump height, with horizontal displacement set at an arbitrary figure of 70 cm (Ross and Guskiwicz, 2008; Ross et al., 2009). Alternatively, hop distance can be normalised to leg length, with recent data indicating significantly longer TTS (p < 0.001) in subjects using this approach in comparison to the predetermined 70 cm protocol (Gribble et al., 2012). The authors subsequently suggested that researchers and clinicians should apply this normalisation procedure to establish a more accurate representation of an athlete s stabilisation capabilities. However, the utilisation of anthropometric measures to determine 105

106 jump distances might subsequently over- or under-estimate an individual s performance. During a maximal single leg hopping task, the same athlete may be capable of much greater jump distances than that of their leg length. Such feats of athleticism are likely to be replicated under the conditions of competitive match play. Therefore, an individual s inherent risk of injury is likely a product of how far they can jump and how well they can attenuate the resultant ground reaction forces on landing; in the standardization of single leg hop stabilisation tests, a more appropriate method may be to utilise a percentage of maximal hop performance to represent their neuromuscular capabilities. The assessment of TTS requires measurement of either ground reaction forces or centre of pressure. Early investigations showed that ground reaction force measures were more sensitive than centre of pressure for identifying stability deficits indicating this method may be more appropriate for accurately screening athletes (Goldie et al., 1989). More recently, ground reaction force profiles determined more accurately the difference between healthy controls and those with a history of ankle injury (Ross et al., 2009). Strong reliability data has also been reported for this approach (ICC = ), whereas, lower values were reported for centre of pressure (ICC = ) (Colby et al., 1999). Furthermore, a single leg hopand-hold task has also shown strong within-session reliability for both dominant (r = 0.82) and non-dominant (r = 0.87) limbs (Myer et al., 2006a). The authors also stated the ground reaction force should be normalised to body mass, with values of < 3 times body mass and a side-side difference of <10% indicative of reduced injury risk (Myer et al., 2006a). This measure has also been used with male youth athletes demonstrating significant reductions in TTS following an injury prevention programmeme, indicating that youth subjects can improve their balance ability following targeted interventions (DiStefano et al., 2010). 106

107 Star excursion or y-balance test Another unilateral balance task used to assess dynamic stability is the Star Excursion Balance test (Plisky et al., 2006; Impellizzeri et al., 2013; Hale et al., 2007). The original version of this test required athletes to stabilise in a unilateral stance and reach in eight specified directions with their opposite limb. The test is graded by marking the reach distance achieved in each direction with scores normalised to leg length. This test has been used as an injury predictor in male youth basketball players, where subjects who recorded an anterior right-left reach difference > 4 cm displayed a 2.5 times greater risk of lower extremity injury (Plisky et al., 2006). Furthermore, in the female group, subjects with a composite reach distance < 94% of their limb length were 6.5 times more likely to sustain a lower extremity injury (Plisky et al., 2006). More recently, a modified version of this assessment has been proposed, namely the Y-Balance test, which only requires athletes to reach in 3 directions: anterior, posteriormedial and posterolateral (Plisky et al., 2009). The posteromedial reach direction has shown equivalent accuracy to all eight reach directions in its ability to identify subjects with chronic ankle instability (Hertel et al., 2006). Significant correlations have also been reported between both posteromedial and posterolateral reach distances and hip abduction and extension strength respectively (Hubbard et al., 2007). In adults, the reliability of the star excursion balance test has been reported previously (Kinzey and Armstrong, 1998; Plisky et al., 2006; Munro and Herrington, 2010). Early investigations demonstrated moderate to strong values (ICC range ) (Kinzey and Armstrong, 1998). The authors suggested that task complexity was responsible for the moderate values, highlighting the need for adequate familiarization. More recent reports confirmed that excursion distances stabilized after 4 trials (Munro and Herrington, 2010), reporting higher levels of reliability with the greater familiarisation provided (ICC range = ; SEM = %, smallest detectable differences = %). To ensure 107

108 time-efficiency in screening a large number of youth athletes, this approach has been modified with practice trials performed in a group setting away from the instrumented device, with an additional practice trial conducted on the Y-Balance kit (Faigenbaum et al., 2014). Moderate to strong reliability was reported in school children of different ages (ICC = ) (Faigenbaum et al., 2014). In youth soccer academies where a large number of athletes must be screened, the prioritisation and use of the anterior reach direction may be more appropriate to detect athletes who demonstrate asymmetrical reach distances and subsequently display a heightened risk of injury (Plisky et al., 2006). Cumulatively, these findings suggest that the Y-balance test may be a reliable and sensitive protocol, which is simple to administer and cost effective for the screening of youth athletes Summary of field-based assessments of neuromuscular control in male youth soccer players In this section of the literature review, overviews of a number of field-based assessments that may be used to screen neuromuscular control in male youth soccer players have been provided. Their suitability for use within the context of a soccer academy development programmeme has also been critically analysed. Due to the paucity of data available in male youth athletes, and in particular soccer players, further investigations are required to examine the reliability and validity of these assessments. The key findings from this review have been outlined below and a suggested evidence-based, test battery that may be used to effectively screen male youth soccer players has been suggested (table 4). In adults, field-based tests of neuromuscular control provide a reliable option for the assessment of injury risk, although limited data are available in male youth soccer players 108

109 Functional hopping tasks can be effectively utilised to screen male youth athletes, and performance should be measured in multiple planes of motion A range of valid and reliable jump-landing based assessments are available using 2 dimensional video analyses. However, data are needed to confirm their validity and reliability in youth male soccer players Measures of balance may predict lower extremity injury in male youth athletes and should include dynamic tasks due to greater ecological validity 109

110 2.6 Overall summary of the literature review Overall injury incidence in elite male youth soccer has ranged from injuries per 1,000h (Junge et al., 2000; Le Gall et al., 2006; Brink et al., 2010), with a player incidence rate of 0.40 injuries per player, per season, and a mean absence of 21.9 days missed per injury (Price et al., 2004). A linear increase in the number of injuries sustained has been observed with advancing age (Price et al., 2004), and these injuries predominantly occur in the lower extremities, with a large proportion classified as non-contact (Price et al., 2004; Le Gall et al., 2006; Rump and Cronin, 2012). The most frequent sites of injury are the upper thigh, knee and ankle (Price et al., 2004; Le Gall et al., 2006), with most major severe injuries occurring at the knee in the form of ligament sprains (Volpi et al., 2003). Therefore, targeted prevention strategies in this cohort should focus on the mechanisms that underpin these types of injury. Risk factors for lower extremity injury in male youth soccer players have recently been proposed (Read et al., 2015); however, a number of these factors are non-modifiable (Grffin et al., 2006; Read et al., 2015) and there is currently a paucity of data to examine the prevalence of more trainable risk factors in this cohort. Altered neuromuscular control has been proposed as a modifiable risk factor in adult and female populations (Williams et al., 2001; Hewett et al., 2005; Myer et al., 2011a), with specific neuromuscular imbalances suggested to increase injury risk (Myer et al., 2011a). Due to the paucity of available data in paediatric male youths, further research is required to investigate these risk factors in this cohort to more accurately identify athletes who may be at high risk (Read et al., in press). The assessment of neuromuscular control is commonly performed using a range of jump-landing tests (Barber et al., 1991; Myer et al., 2008a; Padua et al., 2009; Myer et al., 2011b; Goosens et al., 2015), utilising both kinematic analysis (Myer et al., 2008a; Padua et 110

111 al., 2009; Myer et al., 2011b), and force plate diagnostics (Myer et al., 2006a; Ross and Gusciewicz, 2008; Ebben et al., 2010; Ceroni et al., 2012), in addition to measures of dynamic balance (Plisky et al., 2006; Ross and Gusciewicz, 2008). Studies that investigate the validity and reliability of these screening modalities are required in elite male youth soccer players to aid practitioners in the accurate identification of deficiencies in the various constructs of neuromuscular control. This will further enhance the understanding and prevention of injuries in this target group (Schmikli et al., 2011; Alernton-Geli et al., 2014). Specifically, field-based techniques that are practically viable should be included due to their time-efficiency in screening large the numbers of athletes across the different chronological age groups in a professional soccer academy. 111

112 Chapter 3 STUDY 1: THE RELIABILITY OF FIELD-BASED MEASURES OF NEUROMUSCULAR CONTROL IN ELITE MALE YOUTH SOCCER PLAYERS Includes aspects of the following peer reviewed manuscripts: Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD and Lloyd, RS. Reliability of the tuck jump injury risk screening assessment in elite male youth soccer players. Journal of Strength and Conditioning Research. In Press. Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD and Lloyd, RS. Consistency of field-based measures of neuromuscular control using force plate diagnostics in elite male youth soccer players. Journal of strength and conditioning research. In Press. 3.1 Introduction The sport of soccer imposes high physiological demand and an inherent risk of injury due to frequent repetitions of high intensity movements that involve high internal forces (Daniel et al., 1994). Existing injury incidence data in elite male youth soccer has shown that injuries occur mainly in the lower extremities (71-80%) and are largely non-contact in nature, with a predominance of ligament sprains occurring at the ankle and knee (Price et al., 2004; Le Gall et al., 2006; Cloke et al., 2009; Cloke et al., 2011; Moore et al., 2011). Movements that lead to such injuries include: running, twisting, turning, over-stretching and landing (Price et al., 2004). A linear increase in injury rates has been reported from 9 to 16 years of age in male players (Price et al., 2004) and in particular around the time of peak height velocity (PHV) (van der Sluis et al., 2014). This increased injury incidence may be in part be due to rapid changes in stature and mass but is also associated with altered movement control strategies indicative of an 112

113 increased injury risk (Philapperts et al., 2006; Atkins et al., 2013; van der Sluis et al., 2014). Deficits in neuromuscular control and altered movement patterns are a suggested mechanism (Alternton et al., 2009; Sugio-Moto et al., in press) and a predictor of sport-related injury in male athletes (Gomes et al., 2008; Small et al. 2010). Further, reductions in fundamental movement skills of injured professional male athletes (Kiesel et al., 2007), lower scores on a single leg jump and balance assessment (Goosens et al., 2014) and greater leg asymmetry in dynamic balance tasks (Plisky et al., 2006) have been identified in male and female student athletes as positive predictors of injury. Therefore, it could be argued that the assessment of neuromuscular control is warranted and the ability to successfully perform movements safely in desirable patterns would be an appropriate means of evaluation. Specific imbalances have been identified in young female soccer players (Hewett et al., 2004; Myer et al., 2004; Myer et al., 2008); however limited data are available for males, and in particular male youth soccer players. In order to accurately assess neuromuscular control in male youth soccer players, there is a need for reliable and valid testing protocols. Assessments of neuromuscular control have been analysed previously using a variety of methods, including a predominance of jump-landing assessments using both force plate diagnostics and kinematic analysis (Hewett et al., 2005; Myer et al., 2006; Padua et al., 2009; Ross et al., 2009; Myer et al., 2010). These have been used less frequently in paediatric male subjects, however in adults, single leg power and asymmetry during unilateral horizontal hoping tasks has reported strong reliability (ICC range = ) (Ross et al., 2002; Reid et al., 2007; Munro et al., 2011). In addition to horizontal jumping, single leg vertical jumps should also be considered due to the frequency of such tasks during game play and previously reported strong reliability in youth athletes for variables of peak force and peak power (ICC range = ) (Ceroni et al., 2012). Measuring landing forces are valid because 113

114 high internal loads experienced in sporting actions may be injurious to the lower extremities (Powell and Barber-Foss, 2000). If impact forces on ground contact exceed the force production capabilities of the involved musculature, additional loading will be diverted to other soft tissues such as bones and ligaments heightening the risk of ligamentous injury (Hewett et al., 2010). A common task utilised to assess landing forces with youth athletes is a drop vertical jump (DJ) (Quatman et al., 2005; Larardis et al., 2010; Lazaridis et al., 2013). Reliability of DJ vertical ground reaction force landing measures have reported strong reliability (ICC range = ) in adolescent athletes (Quatman et al., 2005). A measure used frequently to assess dynamic stability is the Star Excursion Balance Test (SEBT) (Plisky et al., 2006; Hale et al., 2007; Impellizzeri et al., 2013), and more recently the Y- Balance test (Plisky et al., 2009; Faigenbaum et al., 2014). This is a modified version of the SEBT which only utilises three reach directions (anterior, posterior-medial and posterior-lateral) as opposed to eight. Early investigations reported moderate-strong reliability of this method (ICC range = ) (Kinzey and Armstrong, 1998). More recently, stronger levels of reliability have been reported with a greater number of familiarisation trials (ICC range = ; SEM = %) (Munro and Herrington, 2010). One available study in youth subjects also reported moderate to -strong reliability (ICC range = ) (Faigenbaum et al., 2014). However, these were school children and the findings may not be representative of more elite populations. Another method to quantify dynamic stability is the measurement of time to stabilisation (Ross and Guskiewicz, 2003; Ebben et al., 2010). Using force vector analysis, the speed in which individuals stabilise within a pre-determined range upon ground contact is quantified. Although a variety of jump-landing tasks have been evaluated (Ebben et al., 2010), the most common task measured is a single leg horizontal jump onto a force plate (Ross and 114

115 Guskiewicz, 2003; Brent et al., 2005; Myer et al., 2006; Gribble et al., 2012). Ground reaction force measures appear to be more reliable than centre of pressure values (ICC range = vs ) (Colby et al., 1999; Ross et al., 2009), and strong within-session reliability has been reported for both dominant (r= 0.82) and non-dominant (r = 0.87) limbs (Myer et al., 2006). In addition to quantifying kinetics during jump-landing tasks, the assessment of kinematic variables may also be of use in the identification of injury risk (Myer et al., 2010). The gold standard assessment of landing kinematics is 3 dimensional motion analysis but this requires specialist equipment and involves significant financial and time implications. A more practical method is the use of standard video cameras positioned in the saggital and frontal planes. Moderate to high agreement has been reported between these measures and 3 dimensional motion analysis (ICC range ) (Myer et al., 2010; (Myer et al., 2011). An example of a clinician friendly diagnostic tool to assess jump-landing kinematics is the tuck jump assessment (Myer et al., 2008). This test involves repeated tuck jumps performed for 10 seconds and is assessed using a ten point rating scale. Initial pilot studies have indicated moderate to strong inter-rater reliability (range = ) (Myer et al., 2008). More recently, intra and intertester reliability have been reported with strong agreement (kappa measurement (k = 0.88)) (Herrington et al., 2012). However, within-subject variation during the tuck jump assessment is currently unknown and this test has not been investigated in youth athletes. In-spite of this growing body of evidence in adult and female populations, such measures have been used less frequently in paediatric athletes and no data are available to confirm their reliability for male youth soccer players. This current paucity of literature does not permit accurate interpretation of results. Further, the predominance of ICC statistics to determine reliability in the absence of other means of analysis limits their usefulness (Hopkins, 2000). This 115

116 is due to reasons of bias in smaller groups, and sensitivity in samples used for estimation. For example, re-test correlations will be low in a group that is homogenous (Hopkins, 2000). Therefore, the primary aim of this study is to determine the reliability of a battery of field-based measures of neuromuscular control in elite male youth soccer players. 3.2 Methods Participants Twenty six pre-phv (age 11.9 ± 0.4 yr; height ± 4.8 cm; body mass 41.1 ± 5.6 kg; maturity offset -2.3 ± 0.4 yr) and twenty five post-phv (age 17.3 ± 0.7; height ± 5.5; body mass 72.3 ± 6.9 kg; maturity offset 2.9 ± 0.8 yr) youth soccer players from the academy of a professional English Championship soccer club volunteered to take part in the study. Participants completed a familiarisation session and two experimental test sessions separated by a period of 7 days. None of the players reported injuries at the time of testing and were all participating regularly in football training and competitions in accordance with the regulations for player contact hours as set out by the Premier League Elite Player Performance Plan (EPPP, 2011). Subjects were familiar with regular performance testing, however, were unfamiliar with a number of the tests included in the current study. Parental consent and participant assent were collected prior to the commencement of testing, in addition to a physical activity readiness questionnaire. Ethical approval was granted by the institutional ethics committee in accordance with the declaration of Helsinki. 116

117 3.2.2 Experimental design This study used a repeated measures design to determine the reliability of a range of field-based neuromuscular control assessments. Participants were required to attend the club training ground on three occasions separated by a period of seven days. The first session was used to familiarise subjects with the test equipment and assessment protocols. In the second and third sessions, data was collected for the reliability study to determine test-retest within subject variation. Six different assessment protocols were used, including: (1) single leg horizontal hop for distance; (2) y-balance test (anterior, posterior-medial, and posterior-lateral reach directions); (3) drop vertical jump onto two separate force plates with measures of peak landing vertical ground reaction force (pvgrf), time to pvgrf, and pvgrf asymmetry; (4) a ten-second repeated tuck jump assessment analysed using two-dimensional video analysis; (5) a single leg 75% horizontal hop and stick, and (6) a single leg vertical jump onto a force plate. All tests were conducted by the same researcher at each test session. Participant instructions were standardised to ensure replication of conditions between days (see appendix d) and removal of the potential for of interrater error. A 10-minute standardised warm up was completed prior to each session consisting of dynamic stretching. The order of testing was randomised using a counterbalanced design to reduce the potential for an order effect. Practice trials were provided for each test until the participants were judged to be competent by the principal researcher. For the purposes of data collection, and three trials were analysed to reduce the influence of a learning effect. One minute of recovery was allowed between trials based on previous recommendations (Ebben et al., 2010), and research which has demonstrated that rest periods greater than 15 seconds produce no further improvements in jump performance (Read and Cisar, 2001). Testing was completed at the same time on each day, and participants were asked to wear the same training kit and footwear, and 117

118 refrain from strenuous exercise at least 48 hours prior to testing. Subjects were also asked to eat according to their normal diet and avoid eating and drinking substances other than water one hour prior to each test session Procedures Anthropometry: Body mass (kg) was measured on a calibrated physician scale (Seca 786 Culta, Milan, Italy). Standing and sitting height (cm) were recorded on a measurement platform (Seca 274, Milan, Italy). Tibia length was also measured to the nearest 0.1cm as the distance between the lateral knee joint line and the lateral malleolus using a standard tape measure. Biological Maturity: Subjects were assessed for stage of biological maturation in a non-invasive manner utilising a regression equation comprising measures of age, body mass, standing height and sitting height (Mirwald, 2002). Using this method, calculation of years from PHV was completed (equation 1). The equation has been used previously to predict maturational status in paediatric research (Lloyd et al., 2009; Lloyd et al., 2011; Lloyd et al., 2012; Lloyd et al., 2014) with a standard error of approximately 6 months. Age at PHV has been shown to occur around 14 (Malina et al., 2004) and 13 years (Nariyama et al., 2001) in average maturing males and male youth soccer players respectively. Therefore, the group classifications used in this study were deemed appropriate to differentiate between participants who were pre and post PHV. 118

119 Maturity offset = [ x leg length and sitting height interaction] [ x age and leg length interaction] + [ x age and sitting height interaction] + [ x weight by height ratio] [equation 1] Y-Balance: Participants placed their hands on their hips and began in a unilateral stance with the most distal aspect of their great toe behind the line on the centre of the Y-Balance test kit (Move2Perform, Evansville, IN). The commercially available device is comprised of a central stance platform with three pieces of polyvinylchloride piping attached displaying incremental measurements of 0.5cm complete with a reach indicator for each pipe. Following four observed practice trials completed in a group format, distances were then recorded by pushing the target reach indicator in the specified direction (either anterior, posterior-medial, or posterior-lateral). Four practice trials were provided to reduce learning effects as it has been reported previously that excursion distances stabilize after four trials (Munro and Herrington, 2010). Due to the large subject numbers and to ensure time-efficiency of testing, three of the specified practice trials were performed in a group setting away from the instrumented device as has been suggested previously (Faigenbaum et al., 2014), with an additional practice trial conducted on the Y- Balance kit. Trials were performed on both legs with the order of testing counterbalanced. 119

120 Throughout, subjects were required to keep the heel of the non-reach leg on the testing platform, maintain balance in a single leg stance, and return the reach foot back to the start prior to attempting the next direction. Also, no visible kicking of the target reach indicator was permitted. Maximal reach distances in each direction were recorded and normalised to leg length, and the composite reach score was calculated by dividing the sum of the maximum reach distance in the anterior (A), posteromedial (PM), and posterolateral (PL) directions by 3 times the limb length (LL) of the individual, then multiplied by 100 {[(A + PM + PL)/(LL 3)] 100}. Tuck jump assessment: Tuck jumps were performed in place for a period of 10 seconds and the technique of each participant were visually graded against a specified 10-point criterion (see chapter 2, figure 2.5) (Myer et al., 2008). Recorded deficits were marked and subsequently tallied to provide a composite score, with higher scores indicative of reduced performance. If a deficit was present on two or more occasions it was marked as previously stated (Myer et al., 2013). To increase accuracy, two-dimensional video cameras were used to capture the test with scores tallied retrospectively. Kinematic data were collected at 50Hz using two HD video cameras (Samsung, New Jersey, USA) positioned in the frontal and sagittal planes at a height 0.70 m, and a triangulated distance of five meters from the capture area. To allow visible tracking of the knees, subjects were required to wear shorts with a line at approximately midthigh. 120

121 Single leg hop for distance: Hop distances were recorded using a tape measure marked out for a length of three metres that was taped to the floor. Subjects began by standing on the designated test leg with their toe on the marked starting line, the hip of the free leg flexed at 90 to avoid contralateral propulsion, and their hands on their hips. When the subjects were ready, instructions were to hop forward as far as possible, landing on the same leg with the hands remaining on their hips throughout. For each test to be recorded, players had to stick the landing and hold for three seconds without any other body part touching the floor in accordance with previous guidelines (Goosens et al., 2015). The test was performed on both legs and the distance in line with the heel was recorded to the nearest 0.1 cm using a ruler stick to increase accuracy of the measurement. Single leg 75% horizontal hop and stick (75%Hop): This test set up involved a tape measure marked out to a three metre distance which was taped to the floor on a horizontal line with the 0 cm mark positioned in line with the centre of a force plate (Pasco, Roseville, California, USA). Participants began by standing in line with the force plate on the designated test leg, hands on their hips, and toe in line with a distance marker on the tape measure representing 75% of their predetermined maximal single leg hop and stick performance. When the subjects were ready, instructions were to hop forward onto the force plate, landing on the same leg with the hands remaining on their hips throughout. For each test to be recorded, players had to stick the landing and hold for seven seconds, remaining as still as possible without any other body part touching the floor. The test was performed on both legs. 121

122 Single leg countermovement jump: Participants began standing on a force plate (Pasco, Roseville, California, USA) in a unilateral stance with their hands on their hips and the opposite hip flexed at 90 to ensure minimal contributions from the contralateral leg. Instructions were to jump as high as possible using a countermovement by dropping into a quarter squat and then immediately triple extending at the ankle, knee and hip in an explosive concentric action. On landing, subjects were required to stick the landing and hold for a period of seven seconds remaining as still as possible. For standardisation, bending of the knees whilst airborne was not permitted, and hands remained in contact with hips throughout the test. Only the landing variables of the jump were used for subsequent analysis. Drop vertical jump: Participants were positioned on top of a box at a height of 30cm with their feet 35cm apart. Instructions were to drop directly down with no vertical elevation onto two separate force plates (Pasco, Roseville, California, USA) positioned 8cm apart. Upon ground contact, players immediately performed a maximum vertical jump as if they were aiming to jump up for a header and then land on the plates and hold the landing which is in line with previous suggestions (Noyes et al., 2005; Myer et al., 2010). Hands were freely available to replicate a natural jump-landing position (Quatman et al., 2005). The use of two force plates as opposed to one is recommended to allow further analysis of leg dominance, a suggested risk factor for lower limb soft tissue injuries (Myer et al., 2011). Only the data from the first landing was used for subsequent analysis. 122

123 Force plate variables: Kinetic data captured from the force platform included: pvgrf recorded in the first 100ms following ground contact, time to pvgrf, and pvgrf asymmetry during the drop vertical jump tests. A cut-off point of 100ms was used to determine peak vertical ground reaction forces due to the reported timing of non-contact injuries which occur within the first 50ms following initial ground contact (Krosshaug et al., 2007). Forces experienced after this point are unlikely to contribute to acute injury risk and were therefore not included in the analysis. The single leg countermovement jump, and 75%Hop was analysed using the same variables as the drop vertical jump test, however, time-to-stabilisation (TTS) was also quantified from the vertical force vector. Vertical TTS was calculated as the time taken from ground contact to the first point when the vertical force component reached and stayed within 5% of body weight for a period of one second. The point of ground contact was then subtracted from this value in accordance with previous guidelines (Ebben et al., 2010). For the drop vertical jump and 75% horizontal hop and stick, initial contact was defined as the point when vertical ground reaction force first exceeded 10 N. In the single leg countermovement jump, the same criteria were used to determine initial contact following the preceding propulsive and time in air phases. For all tests pvgrf was normalised to body weight; however for the purposes of the reliability study the maximum values (N) were reported. All data was recorded at a sampling rate of 1000 Hz and filtered through a fourth-order Butterworth filter. A cut-off frequency of 18, 21, and 26 Hz was used for the single leg countermovement jump, drop vertical jump, and 75% horizontal hop and stick respectively. Asymmetry calculation: Limb asymmetry has been defined as an imbalance in strength, coordination and control between the two lower extremities (Myer et al., 2004), with a reported 123

124 discrepancy >15% an important injury predictor (Crosier and Creelard, 2000). Different classifications of limb dominance have been suggested within the available literature. For example, subjects who subsequently injured their anterior cruciate ligament (ACL) injured their preferred push off leg during a cutting manoeuvre (Zebis et al., 2009). Conversely, epidemiological data in elite male soccer players reported that 74.1% injured their dominant (kicking) leg. However, no studies are available in youth male soccer players. Whilst classifying the performance of the dominant versus non-dominant leg may provide useful information, accurately defining their dominant leg (i.e. kicking leg vs. push off leg) may be challenging for practitioners. Also, factors such as previous injury may result in neuromuscular inhibition (Bullock-Saxton et al., 1994; Friel et al., 2006; Opar et al., 2013), and subsequent performance reductions. Therefore, to quantify asymmetry and determine injury risk, a more appropriate method may be to calculate the percentage difference between the highest vs. lowest performing limb. The value obtained is expressed as the absolute percentage of performance achieved using the higher performing limb as the reference (see equation 2). Asymmetry % =ABS((lowest performing limb - highest performing limb) / highest performing limb * 100) % of Performance achieved = % Asymmetry ABS = Absolute [equation 2] Statistical analysis 124

125 The mean and standard deviations for each test were calculated across the two testing sessions. To determine systematic bias between trials, a series of paired samples t-tests were used with a p value 0.05 indicative of a significant difference between the two test trials. Within-subject variation was determined using mean coefficients of variation (CV %) with acceptable CV values ( 10%). Further reliability statistics also included: change in mean; intra-class correlation coefficient (ICC) to determine rank order repeatability; and typical error of estimates. 95% confidence intervals (95% CI) were used, and all reliability data were computed through Microsoft Excel 2010 using a commercially available spread sheet (Hopkins, 2006). Paired samples t-tests were processed using SPSS (V.21. Chicago Illinois). To determine the accuracy of observations across the two test sessions for each criterion in the tuck jump assessment, percentage of agreement was assessed using Kappa coefficient with threshold classifications: > 0.8 almost perfect agreement; substantial agreement; moderate agreement; slight to fair agreement (Landis and Koch, 1977). The cut-off point selected to determine unsatisfactory agreement was < 0.4 as it has been suggested previously that values below this imply that more variance is attributable to respondent proxy disagreement as opposed to an association with any respondent characteristic (Magaziner et al., 1996). Intra-rater reliability was calculated for the Tuck Jump total score, and Y-Balance test using intra-class correlation coefficient (ICC). 3.3 Results Descriptive statistics and all reliability measures calculated for each test are displayed in table 3.1 (pre-phv group), and table 3.2 (post-phv group). No significant differences were reported 125

126 for the majority of the test variables when the mean scores of the two test trials were analysed using a series of paired samples t-tests (p > 0.05). However, significant differences (p < 0.05) were present in the pre-phv group for the following measures: single leg hop for distance (p = 0.002); single leg countermovement jump time to stabilisation (p = 0.046); Y-Balance anterior reach (p = 0.048); Y-Balance posterior-lateral reach (p = 0.023). Similarly in the post-phv group, significant differences (p < 0.05) were present for the single leg hop for distance (p = < 0.001); Y-Balance composite reach (p = 0.010); Y-Balance anterior reach (p = < 0.001); and Y- Balance posterior-lateral reach (p = 0.050). Table 3. 1 Pre-PHV descriptive results and reliability statistics for all test measures Test Variable Mean Test 1 Mean Test 2 Change in mean Typical Error ICC CV% (95% CI) SLHD (m) 1.51 ± ± 0.15* 0.08 ± ( ) Y-B Comp (cm) ± ± ± ( ) Y-B Ant (cm) ± ± 5.49* ± ( ) Y-B P-M (cm) ± ± ± ( ) Y-B P-L (cm) ± ± 13.94* 3.49 ± ( ) TJ total score 6.88 ± ± ± ( ) 75%Hop pvgrf (N) 1642 ± ± ± ( ) 75%Hop TTS (s) ± ± ± ( ) 75%Hop time to PF (s) ± ± ± ( ) DVJ pvgrf (N) 890 ± ± ± ( ) DVJ time to PF (s) ± ± ± ( ) SLCMJ pvgrf (N) 1221 ± ± ± ( ) SLCMJ TTS (s) ± ± 0.508* ± ( ) SLCMJ time to PF (s) ± ± ± ( ) * Denotes significantly different (p < 0.05) to testing session 1 Note: SLHD = single leg hop for distance; TJ = Tuck Jump; pvgrf = peak vertical ground reaction force; TTS = time to stabilisation; DVJ = drop vertical jump; SLCMJ = single leg countermovement jump; P-M = postero-medial; P-L = postero-lateral; Comp = Composite; Ant = Anterior; Y-B = Y-balance 126

127 Table 3.2 Post-PHV group mean results and reliability statistics for all test measures Test Variable Mean Test 1 Mean Test 2 Change in mean Typical Error ICC CV% (95% CI) SLHD (m) 1.89 ± ± 0.16* 0.10 ± ( ) Y-B Comp (cm) ± ± 17.06* 3.34 ± ( ) Y-B Ant (cm) 74.9 ± ± 15.30* 5.84 ± ( ) Y-B P-M (cm) ± ± ± ( ) Y-B P-L (cm) ± ± 17.77* 3.4 ± ( ) TJ total score 4.70 ± ± ± ( ) 75%Hop pvgrf (N) 2825 ± ± ± ( ) 75%Hop TTS (s) ± ± ± ( ) 75%Hop time to PF (s) ± ± ± ( ) DVJ pvgrf (N) 1781 ± ± ± ( ) DVJ time to PF (s) ± ± ± ( ) SLCMJ pvgrf (N) 2369 ± ± ± ( ) SLCMJ TTS (s) ± ± ± ( ) SLCMJ time to PF (s) ± ± ± ( ) * Denotes significantly different (p < 0.05) to testing session 1 Note: SLHD = single leg hop for distance; TJ = Tuck Jump; pvgrf = peak vertical ground reaction force; TTS = time to stabilisation; DVJ = drop vertical jump; SLCMJ = single leg countermovement jump; P-M = postero-medial; P-L = postero-lateral; Comp = Composite; Ant = Anterior; Y-B = Y-balance Following the analysis of all variables, measures highlighted with acceptable CV values ( 10%) were then further investigated to determine the reliability of lower-limb asymmetry calculations (tables 3 and 4). In both groups, all measures reported acceptable CV values ( 10%) with the exception of drop vertical jump pvgrf and 75%Hop pvgrf in the pre-phv group, and 75%Hop pvgrf for the post-phv cohort. 127

128 Table 3.3 Pre-PHV group asymmetry mean values and reliability statistics expressed as % of performance achieved Test Variable Mean Test 1 Mean Test 2 Change in mean ICC Typical Error CV% (95% CI) SLHD ± ± ± ( ) Y-B Comp ± ± ± ( ) Y-B Ant ± ± ± ) DVJ Pvgrf ± ± ± ( ) SLCMJ pvgrf ± ± ± ( ) 75%Hop pvgrf ± ± ± ( ) *Denotes significantly different (P < 0.05) to testing session 1 Note: SLHD = single leg hop for distance; pvgrf = peak vertical ground reaction force; DVJ = drop vertical jump; SLCMJ = single leg countermovement jump; Comp = Composite; Ant = Anterior; Y-B = Y-balance; Table 3.4 Post-pubertal group asymmetry mean values and reliability statistics expressed as the % of performance achieved Test Variable Mean Test 1 Mean Test 2 Change in mean ICC Typical Error CV SLHD ± ± ± ( ) Y-B Comp ± ± ± ( ) Y-B Ant ± ± ± ( ) DVJ pvgrf ± ± ± ( ) SLCMJ pvgrf ± ± ± ( ) 75%Hop pvgrf ± ± ± ( ) *Denotes significantly different (P < 0.05) to testing session 1 Note: SLHD = single leg hop for distance; PVGRF = peak vertical ground reaction force; DVJ = drop vertical jump; SLCMJ = single leg countermovement jump; Comp = Composite; Ant = Anterior; Y-B = Y-balance Intra-rater reliability for the tuck jump total score (ICC = 0.88), and Y-Balance (ICC = 0.85) was deemed strong based on previous research (Hopkins, 2000). The variance reported for withinsubjects total score on the tuck jump based on the typical error estimate was acceptable in both the pre-phv and post-phv groups. The kappa coefficients used to determine agreement between each of the ten individual tuck jump criteria are displayed in table 5. Knee valgus was the only criterion that reached substantial agreement across the two test trials for both groups suggesting strong reliability. Other measures including unparalleled foot position during landing and a pause 128

129 between jumps also reported substantial agreement for the pre-phv players; however, this was not consistent across groups. Moderate agreement (k = ) was present for foot positioning not being shoulder width during landing (pre-phv group only), thigh motion not reaching parallel and high contact noise (post-phv group only) suggesting these criteria could be measured with acceptable reliability. The agreement between all other criteria was either fair (k < 0.40) or weak (k < 0.20) indicating poor test re-test reliability (Magaziner et al., 1996). Table 3.5. Tuck Jump Kappa Ratings of Agreement Kappa Value Kappa Value Phase of Jump Criterion (pre-phv) (post-phv) knee valgus k = 0.78 k = 0.67 Knee and thigh motion not parallel k = 0.34 k = 0.44 not equal k = 0.11 k = 0.29 not shoulder width k = 0.52 k = 0.25 Foot positioning during landing not parallel k = 0.62 k = 0.20 unequal contact time k = 0.04 k = 0.05 high contact noise k = 0.34 k = 0.43 pause between jumps k = 0.65 k = 0.00 Plyometric technique declines in 10s k = 0.33 k = 0.31 excess flight motion k = k = Discussion The current study assessed the reliability of a field-based neuromuscular control screening battery in elite male youth soccer players who were either pre- or post-phv. In the pre-phv group, single leg hop for distance; all y-balance variables; and pvgrf in both the 75%Hop, and 129

130 single leg countermovement jump demonstrated acceptable reliability (CV < 10%). Similar findings were evident for the same variables in the post-phv cohort. Significant differences between test sessions (p < 0.05) were reported in Y-Balance composite score (post-phv group only), anterior and posterior-lateral reach directions, and single leg hop for distance which may indicate the presence of a learning effect. The within-subject variance of all other assessed variables was above the accepted threshold for acceptable reliability (CV >10%) in both groups. Based on the correlation co-efficient, total score obtained during the repeated tuck jump assessment demonstrated moderate reliability in both groups. However, typical error values indicated a change in total score >1 would be needed to ensure that observed changes are real and fall outside the error range which would be deemed acceptable. When each of the ten individual criterion measures were analysed separately, knee valgus, feet not parallel or shoulder width on landing, and a pause between jumps were the only measures to demonstrate acceptable reliability for pre-pubertal players (range k = ). Observed knee and thigh motion variables including: knee valgus on landing, and thighs not parallel or equal in flight were the only criteria to report acceptable reliability in the post-pubertal group (range k = ). Statistics used to quantify reliability A number of the variables measured in this study demonstrated low ICC statistics. It has been suggested previously that a value > 0.75 is acceptable and values below this provide inadequate reliability (Shrout and Fleiss, 1979; Kovaleski et al., 1997). However, caution should be applied when using this approach for determining the usefulness of a test (Hopkins, 2000). Specifically, a measures of typical error provide information purely on the variation of performance within 130

131 each individual subject. Re-test correlations measure how closely the values of two trials track each other specific to each individual and the reproducibility of the rank order of subjects during the re-test (Hopkins, 2000). If a high ICC value is obtained, this suggests that the subjects measured retained a similar rank order. Lower values indicate that subjects did not retain their order during the re-test. ICC statistics are also affected by the heterogeneity of the values between participants, and the sample used for measurement (Hopkins, 2000). A sample which is of a homogenous nature will demonstrate a low value. The subjects in this study are reflective of a homogenous sample due to their status as elite male youth academy soccer players and this provides a plausible explanation for lower ICC values than those in other studies. Specifically, a number of the test variables in this study reported lower ICC values in the pre-phv players. Due to their status as prepubescent athletes, levels of performance may be more clustered as they have not yet experienced their peak growth spurt. Thus, similar levels of force production and landing abilities may have been present increasing the possibility of players changing their rank order. The post-phv players were likely at different stages of physical development with some players further past the period of PHV than others. In this group changing of rank order may be less frequent due to a wider variation in physical abilities. In addition, younger athletes demonstrate greater variation in jump performance which could also have altered the re-test correlation of the rank order achieved in the pre-phv players. An alternative approach is to use statistical tests which measure the within-subject error as this may be a more appropriate method (Hopkins et al., 1999). Specifically, calculating the coefficient of variation has been suggested as a pertinent indicator of reliability for athletic performance tests (Hopkins, 2000). This method which expresses the typical error as a percentage of the athletes mean score provides a direct measurement of the amount of error 131

132 associated with each test. Using this information, it is possible to determine if responses to training interventions are meaningful, and are greater than the reported error. A somewhat arbitrary interpretation of acceptable CV values (< 10%) has been suggested previously (Cormack et al., 2008; Turner et al., 2015), and has been used in this study. Single leg maximum hop for distance Reliability statistics for this measure demonstrated small (4.1%) to moderate (8.8%) CV s in the post and pre-pubertal groups respectively. However, significant mean differences (p < 0.05) between testing days suggest that systematic variation, most likely attributable to a learning effect, was present. A range of previous studies (Bandy et al., 1994; Paterno and Greenberger, 1996; Bolgla and Kesula, 1997; Brosky et al., 1999) have demonstrated strong reliability based on ICC statistics (ICC range = ) and coefficients of variation (CV range = %) (Maulder and Cronin, 2005; Meylan et al., 2009), albeit in adult populations. Therefore, the results of the current study show greater within-subject variation in youth male soccer players in comparison to previous data on adults. Furthermore, the current study demonstrates that the reliability of pre-phv players is more varied than post-phv players, as indicated by smaller CV s for the post-phv group. These findings are consistent with previous literature that has suggested greater variation in jump performance in younger subjects (Gerodimos et al., 2008). Therefore, when using the current battery of tests with youth male soccer players for future research, greater familiarisation may be necessary to ensure the chances of a learning effect are reduced. 132

133 Differences in task requirements should also be accounted for when attempting to compare the findings of previous work. For example, the majority of previous studies that have used the single leg maximum hop for distance did not specify that subjects had to stick and hold the landing as was required in the current study. Further, the necessity to land in a unilateral stance was not a criteria in earlier investigations, which permitted two footed landings following a single leg take-off (Maulder and Cronin, 2005; Meylan et al., 2009). The requirement to stick the landing in this study was included to provide a greater deceleration component, increasing the eccentric demands of the hamstrings (Augustsson et al., 2006; Goosens et al., in press), and each player s ability to attenuate force during a single limb stance. This method has been used recently and was predictive of a greater risk of hamstring injury (Goosens et al., in press), a frequently occurring injury in male youth soccer (Price et al., 2004). Cumulatively, it is suggested that participants should be required to stick the landing when screening neuromuscular control in male youth soccer players as this will provide a more pertinent measure to determine injury risk. Y-balance test Y-Balance composite score demonstrated acceptable CV s (range %) in all reach directions for both the pre and post-phv groups. Comparative data are limited to equate these values against other youth populations although other unilateral balance tasks have cited similar values as the post-phv group in this study (ICC = ) (Geldhof et al., 2006). The results of the present study noted a significant mean difference between test sessions (p < 0.05) in anterior and posterior-lateral reach directions which may indicate the presence of a learning 133

134 effect. Thus, practitioners using this assessment with youth male soccer players, may wish to increase the number of familiarisation sessions, practice trials or frequency of performance during warm ups to ensure greater accuracy. Early investigations to determine the reliability of this method reported ICC s ranging from (Kinzey and Armstrong, 1998). However, the authors used the star excursion balance test which has reported differences in movement kinematics and test performance to the Y-Balance (Coughlan et al., 2012). Task complexity was cited as a rationale for these moderate values (Kinzey and Armstrong, 1998), highlighting the need for adequate familiarisation. This has been confirmed more recently, where reach directions were assessed over multiple trials to establish learning effects. Reach distances stabilized after 4 trials (Munro and Herrington, 2010). Subsequently, trials 5-7 were used by these authors, reporting higher levels of reliability (ICC range = ; SEM = %) than previous studies which did not allow the same amount of practice trials. The inclusion of multiple practice trials has been utilised in other investigations with high school basketball players (Plisky et al., 2006) and young females (Filipa et al., 2010) resulting in higher ICCs (range = ). Whilst reliability appeared to increase as a consequence of increased familiarisation opportunities, time implications for screening a large number of athletes in youth soccer academies also need to be considered. In the present study, three practice trials were performed during the warm up in a group setting and only one was permitted on the Y-Balance instrumented device. This protocol is similar to that of Faigenbaum et al. (2014) who reported moderate to strong reliability (ICC range = ) in subjects between the ages of 6-11 years. These values were comparable to the post-phv players in the current study but not those in the pre-phv group who were a similar age to the subjects used by Faigenbaum et al. (2014). Whilst no obvious reason is available to explain the 134

135 differences between the two studies, a plausible explanation could be the greater rater-to-child ratio used by Faigenbaum et al. (2014) in comparison to the current study (1:5 versus 1:8 respectively) during the practice trials. Irrespective of the rater-to-child ratio available to practitioners, pre-phv players may require a greater volume of initial feedback on technique than older and more mature players. Findings from previous studies have indicated that an anterior right-left reach difference of greater than 4 cm increases the risk of lower extremity injury (Plisky et al., 2006). This has been confirmed more recently where anterior reach asymmetry was the only Y-Balance variable significantly associated with non-contact injury in college athletes (odds ratio, 2.33) (Smith et al., in press). Due to comparable reliability with other measures and acceptable CV s (< 10%), utilising this single component of the test may provide a more robust assessment tool for screening large groups where time constraints are present. If only the anterior reach direction is selected, practitioners using this assessment are then able to increase the number of familiarisation sessions, practice trials or frequency of performance during warm ups due to analysis of only a single variable. This will increase reliability and accuracy in detection of players with an increased risk of injury due to asymmetrical reach distances (Plisky et al., 2006; Smith et al., in press). Tuck jump assessment A summation of the ten assessed criteria to provide a total score demonstrated low reliability identified from the CV statistics in both the pre-phv (17.7%), and post-phv groups (CV = 28.5%). The typical error range reported ( in pre and post-phv players respectively) 135

136 would be considered acceptable and indicates that changes in total score >1 following an intervention would be classified as a real change in injury risk reduction. Previous investigations have reported large percentage changes following augmented feedback interventions for various criteria in the tuck jump assessment (Stroube et al., 2013), suggesting that a change in total score > 1 is realistic. However, in-spite of the acceptable typical error values reported, when each of the ten individual criterion measures were analysed separately, a number of these demonstrated fair to low agreement. These findings may be explained in part by the greater number of criteria included in the total tuck jump score which may have provided an averaging effect where errors begin to cancel themselves out. This is representative of taking the mean of more trials. However, due to the paucity of literature which has investigated the reliability of this test, further research is warranted. Criterion deficits of knee valgus, feet not parallel or shoulder width on landing, and a pause between jumps were the only measures to demonstrate acceptable reliability for pre-phv players (kappa range = ). Whereas, observed knee and thigh motion variables including: knee valgus on landing, and thighs not parallel in flight and high contact noise were the only criteria to report acceptable reliability in the post-pubertal group (kappa range = ). This suggests that the risk factors within the tuck jump assessment (Myer et al., 2011) that can be reliably assessed as part of a test-retest screen in pre-phv children are ligament dominance (screened via knee valgus and feet not shoulder width apart at landing), leg dominance (screened via foot positioning not parallel on landing) and core dysfunction (screened via pauses between jumps). In the post-phv players, tuck jump assessment related risk factors that can be reliably screened include ligament dominance (screened via knee valgus at landing); core dysfunction (screened via thighs not reaching parallel during flight) and quadriceps 136

137 dominance (screened via high contact noise and evident by the contact of the entire foot and heel on the ground between jumps). Thus, caution should be applied in purely interpreting the composite assessment score due to the high within-subject variation in a number of the individual criteria. Respective of a change in total score > 1 which was identified as a reliable performance modification, it would be uncertain as to how injury risk has changed. A plausible explanation for the variation in performances reported in both groups could be the interrelationship between specific criteria. For example, thighs not reaching parallel in flight may affect plyometric ability or the presence of high contact noise due to reduced landing forces. Further, un-equal thigh motion during flight could impact which foot hits the ground first resulting in un-equal contact times. Whilst this is the only study to analyse test re-test within-subject reliability for each of the individual criteria during the tuck jump assessment screen, knee valgus on landing reported the strongest agreement for both groups. This suggests that screening for this criterion risk factor is reliable for test re-test comparison in male youth soccer players who are either pre- or post-phv. The valgus mechanism has been suggested as a high-risk movement pattern for incidences of both medial collateral ligament (MCL) (Indelicato, 1995; Gardiner et al., 2001), and ACL injury (Hewett et al., 2005; Ford et al., 2003). These injuries occur frequently in male youth soccer (Price et al., 2004; Le Gall et al., 2006). Knee valgus motion has been cited as a key contributing factor to high knee abduction moments and ACL injuries (Myer et al., 2008) and is a predictor of injury assessed during jumping and landing tasks (Hewett et al., 2005). Therefore, practitioners using the tuck jump assessment to screen male youth soccer players may reliably track changes in knee valgus on landing following an intervention. 137

138 There is a paucity of available literature to compare the current findings with other studies and specifically no data are available in male youth of different ages or stages of maturation. In their initial review of this assessment, Myer et al. (2008) stated that unpublished pilot studies indicated moderate-to-strong inter-rater reliability (ICC range = ). More recently, Herrington et al. (2012) assessed the main outcome scores of males and females using two independent assessors demonstrating good intra and inter-tester reliability which is in agreement with the findings of this study (intra rater reliability, ICC = 0.88). This suggests that variation in test scores is due to the subjects and not the rater. However, none of these studies investigated within-subject reliability. In the present study, high within-subject variance was present as indicated by low kappa agreement between trials on a number of the test criteria. This finding is not unexpected, as obtaining high reliability for repeated trials during tasks requiring dynamic jumping and landing activities that are reliant on reflexive muscle responses, proprioceptive and kinaesthetic feedback will likely utilise a range of movement strategies, thus increasing variability (Wikstrom et al., 2006; Ebben et al., 2010). Previous literature has suggested greater variation in jump performance in younger subjects (Gerodimos et al., 2008); lower levels of consistency in technique can be expected in youth players. Poor reliability (CV = 34.1%) of the reactive strength index variable measured in youth male subjects has also been reported during a repeated maximal hopping task (Lloyd et al., 2009). The authors stated this may be attributable to variations in ground contact time and an inability to control forces experienced during landing. Therefore, due to the technical demands of the tuck jump assessment protocol, children may experience difficulty in consistently executing the correct jump-landing pattern. 138

139 Time to stabilisation During the single leg horizontal and vertical jump landing tasks onto a force plate, high CV s were reported for both pre-phv (CV range = %), and post-phv groups (CV range = %). These values indicate large within-subject variance and thus caution should be applied when using this measurement in male youth soccer players. This finding is not unexpected, as obtaining high reliability for repeated trials during tasks requiring dynamic postural stability is difficult (Ebben et al., 2010). Single leg jumping and landing activities that rely on reflexive muscle responses, proprioceptive and kinaesthetic feedback will typically utilise a range of movement strategies and therefore increase variability (Wikstrom et al., 2006; Ebben et al., 2010). No data are available to compare the results of this study to those of similar populations, however, reliability statistics in adults suggest strong reliability (ICC range = ) (Colby et al., 1999; Ross et al., 2009). A plausible explanation could be age related differences due to growth, maturation and skill. Previous literature has suggested that maturation of the neurological, visual, vestibular and proprioceptive systems may lead to enhanced performance during single leg balancing tasks (Mickle et al., 2011). Also, younger subjects demonstrate greater postural sway during single leg balance manoeuvres which may compromise stability (Mickel et al., 2011). Thus, time to stabilisation, a measure of reflex stabilisation (Wikstrom et al., 2005; Ross et al., 2005; Impellizzeri et al., 2013), may be subject to greater variability in youth athletes. Another factor which may explain the differences in reported reliability from this study and those of previous investigations is the task demands (Ebben et al., 2010). In the present study, two single leg landing assessments were used to provide data for both horizontal and vertical jumping tasks. Conversely, the aforementioned studies used horizontal tasks only and a 139

140 standardised distance from the force plate of either leg length (Colby et al., 1999), or an arbitrary distance of 70 cm (Ross et al., 2009). Further, the method to calculate time to stabilisation was different. The utilisation of anthropometric measures or standardised distances may subsequently over or under-estimate an individual s performance. For example, an athlete with short legs may demonstrate a short TTS due to the relatively shorter hopping distance required. However, during a maximal single leg hopping task, the same athlete may be capable of much greater jump distances than that of their leg length. Such feats of athleticism are likely to be replicated under the conditions of competitive match play. Thus, an individual s inherent risk of injury is likely a product of how far they can jump and how well they can attenuate the resultant forces on landing. The present study was comparable to previous investigations (Flanagen et al, 2008; Ebben et al, 2010) which only analysed the vertical force vector, and a stable state was defined as the time taken for a player upon landing to reach and stabilize within a pvgrf range 5% of their bodyweight for a period of one second. Colby et al. (1999) and Ross et al. (2009) used both anterior-posterior and medio-lateral force vectors and a static hold of twenty seconds, scanning the components from the last two windows of the last 10s (i.e. at 10-15s and 15-20s), and the window with the smallest ground reaction force range was accepted as the optimal range variation (Ross and Guskiewicz, 2003). This method, whilst displaying sound reliability, raises concerns of ecological validity when screening youth soccer players. For example, young athletes who are required to spend long periods standing still on a force plate likely demonstrate greater postural sway, thus affecting the ground reaction force range. Also, as a measure to detect injury risk, the requirement to maintain stability for a period of one-second (as in the present study) is likely more reflective of match play situations, as opposed to longer periods of static 140

141 stability. The shorter recording period of seven seconds used in this study as opposed to 20 seconds (Ross and Guskiwicz, 2003) also has implications for testing a large number of athletes, particularly youth athletes who may demonstrate lower levels of concentration. Peak landing forces In both groups, pvgrf was the most reliable kinetic measurement as indicated by the lowest CV s (range = %, and % for the pre-and post-phv players respectively). These findings are commensurate with Cordova et al. (1996) who reported excellent reliability values (ICC = 0.94; SEM = 0.003% body weight) during a single leg countermovement jump onto a force plate. Conversely, vertical impulse (a measure comprised of both force and time) was not reliable (ICC = 0.22; SEM =.24% body weight seconds). In the present study, whilst not a direct measure of impulse, time to pvgrf also showed higher CV values (range %) suggesting greater within-subject variation. Other studies have also reported high within session reliability in adults for pvgrf during a single leg hop and stick for both dominant (ICC = 0.82) and non-dominant (ICC = 0.87) limbs (Brent et al., 2005) and a single leg horizontal drop jump (ICC = 0.84, CV = 5.71) (Stalbom et al., 2007). Additionally, in this study both the single leg countermovement jump and 75% horizontal hop tests demonstrated lower withinsubject variation than the drop vertical jump. Single leg landing assessments may be more ecologically valid due to nature of the movements in soccer competitions and are also indicative of greater injury risk (Sugiomoto et al., 2014). A comparison of double versus single leg landing tasks in male athletes identified that single leg energy dissipation reduced in the sagittal plane but increased in the frontal plane (Yeow et al., 2010). The sagittal plane is the primary motion of 141

142 the knee joint and the associated musculature. In the frontal plane, range of motion and knee joint muscular stability is limited which increases the loading of the passive structures. Cumulatively, the greater risk of alterations in single leg landing technique than double leg landings and larger within-subject variation of the drop vertical jump in this study suggests that an assessment of single leg landing kinetics is a more appropriate measure of injury risk. Limited research is available to compare the reliability of landing force variables in youth males. However, previous data obtained across three sessions to quantify between-session reliability in school children showed strong reliability for measures of landing pvgrf (ICC = 0.89), and take-off (ICC = 0.98) (Quatman et al., 2005). In this study, lower reliability displayed in the pre-pubertal group may be indicative of reduced skill levels, and immature pre-frontal motor cortex activation for cognitive control which results in greater variation in the execution of motor control tasks (Bunge et al., 2002). Additionally, increased jumping skill has been associated with an enhanced ability to absorb landing forces (Prapavessis and McNair, 1999). As males progress through adolescence they appear to display an increased ability to attenuate landing forces, possibly due to the presence of the neuromuscular spurt (Quatman et al., 2005). Conversely, younger children appear to land with greater knee and hip extension and muscle cocontraction on impact which will lead to higher pvgrf (Swartz et al., 2005; Croce et al., 2004). Supporting this notion, lower pvgrf related to body mass during the breaking phase of a drop jump have been reported in adults versus boys (Lazaridis et al., 2013). This may be due to more efficient use of the stretch reflex and greater levels of muscle activation prior to landing and during the breaking phase of the jump (Lazaridis et al., 2010). Data also show that as children mature they become more reliant on supra-spinal feed forward input and short latency stretch reflexes (Lloyd et al., 2012). Cumulatively, the inefficiency in movement and higher landing 142

143 forces may provide a rationale for the greater variability in the pre-pubertal soccer players in this study. Asymmetry The results of this study reported acceptable reliability statistics (CV 10%) for most variables (see tables 5 and 6). The highest within-subject variance was reported for drop vertical jump pvgrf (pre-phv group only). These findings support previous work (Miller and Callister., 2009), with less variability for floor based tests than drop tests when comparing countermovement jumps and drop jumps in male athletes. Limited data is available to confirm the reliability of limb asymmetry statistics despite their high frequency of use during unilateral jumping tasks. One available study in ACL patients determined the limb symmetry index for a range of hopping based tests using ICC and standard error of measurement (SEM) (ICC = ; SEM ) (Reid et al., 2007). The authors assessed differences between the injured and non-injured leg, whereas, previous research has divided the scores of the dominant and nondominant legs and multiplied this value by 100 to provide a percentage asymmetry index (Barber et al., 1990; Maulder and Cronin, 2005). Definitions of lower limb laterality have been suggested, with the foot used in activities such as kicking and jumping the preferred or dominant foot, whereas, the non-preferred foot provides postural support and stabilisation (Peters, 1988). This classification has been supported in other available research (Gentry and Gabbard, 1985; Whittington and Richards, 1987; Hewit and Cronin, 2012). However, purely classifying limbs as dominant or non-dominant may be an over simplification for predicting injury risk as this does not account for neuromuscular inhibition which may be present following an injury (Friel et al., 143

144 2006; Opar et al., 2013). Also, there is a frequent need to jump and land on either leg during the pertubated actions of match play in youth soccer. Therefore, whilst classifying dominant and non-dominant limbs is a valid approach, to quantify injury risk, quantifying the difference between the highest and lowest performing leg study may be more appropriate. This is further supported by the acceptable reliability values for most of the measures included. 3.5 Summary and practical Applications Reliability data are now available for a field-based battery of neuromuscular control assessments to screen youth male soccer players for potential injury risk. This information has not been reported previously within current literature. Practitioners should now be able to more effectively select from the wide range of assessments available in the literature by considering their reproducibility as a basis for test re-test comparison. Furthermore, using the reliability statistics derived from this study, the smallest worthwhile change can be determined by calculating the between-subject standard deviation for each test and multiplying this number by 0.2 or 0.5% of the CV (Hopkins, 2004; Turner et al., 2015). Acceptable reliability values were reported for a variety of measures. In both the pre and post-pubertal groups, single leg hop for distance; all y-balance variables; and pvgrf in both the 75%Hop, and single leg countermovement jump demonstrated acceptable reliability (CV < 10%). These variables should be considered reliable for assessing elite male youth soccer players. However, significant differences between test sessions (p < 0.05) were reported in Y- Balance composite score (post-pubertal group only), anterior and posterior-lateral reach directions, and single leg hop for distance which may indicate the presence of a learning effect. Practitioners using these assessments should ensure adequate familiarisation to ensure 144

145 appropriate reliability is present. Total score obtained during the repeated tuck jump assessment demonstrated acceptable reliability in both the groups. When each criterion were analysed individually, moderate-substantial agreement between-trials was noted in selected criterion but these differed based in each age group apart from knee valgus. This variable can be accurately assessed but caution should be applied is using the other variables and the total score due to uncertainty in how reliable the other identified risk factors are. Overall, the results of this study suggest that a range of field-based tests and variables to determine neuromuscular control abilities and potential injury risk demonstrate acceptable reliability. Additionally, due to their ease of administration, time and cost effectiveness, a collection of these measures could realistically be utilised to form a screening battery for elite male youth soccer players 145

146 Chapter 4 STUDY 2: THE EFFECTS OF CHRONOLOGICAL AGE ON FIELD-BASED MEASURES OF NEUROMUSCULAR CONTROL IN ELITE MALE YOUTH SOCCER PLAYERS 4.1 Introduction Recent data have demonstrated a trend of increasing injury incidence with each level of exposure in junior soccer players, suggesting that male youth players should be a specific target group for injury prevention strategies (Schmikli et al., 2011). A linear increase in injury rates has been reported from 9 to 18 years of age in male youth players (Price et al., 2004), while heightened risk appears around the time of peak height velocity (PHV) (van der Sluis et al., 2014). This increased injury incidence may in part be due to rapid changes in stature and mass but is also associated with altered movement and motor control strategies indicative of an increased injury risk (Philapperts et al., 2006; Atkins et al., 2013; van der Sluis et al., 2014). Movements that lead to injury in elite youth soccer include running, twisting and turning, over-stretching and landing (Price et al., 2004). In adults, altered neuromuscular control during such actions is a key mechanism (Hewett et al., 2004) and a predictor of injury (Gomes et al., 2008; Small et al., 2010). Less information is available in youths and there is currently a paucity of literature available examining neuromuscular control in elite male youth soccer players and the effects of chronological age. Due to the physical demands of youth soccer, the associated injury risk, and the number of children and adolescents who participate in the sport, there is a 146

147 clear need for increased research within male youth soccer players to identify age specific injury risk factors (Alenrton-Geli et al., 2014). Within the available literature, a range of time efficient, non-invasive field-based methods have recently been developed to screen an athlete s level of movement skill and injury risk during jumping and landing tasks (Myer et al., 2008; Ebben et al., 2010). Practitioners should also consider other diagnostic tools to assess a range of movement deficiencies for other prevalent injuries including ankle sprains and hamstring strains. Examples include single leg jumping and landing (Goosens et al., in press) and dynamic balance tasks (Plisky et al., 2006). Study 1 reported acceptable reliability for a range of field-based neuromuscular control assessments. Variables collected within the screen included: (1) peak landing vertical ground reaction force during a maximal single leg vertical countermovement and 75% horizontal jumps; (2) anterior reach distances for the y-balance assessment; (3) hop distance during a maximal single leg horizontal hop; and (4) knee valgus during a repeated tuck jump. Asymmetry between the highest and lowest performing limbs also displayed acceptable reliability in each of the above single leg assessments. To the knowledge of the author, no data are currently available that specifically reports the effects of age on each of these measures in elite male youth soccer players. Two studies have assessed balance in elite male youth soccer players and compared performances to those of adult professional players (Pau et al., 2014; Pau et al., in press). Youth players demonstrated greater postural sway than adults; however, only adolescents (U15-U19 chronological age groups) were included in the youth sample. Consequently, the study does not account for changes in performance due to age, growth and maturation across childhood and adolescence, specifically the influence of rapid growth, which may contribute to, increased 147

148 injury risk (van der Sluis et al., 2014) and momentary disruption in motor control (Philippaerts et al., 2006). Furthermore, the majority of these studies assessed balance using static tasks which are not reflective of the dynamic nature of soccer activities during which injuries occur. Only two studies in the available paediatric literature utilised more dynamic tasks including a functional reach test (Habib and Westcott, 1998) and a single leg landing task (Pau et al., in press). In the latter study, the authors confirmed that the relationship between static and dynamic tasks was insufficient as evidenced by weak correlations. Thus, data are required to report the effects of age on measures of dynamic balance and stability which are considered more functionally relevant in youth male soccer players. Previous research investigating the effects of age on maximal jump performance has demonstrated a general trend of increasing performances with age (Phillippaerats et al., 2006; Figueiredo et al., 2009; Lloyd et al., 2011). This suggests that youth will naturally enhance performance due to growth and maturation. However, within the available literature variation in performance across growth and maturation may be evident (Lloyd et al., 2011), with various peaks and declines based on performance and stage of maturation (Phillippaerts et al., 2006). In male youth soccer players, limited data are available to report the effects of age during single leg jumping tasks. Previous research has primarily used some form of bilateral vertical jumping (Malina et al., 2004; Phillippaerts et al., 2006; Figueiredo et al., 2009; Castro-Pinero et al., 2010) and/or broad jumping (Phillippaerts et al., 2006; Castro-Pinero et al., 2010) tasks to assess jump and landing ability. The outcome variables in the aforementioned research are biased towards performance and power derivatives with less data available to examine kinetic and kinematic landing measures. Additionally, no research to date has reported the effects of age on asymmetry during single leg jump-landing measures; this requires further investigation. 148

149 Within the available literature, limited data are available to describe the effects of age on landing forces in elite male youth soccer players. Sigward et al. (2012) analysed youth players aged 9-22 years, however, the participants were not classified as elite. Also, the majority of studies have utilised bilateral drop vertical jump landing tasks (Quatman et al., 2005; Hewett et al., 2006; Ford et al., 2010; Sigward et al., 2012) which demonstrate reduced ecological validity. The drop jump also poses difficulty in selecting an appropriate drop height that accounts for individual jumping and landing abilities when screening large groups comprising a variety of age ranges and anthropometric profiles (Read et al. in press). Single leg jump-landing assessments may be more ecologically valid due to nature of the movements which occur in training and competitive soccer environments. Single leg landings are also indicative of greater injury risk (Sugiomoto et al., 2014). Furthermore, assessments to compare vertical and horizontal landing forces should be considered due to the inherent nature of soccer activities which involve jumping and landing in multiple planes of motion. Inadequate data are available to analyse the effects of age on landing forces during single leg jumping tasks in both vertical and horizontal directions is warranted; this requires further investigation. Youth males appear to demonstrate kinematic changes around the knee, with reductions in valgus alignment as they progress through maturation (Yu et al., 2005; Swartz et al., 2005; Schmitz et al., 2009). Conflicting findings have also been reported with no normalised differences in valgus alignment between children and older youths (Barber-Westin et al., 2005; Noyes et al., 2005). Reductions in valgus may in part be due to gains in strength of both the knee flexors and extensors (Ahmed et al., 2006), functional hamstring to quadriceps ratio (Ahmed et al., 2006), and increases in hip strength (Hollman et al., 2009), all of which are mediated by maturation. However, the presence of dynamic valgus is not purely strength dependent and is 149

150 likely multifaceted and underpinned by strength, coordination, movement skill, anatomical alignment and arthrokinematic function (Schmitz et al., 2009). Quatman et al. (2006) also demonstrated a concomitant reduction in landing forces with maturation in youth males, which may be a contributing factor. Available research to analyse the effects of age on knee joint kinematics in youth male soccer players is sparse, and in particular, no studies appear to have used elite players. Furthermore, the majority of studies have used both expensive and time in-efficient laboratory techniques (Swartz et al., 2005; Yu et al., 2005; Schmitz et al., 2009) or procedures that require digitisation of joint angles which are time intensive for coaches screening large groups of athletes in the context of a soccer academy (Myer et al., 2011). In contrast to drop jumping, repeated jumping tasks, whereby athletes are required to respond to movement perturbations and forces, may be more ecologically valid in sports such as soccer. An example of a field-based screening tool is the repeated tuck jump assessment (Myer et al., 2008). In study 1, it was identified that the presence of knee valgus may be reliably assessed during this screening task, providing a more time efficient and clinician friendly approach than laboratory techniques or those that require joint digitisation. Quantifying the effects of age on knee valgus motion during repeated jumping tasks will assist coaches in identifying players who demonstrate high risk kinematics and developmental trends associated with age. This data is currently not available and requires further investigation. Limb asymmetry is a risk factor that is inherent to youth male soccer where preferred limb dominance is clearly evident. Available literature has reported isokinetic strength imbalances of the hamstrings and quadriceps combined with reduced dominant leg hip range of motion (Kellis et al., 2001; Daneshjoo et al., 2013). Kinetic differences between limbs in propulsion and force 150

151 absorption during single leg jumping tasks have also been observed (Sannicandro et al., 2012), in addition to contralateral differences in peak ground reaction forces during a deep squat exercise with marked increases during the period of peak height velocity (Atkins et al., 2013). However, only the work of Kellis et al. (2001) and Atkins et al. (2013) assessed the effects of age on these variables. Also, the techniques used to assess asymmetry between limbs in the majority of these studies used isokinetic dynamometry (Kellis et al., 2001; Daneshjoo et al., 2013), which are not appropriate for screening large numbers of players in a soccer academy. Therefore, investigating the effects of age on asymmetry during practically viable and functionally relevant screening tasks is warranted. Cumulatively, there is lack of available evidence to report the effects of age on performance during a range of field-based neuromuscular control assessments in elite male youth soccer players. Specifically, in tasks which require repeated maximal jumps, single leg jumplanding, dynamic balance and quantification of landing force. The aim of the current study is to examine possible age-related differences in a variety of field-based neuromuscular control screening assessments using a cross sectional sample of elite academy male youth soccer players. 4.2 Methods Participants Four hundred elite male youth soccer players from the academies of six professional English Premier League and Championship soccer clubs volunteered to take part in the study and were grouped according to their chronological age for competition. Descriptive statistics for anthropometric measures and predicted maturational status are provided for each group in table 151

152 4.1 Due to the reported measurement error of approximately 6 months in the predictive equation (Mirwald et al., 2002), separation of players into sub-groups reflective of maturational status was deemed inappropriate. Previous research has also identified the following limitations in using this equation: 1) age at PHV for both early and late maturing boys was inaccurately predicted when compared to the criterion measure of skeletal imaging (Malina et al., 2014); 2) variables included in the equation (sitting height and leg length) are subject to ethinic variation and may be likely confounders in maturity estimates (Malina et al., 2004); and 3) the equation has a tendancy to classify boys as average maturers, and this has been shown in youth soccer players (Malina et al., 2012). Thus, while estimating maturation status using this approach displays practical merit and reasonable agreement with skeletal imaging (Malina et al., 2012), due to the aforementioned inaccuracies, chronological age was used to examine between group differences in the current study. Furthermore, the organisational structure within a soccer academy utilises chronological age groupings; therefore, the results of this investigation can be accurately interpreted and applied by practitioners to aid in the development of targeted training programmes. None of the players reported injuries at the time of testing and all were participating regularly in football training and competitions in accordance with the regulations for player contact hours as set out by the Premier League Elite Player Performance Plan. Parental consent, participant assent and physical activity readiness questionnaires were collected prior to the commencement of testing. Ethical approval was granted by the institutional ethics committee in accordance with the declaration of Helsinki. 152

153 Table 4.1 Mean (s) values for participant details per sub-group Age Group N Age (yrs.) Body Mass (kg) Stature (cm) Leg Length (cm) Maturity Offset U ± ± ± ± ± 0.5 U ± ± ± ± ± 0.6 U ± ± ± ± ± 0.7 U ± ± ± ± ± 0.9 U ± ± ± ± ± 0.6 U ± ± ± ± ± 0.6 U ± ± ± ± ± Sample size estimation The estimation of sample size was calculated using equation 1 as suggested by Hopkins (2004). Specifically, the reported coefficient of variation (CV) values from study 1 were used in addition to the smallest worthwhile change ratio, calculated as a factor of 0.2 of the between-subject standard deviation (Hopkins, 2006). Sample size = 8*(CV²/SWC²) [equ. 1] Performing additional trials of a given test has been proven to improve reliability and affect a reduction in CV (Pyne, 2004). Two trials were performed by each subject in this study; a reliability adjustment was completed using equation 2 where the CV is multiplied by a factor of The highest predicted sample size for each test measure obtained from study 1 was N =

154 Therefore, sample sizes greater than this magnitude were deemed necessary to accurately determine any significant between-group differences for any outcome measure. Reliability adjustment = 1/ number of trials [equ. 2] Experimental design This study used a cross sectional design to assess the effects of chronological age on field-based measures of neuromuscular control. Elite male youth soccer players (aged between years) from English Premier League and Championship clubs were required to attend their respective club training grounds on two occasions separated by a period of seven days. The first session was used to familiarise participants with the test equipment and assessment protocols. In the second session, data was collected from five different assessment protocols, including: (1) y- balance test (anterior reach direction only); (2) ten-second repeated tuck jump assessment analysed using two-dimensional video analysis; (3) single leg horizontal hop for distance (SLHD); (4) single leg 75% horizontal hop and stick onto a force plate (75%Hop) and; (5) single leg countermovement and stick (SLCMJ) onto a force plate. A 10-minute standardised warm up was completed prior to each test session consisting of dynamic stretching. The order of testing was randomised using a counterbalanced design to reduce the potential for an order effect. Two trials of each test were performed with the mean score reported. One minute of recovery was allowed between trials based on previous recommendations (Read and Cisar, 2001; Ebben et al., 154

155 2010). Participants were asked to refrain from strenuous exercise at least 48 hours prior to testing. Subjects were also asked to eat according to their normal diet and avoid eating and drinking substances other than water one hour prior to each test session Procedures The protocols for each of the jump-landing tasks and the y-balance test replicated those of study 1. All tests were conducted by the same researcher at each clubs respective test session. Participant instructions were standardised to ensure replication of conditions for each player (see appendix d) and removal of the potential for of inter-rater error. Based on the measurement error results reported, the following modifications were applied: 1) Exclusion of posterior-medial and posterior-lateral y-balance reach directions, thus only the anterior reach direction was used. 2) Retrospective analysis of each participant s tuck jump technique was standardised for 10 jumps per participant and only included observation of the presence of knee valgus, with all other criteria excluded. 3) In the assessment of landing kinetics, the drop vertical jump assessment was removed from the test battery and the number of force variables analysed in both the single leg countermovement jump and 75% horizontal hop was restricted to the assessment of peak landing vertical ground reaction force. In addition to the above modifications, kinetic landing variables were normalised to body weight (N) and single leg hop for distance and y-balance anterior reach scores were normalized 155

156 to leg length (McMahon and Cheng, 1990). This is of particular importance when assessing paediatric subjects due to the non-linear variation in leg length and body mass development which occurs throughout growth and maturation. Normalisation of absolute values will account for both group and individual differences, providing a dimensionless value. Also, to impart a further level of analysis to the repeated tuck jump assessment, valgus angles were subjectively classified as either minor (<10 ), moderate (10-20 ), or severe (>20 ). Whilst no literature is currently available to support these threshold values, the classifications were determined using pilot data and agreement between expert raters (n = 5) including experienced strength and conditioning coaches and rehabilitation specialists. Using these classifications knee valgus in the tuck jump was scored as follows: 0 = no valgus; 1 = minor; 2 = moderate; 3 = severe Statistical analysis Descriptive statistics for each test were calculated for each sub-group. A one-way analysis of variance was performed to determine the existence of any between group differences for all outcome measures except for tuck jump knee valgus where a Kruskal-Wallis test was used due to the data being non-parametric. The level of significance was set at alpha level p Homogeneity of variance was tested by Levene s statistic, and where violated Welch s adjustment was used to calculate the F-ratio. Post-hoc analysis to determine significant mean differences between groups was assessed using Gabriel s test when equal variance was assumed. In situations were equal variance was not assumed, Games-Howell tests were used. Cohens d effect sizes (ES) were calculated to interpret the magnitude of between group differences using the following equation 3, where M1 and M2 represent the means of the two groups being 156

157 compared. To quantify the effect the following classifications were used: standardized mean differences of 0.2, 0.5, and 0.8 for small, medium, and large effect sizes, respectively (Cohen, 1988). [equ. 3] Further analysis included Pearson correlation coefficients to determine relationships between variables and paired samples t-tests for parametric data and Mann-Whitney U tests for non-parametric data respectively to assess differences in performance between limbs. Knee valgus mode scores and frequencies were also calculated on each leg to show the most frequently occurring score for each chronological age group. All data was computed through Microsoft Excel One-way analysis of variance tests were processed using SPSS (V.21. Chicago Illinois). Intra-rater reliability for knee valgus scores in the repeated tuck jump assessment was assessed using intraclass correlation coefficient (ICC). Intra-rater reliability of the y-balance test (ICC = 0.85) was established in study 1 and considered acceptable based on previous guidelines (Hopkins, 2000). To aid the readers and reviewers in clearly visualising the between group difference statistics, a series of tables have been included in appendix b to show the p values and effect sizes of the respective between group comparisons for each test. 157

158 4.4 Results Y-Balance anterior reach distance: Y-balance anterior reach distances for each age group are displayed in figure 4.1 Absolute reach distances were highest in the oldest two age groups (U16 and U18) and the U18s scores were statistically higher than all other age groups apart from the U16 with medium to large effect sizes (ES range d = 0.58 to 1.17). Reach distances for the U16 were statistically greater than all age groups except the U15 and U18 showing small to large effect sizes (ES range d = 0.47 to 1.07). No other group comparisons were statistically significant and effect sizes were small (< 0.45). Within-group analysis showed absolute reach distances were significantly greater on the right versus left leg (p < 0.001) 158

159 % Leg Length Absolute reach distance (cm) U11 U12 U13 U14 U15 U16 U18 Chronological age group right left Figure 4.1. Y-Balance absoute reach scores U11 U12 U13 U14 U15 U16 U18 Chronoloigcal age group (yrs) right left Figure 4.2. Y-Balance relative reach scores 159

160 Mean (± standard deviation) y-balance anterior reach distances normalised to leg length for each age group are displayed in figure 4.2. The highest reach distances were recorded for the U11s and were significantly greater than all age groups except the U14s, U16s and U18s on the left leg. The lowest values were displayed for the U15 age group on both legs and scores were statistically lower than the U11 corresponding to a large effect size (d = to -0.95). Significant differences were evident between the U15 and U18 for their right leg only (P < 0.05; d = 0.74). No other group comparisons reached statistical significance. Within-group analysis showed absolute reach distances were significantly greater on the right versus left leg (p < 0.001). Single leg horizontal hop for distance: Single leg horizontal absolute jump distances for each age group are displayed in figure 4.3, depicting a trend of increasing hop distance with advanced age. A significant increase in hop distances was reported between consecutive groups from the U14 age group with medium to large effect sizes (d = ). Hop distances were lower in the U13 group in comparison to both the U12s and U14s on both legs, however, this reduction was statistically significant compared only to the U14s on the right leg (p < 0.001) and left leg (p < 0.05) with medium to large effect sizes (< 0.65 to 0.84). Effect sizes for all other age group comparisons ranged from small to very large (ES range d = ). Within-group analysis showed no significant difference between limbs. 160

161 Hop distance % leg length Hhoriozntal hop distance (m) 2.5 Right Left U11 U12 U13 U14 U15 U16 U18 Chronological age group (yrs) Figure 4.3. Absolute single leg hop distance 250 Right Left U11 U12 U13 U14 U15 U16 U18 Chronological age group (yrs) Figure 4.4 Relative single leg hop distance 161

162 Hop distances normalised to leg length are displayed in figure 4.4. Scores were lower in the U13 group in comparison to the U12s and U14s on both legs corresponding to a medium effect size (d = to on the left and right respectively between the U13s and U12s). A significant increase in hop distance was present from the U13s to U14s on both the right (p < 0.001; d = 0.58) and left (p < 0.05; d = 0.52) legs. The U18 age group achieved significantly greater hop distances relative to leg length than all other groups on both legs (P < 0.001; ES range d = 0.57 to 1.71) apart from the U16 on the right leg but a medium effect size was recorded (P = 0.170; d = 0.57). Other group comparisons also displayed statistical significance (see appendix B) with effects sizes ranging from small to very large. Within-group analysis showed no significant difference between limbs. Assessment of landing forces: Descriptive statistics for absolute peak landing vertical ground reaction forces experienced by each chronological age group during the 75%Hop and SLCMJ are displayed in figures 4.5 and 4.6 respectively. A trend of increased absolute landing forces with advances in chronological age was evident on both tests. In the 75% horizontal hop, the youngest age group (U11) had significantly lower absolute landing forces than the oldest four age groups (U14 U18) (p < 0.001; ES range d = ). In the SLCMJ, significant differences between the youngest (U11s and U12s) and older age groups were evident from the U13 group on the right leg (p < 0.05) and from the U14s on the left leg (p < 0.001). The U11 and U12 age groups experienced significantly lower landing forces than the U14 to U18 age groups (P < 0.001; ES range d = range 1.06 to 3.18). The within group analysis comparing landing forces on each leg revealed that right leg landing forces were significantly greater during the 75% 162

163 Landing Force (N) horizontal hop (p < 0.001). Conversely, a trend was evident of greater forces experienced on the left leg during the SLCMJ; however, this was not statistically significant right left U11 U12 U13 U14 U15 U16 U18 Chronological age group (yrs) Figure %Hop absolute peak landing forces 163

164 Landing force (N) 3000 right left U11 U12 U13 U14 U15 U16 U18 Chronological age group (yrs) Figure 4.6 SLCMJ absolute peak landing forces When analysed relative to body weight, 75%Hop landing forces displayed a trend of reductions with age (figure 4.7). However, an increase was present in the U18 age group who experienced significantly greater relative forces than the U15 and U16 (p < 0.001; ES range d = 0.71 to 1.05). The lowest relative landing forces were shown in the U16s on both legs and this was statistically significant from all other age groups except the U15s. Within-group analysis showed normalised landing forces were significantly greater on the right versus left leg (p < 0.001). In the SLCMJ, a more variable pattern was observed (figure 4.8). The smallest relative landing forces on the right leg were recorded in the U15s and this reduction was significantly lower than the U11s (p < 0.001; d = ). On the left leg, significantly higher relative forces were shown in the U11s in comparison to the U12s (p < 0.001; d = ) and U15s (p < 0.001; d = ). A trend towards significance was evident with reduced landing forces relative to 164

165 Relative Peak Landing Force (x BW) body weight when comparing the U16 and U11 (p = 0.055; d = ; p = 0.066; d = ) for left and right legs respectively. A non-significant increase in consecutive age groups (U13s and U14s; ES range d = 0.25 to 0.52) and a non-significant reduction in the U15 andu16 age groups (ES range d = to -0.44) was also shown. Within-group analysis showed no significant difference between limbs U11 U12 U13 U14 U15 U16 U18 Chronological age group (yrs) right left Figure %Hop relative landing force 165

166 Relative Peak Lnading Force (x BW) right left 0.0 U11 U12 U13 U14 U15 U16 U18 Chronological age group (yrs) Figure 4.8 SLCMJ relative landing force Knee valgus in the tuck jump assessment: Intra-rater reliability for knee valgus score (ICC = 0.90) was deemed strong based on previous research (Hopkins, 2000). Mode knee valgus scores for each leg are displayed in table 4.2. The U18s reported significantly lower knee valgus scores on the right leg to all other age groups (p < 0.001), except for the U16s with effect sizes ranging from moderate to large (d = ). A trend towards significance was shown with lower right leg knee valgus scores for the U16s versus the U13s (p = 0.058) and U12s (p = 0.083). Effect sizes for all other group comparisons were small (< 0.35). On the left leg, knee valgus scores were significantly higher in the U11s and U12s from all age groups apart from the U13s. Significantly lower scores were recorded in the U18s in comparison to all other age groups (p < 0.001). Asymmetrical scores between limbs were evident in the U14s, U15s (2:1 right vs. left), 166

167 and the U18s (1:0 right vs. left). A Mann-Whitney U test to compare median scores for each leg revealed that knee valgus scores were significantly higher on the right leg (p < 0.001). Table 4.2 Mode knee valgus scores per chronological age group Valgus Right Leg Valgus Left Leg U U U U U U U The distribution of knee valgus scores for each chronological age group is displayed in table 4.3. The greatest frequency of 0 scores was recorded in the U18s. A trend was observed of more 2 and 3 grades in the younger age groups and a reduction in the combined percentage of 2 and 3 scores with age on both legs. The highest percentage of severe classifications (grade 3 scores) was shown in the U13s on the right leg. Table 4.3 Frequency of knee valgus scores for each age group Valgus score right Valgus score left Age group U U U U U U U

168 Asymmetry: Asymmetry scores for each test and respective age group are displayed in figure 4.9. Landing force variables during the 75%Hop and SLCMJ showed greater asymmetries than both the single leg maximum hop and the y-balance anterior reach test. In the SLCMJ, an increased asymmetry was evident with advances in age, reaching statistical significance between the U16s and U11s (p < 0.01; d = -0.68). A trend approaching significance was also shown between the U11s and U18s (p = 0.08; d = 0.60). Asymmetry was also highest in the oldest age group on the 75%Hop, however, no significant differences were reported between groups and effect sizes were small (< 0.49). In the single leg hop for distance, a more variable pattern was observed and the youngest age group displayed the greatest asymmetry. Significant differences were reported between the U11s and U16s (p < 0.05; d = 0.60) with a trend towards significance for the U12 versus U16 group comparison (p = 0.078; d = 0.59). A medium effect size was shown between the U11s and U18s (d = 0.52) and all other effect size calculations were small (< 0.50). In the Y-balance anterior reach test, the lowest and highest asymmetry scores were observed in the U16s and U11s respectively, and this difference approached statistical significance (p = 0.093; d = 0.52). No significant differences were observed between groups and all other effect size comparisons were small (< 0.50). 168

169 Asymmetry (% performance achieved) % Hop SLCMJ SLHD Y-B 0 U11 U12 U13 U14 U15 U16 U18 Chronological age group (yrs) Figure 4.9 Asymmetry scores for 75Hop, SLCMJ, Y-B and SLHD Notes: Y-B = Y-balance; SLHD = single leg hop for distance 4.5 Discussion The current study assessed the effects of age on a variety of field-based neuromuscular control assessments in elite male youth soccer players. Results showed that absolute peak landing forces increased with age in both horizontal and vertical single leg jump-landing tasks. When normalised to body weight, a more variable pattern was evident and inconsistencies were identified between the two tests used. The repeated tuck jump assessment identified reductions in knee valgus grade with advancing age. Knee valgus was significantly greater in the right leg, and asymmetry between legs was also present in specific age groups. The oldest age group (U18s) 169

170 recorded statistically higher anterior reach distances on the y-balance assessment than all other age groups apart from the U16s. However, when distances were reported relative to leg length, the youngest age group (U11s) achieved the highest reach distances. Single leg hop distances also increased with each consecutive age group apart from the U13s where a non-significant reduction in performance was identified. After adjusting the scores relative to leg length, the U18 still achieved the greatest hop distances. However, more variability was noted in respective age group comparisons, particularly around the U13, U14 and U15 age groups, which reflect the typical period of peak height velocity (Malina et al., 2004). Effects of chronological age on y-balance anterior reach scores Absolute y-balance anterior reach distances were highest in the U18s who also had the greatest stature and leg length. This is to be expected as children who are taller have displayed better functional reach scores, and therefore measures should be adjusted for leg length (Habib et al., 1998). Previous research has reported age related improvements in balance and reduced postural sway in recreationally active youths (Nolan et al., 2005; Cumberworth et al., 2007), and elite soccer players (Pau et al., 2014; Pau et al., in press). Shorter stature may challenge balance performance due to a smaller base of support and larger ankle displacements which increases activation of the gastrocnemius musculature and co-contraction of tibialis anterior to stiffen the lower extremity (Habib et al., 1998). Also, children who possess immature stabilisation strategies have been reported to utilise an open-loop, high-velocity, ballistic reactive strategy, with large and rapid adjustments to centre of pressure (Riach and Starkes, 1994; Kirsenbauhm, 2001). With maturation, an integrated open and closed loop strategy is developed which results in more 170

171 controlled and accurate movement strategies (Riach and Starkes, 1994; Habib et al., 1998; Kirsenbauhm, 2001). Normalisation identified that the youngest age group (U11s) displayed the greatest reach distances with effect sizes ranging from trivial to large (range d = ). Greater reach distances relative to leg length in the youngest chronological age group could be linked to heightened ranges of motion occurring via reductions in passive stiffness associated with younger ages (Ochi et al., 2008). Therefore, heightened performances on this test may be constrained more by mobility (predominantly at the ankle) as opposed to dynamic balance. Also, it is plausible that to determine age related changes in balance performance, alternative measures such as postural sway, or the dynamic postural stability index (Wikstrom et al., 2005) may be required. However, due to the paucity of data, this concept remains speculative and warrants further investigation in paediatric populations. Both absolute and normalised reach distances reported similar mean values for the U12, U13, U14 and U16 chronological age groups; however, a decrease was evident between the U14 and U15s ( % on the left and right leg respectively) but this did not reach statistical significance and effect sizes were small (< 0.37). A plateau or decline in flexibility has been associated with the onset of the adolescent growth spurt (Beunen and Malina, 1996; Phillippaearts et al., 2007). This may be related to increased musculotendinous stiffness resulting from increased restriction of titin filaments (Ochi et al., 2008), collagen filaments and cross linking (McCormick, 2003). Specifically, structural changes may present to the ligaments and tendons increasing stiffness of the tissues with the onset of adolescence (McCormick, 2003). However, existing research has utilised largely animal subjects, which may limit interpretation in the present study. Another conceivable explanation is greater leg length differences between 171

172 groups. The greatest mean leg length differences between chronological age groups (8.14%) were present between the U14s and U15s. Bone growth occurring in the absence of muscle length changes may result in momentary reductions in flexibility. Relative reach distances after this point were then increased reaching statistical significance between the U15 and U18 age groups on the right leg (p < 0.05) and approaching significance on the left leg (p = 0.081). The increased relative reach distances in the U18s may be due to increases in strength (as indicated by higher single leg hop distances) which offset possible reductions in flexibility following the adolescent growth spurt. However, due to the lack of isolated flexibility measures within this study, this remains speculative. Effects of chronological age on single leg hop for distance scores Data showed that single leg horizontal hop distances generally increased concomitantly with age on both the right and left legs. These findings are consistent with previous literature in male youth soccer players which has shown significant improvements during functional capacity tests with advances in age (Figueiredo et al., 2009). This may be due to increases in muscular strength as a result of growth and maturation, evidenced by increasing hop distances in each consecutive age group and reports that the standing long jump test has demonstrated strong relationships with lower body muscular strength in youth athletes (Castro-Pinero et al., 2010). Between the U12s and U13s, a reduction in both absolute and relative hop distances was evident. The performance of the U14s was also significantly greater than the U13s for both absolute and relative distances. This could be attributed to a period of adolescent awkwardness, whereby, due to rapid increases in limb length, young soccer players may experience temporary 172

173 decrements in motor skill performance occurring approximately 12 months prior to PHV (Philippaerts et al., 2006). Age at peak height velocity (APHV) in the current study was approx. 14 years, which chronologically aligns this period with the U13 age group although caution should be applied due to the reported error associated with this equation (Mirwald et al., 2002). This may be supported by a large percentage change in leg length between the U12s and U13s (5.36%) suggesting the U13s were at the start of the growth spurt. While adolescent awkwardness does not affect all youth, a disruption in motor control may be present due to disproportionate growth of skeletal and muscle tissue and changes in neuromuscular functioning (De Ste Croix et al., 2012). Relative scores also revealed a significant increase in hop distance in the U14 age group. This may suggest the presence of a neuromuscular spurt in accordance with the onset of puberty (Beunen, 1997: Philippaerts et al., 2006). Greater jump distances / heights will further challenge dynamic stability and dissipation of landing forces. Poor attenuation of ground reaction forces upon landing (as indicated by relative scores in the 75% horizontal hop and SLCMJ) in-spite of improvements in jumping performance may heighten injury risk in youth players around this period which has been associated with a greater risk of injury (van der Sluis et al., 2014). This highlights the need for training interventions that focus on developing appropriate landing mechanics around the time of PHV. Effects of chronological age on landing forces Peak vertical landing forces increased linearly with age in both the 75%Hop and SLCMJ, which is likely due to increases in body weight and muscle mass with each consecutive age group. Across both tests, the U11s displayed a trend of higher ground reaction forces relative to body 173

174 weight. This is supported by previous literature which has suggested that younger children land with different kinematic strategies to adults, characterised by reduced knee and hip flexion angles (Swartz et al., 2005). The described landing position is indicative of a stiffer landing in which ground reaction forces and joint torques may increase (Swartz et al., 2005). However, injury risk is low in this group (Price et al., 2004), and intuitively higher relative ground reaction forces may be offset by inferior body mass, shorter statures and slower intensities of play in comparison to older players. Relative landing forces depicted a variable pattern across chronological age groups and test measures but some general trends were evident. Although statistical significance was not reached, in the SLCMJ, relative forces appeared to reduce from U11 to U13, followed by an increase in the U14s and a subsequent reduction in the U15 age groups. A similar pattern was reported by Hewett et al. (2006) who showed that landing forces increased in youth males around the period associated with the peak growth spurt. High relative forces in the U14 age group may suggest an increased injury risk around the time of PHV, a concept that has recently been reported in youth soccer players (van der Sluis et al., 2014). Also, in the present study, relative landing forces showed that standard deviations were greater in the U14s across both tests suggesting more variance is present around the period of PHV. This trend of greater variation in performance in the U14s was present in a number of the measures used in this study, which may reflect the rapid growth and increased movement variability associated with their stage of maturation (Gerodimos et al., 2008). Another plausible explanation could be that jumping performance is enhanced due to a neuromuscular spurt (as evidenced by significant increases in standing long jump distances between the U13s and U14s), but a developmental lag remains in 174

175 their ability to attenuate ground reaction forces upon landing. This may heighten their risk of injury and highlights the need for interventions focused on dissipating landing forces. Comparison of relative scores in the 75%Hop revealed significant reductions in landing forces between the U14s (left leg only), U15s and U16s respectively. This may be due to a more hip dominant landing strategy, whereby post-phv athletes demonstrate greater ankle and hip joint stiffness which enables more effective and efficient absorption of landing forces (Ford et al., 2010). Greater hip flexion upon ground contact provides a mechanical advantage to the hamstring musculature, heightening their activation and concomitantly reducing quadriceps activation (Schulz, 2007). However, significant increases in landing force were reported when comparing the U18s vs. U16s during the 75% horizontal hop. This could be explained by significantly greater 75% jump distances in the U18s (p < 0.001), however, the increase between these two groups was proportional to other between group comparisons. Previous literature has observed reductions in ground reaction forces relative to bodyweight during a drop jump manoeuvre (Quatman et al., 2006), suggesting that with advanced age and maturation, athletes are better able to attenuate landing forces due to utilisation of more effective movement strategies to dissipate force (Swartz et al., 2005). Conversely, literature has reported that resultant forces are significantly greater in post-phv than pre-phv subjects during a single leg stride jump task (Hass et al., 2003) which may be more representative of the 75% horizontal hop performed during this study. Hass et al. (2003) suggested that with increased age, a more extended knee position on landing is adopted, whereas, greater knee flexion evident in the younger subjects is a protective strategy to protect internal joint structures. The results of the present study do not support this notion due to higher relative landing forces in the both youngest (U11) and oldest (U18) groups. Therefore, it could be inferred that the significant increase in 175

176 landing forces in the U18 age group may be a result of increased training loads following their advanced status as a full time academy scholar, altering their kinematic strategy. Specifically, alterations in the functional hamstring: quadriceps ratio has been shown due to muscle loading patterns which asymmetrically strengthen the quadriceps during training and competitions (Iga et al., 2009). This alters the reciprocal balance of strength and dynamic stabilisation around the knee as indicated by compromised function of the hamstrings during high velocity actions (Iga et al., 2009). Increased exposures in the U18 group who are full time scholars may have resulted in a more quadriceps dominant landing strategy and high ground reaction forces that may increase the risk of injury (Hewett et al., 1996; Hewett et al., 2005). Available data suggest that in elite male youth soccer players, injury risk is greater in U18 age group (Price et al., 2004); a relationship may exist between heightened landing forces and non-contact injuries in this cohort. A pattern of leg dominance was evident on both tests with higher landing forces on the right vs. left leg during the 75% horizontal hop. A recent study by Zebis et al. (2009) identified that during a cutting manoeuvre, female soccer players who subsequently experienced an ACL rupture in the following 2 competitive seasons injured their preferred push off leg. In a retrospective analysis, a significant gender effect for sex differences was reported in the distribution of non-contact ACL injuries, where 74.1% of males injured their dominant (kicking) leg, compared to 32% of females (Brophy et al., 2010). Asymmetry places additional stress on the weaker leg, compromising performance and predisposing athletes to various injuries (Hewit et al., 2012); this risk factor may be present in elite male youth soccer players. However, further research is required in youth populations to investigate the relationship between limb dominance and prevalence of injury. 176

177 There was also a trend of higher landing forces during the 75%Hop versus the SLCMJ. Jump-landing tasks with a greater horizontal motion may elicit differing movement strategies to those performed in a vertical plane. In the present study, correlations between the two tasks reported moderate relationships (r = ). Investigations of leg power indices across three different directions (vertical, horizontal, and lateral) have reported non-significant relationships between tests in the various movement planes (Maulder and Cronin, 2005; Meylan et al., 2009; Hewit et al., 2012). This denotes that testing athletes in multiple directions measures relatively independent qualities of leg-power. Therefore, utilising a range of assessments that target multiplanar actions is recommended in order to establish an accurate and comprehensive injury risk profile of an athlete. Further, comparisons with previous literature to report the effects of age on landing forces in the present study are challenging due to the high reliance on bilateral tasks, specifically drop jumps (Hewett et al., 2006; Quatman et al., 2006; Ford et al., 2010; Sigward et al., 2012) and bi-lateral vertical jumps (Swartz et al., 2005). In determining injury risk, the performance of single limb tasks is preferred to bilateral variations due to their enhanced ecological validity, ability to easily identify asymmetry and sensitivity for determining injury risk due to deficits in neuromuscular control (Yeow et al., 2010; Myer et al., 2011; Sugiomoto et al., 2014). Effects of chronological age on tuck jump knee valgus scores The results displayed in table 4.3 indicate that, knee valgus grade scores greater than 1 were recorded during the repeated tuck jump assessment in all age groups suggesting this risk factor was present to different degrees in the majority of elite male youth soccer players in the study. 177

178 This aberrant movement pattern has been reported as a high risk mechanism for incidences of anterior cruciate ligament (ACL) injury in female athletes (Hewett et al., 2005). Furthermore, it has been suggested that untrained youths who do not undertake strength and plyometric training activities are more likely to demonstrate knee valgus alignment (Noyes et al., 2005). The results of the present study speculatively suggest that strength and motor control deficits may have been present in the sample, which may increase their risk of injury during dynamic jump landing activities. A trend was evident of reduced knee valgus with advancing age. This is consistent with previous literature in young male athletes (Schmitz et al., 2009) and recreational male youth soccer players (Yu et al., 2005). Other investigations have reported no significant age effects for normalised knee and ankle separation distances (Noyes et al., 2005). The results of the present study suggest that elite male youth players will display reductions in knee valgus alignment as they progress through maturation. Strength appears to be an important contributor to the magnitude of dynamic knee valgus alignment (Schmitz et al., 2009). Differences in jumplanding kinematics between males and females have been attributed to greater knee flexor and extensor strength and functional hamstring to quadriceps ratio (Ahmed et al., 2006). The U18 group in the present study displayed significantly lower valgus alignment than all other age groups. Concomitantly, peak landing forces were highest in this age group. This suggests that older players utilise strategies to enhance management of joint positioning which may be due to increases in strength and eccentric control. The U18s also achieved significantly greater hop distances relative to leg length than all other groups apart from the U16s on both legs (p < 0.001; d = ). The standing long jump has been reported as a valid predictor of lower body strength reporting the strongest association with lower body isokinetic muscular strength tests 178

179 (R² = ) (Castro-Pinero et al., 2010). However, valgus alignment during jumplanding tasks is multifaceted and while it is heavily underpinned by strength, coordination, movement skill, anatomical alignment and arthrokinematics are also contributing factors that should be considered (Schmitz et al., 2009). When the data from all groups were pooled, notwithstanding the youngest two groups, knee valgus score was significantly greater on the right leg (p < 0.001). Movement variability during jumping tasks is more evident in younger athletes (Gerodomis et al., 2008), which may explain the discrepancy between the U11s and U12s in comparison to the older groups in the present study. Greater knee valgus scores on the right leg may suggest the emergence of limb asymmetry from the onset of the U13 chronological age group. Asymmetry has been reported to increase during the period of PHV and the early stages of adolescence (Atkins et al., 2013), due to physiological adaptations on the dominant leg in youth soccer players (Daneshjoo et al., 2013). Thus, increases in leg dominance with age may be evident. A plausible explanation could be that the majority of participants in this study were right footed, and with greater exposure to soccer specific practice and competitions, players may become more accustomed / competent at landing on their left leg. While no data is available to confirm this in youth athletes, in elite male soccer players, the distribution of non-contact ACL injuries has been reported where 74.1% of males injured their dominant (kicking) leg (Brophy et al., 2010). However, further research is required to analyse prospective relationships between leg dominance and injury risk in elite male youth players. A trend for symmetry between limbs was present in the majority of age groups. However, the U14, U15 and U18 age groups displayed asymmetrical scores between the right and left limbs. Of particular note, the asymmetry score for the U14s and U15s may reflect an increased 179

180 injury risk (2:1 right vs. left comparison). Also, the highest frequency of severe knee valgus scores (grade 3) were shown in the U13s. Recent data have shown that elite male youth soccer players are particularly susceptible to injury between the ages of 13.5 to 14.5 years (van der Sluis et al., 2014; van der Sluis et al., 2015). The U13s, U14s and U15s in the current study approximately align with this time period due to the completion of testing during the pre-season period. Subsequently, the player s ages would correspond to this period of heightened risk and a greater incidence of overuse and /or traumatic knee injury may also be present during this period (van der Sluis et al., 2014; van der Sluis et al., 2015). However, this requires further investigation to determine if greater valgus scores and asymmetry between limbs increases the risk of injury in this cohort. Effects of chronological age on limb asymmetry In the present study, landing force variables during the 75% horizontal hop and SLCMJ reported greater asymmetries than both the single leg maximum hop and the y-balance anterior reach test. This may be due to task complexity but also highlights that kinetic measures, specifically peak landing forces, are more sensitive than measures of distance in their ability to identify asymmetry. Previous literature has reported differing asymmetry values for a range of variables; distance ( %), peak force ( %), and peak power ( %) for the same jumps across different directions (Meylan et al., 2009). The authors also reported that measures of jump height and distance may be less sensitive for determining limb asymmetry. Such contralateral imbalances are an important component of predicting subsequent injury risk (Baumhauer et al., 1995), and are inherent to soccer where preferred limb dominance is evident. An asymmetry 180

181 threshold of 15% has been identified as a suitable threshold for heightened injury risk prediction (Crosier et al., 2000). A more recent study also used functional hopping tests, reporting distance as the outcome variable in recreationally active students (Munro and Herrington, 2011). All subjects achieved a limb symmetry index of <10%; the authors suggested a minimum symmetry of 90% is more relevant than previous investigations. In the present study, single leg hop for distance and anterior y-balance reach distances displayed asymmetry scores <10%, whereas, peak landing force asymmetries ranged from %. Task complexity and sensitivity of outcome measures used could be cited as plausible explanations for the reported differences in asymmetry. Also, kinetic asymmetries in youth athletes are to be expected, with horizontal ( %) and vertical force ( %) force discrepancies identified between limbs during a maximal running task (Rumpf et al., 2014). The authors stated that asymmetries between 15 20% appear typical in developmental athletes. Therefore, further investigations are warranted to determine if an asymmetry threshold can be identified that increases injury risk for the outcome variables used on each respective test included in this study. This will be investigated in study 5. During the SLCMJ, a trend was evident of increasing asymmetry with advances in age. This may suggest the emergence of increased leg dominance due to accumulated exposure to sport-specific training and competition. Elite soccer in the United Kingdom has recently adopted an early sport specialisation approach, whereby, youth boys participating in academy programmemes are now required to attend multiple weekly training sessions and competitions, with formal registration commencing at the age of 9 years. There is a considerable risk of injury in early soccer specialisation programmemes (Brink et al., 2010; Tak et al., 2015), and recent data has indicated early sport specialisation as an independent injury risk factor even after controlling for age and hours of total training and competitions completed each week (Reudi et 181

182 al., 2012). However, available evidence examining the relationship between early soccer specialisation, asymmetry and injury risk is sparse; further investigation is warranted (Read et al., in press). In the present study, asymmetry was significantly greater in the U16s vs. the U11s during the SLCMJ and a trend approaching significance was evident between the U11s and U18s. Asymmetry was also highest in the oldest age group on both the 75%Hop and SLCMJ. Conversely, previous literature showed the highest asymmetries in peak force were in the U15 age group during an overhead squat task (Atkins et al., 2013). Also, significant differences were identified between limbs (p < 0.05) in all age groups except for the U13s and U17s (the youngest and oldest groups respectively). The authors suggested that asymmetry increases during the period of PHV and the early stages of adolescence. Task differences could be cited as a potential reason for disparity between the work of Atkins et al. (2013) and the current investigation. In professional players, asymmetry has been reported to reduce with a greater training age (Fousekis et al., 2010). However, in their study the number of years of involvement was classified as either; 5-7 yrs.; 8-10 yrs; and 11+ yrs, making comparisons difficult with the present study. In the current study, the increased peak landing force asymmetry in the U18 age group could be due to greater volumes of training required to meet the playing requirements of a full time academy scholar. Total injury incidence in professional English soccer academy players is also highest in the U18 age group although no delineation was provided for contact vs. noncontact injury mechanisms (Price et al., 2004). Therefore, greater training volumes and accumulated training history associated with those undertaken by older players in soccer academies may increase their risk of injury due to heightened asymmetries in force absorption. 182

183 4.5 Summary and practical applications This is the first study to provide cross sectional data for a range of field-based tests of neuromuscular control in a large sample of elite male youth soccer players. The findings of this study may assist practitioners by providing normative data for a range of chronological age groups, from which reduced performances can be identified. Previous literature pertaining to elite male youth soccer players has largely investigated single measures and isolated tests, whereas, this study utilised a range of neuromuscular control field tests, enhancing practical application. The current study provides an original and significant contribution to the paediatric literature through the identification of potential injury risk factors present at different stages of development. The key findings from the current study are summarised below: Anterior reach distances normalised to leg length revealed that the youngest age group (U11s) achieved greater reach distances, which may be linked to reductions in passive stiffness and increased flexibility. It was therefore suggested that the y-balance anterior reach test is constrained more by flexibility than dynamic stability as has been previously suggested. Single leg maximum hop for distance increased linearly with age but when analysed relative to leg length, a reduction in performance was reported in the U13 age group. This could be linked to a period of adolescent awkwardness during which motor control is compromised. A trend of increased absolute peak landing forces was observed with each incremental age group. Absolute landing forces were also higher on the right leg in the 75% horizontal hop from the U13 chronological age group, suggesting the emergence of limb dominance from an early age which may be linked to early soccer specialisation. In the 183

184 SLCMJ, a trend of higher absolute landing forces was evident on the left leg, which highlights discrepancies in vertical and horizontal tasks; profiling should be completed across multiple planes of motion. Relative landing forces demonstrated a more variable pattern across chronological age groups and test measures. During the 75%Hop the highest forces were reported in the youngest and oldest groups respectively. High forces in the older players may be due to high volumes of accumulated soccer training and competitions, which lead to altered kinematic strategies indicative of a stiffer landing normally associated with younger athletes. The presence of high risk landing kinematics in the U18s may subsequently increase their risk of injury due to increased body mass, limb lengths and velocities of play. In the SLCMJ, increased landing forces were evident in the U14s, which may be aligned to greater injury incidence during this period associated with rapid growth in skeletal structures. Knee valgus during the tuck jump assessment appears to reduce with advancing age suggesting better control of joint positioning which is probably due to increases in strength and motor control. Also, during specified chronological age periods, greater asymmetries were present between valgus grade scores on each leg. This may increase the risk of traumatic injury during this period due to asymmetrical loading of passive knee structures. Asymmetry scores were greater on both landing force assessments than measures of distance during the single leg maximum hop and y-balance anterior reach test. This suggests force measures display greater sensitivity in their ability to identify disparity 184

185 between limbs and may be more useful in detecting players who are at a greater risk of injury. Asymmetry in landing forces also increased in the oldest age group in this study, which may provide a plausible explanation for their increased risk of injury due to an inability to effectively attenuate landing forces during a single leg landing. 185

186 Chapter 5 STUDY 3: SEASONAL VARIATION IN FIELD BASED MEASURES OF NEUROMUSCULAR CONTROL IN ELITE MALE YOUTH SOCCER PLAYERS 5.1 Introduction Elite male youth soccer players who regularly participate in training and competitions display an inherent risk of injury (Price et al., 2004; Le Gall et al., 2006; Brink et al., 2010). Throughout the course of a season, injury risk may also be greater at certain time points. In adult male professional players, peak injury rates have been shown during the pre-season period and following a mid-season break (Hawkins et al., 2001). A similar pattern has been observed in elite male youth players (Price et al., 2004; Le Gall et al., 2006; Cloke et al. 2009; Cloke et al. 2011), although some variation is apparent when comparing injuries that occur in competition vs. training (Price et al., 2004). Heightened incidence of injury during early season periods could be linked to environmental factors such as, firm ground and climate, however, accumulated fatigue and reduced neuromuscular control could also be contributing factors. A spike in injury rates following the mid-season break could indicate a lack of conditioning, whereby, players are not physically prepared for the demands of training after a period of de-training (Price et al., 2004). In sports with long seasons, frequent training and competitions may limit opportunities for recovery and decrease performance, leading to changes in neuromuscular control and altered movement mechanics (Caldwell et al., 2009). Therefore, inclusion of in-season monitoring will enable practitioners to objectively determine critical periods of heightened injury risk that may be associated with reductions in performance. 186

187 Currently, there are limited data available to examine the seasonal variation in performance of elite soccer players. A plausible explanation could be limited opportunities for repeated testing with the same players and the difficulty of controlling for confounding variables (Reilly, 2005). For example, in-season injuries, varying competition schedules and transfers between clubs can contribute to high attrition rates. Of the few existing studies in elite male players, outcome variables have included anthropometrics and performance based assessments. Specifically, measures of body composition, (Casajus, 2001; Ostojic, 2003), sprint speed (Ostojic, 2003), running based endurance tests of both aerobic (Casajus, 2001; Clark et al., 2008) and anaerobic capacity (Clark et al., 2008), and jumping tasks (Casajus, 2001; Clark et al., 2008). However, limited data is available to report the seasonal variation in measures of neuromuscular control and injury risk screening. There is a paucity of existing literature to observe seasonal performance changes in youth athletes. Of the few studies available, jump height and selected markers of isokinetic leg strength have been shown to increase from the pre-season to competition period in adolescent female volleyball players (Rousanoglou et al., 2013). The authors attributed heightened jump performance to improvements in coordination, effective utilisation of the stretch-shortening cycle and more powerful force production. Performance markers have also been monitored at 6 month intervals over a 3 year period in elite male youth soccer players (Williams et al., 2011). A strong linear relationship was observed in both jump height (yearly rate of change = 6.9%) and sprint speed (3.1% and 2.7% for 10m and 30m sprint respectively). However, large individual variability was present that could be attributed to the effects of growth and maturation (Williams et al., 2011). Another study utilising elite male youth soccer players examined a range of physiological variables pre and post season including, sprint speed, aerobic capacity and jumping 187

188 tasks (Gravina et al., 2008). Noted changes in performance on certain tests were associated with growth and maturation, specifically, height, weight and salivary testosterone. Additionally, a significant decline in drop jump height was evident at the end of the season, which may have indicated a reduced ability to attenuate landing forces and ineffective utilisation of neurophysiological stretch-shortening cycle mechanisms. Due to the shortage of data in youth subjects and the absence of investigations that have included seasonal measures of neuromuscular control to indicate injury risk, further research is warranted. In order to accurately monitor seasonal changes in neuromuscular control, consideration must be given to both the short and long term reliability of the measures used. Short term reliability was established in study 1, and the data was used to inform the selection of tests and variables showing acceptable reliability for future investigations. Due to both variations in performance, and the growth and maturity related changes that can occur throughout a soccer season (Gravina et al., 2008), measurement of long term reliability in elite male youth soccer players is also required. Buchheit and Mendez-Villaneuva (2013) analysed the long term stability ranking of anthropometrics, sprinting, jumping and maximal aerobic speed testing in elite male youth soccer players over a 4-year period. The level of stability was measure-dependent ranging from moderate to strong (ICC = ). However, no studies have examined changes in neuromuscular control throughout a soccer season in male youth players. Therefore, investigations are needed to determine long term within-subject variation and establish the inseason stability of such measures so that meaningful changes can be recognised which may relate to their relative risk of injury. Cumulatively, there is a lack of available evidence to report seasonal changes in neuromuscular control during field-based assessments in elite male youth soccer players. Thus, 188

189 the aim of the current study is to examine seasonal changes and variability in tasks that require repeated maximal jumps, single leg jump-landing, dynamic balance and quantification of landing force at three time points during an academy soccer season. 5.2 Methods Participants Forty three elite male youth soccer players (age 13.1 ± 2.2 yr; height ± 15.7 cm; body mass 49.4 ± 14.3 kg; maturity offset 0.2 ± 1.9 yr) from the academy of an English Premier League soccer club volunteered to take part in this study. Participants were familiar with regular performance assessments, reported no injuries at the time of testing and were all participating regularly in football training and competitions in accordance with the regulations for player contact hours as set out by the Premier League Elite Player Performance Plan. Parental consent and participant assent were collected prior to the commencement of testing, in addition to a physical activity readiness questionnaire. Ethical approval was granted by the institutional ethics committee in accordance with the declaration of Helsinki Experimental design This study used a repeated measures design to determine the seasonal variation in a range of field-based neuromuscular control assessments. Participants were required to attend the club training ground and complete three experimental test sessions throughout the course of the season. The schedule of testing was as follows: test 1 (July) during pre-season (PRE); test 2 189

190 (January) during mid-season (MID); test 3, (May) at the end of season (END). Anthropometric measures (standing height, leg length and body mass) and four different jumping protocols were analysed, including: (1) single leg hop for distance (SLHD); (2) a ten-second repeated tuck jump (TJ) assessment analysed using two-dimensional video analysis; (3) a single leg 75% horizontal hop and stick (75%Hop), and (4) a single leg countermovement jump (SLCMJ) onto a force plate. Prior to each testing session, a 10-minute standardised warm up was completed consisting of dynamic stretching. The order of testing was randomised using a counterbalanced design to reduce the potential for an order effect. Practice trials were provided for each test and for the purposes of data collection, three trials were analysed. One minute of recovery was allowed between trials based on previous recommendations (Ebben et al., 2010). Testing was completed by the same research group, at the same venue, on the same day of the week and at the same time on each day. Participants were asked to wear the same training kit and footwear, and refrain from strenuous exercise at least 48 hours prior to testing. Participants were also asked to eat according to their normal diet and avoid eating and drinking substances other than water one hour prior to each test session Procedures The protocols for each jumping task replicated those of study 2; please refer to the procedures section in that chapter. All tests were conducted by the same researcher at each test session. Participant instructions were standardised to ensure replication of conditions between days (see appendix d) and removal of the potential for of inter-rater error. However, use of the same rater on the Y-balance assessment was not possible as the researcher who completed the tests in pre- 190

191 season was unable to be present at the subsequent mid and end of season test sessions. Therefore, the Y-balance assessment was removed from the testing battery due to unacceptable inter-rater reliability Statistical analysis Data was checked for normality and descriptive statistics for each variable were determined at all test sessions occurring throughout the season (PRE/MID/END). Percentage change in test scores between consecutive test sessions was calculated from the mean change across all participants. Pearson correlations were also used to examine the relationships between changes in anthropometrics and each outcome measure from test 1 (PRE) and test 3 (END). Association between variables were interpreted using the following classifications: trivial < 0.2; small ; moderate; > 0.8 large based on previous recommendations (Cohen, 1988). A series of repeated measures ANOVA were also used to test for significant differences between each test session. Sphericity of the data was checked by Maulchy s statistic, and where violated Greenhouse-Geiser adjustment was applied. Bonferroni post-hoc tests were used to identify the origin of any significant differences between test sessions, with the level of statistical significance set at alpha level p All data was computed through Microsoft Excel Pearson product moment correlations and repeated measures ANOVA were processed using SPSS (V.21. Chicago Illinois). Long term reliability of the neuromuscular control tests was measured throughout the season using intra-class correlation coefficients (ICC) to determine rank order repeatability; and typical error of estimates. Within-subject variation was reported using mean coefficients of 191

192 variation (CV %). 95% confidence intervals (95% CI) were used, and all reliability data was computed through Microsoft Excel 2010 using a freely available spread sheet (Hopkins, 2006). 5.3 Results Descriptive statistics and percentage change in test scores for each variable are displayed in table 5.1. All anthropometric variables showed significant increases at each test session, whereas, a less consistent pattern was reported for the neuromuscular control tests. A trend was observed of increased peak vertical ground reaction force (pvgrf) throughout the season in both the 75% hop and SLCMJ and this was more evident on the left leg. SLHD increased linearly on both legs with percentage change scores reaching statistical significance between PRE-END and MID- END on the right and left legs respectively. SLHD asymmetry significantly reduced throughout the season and a trend was observed of reduced asymmetry in pvgrf for both the 75% hop and SLCMJ but this did not reach statistical significance. In the assessment of knee valgus during the tuck jump, scores ranged for 0 3 across all players at each test session. Pre-season knee valgus mode scores were 2 on both legs. From PRE- MID season, a reduction in the mode score from 2 to 1 was shown on the right leg, with no further changes in mode score observed from MID-END. A mode score of 2 was recorded on the left leg at each test session, as indicated no change throughout the season. 192

193 Table 5.1 Descriptive statistics and percentage change in test scores across the season Test Variable Mean (PRE) Mean (MID) Mean (END) % Change in Mean Right PRE-MID MID-END PRE-END Height (cm) ± ± ± ± 1.2* 0.9 ± 0.7* 2.3 ± 1.3** Mass (kg) 49.4 ± ± ± ± 4.7** 3.1 ± 2.8** 8.2 ± 5.1** Leg Length (cm) 83.0 ± ± ± ± 2.1** 0.9 ± 1.3* 4.2 ± 2.3** 75%Hop pvgrf R (N) 1690 ± ± ± ± ± ± %Hop pvgrf L (N) 1619 ± ± ± ± ± 12.9** 6.6 ± 15.4** 75%Hop pvgrf Asym (%) 86.5 ± ± ± ± ± ± 13.4 SLCMJ pvgrf R (N) 1550 ± ± ± ± ± ± 14.3* SLCMJ pvgrf L (N) 1286 ± ± ± ± 15.1** 11.6 ± 12.8** 27.8 ± 13.8** SLCMJ pvgrf Asym (%) 81.9 ± ± ± ± ± ± 13.8 SLHD R (m) 1.43 ± ± ± ± ± 9.6* 5.8 ± 10.9** SLHD L (m) 1.42 ± ± ± ± ± 10.1** 5.7 ± 10.4** SLHD Asym (%) 92.9 ± ± ± ± ± 6.6* 2.9 ± 5.1* * Significant at the level of p 0.05 ** Significant at the level of p 0.01 Note: SLHD = single leg hop for distance; pvgrf = peak vertical ground reaction force; SLCMJ SLCMJ = single leg countermovement jump; Asym = asymmetry 193

194 With the exception of asymmetry scores, all variables recorded at each test session showed moderate to very strong relationships (r = ; p < 0.01). However, while some significant correlations were reported between the change in anthropometrics and change in each neuromuscular test measure, the strength of these relationships was consistently weak (r.37). All reliability measures calculated for each test across the season are displayed in table 5.2. Strong to very strong relationships were shown between test sessions for all variables (ICC range = ) indicating maintenance of rank order. Within-subject variation also displayed largely acceptable values (range = %). Typical error values for tuck jump knee valgus score were deemed acceptable. 194

195 Table 5.2 Seasonal reliability statistics for all variables ICC Typical Error CV % (95% CI) Test Variable PRE-MID MID-END PRE-END PRE-MID MID-END PRE-END PRE-MID MID-END PRE-END 75% Hop pvgrf R ( ) 13.9 ( ) 16.6 ( ) 75% Hop pvgrf L ( ) 10.4 ( ) 12.5 ( ) 75%Hop PVGRF Asym ( ) 7.9 ( ) 10.8 ( ) SLCMJ pvgrf R ( ) 11.1 ( ) 11.3 (9.6-14) SLCMJ pvgrf L ( ) 10.8 ( ) 15 ( ) SLCMJ pvgrf Asym ( ) 8.5 ( ) 12 ( ) SLHD R ( ) 6.6 ( ) 7.6 ( ) SLHD L ( ) 6.9 ( ) 7.7 ( ) SLHD Asym ( ) 5.2 ( ) 3.8 ( ) TJ Knee Valgus R TJ Knee Valgus L Note: SLHD = single leg hop for distance; TJ = Tuck Jump; SLCMJ = single leg countermovement jump; 195

196 5.4 Discussion The current study assessed the seasonal variation in performance and long term reliability of a field-based neuromuscular control screening battery in elite male youth soccer players. This data is novel as it is the first investigation that has used elite male youth soccer players to determine seasonal changes in neuromuscular control. Results showed that height, weight and leg length increased throughout the season, whereas, changes in neuromuscular control were more variable. Increased landing forces were displayed throughout the season, in addition to greater distances achieved during the SLHD. Significant relationships were reported between all test sessions, and within-subject variation was considered largely acceptable (range = %). However, the seasonal variation on a number of neuromuscular control tests included was often considerably lower than the random variation, thus observed changes may not be meaningful. Assessment of landing forces Peak vertical landing forces increased significantly throughout the season in both the 75%Hop and SLCMJ with a more pronounced increase on the left leg. In particular, the observed seasonal changes in landing forces on the left leg during the SLCMJ were greater than the random variation at each respective test session (as shown by the coefficient of variation values), indicating a real change in performance that could heighten the risk of injury. Increases in body mass could be a contributing factor; however, weak relationships were shown between changes in landing force and changes in body mass. Another plausible explanation could be increases in jump performance as indicated by improvements during the single leg hop for distance. Greater displacement of centre of mass during any jump task will further challenge dynamic stability and 196

197 dissipation of landing forces (Goosens et al., 2015). Thus, a developmental lag over the course of the season may be present in the attenuation of ground reaction forces upon landing, in-spite of improvements in jumping performance. This may heighten injury risk in youth players and highlights the need for in-season monitoring and year round training interventions that focus on neuromuscular control during single leg jumping tasks. There is a paucity of literature to examine seasonal changes in landing forces, and no data available for youth soccer players. A season long investigation in collegiate male soccer players observed reductions in drop jump performance that were associated with heightened fatigue and perceived stress measured by recovery stress states (RESTQ-Sport) and increases in the global stress score (Rossi et al., 2011). Fatigue may induce changes in biomechanical and neuromuscular function such as muscle activation sequences, kinetics and kinematics (Padua et al., 2006). These decrements have been associated with reductions in joint stability and increased injury risk due to greater stress placed on soft tissue structures (Padua et al., 2005; Hughes and Watkins, 2006). While not representative of long-term changes, the acute effects of soccerspecific fatigue on landing forces have been measured in male youth soccer players (Oliver et al., 2008). Reductions in jump height were present during squat jumps, countermovement jumps and drop jumps, however, impact forces during the drop jumps were the only landing force variables to show a significant change in response to fatigue. The current study used single leg tasks which are more representative of the loading patterns experienced during game play, and may be more sensitive in identifying fatigue than bilateral variations including the squat jump and countermovement jump. Increased impact forces and skeletal loading associated with decrements in landing performance might be of relevance for identifying injury risk. However, the aforementioned study of Oliver et al. (2008) measured acute fatigue, whereas, the present study 197

198 utilised a repeated measures seasonal long design and additional measures to quantify fatigue were not included. Thus, this concept is speculative and further investigations are required to determine changes in neuromuscular control and their relationships to the accumulation of chronic fatigue. Single leg hop for distance Data showed that single leg hop for distance increased significantly throughout the season on both the right and left legs respectively. These findings are consistent with previous literature which has observed in-season increases in jump performance in female volleyball players (Rousanoglou et al., 2013) and amateur adult male soccer players (Caldwell et al., 2009). Conversely, in professional male soccer players, peak performances have been noted mid-season (Clark et al., 2008), followed by a decline towards the end of the season (Casajus, 2001; Clark et al., 2008). A contributing factor could be the congestion of matches in the latter part of the season and the accumulation of high amounts of fatigue. Also, reductions in training frequency due to a higher number of competitions may limit further performance adaptations. Performance increases in the present study are consistent with previous data in elite male youth soccer players, whereby, sprint and jump performances increased consistently over a three year period assessed at 6-month intervals (Williams et al., 2011). A plausible explanation for increases in jump performance could be due to a deconditioned state in the pre-season period and post mid-season break. This is supported by the highest percentage change in performances occurring from MID-END of season testing, and this timeperiod represents the longest sequence of soccer training and competitions without a scheduled 198

199 break. This also indicates the need to include integrative neuromuscular training (INT) during off-season periods to ensure performance reductions are minimised. This is supported by increases in physiological performance in response to a 12-week pre-season strength and power training programmeme in male youth soccer players (Wong et al., 2010). Such interventions should also be maintained in-season as improvements in jumping tasks are possible in male youth soccer players (Meylan and Malatesta, 2009b), as are increases in vertical jump performance during a soccer season when resistance and plyometric training is performed regularly in adult professional players (Wisloff et al., 2004). Finally, increases in muscular strength as a result of growth and maturation could have contributed to improved jump performance. This can be evidenced by increasing hop distances in each consecutive age group as identified in study 2, and reports that the standing long jump test has demonstrated strong relationships with lower body muscular strength in youth athletes (Castro-Pinero et al., 2010). Tuck jump knee valgus tuck jump score To the knowledge of the authors, no research is available to examine seasonal variation in knee valgus scores which limits comparisons with the current study. Absolute knee valgus mode scores reduced from those reported during pre-season screening on the right leg, and this remained consistent at the end of season testing. The typical error scores for this test showed a change in score > 1 exceeds the random variation and could therefore be considered real. Conversely, absolute knee valgus mode scores remained the same at each respective test session throughout the season on the left leg. This finding can also be linked to significant increases in landing forces on the left leg during the SLCMJ. These neuromuscular alterations may lead to an 199

200 overall reduction in dynamic stabilisation upon ground contact (Fukushi Yamada et al., 2012), thus placing the lower limb at increased risk of injury. Muscle loading patterns which asymmetrically strengthen the quadriceps during soccer training and competitions may alter the reciprocal balance of strength and dynamic stabilisation around the knee as indicated by compromised function of the hamstrings during high velocity actions in male youth players (Iga et al., 2009). The hamstring musculature plays a key role in compressing the medial knee joint, limiting knee valgus and tibial external rotation, subsequently reducing the stress on the ACL (Zebis et al., 2009). Speculatively, in the current study, changes in neuromuscular control due to repeated exposures of soccer activities may have induced greater quadriceps dominance in the left leg from recurring kicking motions with the right leg, thus loading the left leg in asymmetrical volumes during the stance phase that could have led to overall changes in motor control. However, due to the paucity of data to support this notion, caution should be applied when interpreting these findings and further research is warranted. Asymmetry A general trend was evident of reductions in asymmetry throughout the season. This finding could be considered unexpected due to the accumulated exposure of sport-specific training and competition that may result in the emergence of increased leg dominance (Daneshjoo et al., 2013). However, asymmetry has been reported to reduce with a greater training age in professional soccer players (Fousekis et al., 2010), that may in part explain the findings of the current study. Also, the ICC values for each asymmetry variable were considerably lower than all other test measures included in the current study. This suggests a sizeable change in the 200

201 maintenance of rank order between players; increased variability could be present with this test. However, the typical error displayed acceptable values; this measure by be accurately tracked at an individual level across the season. Thus, longitudinal tracking of asymmetry is recommended to identify individuals who demonstrate large changes in this risk factor, due to its reported associations with injury (Crosier and Creelard, 2000). Relationships between changes in anthropometrics and neuromuscular control Relationships between the change in anthropometrics and change in each neuromuscular control test were consistently low (r.37). This suggests that alterations in performance were weakly related to anthropometric changes and other factors need to be considered. These findings are consistent with previous literature assessing seasonal changes in performance measures with elite male youth soccer players (Gravina et al., 2008; Williams et al., 2011). Longitudinal tracking of sprint and jumping tasks reported weak associations between changes in body size and performance (Williams et al., 2011). Also, although significant correlations have been reported between both anthropometrics and changes in salivary testosterone and jumping performance (r =.39 to.48) (Gravina et al., 2008), the strength of these relationships should be classified as weak-moderate. In the present study, additional factors contributing to negative performance changes could include team selection motivation, task-related motivation, accumulated fatigue, and subsequent changes in neuromuscular control. Long term reliability 201

202 In contrast to study 1, all variables (with the exception of asymmetry) showed acceptable maintenance of rank order between test sessions as indicated by strong to very strong intra-class correlation coefficients (ICC) (range =.70 to.90). In study 1, lower ICC values were reported between test sessions separated by a week which is often the case in homogenous samples (Hopkins, 2000). While the participants in the present study are reflective of a homogenous group due to their status as elite male youth academy soccer players, the current study included a more heterogeneous sample that contains a larger chronological age and maturation range. In pre-phv athletes, levels of performance may be more clustered as they have not yet experienced their peak growth spurt; similar levels of force production and landing abilities can be expected increasing the possibility of players changing their rank order, resulting in a lower ICC. Also, table 1 showed that distinct variability in performance was present on selected test measures across the sample as indicated by standard deviation values that were greater than the mean. In this study, the decision to analyse the participants in one group was based on sample size estimations, whereby, groups of > 40 were deemed necessary to accurately determine any significant between-group differences for some of the outcome measures included in this study. Future investigations should consider examining the effects of seasonal variation on long term reliability in youth soccer players at different stages of growth and maturation with clear group divisions. The within-subject variation recorded across the season could also be considered largely acceptable (range = %). To determine if a worthwhile change has occurred, the precision of the measures used must be considered. If the observed percentage change score is greater than the reported error (in this case the typical error measurement) this could be considered a meaningful effect. In the present study, with the exception of the SLCMJ, changes 202

203 in performance were predominantly lower than the random variation (CV %) across the season, indicating that the observed percentage changes unlikely to be real. The moderate to low within-subject variation in neuromuscular control test scores suggests that their relative risk of injury remains consistent (and in some cases increases) throughout the season highlighting the need for targeted interventions to improve neuromuscular control upon landing. Examples include plyometric exercises to enhance functional joint stability (Chimera et al., 2004), and a periodised INT programmeme which may improve lower extremity kinematics (Myer et al., 2006a; Myer et al., 2011a), in addition to their regular soccer practice Summary and practical applications This is the first study to provide data on the seasonal variation for a range of field-based neuromuscular control assessments in elite male youth soccer players. The findings of this study suggest that neuromuscular control is likely to reduce over the course of a competitive season, subsequently increasing the relative risk of lower extremity injury. Data are also now available to report the long-term reliability and stability of field based neuromuscular control assessments across a soccer season. This information can be used to identify meaningful changes in performance following a training intervention, and potential for increased injury risk utilising seasonal test re-test comparisons. The key findings are summarised below: pvgrf increased significantly throughout the season in both the 75%Hop and SLCMJ, and this was more pronounced on the left leg. TJ knee valgus scores showed that scores obtained on the left leg could be indicative of greater injury risk as the season progresses, whereas an improvement was displayed on 203

204 the right leg. Contributing factors could include; alterations in neuromuscular control, greater jump distances / heights and the presence of accumulated fatigue. SLHD increased significantly throughout the season on both the right and left legs respectively. This may be due to reduced levels of conditioning in the pre-season period and post mid-season break as supported by the highest percentage change in performances occurring from MID-END of season testing, and this time-period represents the longest sequence of soccer training and competitions without a scheduled break. Nearly all variables showed acceptable maintenance of rank order between test sessions and within-subject variation recorded across the season was also considered low although some variation was present based on the test measure analysed. Moderate to low within-subject variation in neuromuscular control test scores and percentage change score below the level of random variation suggests that in this sample of elite male youth soccer players, performance remains largely unchanged throughout the season. This highlights the need for in-season training interventions to improve functional joint stability landing mechanics in addition to regular soccer practice. Relationships between the change in anthropometrics and change in each neuromuscular control test were consistently low suggesting that other factors need to be considered to when interpreting alterations in performance. 204

205 Chapter 6 STUDY 4: INJURY OCCURRENCE IN ELITE ENGLISH MALE YOUTH SOCCER PLAYERS 6.1 Introduction In the sport of soccer, ensuring that players are free from injury is paramount for successful performance (Eriale et al., 2012). Specifically, lower injury incidence has shown strong relationships with higher league positions, greater number of games won, increased goals scored, goal difference, and total number of points achieved (r = 0.80) (Eriale et al., 2012). In youth soccer players, reducing injury risk is of great importance due to an inherent risk to the immature skeleton in response to high training loads (Mafulli and Pintore, 1990), the frequency of overuse injuries (Schmikli and Bol, 1995) and their status as developmental athletes with the aim of achieving a professional contract (Le Gall et al., 2006). The first step towards injury prevention is to establish the extent of the problem and this involves assessing the incidence, severity and profile of injuries that occur in the sport (Le Gall, 2006). Such information provides a greater understanding of the demands of the game and consequently, targeted prevention measures can be implemented. Variability in the number and characteristics of soccer injuries reported within the available literature may be reflective of methodological differences, player characteristics and geographical location (Brito et al., 2012). Also, current data is required to ensure accurate interpretation of results and identification of trends that may emerge due to changes in competition demands. 205

206 A growing body of evidence has examined the injury incidence in elite male youth soccer players (Junge et al., 2000; Volpi et al., 2003; Price et al., 2004; Le Gall et al., 2006; Cloke et al., 2009; Brink et al., 2010; Cloke et al., 2011; Moore et al., 2011; van der Sluis et al., 2015). Prospective analysis has shown a linear increase in injury with age (Price et al., 2004), conversely, more recent data has shown periods of heightened risk around the time of peak height velocity (PHV) (Rumpf and Cronin, 2012; van der Sluis et al., 2014). Overall injury incidence rate has ranged from injuries per 1000 hours of exposure (Junge et al., 2000; Le Gall et al., 2006; Brink et al., 2010) or a frequency of 0.40 injuries per player, per season (Price et al., 2004). However, variation appears to be evident when incidence rates are analyzed by chronological age (Price et al., 2004; Le Gall et al., 2006), with few studies reporting injury occurrence across a wide range of chronological age groups reflective of the structure within a professional soccer academy. When interpreting existing research, it should be acknowledged that data collection in each of these studies (Price et al., 2004; Cloke et al., 2011; Cloke et al., 2011; Moore et al., 2011) occurred prior to the implementation of the Elite Player Performance Plan (EPPP) devised by the English Premier League. Under their enforced regulations, a substantial increase in the volume of soccer specific training has occurred in boys as young as nine years of age (EPPP, 2011). Of concern, the EPPP adopts a linear approach to increases in training volume from aged years, which coincides with the aforementioned period of PHV and associated heightened injury risk (Rump and Cronin, 2012; van der Sluis et al., 2014). The impact of such a significant increase in training volume and the related injury risk for young athletes that are experiencing a range of growth and maturational processes is currently unknown (Read et al., in press c). Therefore, the purpose of this study was to assess the player incidence rate, type, location, 206

207 severity and seasonal variation of injuries sustained by elite male youth soccer players aged years over the course of a competitive soccer season. 6.2 Methods Experimental design This study employed a prospective cohort design, tracking injury occurrence in elite male youth academy soccer players in the United Kingdom. Data was recorded between July 2014 and May 2015, encapsulating the period of the competitive soccer season. Six professional soccer academies agreed to participate in the study, which involved all registered players with an age range from years. The total sample of players included in the study was N = 608. All players were participating regularly in football training and competitions in accordance with the regulations for player contact hours as set out by the Premier League Elite Player Performance Plan (EPPP). Parental consent and participant assent were collected prior to the commencement of the study. Ethical approval was granted by the institutional ethics committee in accordance with the declaration of Helsinki Procedures Injuries experienced during the study period were diagnosed and prospectively recorded by medical personnel of each club in accordance with the guidelines set out by the Premier League EPPP. A prospective design was used to reduce the risk of recall bias and incorrect injury interpretation (Le Gall et al., 2006). Injuries were recorded if they were experienced during 207

208 soccer-related activities and if the player was subsequently unable to participate in training or competition for a minimum of 48 hours following the incident, not including the day of injury (Price et al., 2004; Le Gall et al., 2006). Players were classified as injured until the medical staff of their respective clubs deemed they were fit to resume full training. Time absent due to illness was not included. The information reported for each injury included: player incidence rate, anatomical location, injury type, mechanism (contact vs. non-contact), severity, date of injury and number of days absent. Player incidence rate was calculated by dividing the number of injuries sustained by the total number of registered players. Injury severity was classified based on the number of days missed as per previous recommendations (Price et al., 2004; Le Gall et al., 2006; Brito et al., 2012). Categories of injury severity included: slight (2-3 days), minor (4-7 days), moderate (1-4 weeks) and severe (> 4 weeks) (Price et al., 2004; Le Gall et al., 2006; Brito et al., 2012). Injury mechanism was defined using previous guidelines (Brito et al., 2012), whereby a contact or non-contact injury was indicated when an incident with clear contact or collision from another player, the ball or another object either did, or did not, occur respectively. An overuse injury was defined as a condition with a gradual onset associated with repetitive micro trauma and where no clearly identifiable acute incident was present (Brito et al., 2012). The date of injury was also recorded to examine the effects of seasonal variation on the number of injuries experienced during each calendar month. Data that may be of interest to practitioners but was not reported in this study included: the specific mechanism of injury (i.e., jumping, cutting, running etc.), which leg was injured (dominant vs. non-dominant), the time in the match when the injury occurred, and if injuries were experienced in training or competition. This information was omitted as it is not mandatory 208

209 to report these details in accordance with the regulations set out by the Premier League EPPP. Thus, concerns regarding the consistency, reliability and comprehensiveness of the dataset would be indicated Statistical analysis Descriptive data were analysed using frequencies and percentages. Non-parametric data were analysed using Chi-squared (χ 2 ) and Kruskal-Wallis tests to investigate between-group differences for seasonal variation, number of injuries and the time loss per injury. Statistical significance was set at an alpha level of p In the Chi-squared test, analysis of the standardized residuals was completed to identify frequencies that would be considered larger in magnitude than might be expected by chance (Sharp, 2015). Standardized residual values were interpreted using the +/- 2 criteria, whereby cell residuals that are greater than what might be expected by chance were deemed significant (Haberman, 1973; MacDonald and Gardner, 2000). Descriptive data was computed through Microsoft Excel 2010 and Chi-squared and Kruskal- Wallis tests were calculated using SPSS (V.21. Chicago Illinois). 6.4 Results Total injuries and player injury rate: From the cohort of 608 players included in this study, 804 injuries were recorded during the course of the season across all age groups, equating to an average injury rate of 1.32 injuries per player and a mean time loss of 21.9 days per injury. The number of injuries, mean time loss per injury and injury rates per player for each respective age group are shown in table 6.1. The U11s sustained significantly fewer injuries than all other age 209

210 groups (p <.001), whereas the U18s incurred significantly more injuries than all age groups, apart from the U15s (p <.05). A trend was evident of greater time loss per injury in the U14s and U15s and these differences were statistically significant when comparing the group with the lowest time loss per injury (U11s). No other group comparisons were statistically significant. Table 6.1 Number of injuries, player incidence rate and mean time loss per age group Age Group # registered players # Injuries Incidence Rate Time Loss (days) U * U U U U U U ** * significantly different form all age groups (p < 0.001) ** significantly different from the U11s (p < 0.001), U12s, U13s, U14s, U16s (p < 0.05) Injury location and type: Injuries were largely non-contact in nature (62.1%), and occurred predominantly in the lower extremities (78%). The anatomical location of all injuries sustained is displayed in table 6.2. With all the age groups combined, the knee and ankle were the most frequent sites of injury (20% and 18.3% respectively), followed by the quadriceps (9.5%). Knee and ankle injuries were also consistently the highest in all age groups; however, some variation was evident between groups (table 6.3). 210

211 Table 6.2 Anatomical location of injuries sustained Injury Location # % Knee Ankle Quad Foot Groin Head Hamstring Hip Lower back Calf Hand Shin Shoulder Arm Pelvis Other Wrist Abdomen Table 6.3 Anatomical location of injuries per age group U11 U12 U13 U14 U15 U16 U18 Location % % % % % % % Ankle Calf Foot Hamstring Hip Knee Quad Groin Shin

212 The overall types of injury experienced and sub-divisions by age group are displayed in tables 6.4 and 6.5. Muscle strains (20.9%) were the most frequently reported injury and there was a high proportion of ligament sprains (16.9%). A large number of injuries were also due to unknown causes (16.3%) or other diagnoses (10.7%). Muscular strains and ligament sprains were consistently the most recurrent diagnosis; however, some variation was evident between age groups. Muscle strains were most prominent in the U15 and U18 age group, whereas, ligament sprains occurred most frequently in the oldest two age groups (U16 and U18). The greatest number of muscle strains was present in the quadriceps (32%), with a relatively equal proportion occurring at the hamstrings, groin and hip (figure 6.1). The majority of ligament sprain injuries were sustained at the ankle (65%) and knee (32%). The knees were the most frequent site of tendinopathy (45%) and overuse symptoms (61%). Growth related injuries were less common (6.6%), with the highest number recorded in the U13 and U14 age groups, predominantly diagnosed as Osgood-Schlatter s or Sever s disease. The distribution by age group of these two growth related injuries is displayed in figure

213 Table 6.4 Overall types of injury sustained Injury Type # Injuries % Muscle Strain Unknown Cause Ligament Sprain Other diagnosis Growth/Overuse Tissue Bruising Overuse Tendinopathy Low Back Pain Fracture Muscle Contusion Cut Inflammatory Synovitis Meniscal Tear Ligament Rupture Periostitis Dislocation Table 6.5 Types of injury sustained per age group U11 U12 U13 U14 U15 U16 U18 Injury Type % % % % % % % Cut Dislocation Fracture Growth/Overuse Inflammatory Synovitis Ligament Rupture Ligament Sprain Low Back Pain Meniscal Tear Muscle Contusion Muscle Strain Other diagnosis Overuse Periostitis Tendinopathy Tissue Bruising

214 Number of Injuries Number of Injuries Calf Groin Hamstring Hip Lower Back Anatomical Location Quad Figure 6.1 Anatomical location of muscle strain injuries Sever's Osgood-Schlatter's U11 U12 U13 U14 U15 U16 U18 Chronological Age Group Figure 6.2 Growth overuse injuries per chronological age group 214

215 Player Incidence Rate: Severe Injuries Injury severity: Injuries classified as moderate were most frequently reported (42.9%). Severe injuries accounted for 22% of the total injuries, with minor and slight injuries occurring less frequently (20.4% and 14.7% respectively). Sub-division of severe injuries for each age group is displayed in figure 6.3. A general trend was evident of an increase in the player incidence rate of severe injuries from the U13 age group; however, a peak was shown in the U15s U11 U12 U13 U14 U15 U16 U18 Chronological Age Group Figure 6.3 Player incidence rate for severe injuries in each chronological age group Seasonal variation: The number of injuries sustained during each calendar month is displayed in figure 6.4. There was a linear increase through the pre-season period (July and August), and two distinct injury peaks were evident where the observed frequencies were greater than what might be expected by chance, specifically, in September (p <.001) and January (p <.001). The highest 215

216 Number of injuries number of injuries was recorded in January following the winter break. Significantly fewer numbers of injuries were also shown in May and June (p <.001). A similar pattern was observed for the seasonal variation in the number of severe injuries experienced. Age group comparisons generally confirmed this trend, however; some variation was present Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun All Injuries Severe Injuries Figure 6.4 Seasonal variation of injuries sustained 6.4 Discussion This is the first study since the inception of the EPPP to examine the injury occurrence and associated trends in elite male youth soccer players. A general linear increase was observed in the number of injuries recorded for each respective chronological age group, with the youngest 216

217 and oldest age groups sustaining the lowest and highest number of injuries respectively. However, in-spite of a high player incidence rate in the U18s, the time loss experienced for each injury was comparable to the other age groups. A trend was evident of injuries classified as severe increasing with age, peaking in the U15 chronological age group. The time loss for each injury was also greatest in the U14s and U15s. The most frequent anatomical sites of injury were the knee and ankle. Muscle strains accounted for the greatest percentage of all injuries (most frequently the quadriceps), followed by ligament sprains (predominantly at the ankle), although this was not consistent across all chronological age groups. Seasonal variation indicated two peaks in incidence, specifically in September and January. The overall player injury incidence rate in this study was 1.32 injuries per player over the course of the season. In general, injury incidence rate increased with age, with highest player incidence recorded in the U15 and U18 age groups. Previous comparative injury incidence data in elite male youth players is sparse. Also, because the majority of studies have reported injury incidence rate per 1000 hours of exposure (Junge et al., 2000; Le Gall et al., 2006; Cloke et al., 2011), it is somewhat challenging to directly compare the results of the current investigation. However, the most notable study that has examined injury risk in male youth soccer players within the United Kingdom adopted the same analytical approach as the current study (Price et al., 2004). The authors audited injuries in elite male academy soccer players between 1999 and 2001 and reported a mean player incidence rate of 0.40 injuries per player, per season (Price et al., 2004). The cumulative findings of the current study and that of Price et al. (2004) suggests that during the 13 years between the respective studies there has been a three-fold increase in player incidence rate sustained by elite male youth soccer players during a season. A plausible explanation could be the dramatic increase in player exposure hours that are required to meet the 217

218 specifications within the EPPP (Read et al., in press c). While anecdotally, the impact of such a significant increase in training volume for these young athletes was of concern to practitioners; this study appears to be novel in the provision of evidence to report the potentially deleterious effects of the EPPP on injury risk in male youth soccer players (Read et al., in press c). Also, a high proportion of injuries were non-contact in nature (62.1%), and therefore, may be preventable with appropriate management of training loads and the inclusion of individualised training programmemes that target deficits in neuromuscular control (Read et al., in press). A trend depicting a linear increase in the number of injuries was observed with each respective age group, whereby the highest and lowest numbers of injuries were recorded in the oldest and youngest age groups respectively. However, the time loss experienced for each injury in the U18 s was comparable to the other age groups. The greater frequency of injuries in the U18s may be due to heightened intensities of play and increased exposure, reflective of their status as full time academy scholars. This would likely increase the occurrence of slight and minor incidences. The greatest incidence of severe injuries was recorded in the U15 chronological age group and the longest time loss per injury was identified in the U14s and U15s. This period coincides with a stage of rapid growth (Malina et al., 2004), whereby, potential alterations in motor control may be present (Philapaerts et al., 2006). A greater risk of overall and severe injury has also been reported during this period (van der Sluis et al., 2014; van der Sluis et al., 2015). Cumulatively, the results of the present study combined with those of previous research (van der Sluis et al., 2014; van der Sluis et al., 2015) would suggest that these age groups should be targeted for screening and prevention strategies. However, despite the apparent need for focused attention on those players around or just after peak height velocity, practitioners should be cognizant that interventions applied during the pre-pubertal years are 218

219 deemed critical due to the accelerated periods of neural plasticity associated with pre-pubescence (Gallahue, 1982; Borms, 1986; Hirtz and Starosta, 2002). Speculatively this may reduce the risk of injury during the later stages of maturation. The most frequent anatomical sites of injury were the knee and ankle, followed by the quadriceps. A high proportion of knee and ankle injuries are consistent with previous literature (Junge et al., 2000; Price et al., 2004; Le Gall et al., 2006), however, the upper thigh (Price et al., 2004; Le Gall et al., 2006) and ankle (Junge et al., 2000) have often been cited as the most common anatomical injury location. A plausible explanation for the findings of this study could be the increased exposures to soccer specific practice required following new EPPP regulations (Read et al., in press c) heightening the frequency of jumping and rapid change of direction activities that may amplify injury risk (Daniel et al., 1994). Also a greater proportion of growth (in particular Osgood-Schlatter s) and overuse injuries were reported in the present study compared to previous investigations (Price et al., 2004; Le Gall et al., 2006). Thus, it could be inferred from these data that the knee and ankle present the greatest risk of injury for elite male youth soccer players, which may be exacerbated by higher training volumes. Further research is required to investigate age-related risk factors, appropriate screening techniques and targeted injury prevention strategies with a specific focus on these anatomical locations. Muscle strains and sprains were the most commonly recorded types of injury which corresponds with previous data (Price et al., 2004; Le Gall et al., 2006). The incidence of these types of injury was also shown to increase in older age groups, whereas, growth related injuries peaked in the U13s and U14s. Research indicates that with maturation there is an increased risk of ligament sprains and a concomitant reduction in fractures and growth related injuries (Adrim and Cheng, 2003). This may be due to heightened intensities of play, enlarged body mass and 219

220 changes in lever lengths resulting from a higher centre of mass in older players that will increase joint torques (Adrim and Cheng, 2003; Ford et al., 2010; Ford et al., 2010). The greater frequency of growth related injuries observed in younger players could be associated with the maximal accelerated growth spurt, during which limb length changes may result in altered force dissipation and changes in biomechanical characteristics (Hewett et al., 2004; Shea et al., 2004). The presence of musculoskeletal growth lags following the onset of a growth spurt up to, and around the period of peak height velocity may also increase the risk of overuse and apophyseal injuries during these key periods of growth (Mountjoy et al., 2008; Xu et al., 2009) as indicated in the results of the present study. In light of these findings, it is suggested that practitioners should implement methods to longitudinally monitor the rate of growth to identify periods of rapid change that may increase injury risk in male youth soccer players (Lloyd et al., 2014; Read et al., 2015). In addition to the monitoring of growth, developmentally appropriate integrative neuromuscular training is fundamental for youth soccer players, whereby prescription is commensurate with the technical competency, training age and stage of maturation of the individual (Lloyd et al., 2014). This may offset the injury risk associated with specialised soccer practice from a young age, where exposure to a wide range of developmental motor skill activities is limited (Myer et al., 2011). A high proportion of muscle strains were recorded in the present study, occurring most frequently in the quadriceps. Previous data in this cohort has shown a propensity for hamstring strain injuries, whereby, of all thigh muscle injuries, 43% and 57% of strains were sustained in the quadriceps and hamstrings respectively (Price et al., 2004). This disparity could be attributed to the greater volume of specialised soccer practice that is now required under new EPPP regulations (Read et al., in press c). Increased exposures to repetitive actions such as kicking 220

221 (requiring hip flexion and knee extension) and rapid eccentric loading of the quadriceps to control knee flexion and hip extension, may subsequently increase injury risk (Kary, 2010). Additionally, eccentric strength and flexibility asymmetries in the quadriceps have been reported as risk factors for injury to this muscle group in professional soccer players (Fousekis et al., 2011). In the present study, a peak in the number of quadriceps strains was recorded in the U15 chronological age group. This period in a child s development occurs shortly after the peak growth spurt and has been associated with a plateau or decline in flexibility (Beunen and Malina, 1996; Phillippaearts et al., 2006). Prior growth in skeletal structures provides a stimulus for adaptation of muscle tissue, whereby a time lag exists between the rate of bone growth and subsequent changes in muscle length (Xu et al., 2009). Further, alterations in cross-sectional area occur following growth in muscle length (Xu, 2009). Cumulatively, exposure to high frequencies of intensive soccer specific training, possible reductions in flexibility and a developmental lag in muscle force-producing capabilities following the onset of the adolescent growth spurt could indicate an increased risk of quadriceps strain injuries during this state of growth and maturation. Seasonal variation of injury incidence was present in the current study, with an initial peak in injury incidence following the completion of pre-season (September), which is commensurate with previous research (Price et al., 2004; Le Gall et al., 2006). The linear increase shown from July to September in the present study is likely due to accumulated fatigue following intensive periods of soccer-specific training and the beginning of the competitive season. There is currently a paucity of literature available to examine the relationship between workloads and injury incidence in elite male youth soccer players, however, heightened training loads have been associated with greater injury risk (Gabbett et al., 2012). Physical stress has also demonstrated relationships with injury in elite male youth soccer players (Brink et al., 2010), with a training 221

222 threshold of more than 16 hours per week significantly increasing the risk of injury (Rose et al., 2008). This risk is further increased in youth athletes undertaking specialised sports practice (Jayinthi et al., 2015). A second peak in injury incidence was also recorded in the current study, which occurred immediately after the mid-winter break (January). In a previous study, Price et al. (2004) hypothesized that the increased incidence of injury following the mid-winter break could be a result of decreased levels of conditioning, and/or inappropriate training loads following a period of reduced activity. This pattern was not observed in other investigations (Le Gall et al., 2006), with the authors speculating that the inclusion of shorter but more frequent periods in which players are allowed to rest and recuperate serving as a more appropriate strategy to reduce injury risk (Le Gall et al., 2006). Subsequently, it may be advisable for governing bodies to organize the competitive season with the inclusion of more regular breaks, allowing for a greater focus on recovery and a more gradual accumulation of the required training volume to maximize their development. When interpreting the data of the current study, practitioners should be cognisant of the inherent limitations. While previous research has utilised a similar sample size to the current study (Junge et al., 2000; Cloke et al., 2011; Brito et al., 2012), it should be noted that the sample size was smaller than those that have previously examined injury incidence in elite male youth soccer players within the United Kingdom (Price et al., 2004; Le Gall et al., 2006; Moore et al., 2011). Also, injury data was recorded over a single season and while this has been indicated as the minimum reporting period required (Fuller et al., 2006), seasonal comparison and the identification of longitudinal trends is not possible. Nonetheless, due to the paucity of data since the EPPP, the current experimental design was deemed appropriate and can be supported by longitudinal investigations in the future. Finally, the present study did not account for player 222

223 exposures, whereas, this has been reported in previous investigations using elite male youth soccer players (Junge et al., 2000; Le Gall et al., 2006; Cloke et al., 2011). This information provides an indication of which chronological age groups are most susceptible to injury, accounting for their relative exposure to training and competition (Price et al., 2004). However, this requires accurate quantification of the number and duration of each match and training session participated in by all the players in the cohort (Junge and Dvorak, 2000). In the current study, due to the variation in reporting procedures between clubs and subsequent inaccuracies in measurement, this approach was deemed impractical. Also, solely reporting exposures based on the hours of participation does not account for the nature of the activities performed and disparity between activity patterns of clubs. For example, a training philosophy focusing on small-sided games will increase the frequency of rapid accelerations, decelerations and changes of direction. Conversely, soccer practices that include higher frequencies of phase of play and technical work will bestow different demands. In this study, player incidence rates were measured based on the number of registered players and the number of injuries sustained as this is not affected by the aforementioned variables and provides an indication of injury risk in each age group. Future investigations should consider quantifying the demands of the training and competition activities undertaken to more accurately identify injury incidence and contextualize the exposure data. 6.5 Summary and practical applications This is the first study to examine injury occurrence in elite male youth soccer players in the United Kingdom since the introduction of the EPPP. A considerable increase in player incidence rate was reported compared to earlier research (Price et al., 2004); indicating that the greater training volumes associated with modern-day elite male youth soccer have heightened the 223

224 number of injuries sustained per player during the course of a season. A key tenant of the EPPP is to enhance technical proficiency and aid progression towards higher levels of performance (EPPP, 2011). However, a greater rate of injury will subsequently reduce developmental opportunities and could negatively impact the long term health of these young athletes. The results of this study also indicate a trend whereby the risk of experiencing moderate and severe injuries is greater within U14 and U15 age groups, timeframes that follow rapid growth in skeletal structures. Therefore, assessments to identify if alterations in neuromuscular control exist in this cohort are warranted, and these players should be considered an important focus group for injury prevention strategies targeting potential neuromuscular deficits. Further, due to the large number of non-contact strain and sprain type injuries occurring at the knee and ankle, injury screening and prevention programmes should focus on the mechanisms related to these injuries. Practitioners should also consider stages within the season where players may be at greater risk, specifically at the end of the pre-season period (September) and following the midwinter break (January) and adopt appropriate training load monitoring strategies to identify players who demonstrate aberrant movement patterns and/or the signs and symptoms of overuse type injuries. 224

225 Chapter 7 STUDY 5: VALIDATION OF A NOVEL MOVEMENT SCREEN TO PREDICT LOWER EXTREMITY INJURY IN MALE YOUTH SOCCER PLAYERS 7.1 Introduction Elite male youth soccer players display an inherent risk of injury and should be considered a target group for injury prevention (Price et al., 2004; Schmikli et al., 2011). Prospective assessment of modifiable risk factors is critical to aid in the identification of injury risk prior to their occurrence and the development of targeted strategies for risk reduction. While growth and maturation are confounding factors that may contribute to injury (Kemper et al., 2015; Read et al., 2015), arguably a greater focus should be applied to modifiable risk factors (Bahr and Holme, 2003). Neuromuscular control may be the most modifiable risk factor (Griffin et al., 200; Hewett et al., 2005) and is a predictor of injury in adult athletes (Padua et al., 2009; Small et al., 2010). Less information is available in male youth soccer players and there is currently a paucity of literature to examine the sensitivity of measures of neuromuscular control to predict lower extremity injuries in this cohort. Previous literature has investigated risk factors for injury in adult male and female athletes. These have included assessments of flexibility (Witvrouw et al., 2003; Engebretsen et al., 2010; Henderson et al., 2010), joint hypermobility (Ostenberg and Roos, 2000; Nilsatd et al., 2014), strength diagnostics (Hendersen et al., 2010; Nilstad et al., 2014), three-dimensional landing kinematics (Goerger et al., 2014), and field-based measures of strength and neuromuscular 225

226 control (Engebretsen et al., 2010; Henderson et al., 2010; Nilstad et al., 2014; Lundgren et al., 2015; Goosens et al., 2015). The data indicate varied findings; some studies showing neuromuscular factors were not predictors of lower extremity injury (Engebretsen et al., 2010; Nilstad et al., 2014), while conversely, reduced eccentric hamstring strength and single leg hop for distance (Goosens et al., 2015), heightened landing force asymmetry (Georger et al., 2014) and non-countermovement jump performance (Henderson et al., 2010) were able to identify subjects at a greater risk of lower extremity injury. These differences may in part be due to inconsistencies in the reliability and construct and criterion validity of commonly used testing protocols (Hegedus et al., 2014). This highlights the need for greater consistency and standardisation of the approaches used to screen athletes and the need for further research in a wider range of populations (McCunn et al., in press). A recent systematic review of the measurement properties of commonly used physical performance tests to assess knee function identified that single leg hop for distance was able to differentiate between injured and non-injured athletes (Hegedus et al., 2014). Single leg countermovement jump performance was also related to patient self-reported knee function, however; the predictive ability of all tests reviewed was limited (Hegedus et al., 2014). More recently, Goosen et al. (2015) showed single leg hop for distance scores were a predictor of hamstring injury, although knee injuries were not recorded, thus making comparisons difficult. Hegedus et al. (2015) completed a follow up systematic review of clinician friendly tests for lower extremity function to determine their measurement properties and relationships with injury (Hegedus et al., 2015). Single leg hop for distance was again able to differentiate between injured and non-injured ankles, whereas, star excursion balance reach distance was identified as a predictor of injury. However, the predominance of studies included in these two reviews 226

227 (Hegedus et al., 2014; Hegedus et al., 2015) utilised adult subjects and therefore further research is required to elucidate relevant risk factors for injury in paediatric populations. Analysis of injury risk factors in elite male youth soccer players is sparse, despite the frequency of injuries reported in this cohort (Junge et al., 2000; Price et al., 2004; Le Gall et al., 2006). In male youth athletes, asymmetrical dynamic balance scores have been shown to predict lower extremity injuries, most frequently occurring at the knee and ankle (Plisky et al., 2006). Specifically, an anterior right-left reach difference of greater than 4 cm during the performance of the star excursion balance test increased the risk of sustaining a lower extremity injury. Anthropometrics (Kemper et al., 2015) and maturity status (Johnson et al., 2009) have been identified as predictors of injury; however, these can be considered non-modifiable risk factors. To the knowledge of the authors, only one study has used measures of neuromuscular control and prospectively tracked injuries throughout the course of a soccer season in a sample that included male youth soccer players (Padua et al., 2015). Altered landing kinematics were reported in players who sustained an anterior cruciate ligament (ACL) injury versus non-injured controls. However, a small number of injuries were recorded during the study period (7 injuries from a sample of 829 players), with only one of the injuries sustained to a male player, and their analysis was restricted to ACL injuries only. Assessments of landing kinematics in male youth soccer players are required to determine their sensitivity as predictors of lower extremity injury in this cohort. Due to the multi-factorial nature of commonly experienced injuries in soccer, risk factors should be examined using a variety of tests to determine their relative contributions (Meeuwise, 1994). For example, using a range of single leg jumping tests in previously injured ACL patients, 60% demonstrated abnormal levels of asymmetry (> 15%) on at least one out of two tests. 227

228 However, when only one test was evaluated, the prevalence of abnormality decreased to 42-50% depending on the test used. Furthermore, assessments of leg power across 3 directions (vertical, horizontal, and lateral) have reported non-significant relationships between tests in the various movement planes (Maulder and Cronin, 2005; Meylan et al., 2009; Hewit et al., 2012). Thus, practitioners are encouraged to utilise a range of assessments that target multi-planar actions and different constructs of neuromuscular control (Read et al., in press b). A multifactorial model to indicate injury risk has also recently been validated in junior surfers, with injured athletes demonstrating lower cumulative performance scores than their non-injured counterparts using a predominance of field-based measures of neuromuscular control (Lundegren et al., 2015). The authors concluded that the screening battery was sensitive, easy to implement and interpret, and could be adapted for different sports. Field-based assessments are more appropriate in the context of a soccer academy due to the cost implications and time efficiency required for screening large numbers of athletes (Read et al., in press a). To the knowledge of the authors, the predictive ability of field-based neuromuscular control testing batteries have not been validated in elite male youth soccer players, therefore, further investigation is required to aid in the identification of at risk athletes in order to reduce their risk of lower extremity injuries. Cumulatively, there is a lack of available evidence examining the sensitivity of field-based measures of neuromuscular control to predict injury risk in elite male youth soccer players. Additionally, no validated field-based testing batteries are available that measure different constructs of neuromuscular control. Therefore, the aim of this study was to examine if a novel field-based assessment battery can discriminate between injured and non-injured elite male youth 228

229 soccer players, and determine the sensitivity of individual tests in their ability to prospectively predict lower extremity injury. 7.2 Methods Participants Three hundred and fifty five elite male youth soccer players from the academies of six professional English Premier League and Championship soccer clubs volunteered to take part in the study. Descriptive statistics for anthropometric measures and predicted maturational status are provided in table 7.1. None of the players reported injuries at the time of testing and all were participating regularly in football training and competitions in accordance with the regulations for player contact hours as set out by the Premier League Elite Player Performance Plan (EPPP). Parental consent, participant assent and physical activity readiness questionnaires were collected prior to the commencement of testing. Ethical approval was granted by the institutional ethics committee in accordance with the declaration of Helsinki. Table 7.1 Mean (s) values for participant details for all players combined and per sub-chronological group Age Group Age (yrs.) Mass (kg) Height (cm) BMI (kg/m²) Leg Length (cm) Maturity Offset All combined 14.3 ± ± ± ± ± ± 1.9 U11-U ± ± ± ± ± ± 0.6 U13-U ± ± ± ± ± ± 0.9 U15-U ± ± ± ± ± ± 0.6 U ± ± ± ± ± ±

230 Experimental design This study used a prospective cohort design to evaluate the sensitivity of a field-based neuromuscular control assessment to predict non-contact lower extremity injuries. Elite male youth soccer players (aged between years) were required to attend their respective club training grounds during the pre-season period on two occasions (inclusive of familiarisation and data collection) separated by a period of seven days. The screening battery was comprised of the following assessments: (1) y-balance test (anterior reach direction only); (2) knee valgus in repeated tuck jump assessment; (3) single leg horizontal hop for distance; (4) single leg 75% horizontal hop and stick peak landing force and; (5) single leg countermovement jump and stick peak landing force. Two trials of each test were performed and the mean score was used to report each player s baseline performance. All other standardisation procedures were consistent with those reported in previous chapters. Following baseline testing, players were tracked longitudinally during the season to record all injuries sustained in soccer training and competitions at their respective clubs in accordance with the guidelines outlined by the EPPP Procedures Field-based neuromuscular control assessments: The protocols and equipment specification for each of the jump-landing tasks and the y-balance test included in the baseline field-based neuromuscular control assessment replicated those of study 2; please refer to the procedures section in that chapter. The reliability of each of these measures and the expected seasonal variation has also been reported in studies 1 and 3 respectively. The tests included showed acceptable test-retest reliability (see the results of study 1), while statistics comparing within- 230

231 subject variation between pre-season and end of season performances (see study 3) indicated largely acceptable values (CV 10%). However, test measures demonstrating higher variability (75%Hop and SLCMJ landing forces) displayed a trend of reductions in performance which may suggest a greater risk of injury. Therefore, baseline test scores reported in this study could be considered an accurate reflection of each player s level of neuromuscular control later in the season; it is conceivable that injury data for a period of one season could be used to examine the sensitivity of the administered testing battery to predict non-contact lower extremity injuries. All tests were conducted by the same researcher at each test session. Participant instructions were standardised to ensure replication of conditions between days (see appendix d) and removal of the potential for of inter-rater error Injury reporting: Injuries experienced during the season were diagnosed and prospectively recorded by each clubs respective medical personnel in accordance with the guidelines set out by the Premier League EPPP and as described in methods outlined in study 4. All trunk and upper extremity injuries and those that reported a contact mechanism were removed from the analysis in this study to ensure relevant risk factors for non-contact lower extremity injuries could be more accurately identified. Specifically, a non-contact injury was noted when an incident occurred without clear contact or collision from another player, the ball, or another object (Brito et al., 2012). A lower extremity injury was defined as any non-contact incident sustained to the hip or leg, excluding the lumbar spine and sacroiliac joints (Plisky et al., 2006). Only injuries reported that incurred a time loss from training and/or competition for a minimum of 48 hours from the time of injury were included. Due to the confounding injury risk associated with previous injuries (Arnason et al., 2004; Kucera et al., 2005; Hagguland et al., 231

232 2006), only the first incident experienced by each player was used in the subsequent analysis as has been suggested previously (Nilstad et al., 2014) Statistical analysis Descriptive statistics for each test were calculated for all participants. To identify if any of the anthropometric or neuromuscular variables could be identified as valid risk factors for lower extremity injury, a three-step approach was used. (i) The first step included a univariate binary logistic regression analysis of each injury risk factor and this was used to reduce the number of risk factors. A greater number of predictor variables included in logistic regression models increases the error degree of freedom and results in poor power of testing (Han et al., 2000). Neuromuscular and anatomical risk factors that displayed a p value < 0.1 were considered for further analysis. (ii) The second step included tests of multicolinearity for the risk factors identified in step 1. Multicolinearity can lead to the development of models that show poor predictor performance due to high prediction variance (Han et al., 2000). This was completed via linear regression analysis; mulitcolinearity was confirmed where the Variance Inflation Factor was > 10 based on previous guidelines (Myers, 1990). In such cases, variables identified with the most clinical significance were investigated further in a multivariate binary logistic regression. (iii) Finally, the odds ratio (OR) for each risk factor in the univariate and multivariate analysis was calculated, with 95% confidence intervals (CI). A p value 0.05 was indicative of a significant effect. Initially, all subjects were combined to examine risk factors for lower extremity. However, to remove confounding factors such as age and their effect on the dependent variables as identified in study 2, further analysis was completed which involved sub-division of players by chronological age group. The following age group sub-categories were used: U11 and 232

233 U12, U13 and U14, U15 and U16, and U18s. The same three-step statistical procedure described above was repeated for chronological age group analysis. 7.3 Results Injury incidence Of the total sample of players included in this study, 99 sustained a non-contact lower extremity injury during the period of the competitive soccer season. The knee was the most frequently injured anatomical site (31%), followed by the ankle (19%). There were a high proportion of strain type injuries (35%), with ligament (17%) and growth/overuse (14%) the most prevalent diagnosis thereafter. Half of the injuries (50%) were moderate (1-4 weeks), however severe injuries (> 4 weeks absent) were also frequently reported (32%). Relationship between injury risk factors and injury occurrence (all players combined) Descriptive statistics for anthropometric data of both the injured and non-injured players are shown in table 7.2. Injured players were older, with an OR signifying a 12% increased risk of lower extremity injury and this was the only anthropometric variable to reach statistical significance. Descriptive statistics and univariate ORs for all field-based neuromuscular control tests are presented in table 7.3. Players who recorded greater absolute landing forces during the 75%Hop on both legs were more likely to sustain an injury, although this corresponded to a small OR. Conversely, relative body weight forces were significantly lower during the SLCMJ on the right leg for the injured group in comparison to non-injured players, indicating a 34% 233

234 increased risk of lower extremity injury. Greater landing force asymmetry scores during the 75%Hop and SLCMJ were also associated with a 3% and 6% increased risk of injury respectively. No other neuromuscular risk factors were shown as significant predictors of lower extremity injury. Table 7.2 Anthropometrics of all injured and non-injured players Anthropometrics Injured Players Non-injured Players Odds Ratio P Value Age (y) 14.7 ± ± ( ).05* Height (cm) ± ± ( ).18 Mass (kg) 55.7 ± ± ( ).23 BMI (kg/m²) 19.9 ± ± ( ).63 Leg Length (cm) 86.9 ± ± ( ).25 Maturity-Offset 0.3 ± ± ( ).37 *Significant at the level of p <.05 BMI = body mass index 234

235 Table 7.3 Descriptive statistics and univariate odds ratios for all injured and non-injured players Neuromuscular Risk Non-injured Factors Injured Players Players Odds Ratio P Value 75%Hop L Pvgrf 1805 ± ± ( ) <.001** 75%Hop L (%BW) 3.37 ± ± ( ).12 75%Hop R pvgrf 1836 ± ± ( ).02* 75%Hop R (%BW) 3.43 ± ± ( ).32 75%Hop Asym ± ± ( ).05* SLCMJ pvgrf L 1649 ± ± ( ).23 SLCMJ L (%BW) 3.07 ± ± ( ).64 SLCMJ pvgrf R 1594 ± ± ( ).82 SLCMJ R (%BW) 2.96 ± ± ( ).05* SLCMJ PVGRF Asym ± ± ( ) <.001** SLHD L (m) 1.49 ± ± ( ).13 SLHD L (% leg length) 1.72 ± ± ( ).42 SLHD R (m) 1.50 ± ± ( ).15 SLHD R (% leg length) 1.74 ± ± ( ).45 SLHD Asym ± ± ( ).19 TJ Knee Valgus R 1.43 ± ± ( ).39 TJ Knee Valgus L 1.07 ± ± ( ).26 Y-B L (cm) ± ± ( ).49 Y-B R (cm) ± ± ( ).15 Y-B (% leg length) L ± ± ( ).79 Y-B (% leg length) R ± ± ( ).47 Y-B Asym ± ± ( ).92 * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym = asymmetry; BW = body weight; SLCMJ = single leg countermovement jump; SLHD = single leg hop for distance; TJ = Tuck Jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick After the removal of variables with a p value > 0.1 identified in the univariate analysis and one additional variable that displayed mutlicolinearity, five predictor variables were then assessed in a multivariate model (table 7.4). The model predicted 95.7% of non-injured players; however, only 15.2% of the injured players were correctly classified. The overall prediction accuracy of the model was 73.2%. Furthermore, a low amount of variance was explained (Nagelkerkes R² = 235

236 0.119). Greater landing force asymmetry during the SLCMJ was the only risk factor significantly associated with an increased risk of lower extremity injury, indicating players with larger asymmetries during this test are at a 6% greater risk of injury. Lower relative landing forces on the right leg were also associated with an increased risk of injury (OR, 0.73; ), however, the wider confidence intervals reduced the precision in estimation of this effect and was therefore not statistically significant. Table 7.4 Multivariate analysis of risk factors for all players Risk Factors Odds Ratio (95% CI) P Value Age 1.01 ( ).88 75%Hop pvgrf L 1.00 ( ).07 75%Hop pvgrf Asym 0.97 ( ).10 SLCMJ R (%BW) 0.73 ( ).14 SLCMJ pvgrf Asym 0.94 ( ) <.001** * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym; asymmetry; BW = body weight; SLCMJ = single leg countermovement jump; SLHD = single leg hop for distance; TJ = Tuck Jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick Relationship between injury risk factors and injury occurrence per chronological age group (combined U11 and U12 players) In total, eighty players were analysed in these two chronological age groups, with eighteen sustaining a lower extremity injury. Descriptive statistics for anthropometric data of both the injured and non-injured U11-U12 players are shown in table 7.5. Injured players were older and this corresponded to a 64% increase in injury risk; however, differences between groups were not statistically significant owing to a wide confidence interval. No other anthropometric 236

237 variables were associated with an increased injury risk. Descriptive statistics and univariate ORs for all field-based neuromuscular control tests for injured and non-injured players in these chronological age groups are presented in table 7.6. Players with greater knee valgus on the right leg during the tuck jump assessment and reduced left leg anterior reach distances on the y- balance test were more likely to sustain a lower extremity injury. Heightened levels of asymmetry in all unilateral jump-landing assessments were also associated with a greater risk of injury, and this was most prevalent for SLCMJ landing force asymmetry with a 15% increased risk. No other neuromuscular risk factors were shown as predictors of injured status in these chronological age groups. Table 7.5 Anthropometrics of U11 & U12 injured and non-injured players Anthropometrics Injured Mean Non-Injured Mean Odds Ratio P Value Age 11.8 ± ± ( ).19 Height ± ± ( ).66 Weight 38.9 ± ± ( ).71 BMI (kg/m²) 23.1 ± ± ( ).44 Leg length 76.7 ± ± ( ).48 Maturational offset -2.4 ± ± ( ).97 * Significant at the level of p <.05) ** Significant at the level of p <.001 BMI = body mass index 237

238 Table 7.6 Descriptive statistics and univariate odds ratios for injured and non-injured U11 & U12 players Neuromuscular Risk Factors Injured Mean Non-Injured Mean Odds Ratio P Value 75%Hop pvgrf R 1347 ± ± ( ).94 75%Hop PVGRF L 1417 ± ± ( ).56 75%Hop L (%BW) 3.58 ± ± ( ).66 75%Hop R (%BW) 3.76 ± ± ( ).80 75&Hop pvgrf Asym ± ± ( ) <.001** SLHD R 1.28 ± ± ( ).54 SLHD L 1.22 ± ± ( ).22 SLHD (% leg length) R 1.66 ± ± ( ).69 SLHD (% leg length) L 1.59 ± ± ( ).32 SLHD Asym ± ± ( ).04* SLCMJ pvgrf L 1129 ± ± ( ).30 SLCMJ pvgrf R 1127 ± ± ( ).46 SLCMJ L (%BW) 3.02 ± ± ( ).42 SLCMJ R (%BW) 3.03 ± ± ( SLCMJ pvgrf Asym ± ± ( ) <.001** TJ Knee Valgus R 1.88 ± ± ( ).03* TJ Knee Valgus L 1.44 ± ± ( ).46 Y-B R ± ± ( ).32 Y-B L ± ± ( ).04* Y-B (% leg length) R ± ± ( ).43 Y-B (% leg length) L ± ± ( ).07 Y-B Asym ± ± ( ).32 * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym = asymmetry; BW = body weight; SLCMJ = single leg countermovement jump; SLHD = single leg hop for distance; TJ = Tuck Jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick Variables with a p value > 0.1 identified in the univariate analysis were removed and one additional variable that displayed mutlicolinearity, leaving five predictor neuromuscular variables to be included in the multivariate model (table 7.7). The model predicted 96.8% and 55.6% of non-injured and injured players respectively. The overall prediction accuracy of the model was 87.5%. The model explained a moderate amount of variance in predicting injury status (Nagelkerkes R² = 0.432). Greater landing force asymmetry during the SLCMJ was the 238

239 only risk factor significantly associated with an increased risk of lower extremity injury, indicating players with larger asymmetries during this test are at a 10% greater risk of injury. Table 7.7 Multivariate analysis of risk factors for U11 & U12 age groups Risk Factors Odds Ratio (95% CI) P Value 75%Hop pvgrf Asym 0.93 ( ).17 TJ Knee Valgus R 1.93 ( ).10 SLCMJ PVGRF Asym 0.90 ( ).04* SLHD Asym 0.92 ( ).09 Y-Balance (Anterior) L 0.94 ( ).10 * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym = asymmetry; SLCMJ = single leg countermovement jump; SLHD = single leg hop for distance; TJ = Tuck Jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick Relationship between injury risk factors and injury occurrence (combined U13 and U14 players) One hundred and fourteen players were analysed in these chronological age groups, with thirty one sustaining a lower extremity injury. Descriptive statistics for anthropometric data of both the injured and non-injured U13-U14 players are shown in table 7.8. Injured participants were further from their age at peak height velocity as indicated by a larger maturational offset. No other anthropometric variables were able to identify significant differences between injured and non-injured players. Descriptive statistics and univariate ORs for all players in the U13s and U14s are presented in table 7.9. There was a trend for increased landing forces relative to body weight on the left leg during the 75%Hop; however no neuromuscular risk factors were associated with a greater risk of sustaining a lower extremity injury in these age groups. 239

240 Table 7.8 Anthropometrics of U13 & U14 injured and non-injured players Anthropometrics Mean Injured Mean Non-Injured Odds Ratio P Value Age 13.5 ± ± ( ).44 Height ± ± ( Weight 47.2 ± ± ( ).29 BMI (kg/m²) 18.2 ± ± ( ).82 Leg Length 85.0 ± ± ( ).82 Maturational Offset -0.8 ± ± ( ).04* * Significant at the level of p <.05) ** Significant at the level of p <.001) BMI = body mass index Table 7.9 Descriptive statistics and univariate odds ratio s for injured and non-injured U13 & U14 players Neuromuscular Risk Factor Mean Injured Mean Non-Injured Odds Ratio P Value 75%Hop pvgrf L 1623 ± ± ( ).26 75%Hop pvgrf R 1580 ± ± ( ).63 75%Hop (%BW) L 3.59 ± ± ( ).09 75%Hop (%BW) R 3.51 ± ± ( ).41 75%Hop Asym ± ± ( ).10 SLHD L 1.37 ± ± ( ).37 SLHD R 1.38 ± ± ( ).30 SLHD (leg length) L 1.67 ± ± ( ).65 SLHD (leg length) R 1.69 ± ± ( ).61 SLHD Asym ± ± ( ).86 SLCMJ pvgrf L 1477 ± ± ( ).75 SLCMJ pvgrf R 1403 ± ± ( ).46 SLCMJ (%BW) L 3.30 ± ± ( ).42 SLCMJ (%BW) R 3.08 ± ± ( ).48 SLCMJ PVGRF Asym ± ± ( ).74 TJ Knee Valgus R 1.68 ± ± ( ).36 TJ Knee Valgus L 1.19 ± ± ( ).75 Y-B L ± ± ( ).69 Y-B R ± ± ( ).27 Y-B L ± ± ( ).91 Y-B (leg length) R ± ± ( ).50 Y-B Asym ± ± ( ).77 * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym = asymmetry; BW = body weight; SLCMJ = single leg countermovement jump; SLHD = single leg hop for distance; TJ = Tuck Jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick 240

241 Only three variables with a p value 0.1 were identified in the univariate analysis and mutlicolinearity was not detected in any of these risk factors; three predictor variables were included in the multivariate model for these age groups (table 7.10). The model accurately predicted 96.4% of non-injured players but only 16.1% of the injured players. The overall prediction accuracy of the model was 74.6% with a low amount of variance explained in predicting injury status (Nagelkerkes R² = 0.117). No neuromuscular variables were associated with a greater injury risk and maturational offset was the only risk factor significantly associated with an increased risk of lower extremity injury, indicating that players with a larger maturational offset are at a 42% greater risk of injury. Table 7.10 Multivariate analysis of risk factors for U13 & U14 age groups Risk Factor Odds Ratio P Value Maturational Offset 0.58 ( ).04* 75%Hop (%BW) L 1.55 ( ).18 75%Hop pvgrf Asym 0.96 ( ).08 * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym = asymmetry; BW = body weight; PVGRF = peak vertical ground reaction force; 75%Hop = 75% horizontal hop and stick Relationship between injury risk factors and injury occurrence (combined U15 & U16 players) One hundred and eighteen players were analysed in these chronological age groups, with thirty four sustaining a lower extremity injury. Descriptive statistics for anthropometric data of both the injured and non-injured U15 & U16 players are shown in table A trend was evident for injured players to be taller but no variables were significantly associated with a greater injury 241

242 risk. Descriptive statistics and univariate ORs for all field-based neuromuscular control tests for all players in the U15s and U16s are presented in table Greater landing forces during the 75%Hop were significantly associated with a heightened risk of injury but this corresponded to a small OR. Conversely, lower right leg relative landing force scores in the SLCMJ were recorded for injured players, indicating a 66% increase in the likelihood of experiencing a lower extremity injury. Superior anterior reach distances were shown in the injured players and this was significant on the left leg with a trend evident on the right leg. SLCMJ landing force asymmetry was also more pronounced in the injured players. Table 7.11 Anthropometrics of U15 & U16 injured and non-injured players Anthropometrics Mean Injured Mean Non-Injured Odds Ratio P Value Age 15.7 ± ± ( ).55 Height ± ± ( ).06 Weight 64.4 ± ± ( ).26 BMI (kg/m²) 20.7 ± ± ( ).80 Leg Length 92.5 ± ± ( ).23 Maturational Offset 1.4 ± ± ( ).80 * Significant at the level of p <.05) ** Significant at the level of p <.001) BMI = body mass index 242

243 Table 7.12 Descriptive statistics and univariate odds ratio s for injured and non-injured U15 & U16 players Neuromuscular Risk Factor Mean Injured Mean Non-Injured Odds Ratio P Value 75%Hop pvgrf R 1989 ± ± ( ).03* 75%Hop pvgrf L 1907 ± ± ( ) <.001** 75%Hop (%BW) R 3.15 ± ± ( ).22 75%Hop (%BW) L 3.02 ± ± ( ).05* 75%Hop pvgrf Asym ± ± ( ).62 SLHD R 1.55 ± ± ( ).57 SLHD L 1.56 ± ± ( ).78 SLHD (% leg length) R 1.68 ± ± ( ).33 SLHD (% leg length) L 1.69 ± ± ( ).80 SLHD Asym ± ± ( ).43 SLCMJ pvgrf L 1853 ± ± ( ).88 SLCMJ pvgrf R 1724 ± ± ( ).18 SLMCJ (%BW) L 2.89 ± ± ( ).14 SLMCJ (%BW) R 2.71 ± ± ( ).01* SLCMJ pvgrf Asym ± ± ( ) <.001** TJ Knee Valgus R 1.29 ± ± ( ).90 TJ Knee Valgus L 1.03 ± ± ( ).97 Y-B R ± ± ( ).06 Y-B L ± ± ( ).05* Y-B (% leg length) R ± ± ( ).25 Y-B (% leg length) L ± ± ( ).19 Y-B Asym ± ± ( ).20 * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym = asymmetry; BW = body weight; SLCMJ = single leg countermovement jump; SLHD = single leg hop for distance; TJ = Tuck Jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick From the univariate analysis, eight variables with a p value 0.1 were identified. Mutlicolinearity was detected in three of these risk factors; five predictor variables were included in the multivariate model for further examination (table 7.13). The model accurately predicted 92.9% of non-injured and 44% of injured players respectively. The overall prediction accuracy of the model was 78.8%. A low to moderate amount of variance was explained in the prediction model (Nagelkerkes R² = 0.312). Neuromuscular variables associated with greater injury risk 243

244 were greater SLCMJ landing force asymmetry and lower right leg relative body weight landing forces during the SLMCJ. A trend was evident of an association between higher left leg landing forces during the 75%Hop, however, this corresponded to a low OR. Table 7.13 Multivariate analysis of risk factors for U15 & U16 age groups Risk Factor Odds Ratio P Value Height 1.01 ( ).76 75%Hop pvgrf Force L 1.00 ( ).06 SLCMJ (%BW) R 0.36 ( ).03* SLCMJ pvgrf Asym 0.91 ( ) <.001** Y-B L 1.01 ( ).57 * Significant at the level of p <.05) ** Significant at the level of p <.001) Asym = asymmetry; BW = body weight; SLCMJ = single leg countermovement jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick Relationship between injury risk factors and injury occurrence (U18 players) Forty four players were analysed in this chronological age group, with seventeen sustaining a lower extremity injury. Descriptive statistics for anthropometric data of both the injured and noninjured U15 and U16 players are shown in table Odds ratios indicated more than three-fold likelihood that injured players were older. No other anthropometric variables were associated with injury status. Descriptive statistics and univariate ORs for all field-based neuromuscular control tests for all players in the U18s are presented in table Greater relative right leg hop distances in SLHD were the only neuromuscular variable associated with a greater risk of injury and a similar trend was also evident on the left leg but this association was not statistically significant. 244

245 Table 7.14 Anthropometrics of U18 injured and non-injured players Anthropometrics Mean Injured Mean Non-Injured Odds Ratio P Value Age 17.9 ± ± ( ).02* Height ± ± ( ).09 Weight 72.3 ± ± ( ).68 BMI (kg/m²) 23.1 ± ± ( ).44 Leg Length 90.4 ± ± ( ).33 Mat Offset 2.9 ± ± ( ).51 * Significant at the level of p <.05) ** Significant at the level of p <.001 BMI = body mass index Table 7.15 Descriptive statistics and univariate odds ratio s for injured and non-injured U18 players Neuromuscular Risk Factor Mean Injured Mean Non-Injured Odds Ratio P Value 75%Hop pvgr R 2478 ± ± ( ) %Hop pvgr L 2457 ± ± ( ) %Hop (%BW) R 3.52 ± ± ( ) %Hop (%BW) L 3.47 ± ± ( ) %Hop pvgrf Asym ± ± ( ) 0.10 SLHD R 1.85 ± ± ( ) 0.09 SLHD L 1.83 ± ± ( ) 0.15 SLHD (% leg length) R 2.05 ± ± ( ) 0.05* SLHD (% leg length) L 2.03 ± ± ( ) 0.08 SLHD Asym ± ± ( ) 0.83 SLCMJ pvgrf R 2212 ± ± ( ) 0.51 SLCMJ pvgrf L 2129 ± ± ( ) 0.62 SLCMJ (%BW) R 3.14 ± ± ( ) 0.53 SLCMJ (%BW) L 3.04 ± ± ( ) 0.57 SLCMJ pvgrf Asym ± ± ( ) 0.16 TJ Knee Valgus R 0.75 ± ± ( ) 0.41 TJ Knee Valgus L 0.51 ± ± ( ) 0.93 Y-B R ± ± ( ) 0.39 Y-B L ± ± ( ) 0.29 Y-B (% leg length) R ± ± ( ) 0.92 Y-B (% leg length) L ± ± ( ) 0.38 Y-B Asym ± ± ( ) 0.74 * Significant at the level of P <.05) ** Significant at the level of P <.001) Asym = asymmetry; BW = body weight; SLCMJ = single leg countermovement jump; SLHD = single leg hop for distance;tj = Tuck Jump; PVGRF = peak vertical ground reaction force; Y-B = y-balance; 75%Hop = 75% horizontal hop and stick 245

246 The univariate analysis identified two anthropometric variables and four neuromuscular variables with a p value 0.1; however, maximal right leg SLHD and left leg SLHD relative performances were removed due to mutlicolinearity. Subsequently, four predictor variables were included in the multivariate model (table 7.16). The model accurately predicted a large number of injured players (85.2%) but was less accurate in predicting non-injured players (62.5%). The overall prediction accuracy of the model was 76.7%. A moderate amount of variance was explained in the prediction model (Nagelkerkes R² = 0.442). The only variable associated with greater injury risk was age demonstrating a three-fold increase in injury risk. Table Multivariate analysis of risk factors for U18 age group Risk Factor Odds Ratio P Value Age 3.62 ( ).04* Height 0.84 ( ).09 SLHD (% leg length) R 1.54 ( ).83 75%Hop pvgrf Asym 1.10 ( * Significant at the level of p <.05) Asym = asymmetry; SLHD = single leg hop for distance; 75%Hop = 75% horizontal hop and stick 7.4 Discussion The current study is the first to examine injury risk factors in elite male youth soccer players using a comprehensive field-based neuromuscular control screening battery. The results showed that combinations of anthropometric and neuromuscular risk factors were predictive of lower extremity injury, but there was variability in which factors were the main predictors across the different chronological age groups. It was also shown that a number of the measured variables were not related to increased injury risk. Of all the neuromuscular control variables analysed, 246

247 SLCMJ landing force asymmetry was the most frequently reported risk factor; however, in the combined U13-U14 group and the U18 age group no association was present between any of the neuromuscular control measures and injury status. Univariate analysis also identified a number of significant predictors, with further trends towards significance combined with noteworthy odds ratios for selected neuromuscular control tests in respective chronological age groups. The relationship between anthropometric variables and injury risk Anthropometric variables reported in this study indicated some association with injury risk but this was not consistent across all sub-groups. With all the players combined, univariate analysis showed that older players were more likely to sustain an injury, although this finding was not replicated in the subsequent multivariate model. Conversely, age demonstrated a three-fold increase in injury risk for the U18s in both the univariate and multivariate analyses. Injury epidemiological data in elite male youth soccer players has shown a linear increase in the number of injuries sustained with age (Price et al., 2004). Furthermore, study 4 of this thesis displayed a trend of heightened player incidence ratios with advances in chronological age in this cohort. It is suggested that the greater frequency of injuries in older players may be due to heightened intensities of play and increased exposures. Additionally, greater muscular power and reductions in flexibility could be confounding factors (Henderson et al., 2010). Specifically, in young male professional soccer players, a 1.78 times greater chance of hamstring strain injury was reported for each annual age increment and the authors reported that older, more powerful and less flexible players are at a greater risk of injury (Henderson et al., 2010). Thus, to reduce injury risk, conditioning programmemes focusing on increasing the capacity to control activities 247

248 where this power is expressed while enhancing flexibility could be recommended. However, the study by Henderson et al. (2010) measured factors associated with injury in young male professional soccer players (22.6 ± 5.2 years of age). Due to the lack of association between the neuromuscular control screening tests used in the current study and injury risk in the U18 age group, further research is warranted to examine if older players who display heightened force production capabilities but reduced flexibility are at a greater risk of injury. Apart from chronological age and maturational offset, no other anthropometric variables were associated with injury. This finding is in contrast with previous data that has shown body mass index (BMI) as the only factor associated with the occurrence of sustaining a new lower extremity injury in young female soccer players (Nilstad et al., 2014). Heavier players will experience greater landing forces on impact which have to be absorbed by soft tissue structures (Emery, 2005). This may stress joint and ligamentous structures to a greater extent than lighter players, ultimately increasing their risk of injury (Emery, 2003; Nilstad et al., 2014). The differences reported between the current study and those of previous research could be due to the homogenous nature of elite male youth soccer players (as indicated by the lack of significant differences between groups), whereby, to gain selection into a professional soccer academy, coaches will likely select boys who are taller and heavier that they deem to be more physically able in addition to displaying high levels of technical skill (Reilly et al., 2000). Also, recent data has signified the importance of the rate of change in anthropometric factors (Kemper et al., 2015). In elite male youth soccer players, growth in stature 0.6 cm, and an increased BMI 0.3kg/m 2 per month were significant predictors of injury (Kemper et al., 2015). Thus, measuring the rate of change could be deemed important for practitioners and could more effectively indicate a player s level of injury risk. 248

249 The relationship between field-based measures of neuromuscular control and injury risk Some variation was evident across different chronological age groups; however, the most frequently identified neuromuscular risk factor was SLCMJ peak landing force asymmetry. Both univariate and multivariate analysis showed significant associations with injury when all players were combined indicating a 6% increased risk. In the U11-U12 and U15-16 chronological age groups, this risk was magnified, ranging from 10-15% for injured players who displayed greater asymmetry. Although not as prevalent, greater asymmetry was also indicated as a risk factor in the univariate analysis for the SLHD (U11-12s only), and the 75%Hop with all players combined and for the U11-U12 sub-group, with a trend of greater 75%Hop asymmetry relating to injury status in the multivariate model for the U13-U14s (OR, 0.96; p = 0.08) and U18 players (OR, 1.10; p = 0.07). Particularly in the U18s, a larger sample size would likely increase the clinical significance of this effect, suggesting this may be a worthwhile test in this age group. Asymmetry places additional stress on the weaker leg during cutting and landing activities which may increase injury risk (Hewit et al., 2012). The presence of limb asymmetry has been reported previously in male youth soccer players (Daneshjoo et al., 2013; Atkins et al., in press), although no evidence is currently available to confirm potential relationships with injury. In adult populations, a discrepancy >15% has been deemed a key predictor of injury (Crosier and Creelard, 2000). Also, subjects with prior history of ACL injury demonstrated greater asymmetry in peak internal knee flexor moments and ground reaction forces during a drop jump landing versus non-injured controls, whereas no kinematic differences were reported (Goerger et al., 2014). In the current study, asymmetry values were greater in the measurement of landing forces (SLCMJ and 75%Hop) versus distance (SLHD and y-balance anterior reach scores), cooresponding with previous research (Hewit et al., 2012) and the results of study 2 in this 249

250 thesis. Also, asymmetry values > 15% were commonly shown in the present study for players who sustained a lower extremity injury (SLCMJ asymmetry range = 12.87% %; 75%Hop asymmetry range = 11.85% 16.51%). Therefore, it could be suggested that landing force asymmetry is a risk factor for injury in this cohort and practitioners should include diagnostic assessments that measure a player s ability to absorb landing forces in both vertical and horizontal directions. However, as a caveat asymmetry measures may be less sensitive in predicting injury in the U13-U14 and older (U18) players. In addition to landing force asymmetry, the magnitude of absolute landing forces also appears to show a relationship with increased injury risk. With all players combined, greater landing forces were evident for the injured players in the 75%Hop, reaching statistical significance in the univariate analysis and showing a trend towards significance when analysed within the multivariate model. The same pattern was observed in the U15-U16 sub-group analysis, however a small odds ratio was observed. A trend was also displayed of increased relative landing forces during the 75%Hop in the univariate model for the U13-U14s, indicating a 32% increased risk of injury. If impact forces on ground contact exceed the force production capabilities of the involved musculature, additional loading will be diverted to other soft tissues such as bones and ligaments, thus heightening the risk of lower extremity injury (Hewett et al., 2010). Furthermore, the heightened landing forces displayed in the injured U15-U16 players could also be related to the period of peak weight velocity (PWV), whereby, players will undergo a period of rapid increases in muscle mass due to heightened androgen concentrations (Viru et al., 1999) and this could increase landing forces due to a heavier body mass. Thus, while neuromuscular performance may increase during this period, a developmental lag in neuromuscular control could be reflected by an inability to attenuate ground reaction forces upon 250

251 landing. This may heighten their risk of injury and highlights the need for interventions focused on dissipating landing forces. An unexpected finding in the current study was that lower relative landing forces during the SLCMJ on the right leg were associated with an increased injury risk. This trend was evident in the univariate analysis with all age groups combined and for the U15-U16 sub-group comparison. A significant association was also shown in the multivariate model for the U15-16s relating to a 64% increased risk of injury. Intuitively, a plausible explanation for this could be that the injured players in this study did not achieve equivalent vertical jump heights in comparison to the non-injured players. Previous data show that reduced hop distances are predictive of hamstring strain injury in young male athletes (Goosens et al., 2015). The authors speculated this was due to a reduced ability of injured players to use their hamstrings to effectively decelerate knee and hip flexion (Goosens et al., 2015). Conversely, in adult professional soccer players, greater jump heights have been associated with increased risk of hamstring injuries (Henderson et al., 2010). Although not a predictor variable in the current study, descriptive analysis showed no significant differences in vertical jump height between the injured and non-injured players for the U15-U16 sub-group analysis. Therefore, an alternative and speculative hypothesis is that injured players adopted a different kinematic strategy when performing a vertical jump-landing task on their right leg. It could be inferred that the majority of players would preferentially utilise their right leg for kicking actions due to preferred limb dominance during soccer match play; thus, greater stability and force absorption would be expected on their contralateral limb due to the requirement to repeatedly stabilise on their stance leg during soccer activities. Interestingly, Brophy et al. (2010) reported that 74.1% of anterior cruciate ligament (ACL) injuries in adult soccer players were to their kicking leg. Also, recent 251

252 data has shown that the combination of knee valgus and ipsilateral trunk motion during a single leg drop vertical landing task was predictive of a non-contact knee injury in female athletes (Dingenen et al., 2014). Changes in trunk positioning can alter the resultant ground reaction forces (Blackburn and Padua, 2009) and could in part explain the reduced relative landing scores for the injured players. Future research should consider the assessment of both kinetic and kinematic variables in this cohort to examine if an aberrant movement strategy is displayed in at risk subjects. The current study appears to be the first to examine the relationship between dynamic knee valgus and injury risk in male youth soccer players. The presence of this risk factor could be expected due to the frequency of rapid changes of direction and high force jump-landing activities that occur in the sport (Daniel et al., 1994). In addition, there is a high prevalence of medial collateral ligament injuries, reported as the most common diagnosis for knee ligament injury (Price et al., 2004). In the present study, no association with increased injury risk was shown when all the players were combined, however; univariate analysis reported a significant increase in risk for the U11-U12 sub-group analysis, combined with a large odds ratio and a trend towards significance in the subsequent multivariate model. Cumulatively, this suggests that the assessment of knee valgus during the tuck jump is a worthwhile screen for these chronological age groups. Younger soccer players are more likely to display a greater degree of dynamic valgus than older youths (Yu et al., 2005; Schmitz et al., 2009), which may be attributed to an increased ability to attenuate landing forces with advancing age following the neuromuscular spurt (Quatman et al., 2006). Thus, it could be suggested that knee valgus increases the likelihood of lower extremity injury due to altered joint kinematics in the youngest age groups. No other age group comparisons reported an association between knee valgus and 252

253 increased injury risk. This in part may be due to not segregating the analysis to knee injuries only, which was due to sample size limitations. Another plausible explanation is that knee valgus in the current study was measured using a subjective scoring classification (0, 1, 2, 3) during a repeated tuck jump based on predicted joint angle classifications to maximise time efficiency and aid practical application (Read et al., in press b). Based on the data reported, this system may not have been sensitive enough to detect differences between injured and non-injured players. A further limitation in the assessment of dynamic valgus is that it does not consider the effects of trunk positioning or account for arthrokinematic frontal plane knee rotation (Schmitz et al., 2009). This is supported by a recent prospective cohort study that showed isolated measurement of knee valgus was not a predictor of non-contact knee injury (Dingenen et al., 2014). Furthermore, data from a recent study indicated that elite youth soccer players (inclusive of both boys and girls) who demonstrated a greater number of landing technique errors were at a greater risk of ACL injury (Padua et al., 2015). Cumulatively, it could be suggested that for the assessment of dynamic knee valgus, practitioners should consider both proximal (trunk/hip) and distal (foot) factors to enhance the predictive value of jump-landing assessments in their ability to identify youth players who display high risk kinematics. In this study, univariate analyses identified significant differences between injured and non-injured players during the y-balance anterior reach test for the U11-U12 and U15-U16 subgroup analysis. Injured players achieved lower and higher anterior left leg reach scores in the U11-U12 and U15-U16 age groups respectively. However, these results corresponded to a small odds ratio and no associations were reported in the subsequent multivariate models. Young soccer players who display reduced balance and control during single leg stance demonstrate greater postural sway which may compromise stability (Mickel et al., 2011). Asymmetrical 253

254 anterior reach distances > 4 cm have also been reported as a risk factor for lower extremity injury in high school basketball players (Plisky et al., 2006). Conversely, Nilstad et al. (2014) reported no association between y-balance composite scores and injury risk in elite female youth soccer players. Effective performance of dynamic stability tasks requires the integration of visual, vestibular and proprioceptive inputs which provides an efferent response to control the body s center of mass within its base of support (Guskiewicz et al., 1996). Thus, the lack of association between y-balance anterior reach scores and injury risk in the current study could be explained in part by the complex interaction of numerous factors that comprise dynamic balance, and as such it may not be possible to accurately assess all of these constructs in a single field-based assessment (Nilstad et al., 2014). Recent literature has identified the period of PHV as a time of heightened injury risk in male youth soccer players (Rumpf and Cronin, 2012; van der Sluis et al., 2014). In the present study, no neuromuscular control variables were shown as risk factors for injury in the U13-U14 sub-group who would most closely align to this stage of maturation. This could be attributed to a period of adolescent awkwardness, whereby, due to rapid increases in limb length, young soccer players may experience temporary decrements in motor skill performance occurring approximately 12 months prior to PHV (Philippaerts et al., 2006). Therefore, due to the potential for increased movement variability during this period, accurately detecting differences between injured and non-injured players may be more difficult. Also, injury prevalence data from study 4 suggests that the U15 chronological age group sustain the highest number of severe injuries and resultant time-loss from training and competitions, which could be associated with the period of PWV. The number of risk factors was also the highest in the U15-U16 sub-group in the current 254

255 study, indicating further that this is a key period of heightened injury risk for male youth soccer players. While no neuromuscular variables were significantly associated with injury risk for the U13s and U14s in this study, maturational offset was a significant predictor in both the univariate and multivariate analysis. Players who were further from their reported attainment age at peak height velocity (APHV) were at a 42% increased risk of injury. Recent data show that later maturing players display a significantly greater overuse injury incidence rate than earlier maturing players (van der Sluis et al., 2015), although other studies have reported no difference based on the tempo of maturation (Le Gall et al., 2006). The timing of the adolescent growth spurt (age at peak height velocity) was not measured in this study; however, practitioners should be cognizant that early and late maturing children will need to be treated differently and this may affect the approach to the provision of screening and training strategies for these players (Lloyd and Oliver, 2012). Also, future studies should examine the relationships between the timing and tempo of maturation, neuromuscular control and the associated injury risk in male youth soccer players. Cumulatively, screening to predict injury risk in these age groups may be difficult as indicated by the low prediction accuracy of the multivariate model used identify injured players in the current study. A final point for practitioners to consider when interpreting the findings of the current study is that variation was evident in predictor variables across the different chronological age sub-groups and this may be due to the aforementioned variation in the timing and tempo of maturation. Thus it is crucial that screening modalities contain a degree of flexibility to account for changes in injury risk at different time points during a child s development. Although a core body of tests can be implemented for all age groups to allow for longitudinal monitoring 255

256 purposes, a greater emphasis may be centered on specific risk factors for different age groups and at various stages of growth and maturation. Also, while reductions in neuromuscular control may increase injury risk (Hewett et al., 2004), weak associations between certain test variables included in this study and lower extremity injury suggests that other confounding factors such as previous injury and fatigue should also be considered (Engebretsen et al., 2010; Read et al., 2015). 7.5 Summary and practical applications The current study appears to be the first to examine injury risk factors in elite male youth soccer players using a comprehensive field-based neuromuscular control screening battery. The results of this study will assist practitioners by providing evidenced based diagnostic assessments for their players, from which, at risk players could be identified through the use of selected assessments and targeted with appropriate prevention strategies prior to the occurrence of a lower extremity injury. However, additional assessments may be required at different stages of growth and maturation due to the reported variance in predictor variables across the different chronological age groups, which may be indicative of greater injury risk at these times. The key findings from the current study are summarised below: SLCMJ landing force asymmetry was the most frequently reported risk factor which suggests that vertical landing force asymmetry is a key risk factor for male youth soccer players 256

257 Other risk factors, namely: 75%Hop landing force; 75%Hop landing force asymmetry and tuck jump knee valgus were also shown as clinical predictors of injury in some age groups. The highest number of neuromuscular risk factors was present in the U15/U16 sub-group, indicating that this is a key period of heightened injury risk for male youth soccer players. Selected neuromuscular control assessments included in this study showed a lack of association with increased injury risk which could be due to the multi-factorial nature of soccer injuries. 257

258 Chapter 8 SUMMARY AND PRACTICAL APPLICATIONS, LIMITATIONS AND RECOMMENDATIONS FOR FUTURE RESEARCH 8.1 Summary and practical applications An extensive review of the available literature examining the injury incidence in elite male youth soccer players indicated that a high proportion of injuries were non-contact in nature and occurred predominantly in the lower limb, with the most frequent sites the knee, ankle and upper thigh. However, limited data were available to investigate the effects of recent increases in training volume now required since the inception of the Premier League s early sport specialisation model, the Elite Player Performance Plan (EPPP). Further literature reviews identified that a paucity of data was available to describe risk factors for the most frequent injuries in this cohort. While research conducted in other populations had shown associations with injury, this had yet to be investigated in elite male youth soccer players. Furthermore, data to support the concept that field-based neuromuscular screening assessments can be used to accurately measure performance and predict injury risk in paediatric male athletes were also sparse. Based on the limitations in the existing body of literature, this thesis was designed to provide an original and significant contribution to the fields of study for injury prevention and paediatric exercise science. Subsequently, a number of aims were established in order to fulfil this objective. This section will review these aims, highlighting the effectiveness of this thesis in providing evidence-based data to address these aims and findings that can be applied practically within the context of a professional male soccer academy. 258

259 Aim 1: Synthesise existing literature to determine the occurrence, type and severity of injury in elite male youth soccer players and report the current data to examine the effects of a recently implemented early soccer specialisation programmeme A review of the existing literature to describe the occurrence, severity and types of injuries that occur in this cohort was performed in section 2.1 of the literature review. However, this data does not account for a recent increase in the number of training hours now required as part of the EPPP. The results of study 4 showed a three-fold increase in the player injury incidence rate which could be attributed to a dramatic increase in player exposure hours. Also, the knee and ankle were identified as the most commonly occurring injuries which indicate that screening assessments should be focused on the mechanisms and constructs of neuromuscular control that are associated with these types of injuries. Seasonal variation in injury occurrence was also reported in chapter 5, which reinforces the need to longitudinally track players throughout the season to identify periods of heightened risk. Finally, although the highest numbers of injuries were recorded in the U18s, when reporting the time loss and severity of injury, the U14s and U15s were shown to experience the greatest time loss per injury and the U15s sustained the greatest frequency of severe injuries. This suggests those players who are experiencing, or who are just past their peak growth spurt, and who may be undergoing peak weight velocity are at greater risk of severe injury and should be a key injury prevention target group. Aim 2: Develop a reliable battery of field-based tests to assess the lower extremity neuromuscular control abilities of elite male youth soccer players 259

260 As identified in the literature review, despite the frequency of assessing neuromuscular control in adult male and female populations, limited data are available reporting the reliability of practically viable tests in elite male youth soccer players. Chapter 3 investigated commonly used field-based neuromuscular control tests to determine their reliability in this cohort. In both the pre and post-phv groups, single leg hop for distance; all y-balance variables; and peak landing vertical ground reaction force in both the 75% hop and single leg countermovement jump demonstrated acceptable reliability. However, measures of time to peak vertical landing force and time to stabilisation were highly variable in both groups, which may limit the accuracy of their use in this cohort. An important finding was that although tuck jump total score was considered reliable, it was unclear which of the risk factors comprising the composite score were contributing to the player s performance. This was due to low-moderate agreement between sessions in a number of the individual technique criteria. Thus, based on these findings, knee valgus was proposed as a plyometric technique error that could be reliably screened in this cohort due to high agreement between sessions. The test variables deemed reliable in this study were then used in the remainder of the thesis to examine the effects of chronological age and seasonal variation on injury risk factors, and the ability of the selected tests to identify players at a greater risk of injury. Also, using the reliability data from this study, practitioners are now able to calculate a worthwhile change in performance to use for longitudinal tracking of these metrics throughout the season and following injury prevention training programmemes. Aim 3: Determine the effects of chronological age for a range of lower extremity field-based neuromuscular control assessments in this cohort 260

261 Chapter 4 investigated the effects of chronological age on the measures of neuromuscular control that were deemed to be reliable in study 1. Variation was evident in the different constructs of neuromuscular control across age groups and tests. Greater relative anterior reach scores which are indicative of dynamic balance were shown in the youngest age group (U11s), which may suggest that this test is constrained more by mobility than dynamic balance due to reduced passive stiffness in younger children. The measurement of landing forces displayed a linear increase with advancing age, whereas relative landing forces were more variable across age groups with stiffer landing associated with reduced performance in the youngest and oldest age groups. Knee valgus also reduced with age which may be suggestive of heightened strength and neuromuscular performance. Cumulatively, these data suggest age effects are present in different constructs of neuromuscular control which may assist in the development of training programmemes to target deficits at different stages of a child s development. The findings of this study may also provide normative data for a range of chronological age groups, from which fluctuations in performance can be identified. Aim 4: Examine the seasonal variation in field-based measures of neuromuscular control to determine if periods of heightened risk are indicated within an academy soccer season Study 4 identified seasonal variation in injury incidence and study 3 examined variation and long-term reliability in field-based neuromuscular control performance at three time points (pre, mid and end of season). The results of study 3 showed that acceptable within subject variation were evident on the majority of the tests measured. All anthropometric variables increased during the season, whereas changes in neuromuscular control were more variable. Specifically, a 261

262 number of the neuromuscular control tests included were often considerably lower than the random variation, thus observed changes may not be meaningful. However, single leg countermovement jump forces increased considerably throughout the season indicating that a real change occurred and this may be associated with greater injury risk. The combined results of studies 3 and 4 suggest that seasonal variation in injury risk may be present. Therefore, practitioners should consider longitudinally tracking neuromuscular control throughout the soccer season with a specific focus on during pre-season and post-christmas periods. In addition to monitoring landing forces, the assessment of asymmetry could also be warranted due to high variability in the maintenance of rank order, indicating that some individuals may substantially increase their risk at various stages in the season. Aim 5: Identify relationships between measures of neuromuscular control and injury risk in elite male youth soccer players Study 5 included in this thesis is the first to examine injury risk factors in elite male youth soccer players using a comprehensive field-based neuromuscular control screening battery. Results showed that combinations of anthropometric and neuromuscular risk factors were predictive of lower extremity injury, but there was variability in which factors were prevalent across the different chronological age groups. The multivariate model identified single leg countermovement jump landing force asymmetry as the most frequently reported risk factor, indicated by significant associations with all the players combined and for the U11-U12 and U15-U16 age group analysis. Univariate analysis also identified a number of significant predictors, with further trends towards significance and noteworthy odds ratios in respective chronological age groups. The results of this study suggest that, in addition to anthropometric 262

263 measures, practitioners can now include evidenced-based diagnostic assessments for their players from which at risk players could be identified. In chapter 2, an injury risk factor hierarchical model was proposed (figure 2.2). Based on the findings of study 5, a revised version of this figure has been presented in figure 8.1 as a novel field-based movement screen to assess lower extremity neuromuscular control in male youth soccer players, the injury prevention asymmetry soccer screen (i-pass). However, due to the reported variance across chronological age groups, greater weightings of specific assessments may be required at different stages of a child s development. Finally, the results of this study suggest that predicting injury risk during the period associated with the peak growth spurt is challenging due to the lack of association between neuromuscular control and players who sustained a lower extremity injury. This could be attributed to high movement variability in this group and distinct variation in their stage of growth and maturation. Thus, practitioners should monitor both anthropometrics and neuromuscular control during this period to identify rapid changes in either growth related processes or decrements in their movement skill. 263

264 INJURY PREVENTION ASYMMETRY SOCCER SCREEN - i-pass Lower vertical landing force Greater horizontal hop vertical landing forces Leg dominance (asymmetry)) Ligament dominance (knee valgus) Single leg countermovement jump (dominant leg) 75% hop and stick Single leg countermovement jump and 75% hop and stick Tuck jump Figure 8.1 A novel field-based movement screen to assess neuromuscular control: The i-pass The top tier of the model includes the associated neuromuscular risk factors for lower extremity injuries in elite male youth soccer players. The second tier represents the assessments identified as valid and reliable in their ability to predict functional deficits in players who are at risk. 264

265 8.2 Limitations of the research The author believes that the current thesis has made an original and significant contribution to the available literature in paediatric science and injury risk factor screening. However, to ensure the research process evolves and progresses, it is important to acknowledge some limitations that were present. The key limitations have been outlined below: Confounding factors associated with injury risk: The results of study 5 indicate that there is a lack of association between a number of field-based measures of neuromuscular control and injury risk in elite male youth soccer players. In addition, for some of the chronological age sub-groups, the accuracy of the multivariate model in predicting injury status could be considered low. These findings are in accordance with previous research conducted with elite female soccer players (Ostenberg and Roos, 2000; Nilstad et al., 2014) and adult amateur male players (Engebretsen et al., 2010). Furthermore, despite two recent systematic reviews indicating that frequently used single leg hop and jump tests were able to discriminate between healthy and injured subjects, their ability to prospectively predict injury was limited (Hegadus et al., 2014; Hegadus et al., 2015). Injuries occurring during soccer are multi-factorial in nature (Ekstrand and Gillquist, 1983), with a number of risk factors likely interacting at any given time-point (Meeuwise, 1994). Thus, while reductions in neuromuscular control may increase injury risk (Hewett et al., 2004), other confounding factors such as previous injury and fatigue should also be considered (Engebretsen et al., 2010; Read et al., 2015). 265

266 Singular time point used to quantify neuromuscular control ability for injury tracking: The results of the seasonal variation and injury prevalence studies included in this thesis (studies 3 and 4) showed that injury risk may increase at various time points during the season which could be attributed to accumulated fatigue. Similarly, acute fatigue may be an integral risk factor due to the reported increase in the number of injuries that occur in the latter stages of a soccer match (Price et al., 2004). Therefore, future studies are needed to examine if variation is present in the ability of field-based neuromuscular control assessments to prospectively predict injuries at different time points of the season. Specifically using screening results from January could be recommended based on the findings of study 4 due to the heightened injury risk shown at this time. However, it was also shown that pre-season represented a secondary period of heightened incidence in this cohort. The pre-planned nature of the neuromuscular control assessments: Tasks which are performed in a closed environment are not reflective of the chaotic environments in which many soccer injuries occur. For example, landing and jumping in response to movements from an opponent or ball are characterised by perturbations to the body s centre of mass and perception and decision making skills (Young and Farrow, 2006). This does not permit sufficient time for the neuromuscular system to adequately adjust landing postures which in turn, increases landing forces and may compromise the integrity of joint and soft tissue structures (Besier et al., 2001). However, standardisation of tasks in which athletes are reacting to external stimuli such as a light system does not truly replicate the player s ability to react to the postural cues or the kinematics of an 266

267 opponent, reducing content validity (Abernethy et al., 1998). Also, the use of other players in the testing environment to mirror the role of an opponent will affect the reliability of the measures used. Therefore, the screening battery used in the current study can be considered appropriate for use. The use of chronological age for group classifications: In studies 3, 4 and 5, players were categorised based on their respective chronological age groups. This approach does not account for the variation in the timing and tempo of maturation that can occur in paediatric subjects (Malina et al., 2004; Lloyd et al., 2012), thus classification based on biological stage of maturity would be more advantageous. A commonly applied and accepted method of assessing maturation in field-based environments is the predictive equation proposed by Mirwald et al. (2002). However, due to the reported error associated with this approach (approx. 6 months) and the reduced accuracy the further a player is away from their age at peak height velocity (Gunter et al., 2008), it was deemed a clear and accurate delineation of maturation status was not possible. Thus, separation by chronological age was used in this thesis with maturation reported as a descriptor to infer maturational stage. Further research is required into injury risk factors amongst elite male youth soccer players, using more appropriate and accurate methods of assessing maturation. 267

268 8.3 Directions for future research The available literature examining pertinent risk factors for injury, screening protocols, and risk reduction strategies in elite male youth soccer players is sparse; therefore, there is a wide scope for future research, which can be conducted in this field. Following completion of this thesis, the primary areas that are deemed important to aid further understanding in this area are outlined below: Assessment of injury risk factors under fatigue: Due to the heightened injury risk displayed in response to both acute and chronic fatigue, practitioners may wish to assess their players movement abilities under conditions of fatigue to determine changes in neuromuscular control. Screening players in a non-fatigued state may not accurately identify those individuals whose movement mechanics deteriorate towards the end of a match, affecting their relative risk of injury. Furthermore, longitudinal tracking at selected time points using selected measures should be considered to assess their sensitivity in the detection of both acute and chronic fatigue, which may be associated with a greater injury risk. The use of more sensitive metrics: While a number of the variables assessed in this thesis showed acceptable reliability, were able to identify between group differences and associations with injury risk, other variables were less sensitive. Thus, alternative modes of analysis could be considered for future research. Some examples include: 268

269 - To more clearly determine age related changes in balance performance, alternative measures such as postural sway or the dynamic postural stability index (Wikstrom et al., 2005) may be required. - Based on recent data (Kemper et al., 2015), measuring the rate of change in anthropometric variables could be deemed important for practitioners and could more effectively indicate a player s level of injury risk. In addition, examining the effects of rapid changes in anthropometrics and training volumes and their effects on measures of neuromuscular control may assist practitioners to develop a clearer understanding of the mechanisms and potential aberrant movement mechanics associated with these changes. - In the current study, knee valgus was measured in isolation as recommended using established protocols (Myer et al., 2008a). Recent data has shown that the combination of knee valgus and ipsilateral trunk motion during a single leg drop vertical landing task was predictive of a non-contact knee injury in female athletes (Dingenen et al., 2014). Changes in trunk positioning can alter the resultant ground reaction forces (Blackburn and Padua, 2009), which may increase the loads experienced by the soft tissue structures. Future research should consider the assessment of both kinetic and kinematic variables in this cohort, specifically with consideration of both proximal and distal factors, in addition to the measurement of knee valgus. 269

270 - Knee valgus in the current study was measured using a subjective scoring classification (0, 1, 2, 3) during a repeated tuck jump protocol based on predicted joint angle classifications to maximise time efficiency and aid practical application. This method was piloted with experienced rehabilitation and strength and conditioning specialists showing strong reliability. However, future research could investigate associations with injury risk, comparing the ordinal scoring scale used in this research versus exact knee valgus angle measurements. Analysis of specific lower extremity injuries: While risk factors for lower extremity were shown in the current study, some of the neuromuscular control tests did not display clinical or significant associations. Analysis of specific lower extremity injuries such as knee ligament injuries or ankle injuries was not performed due to limitations in sample size. Future studies should include further sub-division to assess relationships between the neuromuscular control tests in this study and specific lower extremity injuries. For example, knee valgus could be used to analyse relationships with knee injury and the y- balance test for associations with ankle injury based on previous research (Hewett et al., 2005a; Plisky et al., 2006). The effects of different training modalities on measures of neuromuscular control: The results of this study have successfully identified neuromuscular risk factors for lower extremity injury in this cohort. Future research may wish to examine the effects of different intervention strategies and their ability to reduce risk in the metrics used in this 270

271 thesis. This will enable practitioners to include evidenced based training interventions into the i-pass, providing a systematic approach to exercise prescription, neuromuscular screening and risk factor reduction. Some examples include the effects of age appropriate and technique driven strength training, plyometrics and dynamic balance activities in their ability to reduce landing forces and landing force asymmetry or the effects of these modalities on knee valgus in pre-pubertal soccer players. 271

272 Chapter 9 REFERENCES 1. Aagaard P, Simonsen EB, Magnusson SP, Larsson B and Dyhre-Poulsen P. A new concept for isokinetic hamstring:quadriceps muscle strength ratio. Am J Sports Med 26: 231-7, Abernethy B, Wann J and Parks S. Training perceptual motor skills for sport. In: Training for Sport: Applying Sport Science. B. Elliott, ed., Chichester: John Wiley, pp. 1 68, Adrim TA, Cheng TL. Overview of injuries in the young athlete. Sports Med 33: 75-81, Ahmed CS, Neal EA. Elbow valgus instability in the throwing athlete. J Am Acad Orthop Surg 14: , Alentorn-Geli E, Myer GD, Silvers HJ, Samitier G, Romero D, La zaro-haro C, Cugat R. Prevention of non-contact anterior cruciate ligament injuries in soccer players. Part 1: Mechanisms of injury and underlying risk factors. Knee Surg Sports Traum Arthrosc 17: , Alentorn-Geli E, Mendiguchı a J, Samuelsson K, Musahl V, Karlsson J, Cugat R, Myer GD. Prevention of anterior cruciate ligament injuries in sports - Part I: Systematic review of risk factors in male athletes. Knee Surg Sports Traumatol Arthrosc 22: 3-15, Arnason A, Sigurdsson SB, Gudmundsson A, Holme I, Engebretsen L, and Bahr R. Risk Factors for Injury in Football. AM J Sports Med. 32: 4-16, Atkins, SJ, Hesketh, C, Sinclair, JK. The presence of bilateral imbalance of the lower limbs in elite youth soccer players of different ages. Journal Strength & Cond Res. In press. 9. Atkinson G, Nevill AM. Statistical methods for addressing measurement error (reliability) in variables relevant to sports medicine. Sports Med 26: , Augustsson J, Thomee R, Linden C, Folkesson M, Tranberg R, Karlsson J. Single leg hop testing following fatiguing exercise. Reliability and biomechanical analysis. Scand J Med Sci in Sport 16: , Bandy W, Rusche K, Tekulve F. Reliability and limb symmetry for unilateral functional tests of the lower extremities. Isokinetics and Exercise Science 4 (3): , Bangsbo J. The physiology of soccer-with special reference to intense intermittent exercise. Acta, physiologica Scandinavica Supplementum 619, 1-115, Barber S, Frank B, Noyes F, Mangine R, McCloskey J, Hartman W. Quantitative assessment of functional limitations in normal and anterior cruciate ligament-deficient knees. Clin Orth Rel Res 255: ,

273 14. Barber-Westin SD, Galloway M, Noyes FR. Assessment of lower limb neuromuscular control in prepubescent athletes. AM J Sports Med. 33: , Barber-Westin D, Noyes FR, Galloway M. Jump-land characteristics and muscle strength development in young athletes. A gender comparison of 1140 athletes 9 to 17 years. AM J Sports Med. 34: , Bardaglio G, Settanni M, Marasso D, Musella G, Ciairano S. The development and rasch calibration of a scale to measure coordinative motor skills in typically developing children. Advances in Physical Education 22: 88-94, Bates NA, Ford KR, Myer GD, Hewett TE. Impact differences in ground reaction force and center of mass between the first and second landing phases of a drop vertical jump and their implications for injury risk assessment. J Biomech 26: , Beard DJ, Kyberd PJ, Fergusson CM, et al. Proprioception after rupture of the anterior cruciate ligament. An objective indication of the need for surgery? J Bone Joint Surg. 75: , Beck JL, Wildermuth BP. The female athlete's knee. Clin Sports Med 4: , Besier TF, Lloyd DG, Ackland TR, et al. Anticipatory effects on knee joint loading during running and cutting maneuvers. Med Sci Sport Exerc. 33: , Beunen, Ostyn GM, Simons J, Renson R, Van Gerven D. Chronological and biological age as related to physical fitness in boys 12 to 19 years. Annals of Human Biology 8: , Beunen GP, Malina RM. Growth and biologic maturation: relevance to athletic performance in the child and adolescent athlete. H. Hebestreit and O. Bar-Or Eds. Oxford: Blackwell Publishing, 3-17, Beynnon B, Howe JG, Pope MH, et al. The measurement of ACL strain in-vitro. Int Orth.16: 1-12, Beynnon BD, Fleming BC. Anterior cruciate ligament strain in-vivo: A review of previous work. J Biomech 31: , Bieć E, Kuczyński, M. Postural control in 13-year-old soccer players. Eur J Appl Phys Blackburn JT, Padua DA. Sagittal-Plane trunk position, landing forces, and quadriceps electromyographic activity. J Athl Train. 44(2): , Mar-Apr Blume DD, Kennzeichnung Koordinativer Fähigkeiten und Möglichkeiten ihrer Herausbildung im Trainingprozess. In W. Staro- sta (Ed.), The importance of movement co-ordination, its structure and the hierarchy of integrant elements in sport and physical educa- tion. Roma: Centro Studi & Ricerche, Federazione Ital- iana di Atletica Leggera, 14-88, Bohannon R. Hand-held compared with isokinetic dynamometry for measurement of static knee extension torque (parallel reliability of dynamometers). Clin. Phys. Physiol. Meas 11: 217,

274 29. Bolgla L, Keskula D. Reliability of lower extremity functional performance tests. J Orth Sports Phys Ther 26: , Borg GA. Psychophysical bases of perceived exertion. Med Sci Sport Exer 14: , Borms J. The child and exercise: an overview. J Sports Sci 4: 4-20, Bouchard C, Rankinen T. Individual differences in response to regular physical activity. Med Sci Sports Exerc 33(6): , Brink MS, Visscher C, Arends S, Zwerver J, Post WJ, Lemmink K. Monitoring stress and recovery: new insights for the prevention of injuries and illnesses in elite youth soccer players. Br J Sports Med 44: , Brown C, Ross S, Mynark R, Guskiewicz K. Assessing functional ankle instability with joint position sense, time to stabilisation, and electromyography. J Sport Rehabil. 2004;13: Brito J, Malina R, Seabra A, Massada J, Soares J, Krustrup P, Rebelo A. Injuries in Portuguese youth soccer players during training and match play. J Athl Train 4: , Brophy R, Silvers H, Gonzales T, Mandelbaum BR. Gender influences: the role of leg dominance in ACL injury among soccer players. BR J Sports Med 44: , Brosky J, Nitz A, Malone T, Caborn D, Rayens M. Intrarater reliability of selected clinical outcome measures following anterior cruciate ligament reconstruction. J Orth Sports Phys Ther 29: 39-48, Brown C, Ross S, Mynark R, Guskiewicz K. Assessing functional ankle instability with joint position sense, time to stabilisation, and electromyography. J Sport Rehabil. 13: , Bruininks RH, Bruininks BD. Test of motor proficiency. 2nd edition. Manual: AGS Publishing. Circle Pines, Buchheit M, Simpson MB, Al Haddad H, Bourdon PC, Mendez-Villanueva A. Monitoring changes in physical performance with heart rate measures in young soccer players. Eur J Appl Phys 112: , Buchheit M, Mendez-Villanueva A. Reliability and stability of anthropometric and performance measures in highly-trained young soccer players: effect of age and maturation. J Sports Sci 31: , Buchheit M, Al Haddad H, Simpson BM, Palazzi D, Bourdon PC, Di Salvo V, Mendez- Villanueva A. Monitoring accelerations with GPS in football: time to slow down? Int J Sports Physiol Perform 9: 442-5, Bullock-Saxton JE, Janda V, Bullock MI. The influence of ankle sprain injury on muscle activation during hip extension. Int J Sports Med 15: , Bunge SA, Dudukovic NM, Thomason ME, Vaidya CJ, Gabriela JDE. Immature frontal lobe contributions to cognitive control in children. Evidence from fmri. Neuron 31:

275 45. Caldwell BP and Peters DM., Seasonal Variation in Physiological Fitness of a Semiprofessional Soccer Team, J Strength Cond Res, 2009, 23(5), Casajus JA. Seasonal variation in fitness variables in professional soccer players. J Sports Med Phys Fitness 41:463-9, Castro-Pinero J, Ortega FB, Artero EG, Girela-Rejon MJ, Mora J, Sjostrom M, Ruiz J.R. Assessing muscular strength in youth: usefulness of standing long jump as a general index of muscular fitness. J Strength Cond Res 24: , Ceroni D, Martin XE, Delhumeau C, and Farpour-Lambert NJ. Bilateral and gender differences during single-legged vertical jump performance in healthy teenagers. J Strength Cond Res 26: , Chappell JD, Herman DC, Knight BS, Kirkendall DT, Garrett WE, Yu B. Effect of fatigue on knee kinetics and kinematics in stop-jump task. Am J Sports Med 33: , Chimera NJ, Swanik KA, Swanik BC, Straub SJ. Effects of plyometric training on muscle activation strategies and performance in female athletes. J Athl Train 39: 24 31, Chomiak J, Junge A, Peterson L. Severe injuries in football players, influencing factors. Am J Sports Med 28: 58 68, Chorba RS, Chorba DJ, Buillion LE, Overmyer CA, Landis JA. Use of a functional movement screening tool to determine injury risk in female collegiate athletes. N AM J Sports Phys Ther 5: 47-54, Chumanov ES, Heiderscheit BC, Thelen DG. Hamstring musculotendon dynamics during stance and swing phases of high speed running. Med Sci Sports Exerc 43 (3): , Clark NA, Edwards AM, Morton RH, Butterly RJ. Season-to-Season variations of physiological fitness within a Squad of professional male soccer players. J Sports Sci Med 7: , Cloke D, Spencer S, Hodson A, Deehan D. The epidemiology of ankle injuries occurring in English football association academies. BR J Sports Med 43: , Cloke D, Ansell P, Avery P, Deehan D. Ankle injuries in football academies: a threecentre prospective study. Br J Sports Med 45: , Cohen J. Statistical power analysis for the behavioural sciences (2 nd ed.). Hillsdale, NJ: L. Erlbaum Associates, , Colby M, Dawson B, Heasman J, Rogalski B, Gabbett TJ. Training and game loads and injury risk in elite Australian footballers. J Strength Cond Res 28: , Colby S, Hintermeister R, Torry, Steadman R. Lower limb stability with ACL impairment. J Orth Sports Phys Ther 29: , Cook G, Burton L, Hogenboom B. The use of fundamental movements as an assessment of function - Part 1. NAJSPT 1: 62 72,

276 61. Cools W, Martelaer K, Samaey C, Andries C. Movement skill assessment of typically developing preschool children: a review of seven movement skill assessment tools. Journal of Sports Science and Medicine 8: , Coombs R, Garbutt G. Developments in the Use of the Hamstring/Quadriceps Ratio for the Assessment of Muscle Balance. Journal of Sports Science & Medicine 1: 56-62, Cordova ML, Armstrong CW. Reliability of ground reaction forces during a vertical jump: implications for functional strength assessment. J Athl Training 31: , Cormack SJ, Newton RU, Mcguigan M. Neuromuscular and endocrine responses of elite players to an Australian rules football match. International Journal of Sports Physiology and Performance 3: , Coutts A, Reaburn P, Piva T, Rowswell, G. Markers of recovery in team sport athletes. J Sci Med Sport 5: S6, Croce RV, Russell PJ, Swartz EE. Decoster LC. Knee muscular response strategies differ by developmental level but not gender during jump landing. Electromyography and Clinical Neurophysiology 44: , Croisier JL, Crielaard JM. Hamstring muscle tears with recurrent complaints: an isokinetic profile. Isokinetic Exerc Sci 8:175-80, Cumberworth VL, Patel NN, Rogers W, Kenyon GS. The maturation of balance in children. J Laryngology & otology 121; , Daneshjoo A, Rahnama N, Mokhtar AH, Yusof A. Bilateral and unilateral asymmetries of isokinetic strength and flexibility in male young professional soccer players. J Human Kinetics 36: 45-53, Daniel DM, Stone ML, Dobson BE, Fithian DC, Rossman DJ, Kaufman RF. Fate of the ACL injured patient: a prospective outcome study. AM J Sports Med 22: , Davies GJ, Heiderscheit BC. Reliability of the lido linea closed kinetic chain isokinetic dynamometer. J Orthop Sports Phys Ther 25: , De Ste Croix MBA. Advances in paediatric strength assessment: changing our perspective on strength development. J Sports Sci Med. 6: , De Ste Croix M, Deighan M. Dynamic knee stability during childhood in: paediatric biomechanics and motor control: theory and application. Edited by De Ste Croix, M and Korf, T. Routledge, Oxford, England, , De Ste Croix MDS. The effect of football specific fatigue on dynamic knee stability in female youth players. Unpublished report for UEFA, De Ste Croix, MBA, Priestley, A, Lloyd, RS, Oliver JL. ACL Injury risk in elite female youth soccer: changes in neuromuscular control of the knee following soccer specific fatigue. Scandinavian Journal of Science and Medicine in Sports. Doi: /sms ,

277 76. Dingenenan B, Malfaita B, Vanrenterghemb J, Verschuerena SP, Staesa FF. The reliability and validity of the measurement of lateral trunk motion in two-dimensional video analysis during unipodal functional screening tests in elite female athletes. Phys Ther Sport 15: , DiStefano LJ, Padua DA, Blackburn JT, Garrett WE, Guskiewicz KM and Marshall SW. Integrated injury prevention programme improves balance and vertical jump height in children. J Strength Cond Res 24: , Drabik J. Children & sports training: how your future champions should exercise to be healthy, fit and happy. Island Pond, VT: Stadion Publishing, Draganich LF, Vahey JW. An in-vitro study of ACL strain induced by quadriceps and hamstring forces. J Orth Res. 8: 57-63, Dvorak J, Junge A. Football injuries and physical symptoms: a review of the literature. AM J Sports Med 28: 3-9, Ebben WP, VanderZanden T, Wurm BJ, Petushek, EJ. Evaluating plyometric exercises using time to stabilisation. J Strength Cond Res 24: , Eirale C, Tol J, Farooq A, Smiley F, Chalab H. Low injury rate strongly correlates with team success in Qatari professional football. Br J Sports Med 47: , Ekstrand J, Gillquist J, Liljedahl SO. Prevention of soccer injuries: supervision by doctor and physiotherapist. Am J Sports Med 11: , Ekstrand J, Tropp H. The Incidence of ankle sprains in soccer. Foot Ankle 11: 41-44, Ekstrand J, Hägglund M, Waldén M. Injury incidence and injury patterns in professional football: the UEFA injury study. Br J Sports Med 45: , Emery CA. Risk factors for injury in child and adolescent sport: a systematic review of the literature. Clin J Sports Med 13: , Engebretsen,L, Bahr, R. Intrinsic Risk Factors for Hamstring Injuries Among Male Soccer Players: A Prospective Cohort Study. Am J Sports Med 38: , Faigenbaum AD, Farrell A, Fabiano, M, et al. Effects of integrative neuromuscular training on fitness performance in children. Pediatr Exerc Sci 23: , Faigenbaum AD, Myer GD, Fernandez IP, Gomez Carrasco E, Bates N, Farrell A, Ratamess NA, Kang. Feasibility and reliability of dynamic postural control measures in children in first through fifth grades. Int J Sports Phys Ther 9: , Figueiredo AJ, Goncalves CE, Coelho, MJ, Malina RM. Youth soccer players, years: Maturity, size, function, skill and goal orientation. Annals of Human Biology 36: 60-73, Fitzgerald GK, Axe MJ, Snyder-Mackler L. A decision-making scheme for returning patients to high-level activity with non-operative treatment after anterior cruciate ligament rupture. Knee Surg Sports Traumatol Arthrosc 8: 76 82, Ford KR, Myer GD and Hewett TE. Valgus knee motion during landing in high school female and male basketball players. Med. Sci Sports Exerc 35: ,

278 93. Flanagan EP, Ebben WP, Jensen RL. Reliability of the reactive strength index and time to stabilisation during depth jumps. J Strength Cond Res 22: , Ford KR, Myer GD, Hewett TE. Longitudinal effects of maturation on lower extremity joint stiffness in adolescent athletes. Am J Sports Med 38: , Ford KR, Shapiro R, Myer GD, Van Den Bogert AJ, Hewett TE. Longitudinal sex differences during landing in knee abduction in young athletes. Med Sci Sports Exerc 42: , Foster C. Monitoring training in athletes with reference to overtraining syndrome. Med Sci Sports Exerc 30: , Fousekis K, Tsepis E, Poulmedis P, Anthanasopoulos S, Vagenas G. Intrinsic risk factors of non-contact quadriceps and hamstring strains in soccer: a prospective study of 100 professional players. Br J Sports Med 45: , Freckleton G, Cook J, Pizzari T. The predictive validity of a single leg bridge test forhamstring injuries in Australian Rules Football Players. Br J Sports Med 48: , Friel K, McLean N, Myers C, Caceres M. Ipsilateral hip abductor weakness after inversion ankle sprain. J Athl Train 41: 74 78, Fukushi Yamada RK, Gonçalves Arliani G, Peixoto Leão Almeida G, Manrique Venturine A, Veronese dos Santos C, Costa Astur D, Cohen M. The effects of one-half of a soccer match on the postural stability and functional capacity of the lower limbs in young soccer players. CLINICS 67: , Fuller CW, Ekstrand J, Junge A, Andersen TE, Bahr R, Dvorak J, Hägglund M, McCrory P, Meeuwisse WH. Consensus statement on injury definitions and data collection procedures in studies of football (soccer) injuries. Scand J Med Sci Sports 16: 83 92, Gabbe BJ, Finch CF, Bennell KL, Wajswelner H. How valid is a self-reported 12 month sports injury history? Br J Sports Med 37: , Gabbett TJ, Shahid U. Relationship between running loads and soft-tissue injury in elite team sport athletes. J Strength Cond Res 26: , Gallahue DL. Understanding motor development in children. New York: John Wiley & Sons, , Geldhof E, Cardon G, De Bourdeaudhuij I, Danneels L, Coorevits P, Vanderstraeten G, De Clercq D. Static and dynamic standing balance: test-retest reliability and reference values in 9 to 10 year old children. Euro J Pediatrics 165: , Gentry V, Gabbard C. Foot preference behavior: a developmental perspective. J Gen Psychol 122(1): 37 45, Goerger, BM, Marshall, SW, Beutler, AI, Blackburn, JT, Wickens, JH, Padua, DA. Asymmetry of lower extremity biomechanics in patients with prior ACL injury: The JUMP-ACL study. Free communications, rapid fire poster presentations: Biomechanical and neuromuscular consequences of ACL injury. 278

279 108. Gerodimos V, Zafeiridis A, Perkos S, Dipla K, Manou V, Kellis S. The contribution of stretch-shortening cycle and arm-swing to vertical jumping performance in children, adolescents, and adult basketball players. Ped Ex Sci 20: , Gomes JL, de Castro JV, Becker R. Decreased hip range of motion and noncontact injuries of the anterior cruciate ligament. Arthroscopy 24: , Goldie PA, Bach TM, Evans OM. Force platform measures for evaluating postural control: reliability and validity. Arch Phys Med Rehabil 70: , Goossens L, Witvrouw E, Vanden Bossche L and De Clercq D. Lower eccentric hamstring strength and single leg hop for distance predict hamstring injury in PETE students. Eur J Sports Sci 15: , Grahm B, Regahr G, Wright J. Delphi. A Method to establish consensus for diagnostic criteria. J Clin Epidemiology 56: , Gravina, L, Gil, SM, Ruiz, F, Zubero, J, Gil, J, and Irazusta, J. Anthropometric and physiological differences between first team and reserve soccer players aged years at the beginning and end of the season. J Strength Cond Res 22: , Gribble P, Mitterholzer J and Myers A. Normalizing Considerations for Time to Stabilisation Assessment. J Sport Sci Med Sport, 15: , Griffin, LY, Agel, J, Albohm, MJ et al. Non-contact ACL injuries: risk factors and prevention strategies. J AM Academy Orthop Surg 8: , Guskiewicz KM, Perrin DH, Gansneder B. Effect of mild head injury on postural stability in athletes. J Athl Train. 31: , Haberman SJ. The analysis of residuals in cross-classified tables. Biometrics 29: , Habib Z, Westcott S. Assessment of anthropometric factors on balance tests in children. Ped Phys Ther 10: , Hagglund M, Walden M, Bahr R, Ekstrand J. Methods for epidemiological study of injuries to professional football players: developing the UEFA model. Br J Sports Med: 39: , Hagguland M, Walden M, Ekstrand J. Previous injury as a risk factor for injury in elite football: a prospective study over two consecutive seasons. Br J Sports Med 40: , Hagguland M, Walden M, Ekstrand J. Injuries among male and female elite football players. Scand J Med Sci Sport 19: , Haines S, Baker T, Donaldson M. Development of a physical performance assessment checklist for athletes who sustained a lower extremity injury in preparation for return to sport: a Delphi study. IJSPT 8: 44-53, Hamilton RT, Shultz SJ, Schmitz RJ, Perrin DH. Triple-Hop Distance as a Valid Predictor of Lower Limb Strength and Power. J Athl Train. 43: ,

280 124. Han SH, Yun MH, Kim KJ, Kwahk J. Evaluation of product usability: development and validation of usability dimensions and design elements based on empirical models. Int J Indust Erg 26: , Harmon KG, Dick R. The relationship of skill level to anterior cruciate ligament injury. Clin J Sport Med 8: 260-5, Hartman E, Houwen S, Scherder E, Visscher C. The relationship between motor performance and executive functioning in children with intellectual disabilities. J Intel Disability Res 54: , Hass, CJ, Schick, EA, Chow, JW, Tillman, MD, Brunt, D and Cauraugh, J.H. Lower extremity biomechanics differ in prepubescent and postpubescent female athletes during stride jump landings. Journal of Applied Biomechanics 19: , Hawkins D, and Metheny J. Overuse injuries in youth sports: biomechanical considerations. Med Sci Sport Exerc 33: , Hawkins R., Hulse M., Wilkinson C., Hodson A., Gibson M. The association football medical research programmeme: An audit of injuries in professional football. BR J Sports Med 35, 43-47, Hebert LJ, Maltais D, Lepage Saulnier S Crete M, Perron M. Isometric muscle strength in youth assessed by hand-held dynamometry: A feasibility, reliability, and validity study. Pediatr Phys Ther 23: , Hegedus, EJ, McDonough, S, Bleakley, C, Cook, CE, Baxter, DG. Clinician-friendly lower extremity physical performance measures in athletes: a systematic review of measurement properties and correlation with injury, part 1. The test for knee function including hop tests. Br J Sports Med doi: /bjsports Hegedus EJ, McDonough S, Bleakley C, Baxter DE, Cook CE. Clinician-friendly lower extremity physical performance tests in athletes: a systematic review of measurement properties and correlation with injury. Part 2 - the test for hip, thigh, foot and ankle including the star excursion balance test. Br J Sports Med 49: , Henderson S.E, Sugden DA, Barnett AL. Movement assessment battery for children 2 examiner s manual. Harcourt Assessment, London, Henderson, G, Barnes, CA, Portas, MD. Factors associated with increased propensity for hamstring injury in English Premier League soccer players. J Sci Med Sport 13: , Herrington L, Myer GD, Munro A. Intra and inter-tester reliability of the tuck jump assessment. Phys Ther Sport 14: , Hertel J, Braham RA, Hale SA, Olmsted-Kramer LC. Simplifying the star excursion balance test: analyses of subjects with and without chronic ankle instability. J Orthop Sports Phys Ther. 36: , Hewett TE, Paterno MV, Myer GD. Strategies for enhancing proprioception and neuromuscular control of the knee. Clin Orthop 402: 76-94,

281 138. Hewett TE, Myer GD, Ford KR. Decreases in neuromuscular control about the knee with maturation in female athletes. J Bone Joint Surgery AM 86: , Hewett TE, Myer GD, Ford KR et al. Biomechanical measures of neuromuscular control and valgus loading of the knee predict anterior cruciate ligament injury risk in female athletes: a prospective study. AM J Sports Med. 33: , 2005a Hewett TE, Ford KR, Myer GD et al. Effect of puberty and gender on landing force and jump height. Med Sci Sports Exerc. 37: S66, 2005b Hewett TE, Myer GD, Ford KR, Slauterbeck JR. Preparticipation physical examination using a box drop vertical jump test in young athletes: the effects of puberty and sex. Clin J Sport Med 16: , Hewett T and Johnson D. ACL prevention programmes: fact or fiction? Orthopedics 33: 36-39, Hewett TE, Ford KR, Hoogenboom BJ, Myer GD. Understanding and preventing ACL injuries: current biomechanical and epidemiological considerations-update. N AM J Sports Phys Ther 5: , Hewett TE and Myer GD. The mechanistic connection between the trunk, knee, and anterior cruciate ligament injury. Exerc Sport Sci Rev. 39: , Hewit J, Cronin J, Hume P. Multidirectional leg asymmetry assessment in sport. Strength Cond J 34: 82-86, Hill, C, Croce, R, Miller, J, Cleland, F. Muscle torque relationships between hand-held dynamometry and isokinetic measurements in children ages 9 to 11. J Strength Cond Res 10: 77-82, Hill-Haas S, Coutts A, Rowsell G, Dawson B. Variability of acute physiological responses and performance profiles of youth soccer players in small-sided games. J Sci Med Sport 11: , Hill-Haas SV, Dawson BT, Coutts AJ, Rowsell GJ. Physiological responses and time motion characteristics of various small-sided soccer games in youth players. J Sports Sci 27: 1 8, Hirtz P, Starosta W. Sensitive and critical periods of motor co-ordination development and its relation to motor learning. J Hum Kinet 7: 19-28, Hogan KA, Gross RH. Overuse injuries in pediatric athletes. Orthop Clin North Am 34: , Holcomb WR, Rublet MD, Lee HJ, et al. Effect of hamstring-emphasized resistance training on hamstring: quadriceps strength ratios. J Strength Cond Res. 21: 41-47, Hollman JH, Ginos BE, Kozuchowski J, Vaughn AS, Krause DA, Youdas JW. Relationships between knee valgus, hip-muscle strength, and hip-muscle recruitment during a single-limb step-down. J Sport Rehab 18: , Hopkins WG, Hawley JA, Burke LM. Design and analysis of research on sport performance enhancement. Med Sci Sports Exerc 31: ,

282 154. Hopkins WG. Measures of reliability in sports medicine and science. Sports Med 30: 1 15, Hopper DM, Hopper JL, Elliott BC. Do selected kinanthropometric and performance variables predict injuries in female netball players? J Sports Sci 13: , Howatson G. The impact of damaging exercise on electromechanical delay in biceps brachii. Journal of Electromyography and Kinesiology 20: , Hsieh CY, Phillips RB. Reliability of manual muscle testing with a computerized dynamometer. J Manipul Physiol Ther13:72-82, Hubbard TJ, Kramer LC, Denegar CR, Hertel J. Correlations among multiple measures of functional and mechanical instability in subjects with chronic ankle instability. Journal of Athletic Training 42: , Hughes G, Watkins J. A risk-factor model for anterior cruciate ligament injury. Sports Med 36: , Huijgen BCH, Elferink-Gemser MT, Post WJ, Visscher C. Soccer skill development in professionals. Int J Sports Med 30: , Hung WY, Pang MY. Effects of group-based versus individual-based exercise training on motor performance in children with developmental coordination disorder: a randomized controlled pilot study. J Rehab Med 42: , Hurley MV. The effects of joint damage on muscle function, proprioception and rehabilitation. Man Ther 2: 11-7, Iga J, George K, Lees A. Cross-sectional investigation of indices of isokinetic leg strength in youth soccer players and untrained individuals. Scand J Med Sci Sports. 19: , Imai A, Kaneoka K, Okubo Y, et al. Effects of two types of trunk exercises on balance and athletic performance in youth soccer players. Int J Sports Phys Ther. 9: 47-57, Imprezzilini F, Bizzini M, Dvorak J. Physiological and performance responses to the FIFA 11+ (part 2): a randomised control trial on the training effects. J Sport Sci 31: , Ireland ML. The female ACL: why is it more prone to injury? Orthop Clin North AM. 33: , Jaynathi NA, LaBella CR, Fischer D, Pasulka J, Dugas LR. Sports-specialised intensive training and the risk of injury in young athletes: a clinical case-control study. AM J Sports Med 43: , Johnson A, Doherty PJ, Freemont A. Investigation of growth, development, and factors associated with injury in elite schoolboy footballers: prospective study. BMJ 338: 490, Jones MA, Hitchen PJ, Stratton G. The importance of considering biological maturity when assessing physical fitness measures in girls and boys aged 10 to 16 years. Annals of Human Biology 27: 57-65,

283 170. Junge A, Dvorak J. Influence of definition and data collection on the incidence of injuries in football. Am J Sports Med 20: 40-46, Junge A, Chomiak J, Dvorak J. Incidence of football injuries in youth players: comparison of players from two European regions. Am J Sports Med 28: 47-50, Katzmarzyk PT, Malina RM, Beunen GP. The contribution of biological maturation to the strength and motor fitness of children. Annals of Human Biology. 24: , Kazman JB, Galecki JM, Lisman PD, O'Connor PA, Francis G. Factor structure of the functional movement screen in marine officer candidates. Journal of Strength & Conditioning Research 28: , Kellis S, Gerodimos V, Kellis E, Manou V. Bilateral isokinetic concentric and eccentric strength profiles of the knee extensors and flexors in young soccer players. Isokinetics and Exer Sci 9: 31-39, Kemper GLJ, van der Sluis A, Brink MS, Visscher C, Frencken WGP, Elferink-Gemser MT. Anthropometric injury risk factors in elite-standard youth soccer. DOI /s Published online: Int J Sports Med Kiesel K, Plisky PJ, Voight ML. Can serious injury in professional football be predicted by a preseason functional movement screen? North Am J Sport Phys Ther 2: , Kinzey SJ, Armstrong CW. The reliability of the star-excursion test in assessing dynamic balance. Journal Orth Sports Phys Ther 27: , Klugman MF, Brent JL, Myer GD, Ford KR, and Hewett TE. Does an in-season only neuromuscular training protocol reduce deficits quantified by the tuck jump assessment? Clin Sports Med 30: , Komi PV. Stretch-shortening cycle: a powerful model to study normal and fatigued muscle. J Biomech 33: , Kovaleski JE, Heitman RJ, Gurchiek LR, et al. Reliability and effects of leg dominance on lower extremity isokinetic force and work using the closed chain rider system. J Sport Rehabil 6: , Kraus K, Schütz E, Taylor WR, Doyscher R. Efficacy of the functional movement screen: a review. J Strength Cond Res 28: Kray JM. Diagnosis and management of quadriceps strains and contusions. Curr Rev Musciloskeletal Med 3: 26-31, Krosshaug T, Nakamae A, Boden BP, et al. Mechanisms of anterior cruciate ligament injury in basketball: video analysis of 39 cases. AM J Sports Med. 35: , Kubo K, Morimoto M, Komuro T, et al. Influence of tendon stiffness, joint stiffness and electromyographic activity on jump performances using single joint. Eur J Appl Phys 99: 235, Kucera KL, Marshall SW, Kirkendall DT, Marchak PM, Garrett WE. Injury history as a risk factor for incident in youth soccer. Br J Sports Med 39: ,

284 186. Laffaye G, Bardy BG, Durey A. Leg stiffness and expertise in men jumping. Medicine and Science in Sports and Exercise, 37: , Lambert MI, Mbambo ZH, Gibson AS. Heart rate during training and competition for long distance running. J Sports Sci 16: 85-90, Landis JR, Koch GG. Agreement measures for categorical data. Biometrics 33:159-74, Lazaridis S, Bassa E, Patikas P, Giakas G, Gollhofer A, Kotzamanidis C. Neuromuscular differences between prepubescent boys and adult men during drop jump. Eur J Appl Physiol 110: 67-74, Lazaridis S, Bassa E, Patikas D, Hatzikotoulas K, Lazaridis FK, Kotzamanidis CM. Biomechanical comparison in different jumping tasks between untrained boys and men. Ped Ex Sci 25: , Lee MJ, Reid SL, Elliott BC, Lloyd DG. Running biomechanics and lower limb strength associated with prior hamstring injury. Med Sci Sports Exerc 41: , Lee TQ, Anzel SH, Bennett KA, et al. The influence of fixed rotational deformities of the femur on the patellofemoral contact pressures in human cadaver knees. Clin Orthop. 302: 69-74, Leetun D, Ireland M, Wilson J, et al. Core stability measures as risk factors for lower extremity injury in athletes. Med Sci Sport Exerc. 36: , Le Gall, F, Carling C, Reilly T, Vandewalle H, Chruch, J, Rochcongar P. Incidence of injuries in elite French youth soccer players; a 10-season study. AM J Sports Med 34: , Lehance C, Binet J, Bury T, Crosie, JL. Muscular strength, functional performances and injury risk in professional and junior elite soccer players. Scand J Med Sci Sport 19: , Lephart SM, Ferris CM, Riemann BL, Myers JB, Fu FH. Gender differences in strength and lower extremity kinematics during landing. Clin Orthop Relat Res , Li G, Rudy TW, Allen C, et al. Effect of combined axial compressive and anterior tibial loads on in situ forces in the anterior cruciate ligament: a porcine study. J Orthop Res 16: 122-7, Lisman P, O'Connor FG, Deuster PA, Knapik JJ. Functional movement screen and aerobic fitness predict injuries in military training. Med Sci Sport and Exer 45: , Lloyd, RS, Oliver JL, Hughes MG, Williams CA. Reliability and validity of field-based measures of leg stiffness and reactive strength index in youths. J Sports Sci 27: 14, , Lloyd RS, Oliver JL. The youth physical development model: a new approach to longterm athletic development. Strength Cond J 34: 61 72,

285 201. Lloyd RS, Oliver JL, Hughes MG, Williams CA. Age-related differences in the neural regulation of stretch shortening cycle activities in male youths during maximal and submaximal hopping. J Electromy Kinesiol 22: 37-43, Lloyd RS, Oliver JL, Faigenbaum AD, Myer GD, De Ste Croix MBA. Chronological age vs. biological maturation: implications for exercise programmeming in youth. J Strength Cond Res 28: , Lloyd RS, Oliver JL, Radnor JM, Rhodes BC, Faigenbaum AD, Myer GD. Relationships between functional movement screen scores, maturation and physical performance in young soccer players. J Sports Sci 33: 11 19, Lundgren LE, Tran TT, Nimphius S, Raymond E, Secomb JL, Farley, OR, Newton, RU, Steele, JR, Sheppard, JM. Development and evaluation of a simple, multi-factorial model based on landing performance to indicate injury risk among surfing athletes. IJSPP. Epub MacDonald PL, Gardner RC. Type 1 error rate comparisons of post hoc procedures for I J chisquare tables. Educational and Psychological Measurement, 60: , Mafulli N, Pintore E. Intensive training in young athletes. BR J Sports Med 24: , Magaziner J, Bassett S, Hebel R, Gruber-Baldini A. Use of proxies to measure health and functional status in epidemiologic studies of community-dwelling women aged 65 years and older. Am J Epidemiology 143: , Malina RM, Bouchard C, Bar-Or O. Timing and sequence of changes during adolescence: in growth, maturation and physical activity. Champaign ILL: Human Kinetics , 2004a Malina RM, Eisenmann JC, Cumming SP, Riberio B, Aroso J. Maturity-associated variation in the growth and functional capacities of youth football (soccer) players years. Eur J Appl Physiol 91: , 2004b Malina RM, Coelho E Silva MJ, Figueiredo AJ, Carling C, Beunen GP. Interrelationships among invasive and non-invasive indicatorsof biological maturation in adolescent male soccer players. J Sports Sci 30: , Malina RM, Kozieł SM. Validation of maturity offset in a longitudinal sample of Polish boys. J Sports Sci 32: , Malliou P, Gioftsidou A, Pafis G, et al. Proprioceptive training (balance exercises) reduces lower extremity injuries in young soccer players. J Back Musculoskelet Rehabil. 17: , Maulder P, Cronin, J. Horizontal and vertical jump assessment: reliability, symmetry, discriminative and predictive ability. Physical Therapy in Sport 6: 74 82, McCall A, Carling C, Nedelec M, Davison M, Le Gall F, Berthoin S, Dupont G. Risk factors, testing and preventative strategies for non-contact injuries in professional football: current perceptions and practices of 44 teams from various premier leagues. BJSM. doi: /bjsports

286 215. McCall A, Nedelec M, Carling C, Le Gall F, Berthoin S, Dupont G. Reliability and sensitivity of a simple isometric posterior lower limb muscle test in professional football players. J Sports Sci 33: 12, McCormack RJ. The flexibility of the collagen compartment of muscle. Meat Science 36: 79-91, McCunn R, Aus der Fünten K. Fullagar HK, McKeown I, Meyer T. Reliability and association with injury of movement screens: a critical review. Sports Med. In press McGill SM, Childs A, and Liebenson C. Endurance time for low back stabilisation exercises: clinical targets for testing and training from a normal database. Arch Phys Med Rehabil 80: , McGuine TA, Greene JJ, Best T, et al. Balance as a predictor of ankle injuries in high school basketball players. Clin J Sport Med.10: , McGuine TA, Keene JS. The effect of a balance training programme on the risk of ankle sprains in high school athletes. AM J Sports Med. 34: McLean BD, Coutts AJ, Kelly V, McGuigan MR, Cormack SJ. Neuromuscular, endocrine, and perceptual fatigue responses during different length between match microcycles in professional rugby league players. Int J Sports Physiol Perform 5: , Mclean S, Felin R, Suedekum, N, Calabrese G, Passerallo A, Joy S. Impact of fatigue on gender-based high-risk landing strategies. Med Sci Sport and Exerc 39: , Mclellan C, Lovell D. Creatine kinase and endocrine responses of elite players pre, during, and post rugby league match play. J Strength Cond Res 24: , Meeuwisse WH. Assessing causation in sport injury: a multifactorial model. Clin J Sport Med.4: , Stemmons MV, Lewis C. Hip abductor and knee extensor muscle strength of children with and without Down syndrome. Pediatric Physical Therapy 13: 18-26, Meylan C, McMaster T, Cronin J. Single-leg lateral, horizontal, and vertical jump assessment: reliability, interrelationships, and ability to predict sprint and change-ofdirection performance. J Strength Cond Res 23: , 2009a Meylan, C and Malatesta D. Effects of in-season plyometric training within soccer practice on explosive actions of young players. J Strength Cond Res 23: , 2009b Michaud PA, Renaud A and Narring F. Sports activities related to injuries? A survey among 9-19 year olds in Switzerland. Inj Prev 7: 41-45, Mickel KJ, Munro BJ, Steele JR. Gender and age affect balance performance in primary school-aged children. J Sci Med Sport 14:243-8, Miller A, Callister R. Reliable lower limb musculoskeletal profiling using easily operated, portable equipment. Phys Ther Sport 10: 30 37, Mirwald RL, Baxter-Jones ADG, Bailey DA, Beunen GP. An assessment of maturity from anthropometric measurements. Med Sci Sports Exerc 34: ,

287 232. Moore O, Cloke D, Avery P, Beasley I, and, Deehan D. English premiership academy knee injuries: lessons from a 5 year study. J Sports Sci. 29: , Moore SA, McKay HA, Macdonald H, Nettlefold L, Baxter-Jones ADG, Cameron N, Brasher PMA. Enhancing a Somatic Maturity Prediction Model. Med Sci Sport Exerc. In Press Mountjoy M, Armstrong N, Bizzini L, Blimkie C, Evans J, Gerrard D, Hangen J, Knoll K, Micheli L, Sangenis P, Van Mechelen WV. IOC Consensus statement: training the elite child athlete. Br J Sports Med 42: 163-4, Munro, A and Herrington, L. Between-session reliability of the star excursion balance test. Phys Ther Sport 11: , Munro, AG and Herrington, LC. Between-session reliability of four hop tests and the agility T-test. J Strength Cond Res 25: , Myer G, Ford K, Hewett T. Rationale and clinical techniques for anterior cruciate ligament injury prevention among female athletes. J Athletic Train 39: , Myer GD, Ford KR, Palumbo JP, et al. Neuromuscular training improves performance and lower extremity biomechanics in female athletes. J Strength Cond Res. 19: 51 60, Myer GD, Ford KR, McLean SG, et al. The effects of plyometric versus dynamic stabilisation and balance training on lower extremity biomechanics. Am J Sports Med. 34: , 2006a Myer GD, Paterno MV, Ford KR, Quatmen CE, Hewett TE. Rehabilitation after anterior cruciate ligament reconstruction: criteria-based progression through the return-to-sport phase. JOSPT 36: , 2006b Myer GD, Ford KR, Hewett TE. Tuck jump assessment for reducing anterior cruciate ligament injury risk. Athl Ther Today 13: 39 44, 2008a Myer GD, Chu DA, Brent JL, Hewett TE. Trunk and hip control neuromuscular training for the prevention of knee joint injury. Clin Sports Med 27: , 2008b Myer GD, Ford KR, Khoury J, Succop P, Hewett TE. Development and validation of a clinic based prediction tool to identify female athletes at high risk of ACL injury. Am J Sports Med 38: , Myer GD, Brent JL, Ford KR, et al. Real-time assessment and neuromuscular training feedback techniques to prevent ACL injury in female athletes. Strength Cond J. 33: 21-35, 2011a Myer GD, Ford KR, Khoury JK, Hewett TE. Three-dimensional motion analysis validation of a clinic based nomogram designed to identify high ACL injury risk in females. Phys Sports Med 1: 19-28, 2011b Myer GD, Ford FR, Best TM, Bergeron MF and Hewett TE. When to initiate neuromuscular training to reduce sport related injuries and enhance health and youth. Curr Sports Med Rep 10: , 2011c. 287

288 247. Myer GD, Stroube BW, DiCesare, Brent JL Ford KR, Heidt RS, Hewett, TE. Augmented feedback supports skill transfer and reduces high-risk injury landing mechanics: a double-blind, randomized controlled laboratory study. Am J Sports Med 41: , Myer GD, Kushner AM, Brent JL, Schoenfeld BJ, Hugentobler J, Lloyd RS, Vermeil A, Chu DA, Harbin J, McGill SM. The back squat: a proposed assessment of functional deficits and technical factors that limit performance. Strength Cond J 36: 4-27, Myers RH, Classical and modern regression with application. 2nd Edition. PWS-KENT, Boston Nariyama K, Hauspie RC, Mino T. Longitudinal growth study of male Japanese junior high school athletes. Am J Hum Biol 13: , Nesser TW, Huxel KC, Tincher JL, et al. The relationship between core stability and performance in division I football players. J Strength Cond Res. 22: , Neville V, Gleeson M, Folland JP. Salivary IgA. As a risk factor for upper respiratory infections in elite professional athletes. Med Sci Sports 40: , Nicol C, Avela J, Komi PV. The stretch-shortening cycle: a model to study naturally occurring neuromuscular fatigue. Sports Med 36: , Nilstad, A, Andersen, TE, Bahr, R, Holme, I and Steffen, K. Risk factors for lower extremity injuries in elite female soccer players. Am J Sports Med 42: , Nolan L, Grigorenko A, Thorstennson A. Balance control: sex and age differences in 9 to 16 year olds. Dev Med Child Neurol 47: , Noyes FR, Barber SD, Mangine RE: Abnormal lower limb symmetry determined by functional hop tests after anterior cruciate ligament rupture. Am J Sports Med 19: , Noyes FR, Barber-Westin SD, Fleckenstein C, Walsh C, West J. The drop jump screening test: difference in lower limb control by gender and effect on neuromuscular training in female athletes. AM J Sports Med 33: , Ochi E, Nakazato K, Song H, Nakajima H. Aging effects of passive resistive torque in the rat ankle joint after lengthening contractions. J Orth Sci 13: , O Connor FG, Deuster PA, Davis JAVIS, Pappas CG and Knapik JJ. Functional movement screening: predicting injuries in officer candidates. Med. Sci. Sports Exerc 43: , Oliver J, Armstrong N, Williams C. Changes in jump performance and muscle activity following soccer-specific exercise. J Sports Sci 1-8: 141-8, Oliver JL, De Ste Croix MBA, Lloyd RS, Williams CA. Altered neuromuscular control of leg stiffness following soccer-specific exercise. Eur J Appl Physiol 114: , Onate JA, Dewey T, Kollock RO, Thomas KS, Van Lunen BL, DeMaio M, Ringleb SI. Real-time intersession and interrater reliability of the functional movement screen. J Strength Cond Res 26: ,

289 263. Onishi, H., Yagi, R., Oyama, M., Akasaka, K., Ihashi, K., & Handa, Y. EMG-angle relationship of the hamstring muscles during maximum knee flexion. J Electromyo Kinesiol 12, , Opar D, Timmins R, Dear N, Williams M, Shield A. The role of neuromuscular inhibition in hamstring strain injury recurrence. Journal of Electromyography and Kinesiology 23: , 2013a Opar, DA, Piatkowski, T, Williams, MD, Shield, AJ. A Novel Device Using the Nordic Hamstring Exercise to Assess Eccentric Knee Flexor Strength: A Reliability and Retrospective Injury Study. J Orth Sports Phys Ther 43: , Opar DA, Williams MD, Timmins RJ, Hickey, J, Duhig SJ, and Shield AJ. Eccentric hamstring strength and hamstring injury risk in Australian footballers. Med. Sci. Sports Exerc 47: 12-20, Osternberg A and Roos H. Injury risk factors in female European football: a prospective study of 123 players during one season. Scand J Med Sci Sport 10: , Osternig LR. Assessing human performance. In E. Brown (Ed.), Isokinetics in human performance, (pp.77-96). Champaign IL: Human Kinetics, 77-96, Ostojic SM. Seasonal alterations in body composition and sprint performance of elite soccer players. JEPonline 6:11-14, Pacey V, Nicholson L, Adams RD, Munn J, Munns, CF. Generalized joint hypermobility and risk of lower limb joint injury during sport: a systematic review with meta-analysis. Am J Sports Med 38: , Padilla S, Mujika J, Orbananos, J, Angulo F. Exercise intensity during competition time trials in professional road cycling. Med Sci Sports Exerc 32: , Padua DA, Carcia CR, Arnold BL, Granata KP. Gender differences in leg stiffness and stiffness recruitment strategy during two-legged hopping. J Mot Behav 37: , Padua D, Arnold B, Perrin D, Gansneder B, Carcia C, Granata, K. Fatigue, vertical leg stiffness and stiffness control strategies in males and females. J Athl Train 41: , Padua DA, Marshall SW, Boling MC, Thigpen CA, Garrett WE, Beutler AI. The landing error scoring system (LESS) is a valid and reliable clinical assessment tool of jumplanding biomechanics. Am J Sports Med 37: , Padua DA, Boling MC, DiStefano LJ, Onate JA, Beutler AI and Marshall SW. Reliability of the Landing Error Scoring System-Real Time, a Clinical Assessment Tool of Jump- Landing Biomechanics. J Sport Rehab 20: , Padua, DA, DiStefano, LJ, Beutler, AI, de La Motte, SJ, DiStefano, MJ, Marshall, SW. The landing errors scoring system as a screening tool for an anterior cruciate ligament injury-prevention programme in elite-youth soccer athletes. J Athl Train 50: ,

290 277. Parchmann CJ, McBride JM. Relationship between functional movement screen and athletic performance. J Strength Cond Res 25: , Paterno M, Greenberger H. The test-retest reliability of a one legged hop for distance in young adults with and without ACL reconstruction. Isokinetics and Exercise Science 6: 1-6, Paterno MV, Myer GD, Ford KR, Hewett TE. Neuromuscular training improves singlelimb stability in young female athletes. J Orthop Sports Phys Ther. 34: , Pau M, Ibba G, Leban Scorcu M. Characterization of static balance abilities in elite soccer players by playing position and age. Research in Sports Medicine: An International Journal 22: , Pau M, Arippa F, Leba, B. Corona F, Ibba G, Todde F, Scorcu M. Relationship between static and dynamic balance abilities in Italian professional and youth league soccer players. Phys Ther in Sports. In Press Peng HT, Kernozek T, Song CY. Quadricep and hamstring activation during drop jumps with changes in height. Phys Ther Sport, 12: , Peterson L, Junge A, Chomiak J. Incidence of football injuries and complaints in different age groups and skill-level groups. Am J Sports Med 28: S51 7, Petschnig R, Baron R, Albrecht M. The relationship between isokinetic quadriceps strength test and hop tests for distance and one-legged vertical jump test following anterior cruciate ligament reconstruction. J Orthop Sports Phys Ther 28: 23-31, Philippaerts RM, Vaeyens R, Janssens M, Van Renterghem B, Matthys D, Craen R, Bourgois J, Vrijens J, Beunen GP, and Malina RM. The relationship between peak height velocity and physical performance in youth soccer players. J Sports Sciences 24: , Plisky PJ, Rauh MJ, Kaminski TW, Underwood FB. Star Excursion Balance Test as a predictor of lower extremity injury in high school basketball players. J Orthop Sports Phys Ther 36: , Plisky PJ, Gorman PP, Butler RJ, Kiesel KB, Underwood FB, Elkins B. The reliability of an instrumented device for measuring components of the Star Excursion Balance Test. North American Journal of Sports Physical Therapy 4: 92-99, Pollard CD, Sigward Ota S, Langford K, and Powers CM. The influence of in-season injury prevention training on lower-extremity kinematics during landing in female soccer players. Clin J Sport Med 16: , Powell JW, Barber-Foss KD. Sex related injury patterns among selected high school sports. AM J Sports Med 28: , Prapavessis H, McNair PJ. Effects of instruction in jumping technique and experience jumping on ground reaction forces. J Orthop Sports Phys Ther. 29: , Price RJ, Hawkins RD, Hulse MA, Hodson A. The football association and medical research programmeme: an audit of injuries in academy youth football. Br J Sports Med 38: ,

291 292. Quatman CE, Ford KR, Myer GD, Hewett TE. Maturation leads to gender differences in landing force and vertical jump performance: a longitudinal study. Am J Sports Med 34: , Read M, Cisar C. The influence of varied rest interval length on depth jump performance. J Strength Cond Res 15: , Read PJ, Oliver JL, De Ste Croix MBA, Myer GD, Lloyd RS. Injury risk factors in male youth soccer players. Strength Cond J. 37: 1-7, Read PJ, Oliver JL, De Ste Croix MBA, Myer GD, Lloyd RS. Reliability of the tuck jump screening assessment in elite male youth soccer players. J Strength Cond Res. In press Read, PJ, Oliver, JL, De Ste Croix, MBA, Myer, GD and Lloyd, RS. Neuromuscular Risk Factors for Knee and Ankle Ligament Injuries in Male Youth Soccer Players. Sports Med. In Press a Read PJ, Oliver JL, De Ste Croix MBA, Myer GD, Lloyd RS. Assessment of injury risk factors in male youth soccer players. Strength Cond J. In Press b Read PJ, Oliver JL, De Ste Croix MBA, Myer GD, Lloyd RS. Early sport specialisation in male youth soccer players: is the risk worth the reward. J Sports Sci. In Press c Reid A, Birmingham TB, Stratford PW, Alcock GK, Griffin RJ. Hop Testing Provides a Reliable and Valid Outcome Measure During Rehabilitation After Anterior Cruciate Ligament Reconstruction. Phys Ther 87: , Reilly T, Bangsbo J and Franks A. Anthropometric and physiological predispositions for elite soccer. J Sports Sci, 18: , Reilly T. An ergonomics model of the soccer training process. J Sports Sci 23: ,, Reinking MF, Bockrath-Pugliese K, Worrell T, Kegerreis RL, Miller-Sayers K, Farr J. Assessment of quadriceps muscle performance by hand-held, isometric, and isokinetic dynamometry in patients with knee dysfunction. Journal Orth Sports Phys Ther 24: , Reudi G, Schobersberger W, Pocecco E, et al. Sport injuries and illness during the first winter youth Olympic Games 2012 in Innsbruck, Austria. Br J Sports Med. 46: , Riach CL, Stark JL. Velocity of centre of pressure excursions as an indicator of postural control systems in children. Gait and Posture 2; , Riemann BL, Lephart SM. The sensorimotor system, part 1. The physiological basis for functional joint stability. J Athl Training 37: 71-79, Robson-Ansley P, Gleeson M, Ansley L. Fatigue management in the preparation of Olympic athletes. Journal of Sports Sciences 27: Rose MS, Emery CA, Meeuwisse WH. Sociodemographic predictors of sport injury in adolescents. Med Sci Sports Exerc 40: ,

292 308. Ross, M.D., B. Langford, and P.J. Whelan. Test-retest reliability of 4 single-leg horizontal hop tests. J Strength Cond Res 16: , Ross S, Guskiwicz K. Assessment tools for identifying functional limitations associated with functional ankle instability. J Athl Train. 43: 44-50, Ross S, Guskiewicz K, Gross M and Bing Y. Balance measures for discriminating between functionally unstable and stable ankles. Med Sci Sport and Exerc, 41(2): , Rossi SJ, McMillan J, and Buckley T. Seasonal changes of drop jump performance and recovery-stress scores in collegiate soccer players doi: /01.JSC b Rousanoglou EN, Barzouka KG, Boudolos KD.: Seasonal changes of jumping performance and knee muscle strength in under-19 women volleyball players. J Strength Cond Res 27: , Rumpf M, Cronin J. Injury incidence, body site, and severity in soccer players aged 6 18 years: implications for injury prevention. Strength Cond J 34: 20-31, Rumpf M, Cronin J, Mohamad I, Mohamad S, Oliver JL, Hughes M. Kinetic asymmetries during running in male youth. Phys Ther in Sport 15: 53-57, Ruan M and Li L. Approach run increases pre-activation and eccentric phases of muscle activity during drop jumps from different heights. J Electromyo Kinesiol 20: , Russell PJ, Croce RV, Swart EE. Knee muscle activation during landings: developmental and gender comparisons. Med Sci Sports Exerc. 39: , Sannicandro I, Rosa RA, De De Pascalis S, Piccino A. The determination of functional asymmetries in the lower limbs of young soccer players using the countermovement jump. The lower limbs asymmetry of young soccer players. Science and Sports 27: , Schmikli SL, de Vries WR, Inklaar H, Backx FJ. Injury prevention target groups in soccer: injury characteristics and incidence rates in male junior and senior soccer players. J Sci Med Sport 14: , Schmitz RJ, Schultz SJ, Nguyen AD. Dynamic valgus alignment and functional strength in males and females during maturation. J Athl Train 44: 26-32, Sconce E, Jones P, Turner E, Comfort P and Graham-Smith P. The validity of the Nordic hamstring lower for a field-based assessment of eccentric hamstring strength. J Sport Rehab 24: 13-20, Sharp D. Your chi-square test is statistically significant: now what? Practical Assessment, Research and Evaluation 20: 1-10, Shea KG, Pfieffer R, Wang JH, Curtin M, Apel PJ. ACL injury in pediatric and adolescent soccer players. An analysis of insurance data. J Pediatric Orthop 24: ,

293 323. Sheehan FT, Sipprell WH, 3rd, Boden BP. Dynamic sagittal plane trunk control during anterior cruciate ligament injury. Am J Sports Med 40: , Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psych Bull 86: 420-8, Sigward S.M, Pollard C.D, Powers C.M. The influence of sex and maturation on landing biomechanics: implications for ACL injury. Scand J Med Sci Sport 22: , Simonsen EB, Magnusson SP, Bencke J, et al. Can the hamstring muscles protect the anterior cruciate ligament during a side-cutting maneuver? Scand J Med Sci Sports 10: Small K, McNaughton L, Greig M, Lovell R. The effects of multidirectional soccerspecific fatigue on markers of hamstring injury risk. J Sci Med Sport 13: , Stark, T, Walker, B, Phillipas, JK, Fejer, R, Beck. Hand-held dynamometry correlation with the gold standard isokinetic dynamometry: a systematic review. PM&R 3: , Steindl R, Kunz K, Schrott-Fischer, Scholtz A.W. Effect of age and sex on maturation of the sensory systems and balance control. Developmental Medicine and Child Neurology 48: , Stheeman S, van t Hof M, Mileman P. Use of the Delphi technique to determine standards for quality assessment in diagnostic radiology. Community Dent Health 12: , Sugimoto D, Myer GD, McKeon JM, Hewett TE. Evaluation of the effectiveness of neuromuscular training to reduce anterior cruciate ligament injury in female athletes: a critical review of relative risk reduction and numbers-needed-to-treat analyses. Br J Sports Med 46: , Sugimoto D, Alernto-Geli E, Mediguchia J, Samuelsso K, Karlsson, Myer GD. Biomechanical and neuromuscular characteristics of male athletes: implications for the development of anterior cruciate ligament injury prevention programmes. Sports Med 45: , Swartz EE, Decoster LC, Russell PJ, Croce RV. Effects of developmental stage and sex on lower extremity kinematics and vertical ground reaction forces during landing. J Athl Train. 40: 9-14, Tak I, Weir A, Langhout R, Waarsing J, Stubbe J, Kerkhoffs G, Agricola R. The relationship between the frequency of football practice during skeletal growth and the presence of a cam deformity in adult elite football players. Br J Sports Med 49: , Timpka T, Risto,O, Bjo M. Boys soccer league injuries: a community-based study of time-loss from sports participation and long-term sequelae. European Journal of Public Health 1: 19 24, Turner A, Stewart P, Bishop C, Chavda S, Edwards M, Read P. Avoid overtraining and monitor fatigue. PSCJ 31: 19-27,

294 337. Turner A, Brazier J, Bishop C, Chavda S, Cree J, Read P. Data analysis for strength and conditioning coaches: using excel to analyze reliability, differences, and relationships. Strength Cond J 37: 76-83, Tursz A, Crost M. Sport s related injuries in children: a study of their characteristics, frequency and severity with comparison to other accidental injuries. AM J Sports Med 14: , van der Sluis A, Elferink-Gemser MT, Coelho-e-Sliva MJ, Nijboer JA, Brink MS and Visscher C. Sports injuries aligned to peak height velocity in talented pubertal soccer players. Int J Sports Med 35: , van der Sluis A, Elferink-Gemser MT, Brink MS, Visscher C. Importance of peak height velocity timing in terms of injuries in talented soccer players. Int. J. Sports Med 36: , Viru A, Loko J, Harro M, Volver A, Laaneaots L, and Viru M. Critical periods in the development of performance capacity during childhood and adolescence. Eur J Phys Educ 4: , Volpi P, Pozzoni R, Galli M. The major traumas in youth football. Knee Surg Sports Traumatol Arthrosc 11: , Waldeck MR, Lambert MI. Heart rate during sleep: implications for monitoring training status. JSSM 3: , Warren GL, Ingallis CP, Lowe DA, Armstrong RB. Excitation contraction uncoupling major role in contraction-induced muscle injury. Exerc Sports Sci Review 29: 82-7, Weir A, Darby J, Inklaar H, Koes B, Bakker E and Tol, J. Core Stability: Inter- and Intraobserver Reliability of 6 Clinical Tests. Clin J Sport Med 20: 34-38, Welsh TT, Alemany JA, Montain SJ, et al. Effects of intensified military field training in jumping performance. Int J Sports Med 29: 45-52, Whittington JE, Richards PN. The stability of children s laterality prevalences and their relationship to measures of performance. Br J Edu Psychol 57: 45 55, Wijdicks, CA, Griffith, CJ, LaPrade, R, Spiridonov, SI, Johansen, S, Armitage, BM and Engebretsen, L. Medial Knee Injury: Part 2, Load Sharing Between the Posterior Oblique Ligament and Superficial Medial Collateral Ligament Am J Sports Med 37: , Wikstrom EA, Tillman MD, Smith AN, Borsa PA. A new force-plate technology measure of dynamic postural stability: the dynamic postural stability index. J Athl Train 40: , Wikstrom EA, Tillman MD, Chmielewski TL, Borsa PA. Measurement and evaluation of dynamic joint stability of the knee and ankle after injury. Sports Med 36: , Williams GN, Chmielewski T, Rudolph KS, Buchanan TS and Snyder ML. Dynamic Knee Stability: Current Theory and Implications for Clinicians and Scientists. J Orthop Sports Phys Ther, 31: 546,

295 352. Williams CA, Oliver JL, Faulkener J. Seasonal Monitoring of Sprint and Jump Performance in a Soccer Youth Academy. J Sports Phys Perf 6: , Williams CA, Wood L, De Ste Croix M. Growth and maturation in childhood in; paediatric biomechanics and motor control: theory and application. Ed; De Ste Croix M, and Korff T. Routledge, Abingdon, UK, 15-17, Wisloff U, Castagna C, Helgerud J, Jones R and Hoff J. Strong correlation of maximal squat strength with sprint performance and vertical jump height in elite soccer players. BR J Sports Med 38: , Witvrouw E, Danneels L, Asselman P. D Have T, Cambier D. Muscle flexibility as a risk factor for developing muscle injuries in male professional soccer players: A prospective study. Am J Sports Med 31: 41-46, Wollom M, Purdam C, Drew MK. Reliability of externally fixed dynamometry hamstring strength testing in elite youth football players. J Sci Med Sport 19: 93-96, Wright MD, Portas MD, Evans E, Weston M. The effectiveness of four weeks of fundamental movement training on functional movement screen and physiological performance in physically active children. J Strength Cond Res 29: , Xu L, Nicholson P, Wang Q, Alen M, and Cheng S. Bone and muscle development during puberty in girls. A seven year longitudinal study. J Bone Mineral Research 24: , Yeow CH, Lee PVS, Goh, JCH. Sagittal knee joint kinematics and energetics in response to different landing heights and techniques. The Knee 17: , Young W, Farrow D. A review of agility: practical applications for strength and conditioning. Strength Cond J 28: Yu B, McClure SB, Onat, JA, Guskiewicz KM, Kirkendall DT, Garrett WE. Age and gender effects on lower extremity kinematics of youth soccer players in a stop-jump task. Am J Sports Med 33: , Zazulak BT, Hewett TE, Reeves NB, et al. The effects of core proprioception on knee injury: a prospective biomechanical epidemiological study. Am J Sports Med 35: , 2007a Zazulak B, Hewett T, Reeves P, Goldberg B, Cholewicki J. Deficits in neuromuscular control of the trunk predict knee injury risk: A prospective biomechanicalepidemiological study. Am J Sports Med 35: , 2007b Zebis MK, Andersen LL, Bencke J. Identification of athletes at future risk of ACL ruptures by neuromuscular screening. Am J Sports Med 37: ,

296 Appendices Appendix A EXAMPLE FORMS FOR ETHICS APPROVAL PARENTAL INFORMATION, PARENTAL CONSENT, PARTICIPANT INFORMATION AND PHYSICAL ACTIVITY READINESS QUESTIONNAIRE Appendix B STUDY 2 CROSS SECTIONAL ANALYSIS BETWEEN GROUP STATISTICS Appendix C PUBLISHED RESEARCH ARTICLES EMANATING FROM THE THESIS Appendix D STANDARDISED PROTOCOL INSTRUCTIONS FOR EACH TEST 296

297 Appendix A Ethics documentation, participant and parental information / consent forms, PAR-Q CARDIFF METROPOLITAN UNIVERSITY APPLICATION FOR ETHICS APPROVAL When undertaking a research or enterprise project, Cardiff Met staff and students are obliged to complete this form in order that the ethics implications of that project may be considered. If the project requires ethics approval from an external agency such as the NHS or MoD, you will not need to seek additional ethics approval from Cardiff Met. You should however complete Part One of this form and attach a copy of your NHS application in order that your School is aware of the project. The document Guidelines for obtaining ethics approval will help you complete this form. It is available from the Cardiff Met website. Once you have completed the form, sign the declaration and forward to your School Research Ethics Committee. PLEASE NOTE: Participant recruitment or data collection must not commence until ethics approval has been obtained. PART ONE Name of applicant: Supervisor (if student project): School: Paul Read Dr Rhodri Lloyd Student number (if applicable): Programmeme enrolled on (if applicable): Project Title: Cardiff School of Sport PhD Assessment and trainability of lower limb neuromuscular injury risk factors in elite male youth soccer players 297

298 Expected Start Date: 01/04/2014 Approximate Duration: Funding Body (if applicable): Other researcher(s) working on the project: Will the study involve NHS patients or staff? 4 Years F-Marc Centre Cardiff School of Sport, St Marys University Dr Rhodri Lloyd, Dr Jon Oliver, Professor Mark De Ste Croix No Will the study involve taking samples of human origin from participants? No In no more than 150 words, give a non technical summary of the project Neuromuscular control has been defined as activation of the dynamic restraints which occur in preparation for, and in response to, joint movements and forces providing functional joint stability (Riemann and Lephart, 2002), with deficits in an athlete s muscle strength, power or activation patterns increasing joint loads (Myer et al., 2004). Specific imbalances have been identified as a high risk factor in young girls, and male and female and adult populations (Hewett et al., 2002; Myer et al., 2011). However, less information is readily available in boys, and in particular, elite male youth soccer players despite their greater relative risk of injury during competition (Soderman et al., 2001). Therefore, the first aim of the proposed project is to identify reliable screening methods to determine deficits in neuromuscular within youth soccer players to determine who may be at a high risk of injury. It has also been suggested that neuromuscular control deficits can only be altered with appropriate interventions (Williams et al., 2001), however few studies to date have investigated the training and de-training effects of injury prevention protocols aimed at improving neuromuscular control in young boys. Thus, the assessment and trainability of lower limb neuromuscular control, and their subsequent effects on identified injury risk factors warrants further investigation in this population and is the basis of this project. Does your project fall entirely within one of the following categories: Paper based, involving only documents in No the public domain Laboratory based, not involving human No participants or human tissue samples Practice based not involving human No participants (eg curatorial, practice audit) Compulsory projects in professional practice No (eg Initial Teacher Education) If you have answered YES to any of these questions, no further information regarding your project is required. If you have answered NO to all of these questions, you must complete Part 2 of this form 298

299 DECLARATION: I confirm that this project conforms with the Cardiff Met Research Governance Framework Signature of the applicant: Date: 05/03/2014 P.READ FOR STUDENT PROJECTS ONLY Name of supervisor: Dr Rhodri Lloyd Signature of supervisor: Date: 05/03/2014 Research Ethics Committee use only Decision reached: Project reference number: 14/2/01R Project approved Project approved in principle Decision deferred Project not approved Project rejected Name: Peter O Donoghue Date: 06/03/2014 Signature: Details of any conditions upon which approval is dependant: Click here to enter text. 299

300 300

301 Participant Information Form (Study 1) Project Title: The Assessment and Trainability of Lower Limb Neuromuscular Control In Elite Male Youth Soccer Players UREC protocol number: Lead researcher: Paul Read Chief Supervisor: Dr. Rhodri S. Lloyd Contact details: Dear participant, Please read this information sheet carefully before deciding whether or not you want to take part. If you decide to volunteer we thank you for your participation. If you decide not to take part there will be no disadvantage to you of any kind and we thank you for considering our request. Aims of the research To find out how reliable a variety of jumping and balance tests are which assess movement skill and identify a football player s risk of injury. Also we hope to find out what are considered good and bad scores on each test for a range of high level academy soccer players so that we can make recommendations for what scores players in each age group should try to achieve. What will happen if you decide to volunteer? You will take part in a total of 4 sessions which will last for approximately 60 minutes. First we will find out your standing height, sitting height, weight, age and the flexibility at your ankles. After this we will then ask you to perform different types of jumping, hopping and balance tests on both one and two legs, and lastly we will also ask you to perform different types of squatting movements both on one and two legs. None of these tests we will ask you to do should make you feel tired as they are very short and there is plenty of rest between tests. The first session is a practice and you will try all of the tests a few times and then the other three sessions will be performed one week apart where we will measure the scores of each test to see how consistent the results are across all of the three recorded sessions. These tests are important as they are similar to a number of the movements that you would do in a game of football and we can tell you if you are currently at risk of getting injured, and if so, what you can do to prevent an injury from happening. Note: On each of the tests you will be watched by an assessor and also some of the tests will be filmed by two video cameras, one filming from the side and the other filming from the front. This 301

302 will allow us to compare which method is the best for assessing movement skills in young people and for us to be more accurate with our testing That s it! What type of participants do we want? We want to recruit boys who are currently at professional academy soccer clubs between the ages of 10, and 17 years of age. What are the risks of participating in the study? The risks of participating in the study are very low. This study is fully covered by Cardiff Metropolitan University s insurance in case you should suffer any adverse effects Benefits to the participant You will be given a record of your performance during the tests, which will help you understand more about your abilities as a soccer player and your relative risk of injury. What will happen to the information collected? Everyone that takes part in the study will receive their own results for the tests that they complete. All information collected will be held securely at the University and will only be accessible to relevant University staff (Paul Read and Dr Rhodri S. Lloyd). The University will securely hold the data recorded for seven years, after which it will be destroyed. Results of this project may be published in scientific journals but any data included will in no way be linked to any specific participant with no names used of either the players or the clubs. What next? Questions are always welcome at any time. If you should have any questions about the research then please contact me (details given at top of page). If you would like to participate in the study then the consent forms need to be signed by your parent/guardian and you the participant, and returned to myself. This project has been approved by the University Research Ethics Committee. Many thanks, Paul Read Paul Read MSc, ASCC, CSCS Senior Lecturer in Strength and Conditioning PhD Researcher (Paediatric Exercise Science) Principal Investigator School of Sport Health and Applied Sciences, St Mary s University Physiology and Health Department, Cardiff School of Sport 302

303 Parent/Guardian Information Form (Study 1) Project Title: Project Title: The Assessment and Trainability of Lower Limb Neuromuscular Control In Elite Male Youth Soccer Players UREC Protocol Number. Principal Investigator: Paul Read Principal Supervisor: Dr. Rhodri S. Lloyd Contact details: or Dear Parent/Guardian, Purpose of this information sheet This information sheet is to let you know about my planned research project in the Cardiff School of Sport at Cardiff Metropolitan University. It should help you decide on whether or not you want your child to join the study. Taking part in the research is entirely voluntary and should your child wish to withdraw from the study at any time, they are entitled to do so without any repercussions. Aims of the research The aim of this study is to determine how reliable and repeatable a range of testing methods are for assessing movement skill, and jumping and balance proficiency in youth elite academy male soccer players. To do so necessitates a range of assessors observing and rating the participant s competence whilst performing a series of balance, jumping and movement tests that have been selected for this study based on an extensive review of the literature. Real time observation will be measured against video analysis which has been validated as an effective method for assessing movement skills in various groups, including children. We will also use selected equipment to measure and record things such as jump height, distances and balance scores. The tests chosen for this study consist of movements that reflect the type of physical activities that players will undertake during a football match, i.e. jumping and landing, and thus, these tests are therefore considered important for them to master. The assessments consist of a range of hopping, jumping and balance tasks which they will perform on both one and two legs. In addition, basic anthropometrics (height, weight, sitting height and leg length) will be measured as well as a series movement skills consisting of activities such as squatting on both two legs and a single leg stance, and also balancing on one leg while trying to reach as far as possible in a variety of directions. The results from this battery of tests will provide vital information as to the movement and stabilisation characteristics of elite youth male soccer players which are not yet currently known, and also may assist in identifying a players relative risk of injury, whereby, lower performances on the series of tests may predispose players to an increased chance of sustaining an injury during training practices or match play. Further, by confirming which tests are reliable and the levels of performance on each test from a range of players in each age group, coaches and physiotherapists will be able to assess players more effectively and accurately to prevent a number of injuries which commonly occur in soccer, such as, knee injuries and ankle sprains. 303

304 Note: Whilst performing the various movement tests participants will be watched and rated by a range of assessors, whilst simultaneously filmed by two video cameras, one filming from the side and the other filming from the front. This will enable a comparison between video analyses which is recognised as a valid method for evaluating movement skills in youth, versus real time observation of the assessors. What will happen once you agree to participate in the study? Participants will be asked to complete an initial familiarisation session where they will practice all of the tests lasting roughly 60 minutes, and this will be followed by three further testing sessions where they will complete each test in turn so that we can assess how reliable each test it. The testing sessions will take place at the venue of their football club during scheduled training sessions and we will be one week apart. During the testing sessions participants will perform; A comprehensive, standardized warm-up The required test battery, including a range of movement skill, balance and jump-landing tasks performed on both one and two legs We will also collect some personal information from them (i.e. age, weight, standing height, sitting height, leg length and ankle flexibility) What type of participants are we hoping to use in the study? We are aiming to recruit approx from two different groups: Group 1) Players under 12 which will encompass the following age groups: U/12 s, U11 s; and Group 2) Players who are under 17 encompassing players from the following age groups; U/16 s and U/17 s. This is so we can look at a range of players at different stages of growth and maturation, i.e. pre and post puberty. What are the risks of participating in the study? The risks associated with the study are minimal. Each test lasts for only a few seconds, and participants will be allowed plenty of time to recover in between each test. There are some inherent risks when participating in any form of exercise; however the risks associated with the current study are no more likely to occur than if participants were taking part in any sort of match competition, training session, or PE lesson. The testing battery has been devised by the principal investigator who is an experienced senior lecturer and accredited strength and conditioning coach with the UK Strength and Conditioning and National Strength and Conditioning Associations (ASCC, CSCS). Benefits to the participant As a member of the research group, participants will be given a written record of their performances in the tests. This will provide them with information about their relative risk of injury and movement skills, which may help reduce the likelihood of experiencing an injury at their football club, and during any physical activity. It will also provide them with a hands-on experience of modern-day sport science fitness testing. Benefits to us, the research team 304

305 By completing the research, participants will provide the research team with relevant data which will be used to complete Paul Read s PhD thesis (the principal investigator). More importantly the findings of the study will provide the research team with important new information to publish in Internationally-renowned sport science journals. What will happen to the data and information collected during the study? The information and data we have about each participant will be coded so that players cannot be identified individually. Each participant will receive a copy of their test performances at the end of the testing period. Their performance data, including video performance analysis footage will only be seen by themselves and the research team. Copies of all data collected during the testing period will be stored centrally within a secure holding location in Cardiff Metropolitan University for up to a period of 7 years. Only the principal researcher and his supervisory team will be able to access the data once stored in Cardiff Metropolitan University. Results of this study may be published in scientific journals but participants will not be identified in a publication, and data included will in no way be associated with any named individual or club. What next? Please feel free to ask any question to a member of the research team at any time. You may contact either me or Dr Rhodri S. Lloyd on the above addresses should you have any concerns about the study. Having discussed this matter with your child, and if you would like them to take part in the study, please complete the Parental Consent Form, Participant Assent Form, and Physical Activity Readiness Questionnaire included with this information sheet and return to the most suitable point of contact This project has been approved by UREC (University Research Ethics Committee). Many thanks, Paul Read Paul Read MSc, ASCC, CSCS Senior Lecture Strength & Conditioning at St Marys University PhD Researcher (Exercise Physiology) and Principal Investigator Physiology Department, Cardiff School of Sport 305

306 Physical Activity Readiness Questionnaire (PAR-Q) Name of Participant: Please circle YES or NO to the following questions as appropriate: 1 Do you currently suffer from asthma or any breathing-related condition? YES / NO 2 Have you ever consulted your doctor as a result of suffering from a heart-related YES / NO condition? 3 Have you/do you suffer from any chest pains which may be aggravated by exercise? YES / NO 4 Do you suffer from bouts of dizziness or from feeling faint? YES / NO 5 Have you ever been told by a medical consultant that you suffer from a bone and/or joint condition which might be further aggravated by exercise? YES / NO 6 Have you ever been diagnosed with high blood pressure? YES / NO 7 Have you ever been diagnosed with diabetes? YES / NO 8 Are you unaccustomed to regular vigorous exercise? YES / NO 9 Is there a significant physical reason not mentioned above why you should not take part in the research project? YES / NO If you have circled YES to any of the questions, or, if there are any health and fitness related conditions that could affect your participation in the research which are not covered in questions 1-8, please provide further information below: Should your situation change regarding any of the conditions mentioned above, please notify one of the Researchers / member of staff/teacher as soon as possible. Signed (participant): Signed (parent / guardian): Signed (principal investigator) Date: 306

307 Appendix B Study 2 Cross sectional analysis between group statistics Y-balance right leg absolute reach between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Y-Balance right leg absolute reach between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Y-balance left leg absolute reach between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

308 Y-balance left leg absolute reach between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Y-balance right leg relative between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Y-balance right leg between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Y-balance left leg relative between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

309 Y-balance right leg between group effect sizes (d) U11 U U11 U12 U13 U14 U15 U16 U18 U U U U U Single leg hop for distance right leg absolute between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Single leg hop for distance right leg absolute between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg hop for distance left leg absolute between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

310 Single leg hop for distance right leg absolute between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg hop for distance left leg absolute between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Single leg hop for distance left leg absolute between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg hop for distance left leg relative between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

311 Single leg hop for distance left leg relative between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg countermovement jump right leg absolute between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Single leg countermovement jump right leg absolute between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg countermovement jump left leg absolute between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

312 Single leg countermovement jump left leg absolute between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg countermovement jump right leg relative between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Single leg countermovement jump right leg relative between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg countermovement jump left leg relative between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

313 Single leg countermovement jump left leg relative between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U %Hop right leg absolute between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U %Hop right leg absolute between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U %Hop left leg absolute between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

314 75%Hop left leg absolute between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U %Hop right leg relative between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U %Hop right leg relative between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U %Hop left leg relative between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

315 75%Hop right leg relative between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Single leg hop for distance asymmetry between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Single leg hop for distance asymmetry between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U Y-balance asymmetry between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U

316 Y-balance asymmetry between group effect sizes (d) U11 U U11 U12 U13 U14 U15 U16 U18 U U U U U Single leg countermovement jump asymmetry between group statistics (p values) U11 U12 U13 U14 U15 U16 U18 U U U U U U U Single leg countermovement jump asymmetry between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U %Hop asymmetry between group effect sizes (d) U11 U12 U13 U14 U15 U16 U18 U11 U U U U U U

317 Appendix C PUBLISHED RESEARCH ARTICLES EMANATING FROM THE THESIS 317

318 318

319 319

320 320

321 321